Documente Academic
Documente Profesional
Documente Cultură
4
DISEASE CONTROL PRIORITIES THIRD EDITION
Mental, Neurological,
and Substance Use
Disorders
DISEASE CONTROL PRIORITIES THIRD EDITION
Series Editors
Dean T. Jamison
Rachel Nugent
Hellen Gelband
Susan Horton
Prabhat Jha
Ramanan Laxminarayan
Charles N. Mock
Dean T. Jamison
Rachel Nugent
Hellen Gelband
Susan Horton
Prabhat Jha
Ramanan Laxminarayan
Charles N. Mock
VOLUME
4
DISEASE CONTROL PRIORITIES THIRD EDITION
Mental, Neurological,
and Substance Use
Disorders
EDITORS
Vikram Patel
Dan Chisholm
Tarun Dua
Ramanan Laxminarayan
Mara Elena Medina-Mora
2015 International Bank for Reconstruction and Development / The World Bank
1818 H Street NW, Washington, DC 20433
Telephone: 202-473-1000; Internet: www.worldbank.org
Some rights reserved
1 2 3 4 19 18 17 16
This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and
conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors,
or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The
boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the
part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries.
Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World
Bank, all of which are specifically reserved.
This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://creativecommons.org
/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this
work, including for commercial purposes, under the following conditions:
AttributionPlease cite the work as follows: Patel, V., D. Chisholm., T. Dua, R. Laxminarayan, and M. E. Medina-Mora, editors.
2015. Mental, Neurological, and Substance Use Disorders. Disease Control Priorities, third edition, volume 4. Washington, DC:
World Bank. doi:10.1596/978-1-4648-0426-7. License: Creative Commons Attribution CC BY 3.0 IGO
TranslationsIf you create a translation of this work, please add the following disclaimer along with the attribution: This
translation was not created by The World Bank and should not be considered an official World Bank translation. The World
Bank shall not be liable for any content or error in this translation.
AdaptationsIf you create an adaptation of this work, please add the following disclaimer along with the attribution:
This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole
responsibility of the author or authors of the adaptation and are not endorsed by The World Bank.
Third-party contentThe World Bank does not necessarily own each component of the content contained within the work.
The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in
the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely
with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed
for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited
to, tables, figures, or images.
All queries on rights and licenses should be addressed to the Publishing and Knowledge Division, The World Bank, 1818 H
Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org.
Softcover
ISBN (paperback): 978-1-4648-0426-7
ISBN (electronic): 978-1-4648-0428-1
DOI: 10.1596/978-1-4648-0426-7
Hardcover
ISBN (hardcover): 978-1-4648-0427-4
DOI: 10.1596/978-1-4648-0427-4
Cover photo: Curt Carnemark / World Bank. Further permission required for reuse.
Cover design: Debra Naylor, Naylor Design, Inc., Washington, DC.
Library of Congress Cataloging-in-Publication Data
Names: Patel, Vikram, editor. | Chisholm, Dan, editor. | Dua, Tarun, editor. | Laxminarayan, Ramanan, editor. | Medina-Mora,
Maria Elena, editor. | World Bank, issuing body,
Title: Mental, neurological, and substance use disorders / editors, Vikram Patel, Dan Chisholm, Tarun Dua,
Ramanan Laxminarayan, Maria Elena Medina-Mora.
Other titles: Disease control priorities ; v. 4. Description: Washington, DC : International Bank for Reconstruction and
Development /The World Bank, [2015] | Series: Disease control priorities ; volume 4 | Includes bibliographical references
and index.
Identifiers: LCCN 2015041175 (print) | LCCN 2015041905 (ebook) | ISBN 9781464804267 (alk. paper)
| ISBN 9781464804274 (alk : paper : hc) | ISBN 9781464804281 (ebook)
Subjects: | MESH: Mental Disorders. | Developing Countries. | Public Health. | Substance-Related Disorders.
Classification: LCC RA790.5 (print) | LCC RA790.5 (ebook) | NLM WA 395 | DDC
362.19689dc23
LC record available at http://lccn.loc.gov/2015041175
Contents
Foreword xi
Preface xiii
Abbreviations xv
1. Global Priorities for Addressing the Burden of Mental, Neurological, and Substance
Use Disorders 1
Vikram Patel, Dan Chisholm, Rachana Parikh, Fiona J. Charlson, Louisa Degenhardt,
Tarun Dua, Alize J. Ferrari, Steven Hyman, Ramanan Laxminarayan, Carol Levin, Crick Lund,
Mara Elena Medina-Mora, Inge Petersen, James G. Scott, Rahul Shidhaye, Lakshmi Vijayakumar,
Graham Thornicroft, and Harvey A. Whiteford, on behalf of the DCP MNS authors group
PART 1 BURDEN
2. Global Burden of Mental, Neurological, and Substance Use Disorders: An Analysis from
the Global Burden of Disease Study 2010 29
Harvey A. Whiteford, Alize J. Ferrari, Louisa Degenhardt, Valery Feigin, and Theo Vos
3. Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global
Burden of Disease Study 2010 41
Fiona J. Charlson, Amanda J. Baxter, Tarun Dua, Louisa Degenhardt, Harvey A. Whiteford, and Theo Vos
PART 2 INTERVENTIONS
5. Neurological Disorders 87
Kiran T. Thakur, Emiliano Albanese, Panteleimon Giannakopoulos, Nathalie Jette, Mattias Linde,
Martin J. Prince, Timothy J. Steiner, and Tarun Dua
ix
8. Childhood Mental and Developmental Disorders 145
James G. Scott, Cathrine Mihalopoulos, Holly E. Erskine, Jacqueline Roberts, and Atif Rahman
9. Suicide 163
Lakshmi Vijayakumar, Michael R. Phillips, Morton M. Silverman, David Gunnell, and Vladimir Carli
12. Cost-Effectiveness and Affordability of Interventions, Policies, and Platforms for the
Prevention and Treatment of Mental, Neurological, and Substance Use Disorders 219
Carol Levin and Dan Chisholm
13. Universal Health Coverage for Mental, Neurological, and Substance Use Disorders:
An Extended Cost-Effectiveness Analysis 237
Dan Chisholm, Kjell Arne Johansson, Neha Raykar, Itamar Megiddo, Aditi Nigam,
Kirsten Bjerkreim Strand, Abigail Colson, Abebaw Fekadu, and Stphane Verguet
x Contents
Foreword
I personally felt mental healths deep-rooted importance Although these steps may seem daunting, there is
when I returned home to Rwanda in 1996, just after my reason for hope. We can build on the lessons from the
people were traumatized by the 1994 Tutsi genocide. At worlds 15-year fight against HIV/AIDS. Across low- and
a time when we needed mental health services the most, middle-income countries (LMICs) in the 1990s, both
there was only one psychiatrist in the entire country. supply and demand for HIV/AIDS services were absent
In an act to survive and rebuild, we turned to our because there were no delivery platforms. No money or
communities for healing. Giving a voice to the people support was given to create a delivery structure. No laws
and collectively finding a solution to the mental health were written to protect the human rights of those stig-
challenges that we faced at that time has helped Rwanda matized by HIV/AIDS.
to resiliently move forward on a path toward recovery. Today, it is a drastically different story. Progress
This volume of Disease Control Priorities, third against HIV/AIDS for the past 15 years tells us that no
edition (DCP3), is thus a welcome call to action for evidence-based, multisectoral, holistic, and rights-based
augmenting the response needed to address the growing approach is too sophisticated for LMICs. It demonstrates
challenge of mental, neurological, and substance use that specialized referral service systems are possible,
(MNS) disorders. Such illnesses lurk in the shadows. even for one of the most complicated and stigmatized
Although they account for 10 percent of the global of conditions. It illustrates that as bidirectional supply
disease burden, they are left underestimated and unsup- and demand is created, the much-needed link between
ported worldwide. patients needs and an effective global care response will
In the pages that follow, the world has in its hands grow stronger.
a series of evidence-based approaches, cost-effective I challenge global leaders to build upon these lessons
strategies, and implementation guidelines for MNS learned from the HIV/AIDS response and apply it pos-
disorders. This comes at an opportune time. Changing itively to the challenge of MNS disorders. We must no
epidemiological and social determinant health profiles longer overlook the deleterious effects that the lack of
show the worlds readiness for sustainable development quality MNS services has upon our communities. We
goals (SDGs) to aim for universal health coverage. We, should strive to build universal health care systems spe-
as global leaders, have a moral obligation to advocate cifically recognizing MNS disorders genetic, biological,
for comprehensive, effective services backed by human- and cultural roots. And as a global community, I implore
rights-oriented legal frameworks to protect those living us to create enabling environments to address the social
with MNS disorders as part of this quest toward mean- determinants of health affecting MNS disorders.
ingful universal health coverage. Prioritizing the supply This call to action need not be answered alone; let us
of quality MNS services at the community level while work together as a global team to change the status quo
also improving the demand for such services must come and demand health equity for all.
with this advocacy effort.
Agnes Binagwaho, MD, MPed, PhD
Minister of Health, Rwanda
xi
Preface
Mental, neurological, and substance use (MNS) disor- those conditions that are associated with a significant
ders contribute approximately 10 percent of the global global burden. In doing so, we address the majority of
burden of disease. They often run a chronic course, the burden associated with these disorders. We have
are highly disabling, and are associated with significant organized these heterogeneous groups of disorders into
premature mortality. Moreover, beyond their health five groups: adult mental disorders, child mental and
consequences, the impact of these disorders on the social developmental disorders, neurological disorders, alcohol
and economic well-being of individuals, families, and use disorders, and illicit drug use disorders. The volume
societies is enormous. also addresses suicide and self-harm, which are strongly
Despite this burden, MNS disorders have been sys- associated with MNS disorders.
tematically neglected in most of the world, particu- In addition to providing an up-to-date synthesis of
larly in low- and middle-income countries (LMICs), the burden, prevalence, determinants, and interventions
with pitifully small contributions to prevention and for prevention and care of the selected disorders, the
treatment by governments and development agencies. volume offers a number of novel contributions to the
Systematically compiling the substantial evidence that policy-relevant evidence on MNS disorders.
already exists to address this inequity is the central goal
of volume 4 of Disease Control Priorities, third edition First, we present a systematic analysis of the excess
(DCP3). The evidence presented in this volume will help mortality associated with these disorders, enhancing
to build an evidence-based perspective on which policies our understanding of the true burden of disease
and interventions for addressing MNS disorders should attributable to them.
be prioritized in resource-constrained settings. These Second, the discussion of interventions embraces a
recommendations will be of relevance to ministries of health system perspective, such that, after a review of
health andgiven the intersectoral nature of the inter- the effective interventions for specific disorders, these
ventions and impacts of MNS disordersto ministries are then organized according to how they might be
of health and social welfare, as well as to institutions delivered across three distinct and complementary
and donors concerned with sustainable development. platforms: population, community, and health and
Reaching a broader audience of academics, research social care. This approach allows us not only to reflect
organizations, and public health practitioners is another on how interventions are planned and delivered in
goal of this effort. health systems, but also to highlight the potential
MNS disorders include a large number of discrete opportunities, synergies, and efficiencies for resource
health conditions, each with its own epidemiological allocation.
characteristics and interventions for prevention and Third, in addition to a review of the recent evidence
care. These disorders, like most chronic noncommu- for cost-effectiveness, the efforts to scale up the com-
nicable diseases, are caused by complex interactions munity-based services for mental health in selected
among genetic, biological, social, and psychological LMICsIndia and Ethiopiahave been exam-
determinants. In this volume, we chose to address only ined through the lens of extended cost-effectiveness
xiii
analysis to consider the distribution of costs and The findings of this volume make an emphatic case
outcomes, as well as the extent to which policies offer for a substantially increased investment in the preven-
financial protection to households. tion of and care for MNS disorders. We document highly
cost-effective strategies for the prevention of some MNS
We thank the large international group of authors disorders and affordable models of care for the deliv-
who have contributed to the development of the volume ery of treatment interventions in routine health care
for their time, effort, and thoroughness and for presenta- platforms through nonspecialist health workers. Such
tion of the evidence succinctly. We hope readers will find investments make economic sense for two reasons: the
that the exhaustive information the authors have synthe- interventions we recommend are cost-effective, and the
sized is presented in a manner that is clear and engaging. impact of these interventions on social and economic
We thank the Bill & Melinda Gates Foundation for outcomes is immense. The counterfactual situation of
providing funding support to the DCP3, the Institute of not doing enough, which prevails in most populations,
Medicine for coordinating the peer-review process, and is leading to enormous loss of human capital and will
the World Bank staff who coordinated the publication hinder the ambition of sustainable development. The
of the volume. We are grateful to the DCP3 secretariat, evidence in this volume can be translated into practice
in particular, Dean Jamison and Rachel Nugent, for only with strong political will and commitment from
their expert inputs on various chapters. In addition, we the governments and developmental agencies who now
thank Brianne Adderley, Kristen Danforth, and Elizabeth have to make the necessary investments in their scale-up.
Brouwer for their unstinting support, and Rachana We have the evidence to act. There is a moral case to
Parikh for coordinating the volume. act. The time to act is now.
Vikram Patel
Dan Chisholm
Tarun Dua
Ramanan Laxminarayan
Mara Elena Medina-Mora
xiv Preface
Abbreviations
xv
ECEA extended cost-effectiveness analysis
EEG electroencephalogram
EOD early-onset dementia
ES effect size
FAS Fetal Alcohol Syndrome
FASD Fetal Alcohol Syndrome Disorders
FRP financial risk protection
GBD Global burden of disease
GBD 2010 Global Burden of Disease Study 2010
g/dl grams per deciliter
GDP gross domestic product
GHE Global Health Estimates
GNI gross national income
GRADE Grading of Recommendations Assessment, Development and Evaluation
HCV hepatitis C
HICs high-income countries
HIV/AIDS human immunodeficiency virus/acquired immune deficiency syndrome
HIV human immunodeficiency virus
HMT heroin maintenance treatment
HR hazard ratio
IASC Inter-Agency Standing Committee
ICD International Classification of Diseases
ICT information and communications technology
IHD ischemic heart disease
IHME Institute for Health Metrics and Evaluation
IMAI Integrated Management of Adult and Adolescent Illness
IOM Institute of Medicine
IQs intelligence quotients
INCB International Narcotics Control Board
IOM Institute of Medicine
IQR interquartile range
LICs low-income countries
LMICs low- and-middle-income countries
MCH maternal and child health
MDMA 3,4-methylenedioxy-N-methylamphetamine
MDPV methylenedioxypyrovalerone
MHaPP Mental Health and Poverty Project
mhGAP Mental Health Gap Action Programme
MICs middle-income countries
MMT methadone maintenance treatment
MNS mental, neurological, and substance use
MOH medication-overuse headache
MSIC Medically Supervised Injecting Centre
NIAAA National Institute of Alcohol Abuse and Alcoholism
NCD noncommunicable disease
NICE National Institute for Health and Clinical Excellence
OCD obsessive-compulsive disorder
ONDCP Office of National Drug Control Policy
OOP out-of-pocket
OR odds ratio
OST opioid substitution treatment
PAF population attributable fractions
PC101 Primary Care 101
xvi Abbreviations
PHC primary health care
PRIME Programme for Improving Mental health carE
PSST problem-solving skills therapy
PTSD post-traumatic stress disorder
QA quality assurance
QALYs quality-adjusted life years
QI quality improvement
RR relative risk
RCT randomized controlled trial
SAPS South African Police Service
SAR Special Administrative Region
SDG sustainable development goal
SEL social emotional learning
SHR sustained headache relief
SIFs supervised injecting facilities
SMART Self-Management and Recovery Training
SMDs severe mental disorders
SMR standardized mortality ratio
SNRIs serotonin-norepinephrine reuptake inhibitors
SSRIs selective serotonin reuptake inhibitors
TC therapeutic community
TCA tricyclic antidepressant
TPO Transcultural Psychosocial Organization
TTH tension-type headache
TQ Ten Question
UHC universal health coverage
UI uncertainty interval
UMICs upper middle-income countries
UNDCP United Nations International Drug Control Programme
UNODC United Nations Office on Drugs and Crime
UPF universal public finance
WHO World Health Organization
WMH World Mental Health
WONCA World Organization of Family Doctors
YLDs years lived with disability
YLLs years of life lost
Abbreviations xvii
Chapter
1
Global Priorities for Addressing the
Burden of Mental, Neurological, and
Substance Use Disorders
Vikram Patel, Dan Chisholm, Rachana Parikh, Fiona J. Charlson,
Louisa Degenhardt, Tarun Dua, Alize J. Ferrari, Steven Hyman,
Ramanan Laxminarayan, Carol Levin, Crick Lund, Mara Elena
Medina-Mora, Inge Petersen, James G. Scott, Rahul Shidhaye,
Lakshmi Vijayakumar, Graham Thornicroft, and Harvey A.
Whiteford, on behalf of the DCP MNS authors group
Corresponding author: Vikram Patel, Public Health Foundation of India, the London School of Hygiene & Tropical Medicine, and Sangath, Goa, India,
vikram.patel@lshtm.ac.uk.
1
Box 1.1
Budgets constrain choices. Policy analysis helps platforms and packages, and by offering explicit con-
decision makers achieve the greatest value from sideration of the financial risk protection objective
limited available resources. In 1993, the World Bank of health systems. In populations lacking access to
published Disease Control Priorities in Developing health insurance or prepaid care, medical expenses
Countries (DCP1), an attempt to assess the cost- that are high relative to income can be impover-
effectiveness (value for money) of interventions in a ishing. Where incomes are low, seemingly inex-
systematic way that would address the major sources pensive medical procedures can have catastrophic
of disease burden in low- and middle-income coun- financial effects. DCP3 offers an approach that
tries (Jamison and others 1993). The World Banks explicitly includes financial protection as well as the
1993 World Development Report on health drew distribution across income groups of financial and
heavily on the findings in DCP1 to conclude that health resulting from policies (for example, public
specific interventions against noncommunicable finance) to increase intervention uptake (Verguet,
diseases were cost-effective, even in environments in Laxminarayan, and Jamison 2015).
which substantial burdens of infection and under- The task in all DCP volumes has been to combine the
nutrition persisted. available science about interventions implemented
DCP2, published in 2006, updated and extended in very specific locales and under very specific con-
DCP1 in several respects, including explicit con- ditions with informed judgment to reach reasonable
sideration of the implications for health systems of conclusions about the impact of intervention mixes
expanded intervention coverage (Jamison and oth- in diverse environments. The broad aim of DCP3
ers 2006). One way that health systems expand inter- is to delineate essential intervention packages
vention coverage is through selected platforms that such as the package for mental, neurological, and
deliver interventions that require similar logistics but substance use disorders, in this volumeand their
address heterogeneous health problems. Platforms related delivery platforms. This information will
often provide a more natural unit for investment assist decision makers in allocating often tightly
than do individual interventions, but conventional constrained budgets so that health system objectives
health economics has offered little understanding of are maximally achieved.
how to make choices across platforms. Analysis of DCP3s nine volumes are being published in 2015
the costs of packages and platformsand the health and 2016 in an environment in which serious dis-
improvements they can generate in given epidemio- cussion continues about quantifying the sustainable
logical environmentscan help guide health system development goal (SDG) for health (UN 2015).
investments and development. DCP3s analyses are well-placed to assist in choosing
DCP3 differs substantively from DCP1 and DCP2 the means to attain the health SDG and assessing the
by extending and consolidating the concepts of related costs for scaled-up action.
of specialist services in these settings. Such services have we have considered interventions for five groups
been the hallmark of the health system response to these of disordersadult mental disorders, child men-
conditions in high-income countries (HICs). tal and developmental disorders, neurological dis-
DCP1 had only addressed a few MNS disorders: orders, alcohol use disorder, and illicit drug use
psychosis and bipolar disorder. DCP2 had focused such as opioid dependenceand suicide and self-
on the cost-effectiveness of specific interventions harm-health outcomes strongly associated with MNS
for burdensome disorders, organized separately for disorders. Within each group, we have prioritized
mental disorders, neurological disorders, alcohol use conditions associated with high burden for which
disorders, illicit drug use disorders, and learning there is evidence in support of interventions that are
and developmental disabilities. In this third edition, cost-effective and scalable.
A range of social determinants influences the risk for MNS disorders (the social causation pathway);
and outcome of MNS disorders. In particular, the on the other hand, people living with MNS disor-
following factors have been shown to be associated ders drift into poverty during the course of their life
with several MNS disorders (Patel and others 2009): through increased health care expenditures, reduced
economic productivity associated with the disability
1. Demographic factors, such as age, gender, and of their condition, and stigma and discrimination
ethnicity associated with these conditions (the social drift
2. Socioeconomic status: low income, unemploy- pathway).
ment, income inequality, low education, and low
social support Understanding the vicious cycle of social determi-
3. Neighborhood factors: inadequate housing, over- nants and MNS disorders provides opportunities for
crowding, neighborhood violence interventions that target social causation and social
4. Environmental events: natural disasters, war, drift. In relation to social causation, the evidence
conflict, climate change, and migration. for the mental health benefits of poverty alleviation
5. Social change associated with changes in income, interventions is mixed but growing. In relation
urbanization, and environmental degradation to social drift, the evidence for the individual and
household economic benefits of the prevention and
The causal mechanisms of the social determinants of treatment of MNS disorders is compelling, and sup-
MNS disorders indicate a cyclical pattern. On the one ports the economic argument for scaling up these
hand, socioeconomic adversities increase the risk interventions (Lund and others 2011).
Inevitably, such an approach does not address a Second, we address the question of what by reviewing
significant number of conditions, for example, mul- the evidence on the effectiveness of specific interven-
tiple sclerosis as a neurological disorder and anorexia tions for the prevention and treatment of a selection
nervosa as an adult mental disorder. However, the rec- of MNS disorders.
ommendations in this volume, particularly regarding Third, we consider how and where these interventions
the delivery of packages for care, could be extended to can be appropriately implemented across a range of
other conditions not expressly addressed. In addition, service delivery platforms.
some important MNS disorders or concerns are cov- Fourth, we address the question of how much by
ered in companion volumes of DCP3, notably, nicotine examining the cost of scaling up cost-effective inter-
dependence, early childhood development, neurological ventions and the case for enhanced service coverage
infections, and stroke. and financial protection for MNS disorders.
This volume addresses four overall questions and
themes (box 1.3): This chapter also considers how some countries
have attempted to incorporate this body of evidence
First, we address the question of why MNS disorders into scaled-up programs for MNS disorders. The
deserve prioritization by pointing to and reviewing chapter discusses lessons on barriers and strategies
the health and economic burden of disease attrib- for how these will need to be addressed for successful
utable to MNS disorders. We build on the 2010 scaling-up.
estimates of the Global Burden of Diseases, Injuries, The primary focus of the volumeand DCP3 as a
and Risk Factors Study (GBD 2010) in two important wholeis on LMICs. We include HICs in the section
ways: by examining trends in the burden over time, on global disease burden, and we draw liberally on the
and by estimating the additional mortality attribut- concentration of available evidence on intervention
able to these disorders. effectiveness from these countries.
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders 3
Box 1.3
Key Messages
This volume of the third edition of Disease Control 3. Best practice interventions for MNS disorders
Priorities addresses mental, neurological, and sub- can be appropriately implemented across a
stance use (MNS) disorders. These heterogeneous range of population, community, and health care
conditions share several characteristics, not least that platforms.
they are among the most neglected of diseases glob- At the population-level platform of service
ally. This volume focuses on those conditions asso- delivery, best practices include legislative and
ciated with the greatest burden for which there are regulatory measures to restrict access to means
effective and scalable interventions. The key findings of self-harm/suicide and reduce the availabil-
and messages of the volume are presented in this ity of and demand for alcohol.
overview chapter, as well as an assessment of critical At the community-level platform, best prac-
health system barriers to scaling up evidence-based tices include life skills training in schools to
interventions and how to overcome them. build social and emotional competencies in
The following are the key messages: children and adolescents.
At the health care platform, which covers
self-care, primary health care, and hospital
1. The burden of MNS disorders is large, growing, and care delivery channels, best practices include
underestimated. self-management of migraine; diagnosis and
The public health burden of MNS disorders, as management of epilepsy, headache, depres-
estimated by disability-adjusted life years, is on a sion, anxiety, alcohol and illicit drug use dis-
sharp upward trajectory; it increased by 41 percent orders; and continuing care of schizophrenia
between 1990 and 2010 and now accounts for one and bipolar disorder in primary care.
in every 10 years of lost health globally. Even this 4. Public financing of scaling-up is affordable and
sobering statistic is an underestimate, because it increases financial protection.
does not explicitly take into consideration either The costs of providing a significantly scaled-up
the substantial excess mortality associated with package of specified cost-effective interventions
these disorders, estimated in this volume for the for prioritized MNS disorders is estimated at
first time, or the enormous social and economic US$3US$4 per capita of total population per
consequences of MNS disorders on affected per- year in low- and lower-middle-income countries,
sons, their caregivers, and societies. and at least double that in upper-middle-income
2. Many MNS disorders can be prevented and treated countries. This package includes interventions at
effectively. the population, community, and health care lev-
A wide variety of effective interventions can pre- els. Since a significant proportion of MNS disor-
vent and treat MNS disorders. Although some ders may run a chronic and disabling course and
of these interventions are also supported by adversely affect household welfare, it is important
evidence of cost-effectiveness, significant gaps that intervention costs are largely met by gov-
remain in the availability of evidence to support ernments through increased resource allocation
the scaling-up of many interventions. Some of and financial protection measures. Investment of
these interventions can have significant impacts public resources in the prevention and treatment
on other global health and development prior- of MNS disorders addresses a large and neglected
ities. For example, the effective management of public health concern; if targeted wisely, this
maternal depression can affect child health out- investment will produce substantial economic
comes, and the effective management of conduct as well as health benefits in populations at an
disorders in children can affect adult antisocial affordable cost. A policy of moving toward uni-
and criminal behavior. versal public finance can lead to a far more
equitable allocation of public health resources lack of strong and technically sound leadership to
across income groups. guide the scaling-up effort, the relatively low levels
As many countries and the global community move of demand for care for some of the most common
toward a consensus on the need for universal health conditions, the high levels of stigma attached to
coverage, this volume provides clear recommenda- many conditions, and the continuing reliance on
tions about which interventions should be priori- specialized hospital-based care as the primary deliv-
tized, how they can be delivered, and the expected ery platform.
cost of scaling up these interventions. We provide Realizing the health gains associated with the inter-
evidence from four countries to demonstrate how a ventions recommended in this volume will require
combination of political will and increased financial more than financial resources. Committed and
commitment to support the delivery of cost-effective sustained efforts will be needed to address these
preventive and treatment interventions through barriers. The ultimate goal is massively increasing
public systems can lead to significant improvements opportunities for persons with MNS disorders to
in service coverage and health outcomes. In most access services without the prospect of discrimi-
countries, a range of health system barriers will need nation or impoverishment, and with the hope of
to be addressed to achieve these goals, not least the attaining optimal health and social outcomes.
WHY MNS DISORDERS MATTER FOR 18.6 percent of total DALYs for individuals aged 15 to
GLOBAL HEALTH 49 years, compared with 10.4 percent for all ages com-
bined. Within the 15 to 49 years age group, mental and
The GBD 2010 identified MNS disorders as significant substance use disorders were the leading contributor to
causes of the worlds disease burden (Whiteford and the total burden caused by MNS disorders. For neuro-
others 2013). The DCP3 series as a whole uses the Global logical disorders, DALYs were highest in the elderly.
Health Estimates of disease burden. This volume also There are important gender differences in the
includes data from the 2010 GBD study, which are used in burden of these disorders. Overall, males accounted
the burden calculations presented in chapter 3 (Charlson for 48.1 percent and females for 51.9 percent of DALYs
and others 2015). The broad patterns conveyed are the for MNS disorders. Males accounted for more DALYs for
same across the 2010 GBD study (Whiteford and others mental disorders occurring in childhood, schizophrenia,
2013), the more recent 2013 GBD data (Global Burden substance use disorders, Parkinsons disease, and
of Disease Study 2013 Collaborators 2015), and WHOs epilepsy; whereas, more DALYs accrued to females for
Global Health Estimates (WHO 2014). all other disorders in this group. The relative proportion
In chapter 2 in this volume (Whiteford and others of DALYs for MNS disorders to overall disease burden
2015), we investigate trends in the burden caused by was estimated to be 1.6 times higher in HICs (15.5 per-
MNS disorders. There was a 41 percent increase in cent of total DALYs) than in LMICs (9.4 percent of total
absolute disability-adjusted life years (DALYs) caused by DALYs), largely because of the relatively higher burden
MNS disorders between 1990 and 2010, from 182 million of other health conditions, such as infectious and peri-
to 258 million DALYs (the proportion of global disease natal diseases, in LMICs. However, because of the larger
burden increased from 7.3 to 10.4 percent). With the population of LMICs, absolute DALYs for MNS disor-
exception of substance use disorders, which increased ders are higher in LMICs compared with HICs.
because of changes in prevalence over time, this increase Data from GBD 2010 on burden caused by pre-
was largely caused by population growth and aging. mature mortality may incorrectly lead to the inter-
DALYs are constituted of two components: years of pretation that premature death in people with MNS
life lost (YLLs) and years lived with disability (YLDs). disorders is inconsequential. This interpretation is due
Figure 1.1 summarizes the proportion of all-cause to how causes of deaths are assigned in the International
YLLs and YLDs explained by MNS disorders in 2010. Classification of Diseases (ICD) death coding system
As a group, MNS disorders were the leading cause of used by GBD 2010. Yet, evidence shows that people with
YLDs in the world. In 2010, DALYs for MNS disorders MNS disorders experience a significant reduction in life
were highest during early to mid-adulthood, explaining expectancy, with the risk of mortality increasing with
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders 5
Figure 1.1 Proportion of Global YLDs and YLLs Attributable to Mental, Neurological, and Substance Use Disorders, 2010
a. YLLs b. YLDs
Communicable
Mental diseases
disorders 15.5%
18.9% Injuries
Non communicable 5.9%
Communicable
diseases (excluding diseases
MNS disorders) Neurological
43.7% disorders
40.5%
5.6%
Non communicable
diseases (excluding
MNS disorders)
Injuries 50.2%
13.5%
the severity of the disorder (Chang and others 2011; be interpreted carefully. Table 1.1 summarizes cause-
Lawrence, Hancock, and Kisely 2013; Walker, McGee, specific and excess deaths attributable to each MNS dis-
and Druss 2015). order. Comparative risk analyses have also highlighted
Therefore, chapter 3 in this volume (Charlson and mental and substance use disorders as significant risk
others 2015) explores differences between the GBD factors of premature death from a range of other health
2010 estimates of cause-specific and excess mortality outcomes (Lim and others 2012). For example, an esti-
of these disorders, and potential contributors to life mated 60 percent of suicide deaths can be re-attributed
expectancy gaps. Although reported YLLs accounted for to mental and substance use disorders, elevating them
only 15.3 percent of MNS disorder DALYs, equivalent from the fifth to third leading cause of burden of dis-
to 840,000 deaths, natural history models generated by ease (Ferrari and others 2014). These findings strongly
DisMod-MR (a disease modeling tool) estimate that suggest the importance of continued assessment of the
substantially more deaths are associated with these dis- role MNS disorders play in premature death and as risk
orders. Excess deaths associated with major depression factors for other health outcomes.
alone were estimated at more than 2.2 million in 2010. The estimates of disease burden do not fully take
This figure is significantly higher than other attempts to into account the significant social and economic con-
quantify these deaths (Walker, McGee, and Druss 2015), sequences of MNS disorders, not only for affected indi-
and indicates a potentially higher degree of mortality viduals and households, but also for communities and
associated with MNS disorders than that captured by economies. Notable examples of such impacts include
GBD 2010 YLLs. the effects of maternal mental disorders on the well-
Since these estimates of excess deaths include deaths being of children, contributing to the intergenerational
from causal and non-causal origins, however, they must transmission of ill-health and poverty; the effects of
Major depressive disorder 0 2,224,000 Suicide and comorbid disease such as cardiovascular
(1,900,0002,586,000) disease and infectious disease.
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders 7
substance use disorders on criminal behavior and incar- and young adulthood; and dementia late in life. The
ceration; and the effects of a range of severe conditions epidemiologies of these disorders share some important
on the economic productivity of affected persons and characteristics: with the exception of dementia, the vast
family members engaged in caregiving. majority of cases have their onset before age 30 years
A recent study estimated that total economic out- and most tend to run a chronic or relapsing course. In
put lost to MNS disorders globally was US$8.5 trillion addition, several of the disorders are associated with
in 2010, a sum expected to nearly double by 2030 if a other health concerns. For example, injecting drug use
concerted response is not mounted (Bloom and others is associated with HIV/AIDS, alcohol use disorders are
2011). A separate study estimated the economic costs associated with road traffic injuries and liver cirrhosis,
attributable to alcohol use and alcohol use disorders to depression is associated with cardiovascular disease, and
amount to the equivalent of between 1.3 and 3.3 percent maternal depression is associated with child undernu-
of gross domestic product (GDP) in a range of high- trition and delayed cognitive development (Prince and
and middle-income countries, with over two-thirds of others 2007).
the loss represented by productivity losses (Rehm and The evidence on interventions presented in this
others 2009). section builds on the work published in DCP2 and its
The global cost of dementia in 2010 was estimated findings (Chandra and others 2006; Hyman and others
to be US$604 billion, equivalent to 1 percent of global 2006; Rehm and others 2006). The evidence is derived
GDP (WHO 2012). In addition, a rising tide of social from various sources: the mhGAP guidelines developed
adversities is associated with MNS disorders (box 1.2). by the World Health Organization (WHO) for use in
Moreover, large and growing proportions of the global non-specialized health settings, which used the Grading
population have been affected by conflict or displace- of Recommendations Assessment, Development and
ment because of environmental degradation and climate Evaluation (GRADE) methodology to review the litera-
change, which bodes for a grim forecast on the future ture published up to 2009 (Dua and others 2011); other
burden of these conditions. recent reviews, where appropriate, such as Strang and
Finally, the disease burden estimates do not account others (2012) for illicit drugs; interventions that require
for the significant hazards faced by persons with MNS a specialist for delivery but that were not addressed by
disorders in relation to the systematic denial of basic mhGAP or DCP2, assessed with GRADE; and a review
human rights. These costs range from limited oppor- of all reviews. The review of all reviews includes sys-
tunities for education and employment, to torture and tematic reviews and any type of evaluation evidence
denial of freedom, sometimes within health care institu- from LMICs published since mhGAP and assessed with
tions (Patel, Kleinman, and Saraceno 2012). GRADE. The findings are summarized in table 1.2.
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders
9
10
Table 1.2 Effective Interventions for the Prevention, Treatment, and Care of Mental, Neurological, and Substance Use Disorders (continued)
Type of disorder Preventive interventions Drug and physical interventions Psychosocial interventions
Intellectual disability Significantly impaired cognitive functioning and Psychosocial stimulation of infants Parental education and skills
(idiopathic) deficits in two or more adaptive behaviors and young children;* training*
(0.4% of total MNS perinatal interventions, for Educational support*
NEUROLOGICAL DISORDERS
Migraine Episodic attacks where headache and nausea are Prophylactic drug treatment with Drug treatments, aspirin or one of Behavioral and cognitive
(8.7% of total MNS the most characteristic attack features; headache propranolol or amitriptyline*** several other nonsteroidal anti- interventions*
DALYs) lasting for hours to 23 days, typically moderate inflammatory drugs***
or severe and likely to be unilateral, pulsating,
and aggravated by routine physical activity
Epilepsy A brain disorder traditionally defined as the Population-based interventions Standard anti-epileptic medications
(6.8% of total MNS occurrence of two unprovoked seizures occurring targeting epilepsy risk factors (phenobarbital, phenytoin,
DALYs) more than 24 hours apart with an enduring (preventing head injuries, carbamazepine, valproic acid);***
predisposition to generate further seizures neurocysticercosis prevention)* epilepsy surgery**
Dementia A neuropsychiatric syndrome characterized by a Cardiovascular risk factors Cholinesterase inhibitors and Caregiver education and
(4.4% of total MNS combination of progressive cognitive impairment, management (healthy diet, physical memantine for cognitive functions; support***
DALYs) BPSD, and functional difficulties activity, tobacco use cessation)* medications for management of Behavioral training and
BPSD* environmental modifications **
Interventions to support
caregivers of people with
dementia**
table continues next page
Table 1.2 Effective Interventions for the Prevention, Treatment, and Care of Mental, Neurological, and Substance Use Disorders (continued)
Type of disorder Preventive interventions Drug and physical interventions Psychosocial interventions
SUBSTANCE USE DISORDERS
Alcohol use disorders Harmful use is a pattern of alcohol use that Excise taxes*** Naltrexone, acamprosate* Family support*
(6.9% of total MNS causes damage to physical or mental health Restriction on sales** Motivational enhancement, brief
DALYs) Alcohol dependence is a cluster of physiological, Minimum legal age** advice, CBT**
behavioral, and cognitive phenomena in which Screening and brief
the use of a substance takes on a much higher Drunk driving countermeasures**
interventions***
priority for a given individual than other behaviors Advertising bans*
that once had greater value Self-help groups*
Restrictions on density*
Opening and closing hours and days
of sale**
Family interventions*
Illicit drug use disorders A pattern of regular use of illicit drugs Psychosocial interventions with Opioid substitution therapy Self-help groups, psychological
(7.8% of total MNS characterized by significantly impaired control primary school children, such as the (methadone, buprenorphine)*** interventions, CBT*
DALYs) over use and physiological adaptation to regular Good Behavior Game or Strengthening
consumption as indicated by tolerance and Families Program*
withdrawal
SUICIDE AND SELF-HARM
Suicide and self-harm The act of deliberately killing oneself; suicide Policies and legislation to reduce Effective drug interventions for Social support and psychological
(1.47% of GBD; 22.5 attempt refers to any nonfatal suicidal behavior access to the means of suicide (such underlying MNS disorders** therapies for underlying MNS
million YLLs or 62.1% and intentional self-inflicted poisoning, injury, or as pesticides)*** Emergency management of disorders, Planned follow-up and
of suicide YLLs are self-harm that may or may not have a fatal intent Decriminalization of suicide* poisoning** monitoring of suicide attempters*
attributed to mental and or outcome
Responsible media reporting of
substance use disorders suicide*
in 2010)
Notes: ADHD = Attention Deficit Hyperactivity Disorder; BPSD = behavioral and psychological symptoms; CBT = cognitive behavioral therapy; DALY = disability-adjusted life year; ECT = electroconvulsive therapy; GBD = Global Burden of Diseases;
MNS = mental, neurological, and substance use; YLLs = years of life lost.
*** = evidence of cost-effectiveness; ** = strong evidence of effectiveness but not cost-effectiveness;
* = modest evidence of effectiveness and either no cost-effectiveness or no evidence of cost-effectiveness.
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders
11
Certain preventive interventions that are primarily example, the symptoms associated with depression or
intended to target disorders covered in other DCP3 anxiety disorders are commonly interpreted as being
volumes, for example, to prevent cardiovascular diseases normative consequences of social adversity, and proven
or neurocysticercosis, will also have benefits for disorders biomedical or psychological causal models are rare,
covered in this volume, such as dementia and epilepsy, leading to low demand for care and low visibility of the
respectively. Conversely, some interventions targeting condition from the view of health policy makers and
MNS disorders are also associated with benefits to health providers (Aggarwal and others 2014). It is clear that
outcomes for other disorders. Examples include injury these competing views will affect the societal preference
prevention as a result of reduced alcohol or drug use or for and acceptability of investment in the wider adop-
effective treatment of Attention Deficit Hyperactivity tion of effective interventions for MNS disorders. More
Disorder, reduced antisocial behaviors and associated generally, stigma, lack of awareness, and discrimination
social consequences as a result of treatment of conduct are major factors behind low levels of political commit-
disorders in childhood, improved cardiovascular health ment and the paucity of demand for care for persons
as a result of recovery from depression, and enhanced with MNS disorders in many populations (Saraceno and
early child development as a result of psychosocial others 2007).
stimulation in infancy. Even for those conditions for
which there are currently no highly effective treatments
for the primary disorder, such as autism and dementia, HOW TO DELIVER EFFECTIVE
psychosocial interventions have been shown to be effec-
tive in addressing their adverse social consequences and
INTERVENTIONS?
supporting family caregivers. The implementation of evidence-based interventions
for MNS disorders seldom occurs through the delivery
of single, vertical interventions. More frequently, these
Limited Access to Essential Interventions interventions are delivered via platformsthe level of
Despite this evidence, many persons affected by MNS the health or welfare system at which interventions or
disorders do not have access to the interventions. In packages can be most appropriately, effectively, and effi-
general, severe MNS disorders tend to have higher rates ciently delivered. A specific delivery channel, such as a
of contact coverage, while treatment gaps for less visible school or a primary health care center, can be viewed as
conditions, such as harmful drinking and depression the vehicle for delivery of a particular intervention on a
and anxiety disorders, approach or exceed 90 percent specified platform. Identifying the set of interventions
in many populations. Similarly, the coverage rates tend that fall within the realm of a particular delivery channel
to be much higher for medicines than for psychosocial or platform is of interest and relevance to decision mak-
interventions. Across all disorders, the rates of effec- ers because it enables potential opportunities, synergies,
tive coverage are low. Supply-side and demand-side and efficiencies to be identified. It also reflects how
barriers play a role in explaining these low coverage resources are often allocated in practice, for example, to
rates. The lack of adoption of effective interventions is schools or primary health care services, rather than to
often influenced by concerns about financial resources. specific interventions or disorders. This section identi-
This issue is being addressed by a mounting evidence fies three broad platforms: population, community, and
base demonstrating the effectiveness of the delivery of health care.
these interventions by nonspecialist health workers (van There is a fair amount of good evidence from HICs
Ginneken and others 2013), as well as their costs and for interventions across these platforms and along the
cost-effectiveness (chapter 12 in this volume, Levin and continuum of primary, secondary, and tertiary preven-
Chisholm 2015). tion. However, the evidence base for LMICs is far less
A related resource constraint concerns the low avail- robust. Recommendations for best practice and good
ability of appropriately trained mental health workers. practice interventions for the platforms are shown in
Cultural attitudes and beliefs may also pose specific table 1.3. Best practice interventions were identified on
barriers. For example, the moral model of addiction sees the basis of evidence for their effectiveness and contex-
it as largely a voluntary behavior in which people freely tual acceptability and scalability in LMICs, plus evidence
engage in substance use. By contrast, the medical model of their cost-effectiveness at least in HICs. Good practice
of addiction recognizes that a minority of users will lose interventions were identified on the basis of sufficient
control over their use and develop a mental or physical evidence of their effectiveness in HICs and/or promising
disorderan addictionthat requires specific treat- evidence of their effectiveness in LMICs. The lack of evi-
ment if sufferers are to become abstinent. As another dence of cost-effectiveness in LMICs reflects the absence
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders
Early child enrichment/preschool disruptive behavior disorders*
education programs Improve the quality of antenatal and
Identification of children with perinatal care to reduce risk factors
MNS disorders in schools associated with intellectual disability
13
table continues next page
14
Table 1.3 Intervention Priorities for Mental, Neurological, and Substance Use Disorders by Delivery Platform (continued)
Platforms for intervention delivery
Health care platforms
Target area Population platform Community platform Self-care Primary health care First-level hospital care Specialized care
Neurological Policy interventions to Self-managed Diagnosis and management of Diagnosis of dementia Surgery for
disorders address the risk factors for treatment of migraine epilepsy and headaches and secondary causes of refractory epilepsy
cardio-vascular diseases, Self-identification/ Screening for detection of dementia headache
for example, tobacco management of seizure
control Interventions to support
triggers caregivers of patients with
Improved control of Self-management of dementia
neurocysticercosis
Emergency management of
poisoning
Note: Red type denotes urgent care; blue type denotes continuing care; black type denotes routine care. Recommendations in bold = best practice; recommendations in normal font = good practice.
ADHD = Attention Deficit Hyperactivity Disorder; BAC = blood alcohol concentration; CBT = cognitive behavioral therapy; ECT = electroconvulsive therapy; HIV = human immunodeficiency virus; MNS = mental, neurological, and substance use;
NCDs = noncommunicable diseases.
*There is no fixed time period for the management of these complex conditions; for example, in the management of depression, some individuals need relatively short periods of engagement (for example, 6-12 months for a single episode) at the one end,
while others may need maintenance care for several years (for example, when there is a relapsing course).
of evidence rather than the lack of cost-effectiveness for channels: self-management and care, primary health
most interventions. care (which includes outreach services in the commu-
In addition to bridging the treatment gap for MNS nity), and hospital care (which include MNS specialist
disorders by improving access to evidence-based inter- services and other specialist services, such as HIV or
ventions, it is imperative to enhance the quality of ser- maternal health care).
vice delivery, which together with need and utilization Examples of best or good practice packages for self-
make up the concept of effective coverage. The quality of care include the self-management of conditions, such
care should not be subservient to the quantity of avail- as migraines, and web-based psychological therapy for
able and accessible services, not least since robust quality depression and anxiety disorders, increasingly enabled
improvement mechanisms ensure that limited resources by internet- and smartphone-based delivery.
are utilized appropriately. Good quality services also At the primary health care level, a range of case-
build peoples confidence in care, thereby fueling the finding, detection, and diagnostic measures, as well as
demand for and increased utilization of preventive and the psychological and pharmacological management of
treatment interventions. such conditions, can be effectively performed. The con-
ditions include depression (including maternal depres-
sion), anxiety disorders, migraines, and alcohol and
Population and Community Platforms illicit drug use disorders, as well as continuing care for
Chapter 10 in this volume (Petersen and others 2015) severe disorders such as epilepsy or psychosis.
outlines the intervention packages for delivery through The recommended delivery model is collaborative
the population and community platforms. Population stepped care, in which patient care is coordinated by
platform interventions typically apply to the entire pop- a primary carebased nonspecialist case manager who
ulation and mainly revolve around promoting men- carries out a range of tasks including screening, provi-
tal health, preventing MNS disorders, and addressing sion of psychosocial interventions, and proactive moni-
demand-side barriers. Best practice packages include toring, while working in close liaison with, and acting as
legislative and regulatory measures to restrict access a link between the patient, primary care physician, and
to means of self-harm/suicide (notably pesticides) and specialist services. A robust evidence base supports the
reduce the availability of and demand for alcohol, includ- delivery of psychosocial interventions by appropriately
ing increased taxes and advertising bans. Good practice trained and supervised nonspecialist health workers
packages include interventions aimed at raising mental (van Ginneken and others 2013) and the collaborative
health literacy and reducing stigma and discrimination. stepped care model of delivery (Patel and others 2013).
The criminal justice system offers an important channel At the hospital level, first-level hospitals, typically
for the delivery of interventions for a range of MNS disor- district hospitals, offer a range of medical care services
ders, notably those associated with alcohol and illicit drug focused on providing integrated care for MNS disorders,
use, behavior disorders in adolescents, and psychoses. by implementing the same packages as recommended
Other preventive and promotion interventions do for the primary care channel. In particular, first-level
not require such a populationwide approach. These hospitals offer those services where MNS disorders
interventions are best delivered by targeting a group of frequently co-occur, such as maternal health, other
people in the community that share a certain character- noncommunicable diseases, and HIV/AIDS (Kaaya and
istic or are part of a particular setting, such as children others 2013; Ngo and others 2013; Rahman and others
in school. This platform is referred to as the community. 2013). Specialist health care may be offered in first-
Best practice packages at the community level include level hospitals or separate specialist hospitals, such as
life skills training to build social and emotional com- psychiatric hospitals or de-addiction centers. Specialist
petencies in children and adolescents (school-based health care delivery channels focus on the diagnosis
programs and programs that target vulnerable children). and management of complex, refractory, and severe
Good practice packages at the community level are cases (for example for psychosis, bipolar disorder, or
reported in table 1.3. refractory epilepsy); childhood behavioral disorders;
dementia; severe alcohol or illicit drug dependence and
withdrawal; and severe depression.
Health Care Platform A small minority of individuals with MNS disor-
Chapter 11 in this volume (Shidhaye, Lund, and ders will require ongoing care in community-based
Chisholm 2015) outlines the packages pertaining to residential facilities because of their disability and lack
the health care platform through three specific delivery of alternative sources of care and support. The role of
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders 15
community outreach teams that can provide variable increased risk of MNS disorders that can overwhelm
levels of intensity of care appropriate for individuals the local capacity to respond, particularly if the existing
needs is also crucial as they provide support to enable infrastructure or health system was already weak or may
these individuals to function in an independent way, in have been rendered dysfunctional as a result of the emer-
the community, alongside close liaison with general pri- gency situation. There is a heightened need to identify
mary care services and other social and criminal justice and allocate resources for providing mental health care
services. and psychosocial support in these settings, for those
with disorders induced by the emergency and for those
with preexisting disorders. International humanitarian
Humanitarian Aid and Emergency Response aid and emergency response at the national level can be
In humanitarian contexts and emergency affected pop- a channel for rapidly enabling or supporting the avail-
ulations, such as those arising from conflicts or natural ability of and access to basic or specialist care. In several
disasters, the humanitarian aid and emergency response countries, such emergencies have actually provided
channel is yet another channel for delivering much opportunities for systemic change or service reform in
needed mental health care. These populations are at an public mental health (WHO 2013b; see also box 1.4).
Box 1.4
Country Case Studies on Scaling Up Interventions for Mental, Neurological, and Substance
Use Disorders
The 686 Project: China (Hong 2012) rates of creating disturbances and causing serious
The Central Government Support for the Local accidents.
Management and Treatment of Severe Mental Government investment in the program amounted
Illnesses Project was initiated in China in 2004 to 280 million in 2011. The programs key inno-
with the first financial allotment of 6.86 million vations were the increase in the availability of
(US$829,000 in 2004 dollars). Subsequently it was human resources, including the involvement of
referred to as the 686 Project. Modeled on the non-mental-health professionals and their intensive
World Health Organizations (WHOs) recom- capacity building, which increased the number of
mended method for integrating hospital-based and psychiatrists in the country by one-third.
community-based mental health services, this pro-
gram provides care for a range of severe mental dis-
orders through the delivery of a community-based The National Depression Detection and Treatment
package by multidisciplinary teams. Program: Chile (MHIN)
The National Depression Detection and Treatment
The interventions are functionality oriented and Program in Chile is a national mental health pro-
provide free outpatient treatment through insurance gram that integrates detection and treatment of
coverage (New Rural Cooperative Medical Care depression in primary care. The program is based on
system) along with subsidized inpatient treatment scaling up an evidence-based collaborative stepped
for poor patients. The program covered 30 percent care intervention in which most patients diag-
of the population of China by the end of 2011. nosed with depression are provided medications
Evaluation of the program showed improved out- and psychotherapy at primary care clinics, while
comes for the more than 280,000 registered patients, only severe cases are referred to specialists. Launched
as the proportion of patients with severe mental in 2001, the program operates through a network
illnesses who did not suffer a relapse for five years of 500 primary care centers, and presently covers
or longer increased from a baseline of 67 percent 50 percent of Chiles population.
to 90 percent, along with large reductions in the
The program has added many psychologists in pri- In 2011, funding from the Dutch government
mary care, amounting to an increase of 344 percent enabled HealthNet TPO and the Burundian
between 2003 and 2008. Enrollment of patients in the government to initiate a five-year project aimed
program has grown steadily, with around 100,000 to at strengthening health systems. One of the
125,000 patients starting treatment each year from projects components is the integration of mental
2004 to 2006 and close to 170,000 patients starting health care into primary care using WHO Mental
treatment in 2007. Nationwide implementation of Health Gap Action Programme guidelines. The
the program has led to greater utilization of health government has established a national commission
services by women and the less educated, contribut- for mental health and appropriate steps are being
ing to reduced health inequalities. The programs suc- taken to support the provision of mental health
cess can be attributed to the use of an evidence-based care in general hospitals and follow-up within the
design that was made available to policy makers, community.
teamwork, proactive leadership, strategic alliances
across sectors, sustained investment and ring-fencing Suicide Prevention through Pesticide Regulation:
new and essential financial resources, program Sri Lanka (Gunnell and others 2007)
institutionalization, and sustained development of In Sri Lanka, as well as in other Asian countries,
human resources that can implement the program. pesticide self-poisoning is one of the most commonly
used methods of suicide. Suicide rates in Sri Lanka
Building Back Better: Burundi (WHO 2013a) increased eight-fold from 1950 to 1995, and the
Civil war in the last decade of the 20th century and country had the highest rate of suicide worldwide
first decade of this century resulted in widespread (approximately 47 per 100,000 population) during
massacres and forced migrations and internal this period. A series of policy and legislative actions
displacement of around one million individuals around this time reduced the suicide rate by half
in Burundi. To address this humanitarian crisis, by 2005.
Healthnet Transcultural Psychosocial Organization Gunnell and others (2007) carried out an ecologi-
(TPO) started providing mental health services in cal analysis of trends in suicide and risk factors for
Burundi during 2000 when the then Ministry of suicide in Sri Lanka during 19752005. The analy-
Public Health had no mental health policy, plan, or sis suggests that the marked decline in Sri Lankas
unit, and virtually all the psychiatric services were suicide rate in the mid-1990s coincided with the
provided by one psychiatric hospital. Healthnet TPO culmination of a series of legislative activities that
first conducted a needs assessment and then built a systematically banned the most highly toxic pesti-
network of psychosocial and mental health services cides that had been responsible for the majority of
in communities in the national capital, Bujumbura, pesticide deaths in the preceding two decades. The
and in seven of the countrys 17 provinces. A new Registrar of Pesticides banned methyl parathion
health worker cadre, the psychosocial worker, played and parathion in 1984 and over the following years
a pivotal role in delivery of these services. gradually phased out all the remaining Class I (the
Considerable progress has been made in the past most toxic) organophosphate pesticides, culminat-
decade. The government now supplies essential ing in July 1995 with bans on the remaining Class I
psychiatric medications through its national drug pesticides monocrotophos and methamidophos. By
distribution center, and outpatient mental health December 1998, endosulfan (a Class II pesticide)
clinics are established in several provincial hospitals. was also banned as farmers had substituted Class I
From 2000 to 2008, more than 27,000 people were pesticides with endosulfan.
helped by newly established mental health and By 2005, suicide rates halved to around 25 per
psychosocial services. Between 2006 and 2008, the 100,000 population. This case study underlines the
mental health clinics in the provincial hospitals fact that in countries where pesticides are commonly
registered almost 10,000 people, who received more used in acts of self-poisoning, regulatory controls
than 60,000 consultations. The majority (65 percent) on the sale of the most toxic pesticides may help to
were people with epilepsy. reduce the number of suicides.
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders 17
HOW MUCH WILL IT COST? MOVING This volume reviews existing cost-effectiveness evidence
TOWARD UNIVERSAL HEALTH COVERAGE and new analyses of the distributional and financial pro-
tection effects of interventions (box 1.5).
FOR MNS DISORDERS
For successful and sustainable scale-up of effective
Intervention Costs and Cost-Effectiveness
interventions and innovative service delivery strategies,
such as task-sharing and collaborative care, decision There is a small but growing economic evidence base
makers require not only evidence of an interventions to inform decision making in LMICs, mainly on the
health impact, but also the costs and cost-effectiveness. treatment of specific disorders. Analysis undertaken at
Even when cost-effectiveness evidence is available, there the global level by WHO, updated to 2012 values for
remains the question of whether or how an intervention DCP3, reveals a marked variation in the cost per DALY
might confer wider economic and social benefits on averted, not only between different regions of the world,
households or society, such as restored productivity, but also between different disorders and interventions
reduced medical impoverishment, or greater equality. (Chisholm and Saxena 2012; Hyman and others 2006).
Box 1.5
Economic Evaluation of the Treatment and Prevention of Mental, Neurological, and Substance
Use Disorders
Economic evaluations aim to inform decision making 4. Assessment of the economic benefits, measured
by quantifying the trade-offs between the resource in monetary terms, from investment in a health
inputs needed for alternative investments and the intervention and weighing that benefit against its
resulting outcomes. Four approaches to economic cost (benefit-cost analysis). This analysis enables
evaluation in health are particularly prominent: comparison of the attractiveness of health invest-
ments compared with those in other sectors.
1. Assessment of how much of a specific health
outcome (for example, depressive episodes or Cost-effectiveness analyses predominate among eco-
epileptic seizures averted) can be attained for a nomic evaluations in the care and prevention of
particular level of resource input. mental, neurological, and substance use (MNS)
2. Assessment of how much of an aggregate measure disorders. These types of analysis are reviewed in
of health (for example, averted deaths, disability, the disorder-specific chapters of the volume and,
or quality-adjusted life years) can be attained in a more synthesized format, in chapter 12 (Levin
from a particular level of resource inputs applied and Chisholm 2015). This review shows that the
to alternative interventions. This approach of economic evidence base for mental health policy
cost-effectiveness analysis enables comparison and planning continues to strengthen. Thus, the
of the attractiveness of interventions addressing overgeneralized claim that treatment of MNS disor-
many different health outcomes (such as tuber- ders is not a cost-effective use of scarce health care
culosis or HIV treatment versus prevention of resources can be increasingly debunked.
harmful alcohol use or treatment of psychosis). Extended cost-effectiveness analyses remain a fairly
3. Assessment of how much health and financial new evaluation approach developed for Disease
risk protection can be attained for a particular level Control Priorities, 3rd edition (DCP3). In this volume,
of public sector finance of a particular interven- Chisholm and others (chapter 13) apply extended
tion. This approach (extended cost-effectiveness cost-effectiveness analysis to a range of MNS disor-
analysis) enables assessment not only of effi- der interventions in Ethiopia and India. The chapter
ciency in improving the health of a population, shows that moving toward universal coverage via
but also of efficiency in achieving the other major scaled-up provision of publicly financed services
goal of a health system (that is, protection of the leads to significant financial protection effects as
population from financial risk). well as health gains in the population.
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders 19
Figure 1.2 Cost-Effectiveness of Selected Interventions for Addressing Mental, Neurological and Substance Use Disorders in Low-
income and Middle-income Countries (2012 US$ per DALY averted)
1,511
(Thailand; Prukkanone and others 2012)
Continuation treatment in primary care with newer SSRIs
1,312
(Thailand; Prukkanone and others 2012)
Episodic psychosocial treatment in primary care
914
(Thailand; Prukkanone and others 2012)
Maintenance psychosocial treatment in primary care
437
(Thailand; Prukkanone and others 2012)
Heavy Epilepsy
236
(Nigeria; Gureje and others 2007)
10,000
1,000
100
Schizophrenia: Depression: Episodic Epilepsy: Treatment in Alcohol use disorder:
Community-based treatment in primary care primary care with older Brief physician advice in
treatment with older with (generic) antidepressant anti-epileptic drug primary care
antipsychotic medication medication and psychosocial
and psychosocial treatment treatment
Source: Hyman and others 2006; Chisholm and Saxena 2012; Levin and Chisholm 2015.
Note: In panel a, all reported cost-effectiveness estimates have been converted to 2012 US$. In panel b, previously published findings have been converted to 2012 US$ values, based on
International Monetary Fund inflation estimates for World Bank reporting regions. Bars show the range in cost-effectiveness for six low- and middle-income world regions: Sub-Saharan Africa,
Latin America and the Caribbean, Middle East and North Africa, Europe and Central Asia, South Asia, and East Asia and Pacific. DALY = disability-adjusted life year; SSRI = selective serotonin
reuptake inhibitor; TCA = tricyclic antidepressants.
30 8,000
7,000
25
Public health spending (%)
6,000
15 4,000
3,000
10
2,000
5
1,000
0 0
1 2 3 4 5 1 2 3 4 5
HOW TO SCALE UP? HEALTH SYSTEM disorders. In addition, the following are lacking: techni-
BARRIERS AND OPPORTUNITIES cally sound leadership in designing and implementing
evidence-based programs; adequate absorptive capacity
Despite the need for renewed attention and scaled-up in the existing health care system; competing policy
investment, there is relatively little action on addressing priorities and vested interests; and effective agency and
MNS disorders in most LMICs. There are several reasons advocacy by affected people. And there is a persisting
for this lack of action, perhaps the most important one belief in the importance of hospital-based specialized
being the overall lack of policy commitment to MNS models of care, which continue to absorb disproportion-
disorders, as is evident from the fact that less than 1 ate amounts of the already meager budgetary allocations
percent of the health budget is allocated to mental health for this sector (Saraceno and others 2007).
in most LMICs (Saxena and others 2007). Similarly,
despite the evidence-based calls to action for scaling
up services for almost a decade (Lancet Global Mental Knowledge Gaps
Health Group 2007), less than 1 percent of development There is a lack of evidence from LMICs, especially on the
assistance for health is devoted to mental health (Gilbert cost-effectiveness of many interventions and the inte-
and others 2015). gration of care for MNS disorders in routine health and
social care platforms. This lack continues to represent a
constraint to investment for many stakeholders, and is
Political Will partly a result of low levels of political commitment to
Key contributors to the lack of political will and con- this dimension of health through disproportionately less
sequently low levels of resource allocation include funding for research. The critical knowledge gaps are
the low demand for mental health care interventions, related to implementation science, that is, research to
which is in part caused by low levels of mental health bridge the gap between what we know works and how to
literacy and high levels of stigma attached to MNS implement it at scale (Collins and others 2011).
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders 21
Research that seeks to address the significant knowl- Investing in research across the translational con-
edge gaps on the causes of MNS disorders and the dis- tinuum to improve knowledge on more effective
covery of novel interventions is also urgently needed. An interventions and more effective delivery systems,
empirical approach to analysis of the impact of macro- including innovative financing options such as rais-
economic and structural factors on the burden of MNS ing and diverting income from taxes on unhealthy
disorders, such as global conventions on the regulation products (such as alcohol and tobacco)
of illicit drugs and climate change, is warranted to guide Emphasizing the use of low-cost generic medicines
evidence-based policy making in the wider context. throughout the health care systems, and reallocating
However, these knowledge gaps cannot explain why expenditure on ineffective or low-value interventions,
even known cost-effective interventions have not been such as overprescription of benzodiazepines and vita-
adopted. mins in primary care.
A complicating factor is the limitations of the evi- Finally, it will be important to embed health indicators
dence synthesized in this chapter. In particular, there for MNS disorders within national health information
are significant gaps in the evidence in support of some and surveillance systems so that progress and achieve-
interventions in LMICs and limited effectiveness of ments can be monitored and evaluated (WHO 2015).
the best available interventions for some disorders. To
address these barriers, the scaling-up of interventions The WHO Comprehensive Mental Health Action Plan
for MNS disorders requires an approach that embraces (Saxena, Funk, and Chisholm 2013) offers a clear road
public health principles, systems thinking, and a whole- map for countries at any stage of the journey to scale up.
of-government perspective. Reassuringly, several coun- Some regions (such as the Eastern Mediterranean) have
tries are now demonstrating how a combination of these adapted this new policy instrument to initiate consul-
ingredients can lead to significant increases in the cover- tations with international experts and regional policy
age of evidence-based interventions (box 1.4). makers and develop frameworks for action (box 1.6)
across all four domains of the plan, along with priority
interventions and indicators for evaluation of progress
Strategies for Strengthening the Health System (Gater, Saeed, and Rahman 2015).
Key strategies for strengthening the health system
include the following:
TIME TO ACT NOW
Mainstreaming a rights-based perspective throughout MNS disorders account for a substantial proportion of
the health system and ensuring health policies, plans, the global disease burden. This burden has increased dra-
and laws are updated to be consistent with interna- matically since 1990 and is likely to continue to rise with
tional human rights standards and conventions the epidemiological transition from infectious diseases to
Implementing multicomponent initiatives to address noncommunicable diseases, the demographic transition
stigma, enhance mental health literacy and demand in LMICs, and the increase in the prevalence of several
for care, and mobilize people with the conditions to social determinants associated with these conditions.
support one another and be effective advocates Despite the challenges in quantifying causal mortality
Engaging other key sectors concerned with MNS in these disorders, new analyses presented in this volume
disorders to improve services, notably the social suggest that the mortality-associated disease burden
care, non-governmental organizations, private sector, is very large and was previously underestimated. This
criminal justice, education, and indigenous medical volume also summarizes evidence to document effective
sectors, as they all have complementary roles. treatment and prevention interventions that are feasible
Providing inpatient care through units in general or to implement across diverse socioeconomic and cultural
district hospitals rather than standalone psychiatric settings for a range of priority MNS disorders. A criti-
hospitals cally relevant aspect of these disorders is their propensity
Implementing large-scale or national rollouts of to strike early in life, which is a key factor behind their
training and supervision programs for nonspecialist large contribution to the global burden of disease.
human resource cadres that can perform the roles Populationwide platforms are primarily suited for
of case managers for delivery of collaborative care policy-level interventions for promoting mental health,
in primary care and other health care platforms to preventing MNS disorders, improving mental health
improve treatment coverage literacy, and protecting the human rights of persons
Ensuring the supply of essential medicines at relevant affected by these disorders. The community platform
platforms provides opportunities for leveraging non-health
Proposed Regional Framework to Scale Up Action on Mental Health in the WHO Eastern
Mediterranean Region
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders 23
Box 1.6 (continued)
resources for prevention and promotion interventions necessary resources and provide technical leadership.
targeting particular groups of people or particular set- As also emphasized in the WHO Mental Health Action
tings. The health care interventions primarily comprise Plan, this will and commitment are essential to address
generic medicines, brief psychological treatments, and the avoidable toll of suffering caused by MNS disorders,
social interventions. Interventions for diverse disorders not least among the poorest people and least resourced
can be packaged together to deploy low-cost and widely countries in the world.
available human resources in primary health care and This volume presents strong clinical and economic
non-health care platforms, with appropriate support evidence to back this investment. Ultimately there must
and supervision provided by mental health care profes- also be a moral case for scaling up care for the hundreds of
sionals. In settings with a higher level of resources, as is millions of people whose health care needs have been sys-
the case in many middle-income countries, specialist tematically neglected and whose basic human rights have
platforms offer incremental value in addressing the been routinely denied (Patel, Saraceno, and Kleinman
needs of the relatively small proportion of persons with 2006). The time to act on this evidence is therefore now.
complex, severe, or refractory clinical presentations.
Apart from being effective and feasible and providing NOTE
benefits that improve the lifelong trajectories of indi-
viduals, many of these interventions are also inexpen- Disclaimer: Dan Chisholm and Tarun Dua are staff members
sive to implement and represent a cost-effective use of of the World Health Organization. The authors alone are
resources for health. Furthermore, a policy of moving responsible for the views expressed in this publication and they
do not necessarily represent the decisions, policy, or views of
toward universal public finance for MNS disorders can
the World Health Organization.
be expected to lead to a far more equitable allocation of World Bank Income Classifications as of July 2014 are as
public health resources across income groups. With uni- follows, based on estimates of gross national income (GNI)
versal public finance, the lowest-income groups would per capita for 2013:
benefit most from the value of insurance (used here as a
measure of financial protection). Low-income countries (LICs) = US$1,045 or less
Middle-income countries (MICs) are subdivided:
Country case studies show that the most important
a) Lower-middle-income = US$1,045 to US$4,125
drivers of change are the political will and commitment b) Upper-middle-income (UMICs) = US$4,126 to US$12,735
of countries and development agencies to allocate the High-income countries (HICs) = US$12,736 or more.
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders 25
Lawrence, D., K. J. Hancock, and S. Kisely. 2013. The Gap (third edition): Volume 4, Mental, Neurological, and
in Life Expectancy from Preventable Physical Illness in Substance Use Disorders, edited by V. Patel, D. Chisholm,
Psychiatric Patients in Western Australia: Retrospective T. Dua, R. Laxminarayan, and M. E. Medina-Mora.
Analysis of Population Based Registers. BMJ 346: f2539. Washington, DC: World Bank.
doi:10.1136/bmj.f2539. Prince, M., V. Patel, S. Saxena, M. Maj, J. Maselko, and others.
Levin, C., and D. Chisholm. 2015. Cost and Cost-Effectiveness 2007. No Health without Mental Health. The Lancet 370
of Interventions, Policies, and Platforms for the Prevention (9590): 85977.
and Treatment of Mental, Neurological, and Substance Use Prukkanone, B., T. Vos, M. Bertram, and S. Lim. 2012. Cost-
Disorders. In Disease Control Priorities (third edition): Effectiveness Analysis for Antidepressants and Cognitive
Volume 4, Mental, Neurological, and Substance Use Disorders, Behavioral Therapy for Major Depression in Thailand.
edited by V. Patel, D. Chisholm, T. Dua, R. Laxminarayan, Value in Health 15 (1): S38.
and M. E. Medina-Mora. Washington, DC: World Bank. Rahman, A., P. J. Surkan, C. E. Cayetano, P. Rwagatare, and
Lim, S. S., T. Vos, A. D. Flaxman, G. Danaei, K. Shibuya, and K. E. Dickson. Grand Challenges: Integrating Maternal
others. 2012. A Comparative Risk Assessment of Burden Mental Health into Maternal and Child Health pro-
of Disease and Injury Attributable to 67 Risk Factors grammes. PLoS Med 2013; 10: e1001442.
and Risk Factor Clusters in 21 Regions, 1990-2010: A Rehm, J., D. Chisholm, R. Room, and A. D. Lopez. 2006.
Systematic Analysis for the Global Burden of Disease Alcohol. In Disease Control Priorities in Developing
Study 2010. The Lancet 380 (9859): 222460. doi:10.1016/ Countries, 2nd ed., edited by D. T. Jamison, J. G. Breman,
s0140-6736(12)61766-8. A. R. Measham, G. Alleyene, M. Claeson, and D. B. Evans,
Lindner, L. M., A. C. Marasciulo, M. R. Farias, and G. E. M. Grohs. P. Jha, A. Mills, and P. Musgrove. Washington, DC: World
2009. Economic Evaluation of Antipsychotic Drugs for Bank and Oxford University Press.
Schizophrenia Treatment within the Brazilian Healthcare Rehm, J., C. Mathers, S. Popova, M. Thavorncharoensap,
System. Revista de sade pblica 43: 6269. Y. Teerawattananon, and others. 2009. Global Burden of
Lund, C., M. De Silva, S. Plagerson, S. Cooper, D. Chisholm, Disease and Injury and Economic Cost Attributable to
and others. 2011. Poverty and Mental Disorders: Breaking Alcohol Use and Alcohol-Use Disorders. The Lancet 373
the Cycle in Low-Income and Middle-Income Countries. (9682): 222333.
The Lancet 378 (9801): 150214. Saraceno, B., M. van Ommeren, R. Batniji, A. Cohen, O. Gureje,
MHIN (Mental Health Innovation Network). Program for and others. 2007. Barriers to Improvement of Mental
Screening, Diagnosis and Comprehensive Treatment of Health Services in Low-Income and Middle-Income
Depression. MHIN. http://mhinnovation.net/innovations Countries. The Lancet 370 (9593): 116474.
/program-screening-diagnosis-and-comprehensive Saxena, S., M. Funk, and D. Chisholm. 2013. World Health
-treatment-depression#.VVYpd46qqkp. Assembly Adopts Comprehensive Mental Health Action
Ngo, V. K., A. Rubinstein, V. Ganju, P. Kanellis, N. Loza, and Plan 20132020. The Lancet 381 (9882): 197071.
others. 2013. Grand Challenges: Integrating Mental Health doi:10.1016/S0140-6736(13)61139-3.
Care into the Non-Communicable Disease Agenda. PLoS Saxena, S., G. Thornicroft, M. Knapp, and H. Whiteford.
Med 10: e1001443. 2007. Resources for Mental Health: Scarcity, Inequity, and
Patel, V., G. S. Belkin, A. Chockalingam, J. Cooper, S. Saxena, Inefficiency. The Lancet 370 (9590): 87889.
and others. 2013. Integrating Mental Health Services Shidhaye, R., C. Lund, and D. Chisholm. 2015. Health Care
into Priority Health Care Platforms: Addressing a Grand Delivery Platforms. In Disease Control Priorities (third
Challenge in Global Mental Health. PLoS Med 10: e1001448. edition): Volume 4, Mental, Neurological, and Substance
Patel, V., A. Kleinman, and B. Saraceno. 2012. Protecting the Use Disorders, edited by V. Patel, D. Chisholm, T. Dua,
Human Rights of People with Mental Disorders: A Call R. Laxminarayan, and M. E. Medina-Mora. Washington,
to Action for Global Mental Health. In Mental Health & DC: World Bank.
Human Rights, edited by M. Dudley, D. Silove, and F. Gale. Strang, J., T. Babor, J. Caulkins, B. Fischer, D. Foxcroft, and
Oxford: Oxford University Press. others. 2012. Drug Policy and the Public Good: Evidence
Patel, V., C. Lund, S. Heatherill, S. Plagerson, J. Corrigal, and for Effective Interventions. The Lancet 379 (9810): 7183.
others. 2009. Social Determinants of Mental Disorders. doi:10.1016/s0140-6736(11)61674-7.
In Priority Public Health Conditions: From Learning to United Nations. 2015. Global Sustainable Development Report,
Action on Social Determinants of Health, edited by E. 2015 Ediion. New York: United Nations.
Blas and A. Sivasankara Kurup. Geneva: World Health van Ginneken, N., P. Tharyan, S. Lewin, G. N. Rao, S. Meera, and
Organization. others. 2013. Non-Specialist Health Worker Interventions
Patel, V., B. Saraceno, and A. Kleinman. 2006. Beyond for the Care of Mental, Neurological and Substance-
Evidence: The Moral Case for International Mental Health. Abuse Disorders in Low- and Middle-Income Countries.
American Journal of Psychiatry 163 (8): 131215. Cochrane Database Systematic Reviews 11: CD009149.
Petersen, I., S. Evans-Lacko, M. Semrau, M. Barry, D. Chisholm, doi:10.1002/14651858.CD009149.pub2.
P. Gronholm, C. O. Egbe, and G. Thornicrost. 2015. Verguet, S., R. Laxminarayan, and D. T. Jamison. 2015.
Population Platforms. In Disease Control Priorities Universal Public Finance of Tuberculosis Treatment in
Global Priorities for Addressing the Burden of Mental, Neurological, and Substance Use Disorders 27
Chapter
2
Global Burden of Mental, Neurological, and
Substance Use Disorders: An Analysis from the
Global Burden of Disease Study 2010
Harvey A. Whiteford, Alize J. Ferrari, Louisa Degenhardt,
Valery Feigin, and Theo Vos
Corresponding author: Harvey A. Whiteford, Queensland Centre for Mental Health Research, University of Queensland, the Park Centre for Mental Health, Wacol,
QLD 4076, Australia; h.whiteford@uq.edu.au.
29
and alcohol and drug use disordersaccounted for priority in many LMICs, increasing life expectancies
10.5 percent of the worlds disease burden, as measured due to better reproductive health, childhood nutrition,
by disability-adjusted life years (DALYs). The DALY is a and control of communicable diseases meant that more
health metric that captures the nonfatal component of people in 2010 were living to ages where mental, neuro-
the disease burden as years lived with disability (YLDs), logical, and substance use disorders were most prevalent
and the fatal component as years of life lost (YLLs) to (Whiteford, Degenhardt, and others 2013).
premature mortality (Murray and Lopez 1996). GBD In GBD 2010, the burden of mental and substance use
1990 showed that five of the top 10 causes of disability disorders was estimated separately from that of neuro-
making up more than 25 percent of global YLDs for logical disorders, such as dementia, Parkinsons disease,
1990belonged to the category of mental, neurological, and epilepsy. This approach enabled us to investigate
and substance use disorders (Murray and Lopez 1996). more comprehensively the differences in the epidemiol-
In its update of burden estimates for 200005, the World ogy and burden between these groups of disorders com-
Health Organization (WHO) assigned 31.7 percent of pared with previous GBD studies. Mental and substance
all YLDs to mental, neurological, and substance use use disorders were among the leading causes of disease
conditions; the five main contributors of this burden burden in 2010. They were responsible for 7.4 percent of
were depression (11.8 percent), alcohol use disorders global DALYs and 22.9 percent of global YLDs, making
(3.3 percent), schizophrenia (2.8 percent), bipolar disorder them the fifth-leading cause of DALYs and the leading
(2.4 percent), and dementia (1.6 percent) (WHO 2008). cause of YLDs (Whiteford, Degenhardt, and others
2013). Neurological disorders explained 3.0 percent of
global DALYs and 5.6 percent of global YLDs (Murray,
Global Burden of Disease Study 2010 Vos, and others 2012; Vos and others 2012).
In this chapter we present findings from GBD 2010. The The overarching findings of the study for all 291 dis-
GBD 2010 estimated the burden for 291 diseases and eases and injuries have been presented (Lim and others
injuries and 67 risk factors and was the first comprehen- 2012; Lozano and others 2012; Murray, Ezzati, and oth-
sive re-analysis of the burden since GBD 1990 (Lim and ers 2012; Murray, Vos, and others 2012; Salomon and
others 2012; Lozano and others 2012; Murray, Vos, and others 2012; Vos and others 2012), as have the GBD
others 2012; Salomon and others 2012; Vos and others 2010 results for mental and substance use disorders
2012; Wang and others 2012). GBD 2010 estimated bur- (Degenhardt, Whiteford, and others 2013; Whiteford,
den for three main cause groups: Degenhardt, and others 2013). This chapter presents GBD
2010 burden estimates of mental, neurological, and sub-
Communicable diseases: infectious or transmissible stance use disorders as a group. Specifically, we quantify
diseases the global disease burden attributable to mental, neuro-
Noncommunicable diseases: noninfectious or non logical, and substance use disorders and explore variations
transmissible diseases in burden by disorder type, age, gender, year, and region.
Injuries (accidental or intentional). This approach provides background and context for
chapter 3 in this volume (Charlson and others 2015),
The study included a complete epidemiological reas- which responds to the lack of deaths and fatal burden
sessment of these communicable and noncommunica- estimated by GBD 2010 for mental, neurological, and
ble diseases and injuries across 187 countries; 21 world substance use disorders. Most important, this chapter for
regions; males and females; estimated burden for 1990, the first time presents GBD 2010 burden of disease esti-
2005, and 2010; and 20 age groups. Rather than rely on mates at the aggregated level of mental, neurological, and
a selective sample of data points as previous GBD studies substance use disorders. Analysis of burden estimates at
had, burden estimates were based on a systematic review this aggregated level is important from the clinical and
of the literature to obtain all available epidemiological population health perspectives, given that the organiza-
data. The estimates were also derived through the use of tion of services in many LMICs does not separate neuro-
new statistical methods to model the epidemiological data, logical disorders from mental disorders, something seen
quantify disability, adjust for comorbidity between dis- as a progression of Western medical subspecialization.
eases, and propagate uncertainty to final burden estimates
(Murray, Vos, and others 2012; Vos and others 2012).
GBD 2010 highlighted a shift in burden from com-
METHODOLOGY
municable to noncommunicable diseases and from YLLs Annex 2A summarizes the mental, neurological, and
to YLDs (Murray, Vos, and others 2012; Vos and others substance use disorders investigated in GBD 2010 and
2012). Although communicable diseases remain a health describes how the YLDs, YLLs, and DALYs for each
300
and the Caribbean, the Middle East and North Africa,
South Asia, and Sub-Saharan Africa 250
Disaggregated by developed and developing regions.
200
7
80
7 6 da
3 da
da
Global Burden of Mental, Neurological, and Substance Use Disorders: An Analysis from the Global Burden of Disease Study 2010 31
Table 2.1 Age-Standardized DALY Rates Attributable to Mental, Neurological, and Substance Use Disorders,
1990 and 2010
Age-standardized DALY rates (per 100,000)
Male Female
Disorder 1990 2010 1990 2010
Mental disorders
Major depressive disorder 694.8 689.9 1,171.7 1,161.2
Dysthymia 135.3 135.8 189.7 190.0
Bipolar disorder 172.0 172.1 204.6 204.8
Schizophrenia 230.7 223.0 187.8 180.6
Anxiety disorders 274.3 273.0 508.9 510.3
Eating disorders 4.4 3.9 47.6 59.5
Autism 85.1 85.8 29.5 29.6
Aspergers syndrome 85.2 85.0 20.3 20.3
Attention-deficit hyperactivity disorder 10.8 10.6 3.1 3.1
Conduct disorder 111.9 113.3 47.0 47.6
Idiopathic intellectual disability 25.3 17.7 18.2 11.9
Other mental and behavioral disorders 25.5 23.3 21.5 20.8
Neurological disorders
Alzheimers disease and other dementias 125.7 155.5 153.7 178.6
Parkinsons disease 32.7 36.6 23.2 23.3
Epilepsy 261.6 269.3 226.0 232.9
Multiple sclerosis 16.3 12.3 23.7 19.8
Migraine 233.1 236.6 405.9 415.8
Tension-type headache 24.1 24.0 28.3 28.3
Other neurological disorders 228.0 259.9 200.0 266.7
Substance use disorders
Alcohol use disordersa 431.0 409.9 117.2 106.0
Opioid dependence 139.0 184.4 63.8 78.4
Cocaine dependence 22.5 22.0 10.3 9.7
Amphetamine dependence 45.4 47.3 26.9 27.6
Cannabis dependence 38.8 36.7 22.3 21.3
Other drug use disorders 83.7 97.0 44.6 47.9
Source: http://vizhub.healthdata.org/gbd-compare/.
Note: DALY = disability-adjusted life year.
a. Alcohol use disorders include alcohol dependence and fetal alcohol syndrome.
respiratory infections, meningitis, and other common Overall, in 2010, 124 million mental, neurological,
infectious diseases (explaining 11.4 percent of DALYs). and substance use DALYs occurred among males and
Major depressive disorder was responsible for the high- 134 million among females. Figure 2.2 shows DALY rates
est proportion of mental, neurological, and substance for each mental, neurological, and substance use disorder
use disorder DALYs (24.5 percent); attention-deficit by gender. Females accounted for more DALYs for most
hyperactivity disorder was responsible for the lowest of the mental and neurological disorders, except for
(0.2 percent). mental disorders occurring in childhood, schizophrenia,
Parkinsons disease, and epilepsy, where males accounted in 2010 by the GBD 2010 superregion groupings and
for more DALYs. Males also accounted for more DALYs by developed and developing world regions. Overall,
than females in all substance use disorders. the burden of these disorders as age-standardized rates
Figure 2.3 shows the burden attributable to mental, was approximately 1.6 times higher in developed regions
neurological, and substance use disorders as a group (explaining 15.5 percent of total DALYs) compared
Global Burden of Mental, Neurological, and Substance Use Disorders: An Analysis from the Global Burden of Disease Study 2010 33
Figure 2.2 Age-Standardized DALY Rates Attributable to Individual Mental, Neurological, and Substance Use Disorders, by Gender, 2010
Conduct disorder
Aspergers syndrome
Autism
Eating disorders
Anxiety disorders
Bipolar disorder
Dysthymia
Cannabis dependence
Amphetamine dependence
Cocaine dependence
Opioid dependence
Schizophrenia
Tension-type headache
Migraine
Multiple sclerosis
Epilepsy
Parkinsons disease
Alzheimers disease
Source: http://vizhub.healthdata.org/gbd-compare/.
Note: DALY = disability-adjusted life year.
a. Alcohol use disorders include alcohol dependence and fetal alcohol syndrome.
ia
a
and injuries. Noncommunicable diseases explained a
sia
ific
an
ing
ed
om
ric
ric
As
be
hA
ac
lop
lop
Af
Af
inc
rib
tra
dP
large proportion of YLDs and YLLs in 2010. Within this
ve
ve
ut
th
gh
ra
Ca
en
So
an
De
De
or
ha
Hi
dC
dN
he
group, mental, neurological, and substance use disorders
sia
Sa
dt
an
an
tA
b-
an
Su
pe
were responsible for 28.5 percent of all YLDs, making
st
as
ica
Ea
ro
E
Eu
er
le
them the leading cause of YLDs worldwide.
Am
idd
rn
ste
tin
In comparison, mental, neurological, and substance
Ea
La
use disorders contributed to only 2.3 percent of YLLs. Region
Deaths and YLLs could be assigned to a mental, neu- Mental disorders Neurological disorders
Substance use disorders
rological, or substance use disorder only when the dis-
order was considered as a direct cause of death in the Source: http://vizhub.healthdata.org/gbd-compare/.
ICD-10 cause-of-death directory. Using this approach, Note: DALY = disability-adjusted life year. DALYs were disaggregated by GBD 2010s seven
superregion groupsEast Asia and Pacific, Eastern Europe and Central Asia, high-income regions
the majority of excess deaths in individuals with a mental
(North America, Australasia, Western Europe, high-income Asia Pacific, and southern Latin
disorder, in particular, were coded to the direct physical America), Latin America and the Caribbean, the Middle East and North Africa, South Asia, and
cause of death (for example, suicide deaths were coded Sub-Saharan Africaand by developed and developing regions.
Figure 2.4 Proportion of Global YLDs and YLLs Attributable to Mental, Neurological, and Substance Use Disorders, 2010
a. YLLs b. YLDs
Mental disorders
Neurological disorders Substance use disorders Substance use disorders
0.1%
1.8% 0.4% 3.9%
Communicable
Mental diseases
disorders 15.5%
18.9% Injuries
Noncommunicable 5.9%
Communicable
diseases (excluding diseases
MNS disorders) Neurological disorders
43.7% 5.6%
40.5%
Noncommunicable
diseases (excluding
MNS disorders)
Injuries 50.2%
13.5%
Source: http://vizhub.healthdata.org/gbd-compare/.
Note: MNS = mental, neurological, and substance use; YLLs = years of life lost; YLDs = years lived with disability.
Global Burden of Mental, Neurological, and Substance Use Disorders: An Analysis from the Global Burden of Disease Study 2010 35
under injuries as self-harm) rather than to the disorder. GBD studies (Murray and Acharya 1997) recognizes
An analysis of excess mortality in individuals with men- and attempts to incorporate the social preference for
tal, neurological, and substance use disorders and the avoiding health loss in young adults. In spite of the
implications for burden of disease estimates is presented absence of age weighting in the GBD 2010 estimates,
in chapter 3 in this volume (Charlson and others 2015). the peak impact of mental, neurological, and substance
use disorders in early adulthood remained and demon-
strated the ubiquitous effect of these disorders at a time
IMPLICATIONS OF THE GBD 2010 FINDINGS of life when individuals are starting to make significant
FOR MENTAL, NEUROLOGICAL, AND social and economic contributions to their families and
societies. The peak in the total burden of mental, neuro-
SUBSTANCE USE DISORDERS logical, and substance use disorders was found in young
Mental, neurological, and substance use disorders are a adults. However, unlike many chronic diseases, there is a
leading cause of the disease burden worldwide, substan- significant burden in children, lending further evidence
tially contributing to health loss in individuals of all ages, to the importance of early intervention strategies for
from developed and developing regions. mental, neurological, and substance use disorders.
In GBD 2010, the differences in DALYs between men- The presentation of burden estimates by age in GBD
tal, neurological, and substance use disorders were guided 2010 facilitates the selection and tailoring of intervention
by differences in the prevalence, death, and disability strategies for mental, neurological, and substance use
weights associated with each disorder. The input data that disorders. For instance, it allows us to identify the ages at
were used to estimate burden are presented in greater which interventions would be most beneficial. Historically,
detail elsewhere (Baxter and others 2015; Baxter and oth- mental, neurological, and substance use disorders occur-
ers 2014a; Baxter and others 2014b; Degenhardt, Baxter, ring in childhood have not been well represented in
and others 2014; Degenhardt, Charlson, and others 2014; burden of disease analyses. GBD 2010 was the first study
Degenhardt, Ferrari, and others 2013; Erskine and others to estimate the burden associated with childhood mental
2013; Ferrari and others 2013a; Ferrari and others 2013b; disorders like autism, Aspergers disorder, attention-deficit
Saha and others 2005). Mental disorders, such as anxiety hyperactivity disorder, and conduct disorder. For coun-
and depressive disorders, were associated with high levels tries such as those in Sub-Saharan Africa, where children
of prevalence and disability. In comparison, schizophre- constitute 40 percent of the population (UN 2011), these
nia was associated with low prevalence but high levels of findings highlight the need for prevention and treat-
disability; an acute state of schizophrenia obtained the ment services targeted to children and adolescents. The
highest disability weight in GBD 2010. The same was availability of such services is often more sporadic than
true for opioid dependence, which, although it had lower that of adult services. In addition, the high burden of
prevalence in comparison with other substance use disor- neurological disorders in elderly persons emphasizes the
ders like cannabis dependence, was associated with high need for the development and implementation of more
disability and death. Migraine, in contrast, was associated effective prevention strategies for these disorders, espe-
with high levels of prevalence but low disability. cially given the worldwide aging of the population, as well
Analysis of burden estimates across time illustrated as the need for equitable health care resource allocation
how population growth and a changing age profile for people affected by neurological disorders.
between 1990 and 2010 produced a shift in the global The GBD 2010 burden estimates also underlined
disease burden from communicable to noncommunica- the extent of the challenge faced by health systems in
ble diseases and from YLLs to YLDs (Murray, Vos, and developed and developing regions as a result of mental,
others 2012). With improvements in infant and mater- neurological, and substance use disorders. Mental dis-
nal health and declining rates of mortality caused by order DALYs are highest in the Middle East and North
infectious diseases, particularly in developing regions, Africa, substance use disorder DALYs are highest in
more people are now living to the age where noncom- Eastern Europe and Central Asia, and neurological dis-
municable diseases such as mental, neurological, and order DALYs are highest in South Asia. These regional
substance use disorders are most prevalent. This demo- differences are driven by the global distribution of
graphic and epidemiological transition is contributing disorder prevalence and, in some instances, deaths.
to a rise in the absolute burden of mental, neurological, Analysis of GBD 2010 prevalence data for mental
and substance use disorders (Whiteford, Degenhardt, disorders highlighted the effect of conflict status on
and others 2013). the estimates. The prevalence of major depressive dis-
Although not adopted in GBD 2010, the use of age order and anxiety disorders was highest in countries
weighting in many economic analyses and in earlier with a history of conflict or war, many of which are in
Global Burden of Mental, Neurological, and Substance Use Disorders: An Analysis from the Global Burden of Disease Study 2010 37
languages do not have the words to describe concepts societies. Definitions of mental, neurological, and sub-
such as sadness or depression consistent with how stance use disorders and the subsequent quantification of
they are described in Western countries. Explanations disability may not be fully representative of non-Western
for the onset and progression of mental, neurological, presentations of these disorders. Further research into the
and substance use disorders may be explained through cross-cultural presentations of these disorders is required
mechanisms such the presence of spirits or curses, rather for a more comprehensive analysis of burden.
than as medical disorders (Jorm 2006).
Epidemiological surveys in many LMICs tend to cap-
ture somatic manifestations of disorders such as depres- ANNEX
sion and anxiety, which may not be as relevant to other The annex to this chapter is as follows. It is available at
countries and cultures (Cheng 2001; Whiteford, Ferrari, http://www.dcp-3.org/mentalhealth.
and others 2013; Yang and Link 2009). In their survey
of mental disorders in China, Phillips and others (2009) Annex 2A. Global Burden of Mental, Neurological,
concluded that some cases of minor depression were and Substance Use Disorders: An Analysis from the
likely misdiagnosed cases of major depressive disorder, Global Burden of Disease Study 2010
given that standard diagnostic criteria were not sensitive
to cross-cultural presentations of this disorder. A task for
upcoming GBD analyses will be to explore the extent to NOTE
which certain disorders are misdiagnosed as other men-
tal or physical disorders in developing countries and the This chapter was previously published in an article by
consequence on burden. H. A. Whiteford, A. J. Ferrari, L. Degenhardt, V. Feigin, and
T. Vos, entitled The Global Burden of Mental, Neurological,
Finally, regular updating of burden of disease esti-
and Substance Use Disorders: An Analysis for the Global Burden
mates, using the most up-to-date epidemiological data of Disease Study 2010. PLoS ONE, 2015: 10 (2): e0116820.
and burden estimation methodology is important. After doi:10.1371/journal.pone.0116820. http://www.ncbi.nlm.nih
GBD 2010 was published, the Institute for Health .gov/pmc/articles/PMC4320057/pdf/pone.0116820.pdf.
Metrics and Evaluation at the University of Washington
endeavored to make available yearly updates of burden
of disease estimates. The Global Burden of Disease Study REFERENCES
2013 (GBD 2013) published in 2015 was the first of these Alonso, J., S. Chatterji, and Y. He, eds. 2013. The Burdens
updates (GBD 2013 DALYs Hale Collaborators 2015). of Mental Disorders: Global Perspectives from the WHO
Although high-level findings were largely consistent World Mental Health Surveys. Cambridge, U.K.: Cambridge
between GBD 2010 and GBD 2013, continued updating University Press.
of estimates presented in this chapter is required. APA (American Psychiatric Association). 2000. Diagnostic and
Statistical Manual of Mental Disorders (DSM-IV-TR). 4th
ed. Text Revision ed. Washington, DC: APA.
CONCLUSIONS Baxter, A. J., T. S. Brugha, H. E. Erskine, R. W. Scheurer, T. Vos,
and others. 2015. The Epidemiology and Global Burden of
According to the findings in GBD 2010, mental, neurolog- Autism Spectrum Disorders. Psychological Medicine 45 (3):
ical, and substance use disorders contribute to a significant 60113. doi:21.1017/S003329171400172X.
proportion of the global burden of disease and will con- Baxter, A. J., G. Patton, L. Degenhardt, K. M. Scott, and
tinue to do so as the shift in burden from communicable H. A. Whiteford. 2013. Global Epidemiology of Mental
to noncommunicable diseases continues. Health systems Disorders: What Are We Missing? PLoS One 8 (6): e65514.
worldwide can respond to these findings by implementing Baxter, A. J., T. Vos, K. M. Scott, A. J. Ferrari, and H. A. Whiteford.
proven, cost-effective interventions; where these are lim- 2014a. The Global Burden of Anxiety Disorders in 2010.
ited, it will be important to support the research necessary Psychological Medicine 44 (11): 236374.
to develop better prevention and treatment options. Baxter, A. J., T. Vos, K. M. Scott, R. E. Norman, A. D. Flaxman,
and others. 2014b. The Regional Distribution of Anxiety
Although GBD 2010 represents the most comprehen-
Disorders: Implications for the Global Burden of Disease
sive assessment of the burden due to mental, neurological, Study, 2010. International Journal of Methods in Psychiatric
and substance use disorders to date, some limitations Research 23 (4): 42238. doi:10.1002/mpr.1444.
need to be acknowledged. For instance, the definition Charlson, F. J., A. J. Baxter, T. Dua, L. Degenhardt,
of burden in GBD 2010 does not extend to welfare loss; H. A. Whiteford, and T. Vos. 2015. Excess Mortality
accordingly, it does not capture all the consequences of from Mental, Neurological, and Substance Use Disorders
mental, neurological, and substance use disorders on in the Global Burden of Disease Study 2010. In Disease
Global Burden of Mental, Neurological, and Substance Use Disorders: An Analysis from the Global Burden of Disease Study 2010 39
A Systematic Analysis for the Global Burden of Disease Whiteford, H. A., L. Degenhardt, J. Rehm, A. J. Baxter,
Study 2010. The Lancet 380 (9859): 216396. A. J. Ferrari, and others. 2013. The Global Burden of
Wang, H., L. Dwyer-Lindgren, K. T. Lofgren, J. K. Rajaratnam, Mental and Substance Use Disorders, 2010. The Lancet
J. R. Marcus, and others. 2012. Age-Specific and Sex-Specific 382 (9904): 157586.
Mortality in 187 Countries, 19702010: A Systematic Whiteford, H. A., A. J. Ferrari, A. J. Baxter, F. J. Charlson, and
Analysis for the Global Burden of Disease Study 2010. L. Degenhardt. 2013. How Did We Arrive at Burden of
The Lancet 380: 207194. Disease Estimates for Mental and Illicit Drug Use Disorders
Wang, P. S., S. Aguilar-Gaxiola, J. Alonso, M. C. Angermeyer, in the Global Burden of Disease Study 2010? Current
G. Borges, and others. 2007. Use of Mental Health Services Opinion in Psychiatry 26 (4): 37683.
for Anxiety, Mood, and Substance Disorders in 17 Countries WHO (World Health Organization). 1992. The ICD-10
in the WHO World Mental Health Surveys. The Lancet Classification of Diseases: Clinical Descriptions and
370 (9590): 84150. Diagnostic Guidelines. Geneva: WHO.
Wang, P. S., M. Angermeyer, G. Borges, R. Bruffaerts, W. Tat . 2008. The Global Burden of Disease: 2004 Update.
Chiu, and others. 2007. Delay and Failure in Treatment Geneva: WHO.
Seeking after First Onset of Mental Disorders in the Yang, L. H., and B. G. Link. 2009. Comparing Diagnostic
World Health Organizations World Mental Health Survey Methods for Mental Disorders in China. The Lancet 373
Initiative. World Psychiatry 6 (3): 17785. (9680): 200204.
Corresponding author: Fiona J. Charlson, University of Queensland, School of Public Health, Herston, Queensland, Australia; fiona_charlson@qcmhr.uq.edu.au.
41
use disorders in LMICs is sparse, but these groups underlying causes of deathfor example, suicide as a
are understood to experience reduced life expectancy, direct result of major depressive disorderand likely
although causes of death may vary across regions. underestimates the true number of deaths attributable to
This chapter explores the cause-specific and excess a particular disorder. However, the estimation of excess
mortality of individual mental, neurological, and sub- mortality using natural history models often includes
stance use disorders estimated by GBD 2010 and dis- deaths from causal and noncausal origins and likely
cusses the results. We present the additional burden that overestimates the true number of deaths attributable to
can be attributed to these disorders, using GBD results a particular disorder. The challenge is to parse out causal
for comparative risk assessments (CRAs) assessing men- contributions to mortality, beyond those already iden-
tal, neurological, and substance use disorders as risk fac- tified as cause-specific, from the effects of confounders.
tors for other health outcomes. We focus on the following The quantification of the burden attributable to risk
mental, neurological, and substance use disorders: factors requires approaches such as CRA, which is now
an integral part of the GBD studies. The fundamental
Mental disorders, including schizophrenia, major
approach is to calculate the proportion of deaths or dis-
depressive disorder, anxiety disorders, bipolar dis-
ease burden caused by specific risk factorsfor example,
order, autistic disorder, and disruptive behavioral
lung cancer caused by tobacco smokingwhile holding
disorders (attention-deficit hyperactivity disorder
all other independent factors constant. A counterfactual
[ADHD] and conduct disorder [CD])
approach is used to compare the burden associated to
Substance use disorders, including alcohol use disor-
an outcome with the amount expected in a hypotheti-
ders (alcohol dependence and fetal alcohol syndrome)
cal situation of ideal risk factor exposure, for example,
and opioid, cocaine, cannabis, and amphetamine
zero prevalence. This provides a consistent method
dependence
for estimating the changes in population health when
Neurological disorders, including dementia, epilepsy,
decreasing or increasing the level of exposure to risk
and migraine.
factors (Lim and others 2012).
For the purposes of GBD 2010, countries were
grouped into 21 regions and 7 super-regions based on
geographic proximity and levels of child and adult mor- METHODOLOGY
tality (IHME 2014; Murray and others 2012). Regions
were further grouped into developed and developing Years of Life Lost and Cause of Death
categories using the GBD 2010 method. Details of coun- The GBD uses YLLs to quantify the fatal burden due
tries in each region and super-region can be found on to a given disease or injury (Lozano and others 2012).
the Institute for Health Metrics and Evaluation (IHME) YLLs are computed by multiplying the number of
website (IHME 2014). deaths attributable to a particular disease at each age by
The mortality associated with a disease can be quan- a standard life expectancy at that age. The standard life
tified using two different, yet complementary, methods expectancy represents the normative goal for survival;
employed as part of the GBD analyses. First, cause-specific for GBD 2010, it was computed based on the lowest
mortality draws on vital registration systems and verbal recorded death rates in any age group in countries with
autopsy studies that identify deaths attributed to a single populations greater than five million (Salomon and
underlying cause using the International Classification others 2012).
of Diseases (ICD) death coding system. Second, GBD Cause-specific death estimates in GBD 2010 were
creates natural history models of disease, drawing on produced from available cause-of-death data for 187
a range of epidemiological inputs, which ultimately countries from 1980 to 2010. Data sources included vital
provide epidemiological estimates for parameters registration, verbal autopsy, mortality surveillance, cen-
including excess mortalitythat is, the all-cause mortal- suses, surveys, hospitals, police records, and mortuaries
ity rate in a population with the disorder above the all- (Lozano and others 2012). Because cause-of-death data
cause mortality rate observed in a population without are often not available or are subject to substantial prob-
the disorder. By definition, the estimates of excess deaths lems of comparability, a method of modeling cause-
include cause-specific deaths. of-death estimates and trends was developed. Cause
Although arbitrary, the ICD conventions are a neces- of Death Ensemble Modeling (CODEm) was used for
sary attempt to deal with the multi-causal nature of mor- all mental, neurological, and substance use disorders
tality and avoid the double-counting of deaths. Despite (Foreman and others 2012). CODEm uses four families
the systems clear strengths, cause-specific mortality of statistical models testing a large set of different mod-
estimated via the ICD obscures the contribution of other els using different permutations of covariates. Model
New cases (incidence) Cases recovered (remission) Deaths in people with disease
attributable to other causes
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 43
Counterfactual Burden and Comparative Risk is evident, the comparatively smaller contribution of
Assessment several mental disorders is a finding that requires further
Using counterfactual analysis, the effect of a risk factor explanation.
can be quantified by comparing the burden associ- Examination of age-standardized YLL rates indicates
ated with an outcome with the amount expected in large variations across the seven GBD 2010 geograph-
a hypothetical situation of ideal risk factor exposure. ical super-regions, primarily because of differences in
Prince and others (2007) have summarized the evi- patterns of alcohol use disorders, drug dependence, and
dence where causal relationships between mental and mental and neurological disorder prevalence. Several
substance use disorders and other health outcomes have regions have significant deviations from the global aver-
been proposed. In GBD 2010, reviews were conducted to age YLL rates (figure 3.3).
assess the strength of evidence for mental, neurological, In figure 3.3, amphetamine and cocaine depen-
and substance use disorders as independent risk factors dence have been aggregated under psychostimulant
for other health outcomes (Charlson and others 2011; dependence. Details of which countries are in each super-
Degenhardt and Hall 2012; Degenhardt, Hall, and oth- region can be found on the IHME website (IHME 2014).
ers 2009; Rehm, Baliunas, and others 2010). Risk factor In 2010, YLL rates were highest in Sub-Saharan
studies were identified through systematic searches Africa (604 YLLs per 100,000 population) and Central/
of published and unpublished data, and information Eastern Europe and Central Asia (593 YLLs per 100,000);
on effect sizes and study characteristics was extracted the causes of these high fatal burden estimates vary
and collated (Charlson and others 2013; Degenhardt, considerably (figure 3.3). In Sub-Saharan Africa, the
Whiteford, and others 2013; Ferrari and others 2014). YLL burden was driven by epilepsy, which accounted
Data were metasynthesized to calculate relative risks for 511 YLLs per 100,000 population. This rate is four-
(RR) for mental and alcohol use disorders (the expo- fold higher than the global average and approximately
sures) as risk factors for other health outcomes. These 85 percent of all YLLs attributed to mental, neurological,
included mental and substance use disorders collectively and substance use disorders in the region. Sub-Saharan
as risk factors for suicide, alcohol use as a risk factor for Africa has comparatively lower YLL rates for substance
a range of health outcomes, and injecting drug use as a use disorders; however, illicit drug dependence YLLs
risk factor for blood-borne viruses. The RR was applied increased by 3.0 percent from 1990 to 2010, almost
to prevalence distributions of the specific exposures by double the average global increase and the highest of
gender and age group for each region to derive pop- all regions. The Middle East and North Africa follows
ulation attributable fractions (PAFs). The additional with a 2.6 percent increase (Degenhardt, Whiteford, and
burden (YLLs and YLDs) attributable to mental, neu- others 2013).
rological, and substance use disorders is the product of The high fatal burden in Central/Eastern Europe and
the PAFs and the burden for the health outcome as esti- Central Asia was largely caused by deaths attributed to
mated in GBD 2010. More detail on the calculation of alcohol use disorders. These disorders accounted for 331
PAFs in GBD 2010 is provided by Lim and others (2012). YLLs per 100,000 population, compared with a global
average of 57 YLLs per 100,000 population. High mor-
tality caused by illicit drug use disorders also contributed
MORTALITY AND MENTAL, NEUROLOGICAL, to the YLL rate in Central/Eastern Europe and Central
AND SUBSTANCE USE DISORDERS Asia, with all substance use disorders together explaining
73 percent of YLLs in the region.
Causal Mortality and Years of Life Lost Substance use disorders also explained a high pro-
The seven disorders for which YLLs were estimated in portion of total mental, neurological, and substance
GBD 2010 were directly responsible for 840,000 deaths use YLLs in Latin America and the Caribbean and in
in 2010, or approximately 20 million YLLs (figure 3.2). HICs. In Latin America and the Caribbean, substance
Online annex 3A further summarizes the YLLs allocated use disorders accounted for 142 YLLs per 100,000 pop-
to mental, neurological, and substance use disorders by ulation (54 percent of the regions mental, neurological,
disorder, age, and gender. The YLLs attributable to each and substance use YLLs). In HICs, substance use dis-
disorder as a proportion of total YLLs caused by men- orders accounted for 151 YLLs per 100,000 population
tal, neurological, and substance use disorders highlight (49 percent of the regions mental, neurological, and
several key points. Globally, epilepsy contributed the substance use YLLs). Countries in East Asia and Pacific
greatest proportion of YLLs within this group, followed exhibit very low YLL rates across all mental, neuro-
by dementia. Although the impact of substance use logical, and substance use disorders, with little change
disorders, specifically alcohol and opioid dependence, observed between 1990 and 2010.
200
Excess Mortality from a Natural History Model
100
The GBD cause-of-death modeling translates to a rela-
tively small YLL burden attributable to mental, neurolog- 0
Sa As ,
ia
b- al pe
ca
st
nia
n
es
al
ea
As
ha ia
ob
ea
ibb
nA
h
le
Gl
Oc
t
u
ou
idd
ar
ra e co
C n
dC
S
te
an
Ce com
an
nd
ia
As
aa
ica
-in
er
gh
ric
t
nt
as
Af
t/E
rth
tin
as
he
La
ut
So
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 45
Figure 3.4 Age-Standardized YLL Rates for Mental, Neurological, and These were not included in the estimated cause-specific
Substance Use Disorders, by GBD 2010 Super-Region and Gender, deaths and YLLs, because the method for cause-of-death
per 100,000 Population, 2010 estimation, where death counts are used to calculate
YLLs, can only be attributed to the primary ICD cause
1,000
of death.
900
800
Examination of excess mortality derived from natu-
YLL rates (per 100,000)
ca
Ce rn E s
sia
ia
al
e
ric
ea
fri
an
ob
as ntri
Su tra ro
ha ia
hA
Mental Disorders
Af
ibb
nA
ce
u
Gl
ou
ut
th
dO
ar
ra
ec
So
or
dC
an
dN
m
an
co
sia
l/E
an
b-
tA
ra
st
gh
er
nt
as
Ea
Am
Hi
Ce
tin
as
he
ut
So
Table 3.1 Presence of Cause-Specific Mortality and Excess Mortality Attributed to Mental, Neurological, and
Substance Use Disorders in GBD 2010
Cause-specific mortality attributed Excess mortality attributed
Disorders to disorders in GBD 2010 to disorders in GBD 2010
Mental disorders
Major depressive disorder No Yes
Anxiety disorders No No
Schizophrenia Yes Yes
Bipolar disorders No Yes
Disruptive behavioral disorders: ADHD and CD No No
Autistic disorder No Yes
Substance use disorders
Alcohol use disordersa Yes Yes
Opioid dependence Yes Yes
Cannabis dependence No No
Amphetamine dependence Yes Yes
Cocaine dependence Yes Yes
Neurological disorders
Epilepsy Yes Yes
Migraine No No
Dementia Yes Yes
Note: ADHD = attention-deficit hyperactivity disorder; CD = conduct disorder; GBD = Global Burden of Disease study.
a. Cause-specific deaths for alcohol use disorders include those from alcohol dependence and fetal alcohol syndrome; differentially, excess deaths represent those from alcohol
dependence only.
0.30
0.15
0.25
0.20
0.10
0.15
0.05 0.10
0.05
0 0
+
4
4
4
9
4
4
4
4
9
4
4
4
4
75
75
2
5
1
5
1
3
4
1
5
2
3
6
7
20
55
15
45
10
15
25
35
45
65
10
20
25
35
55
65
Age Age
0.30
0.020
Global deaths (millions)
0.25
Global deaths (millions)
0.015
0.20
0.010 0.15
0.10
0.005
0.05
0 0
+
+
4
4
4
9
25 4
4
4
4
9
4
75
85
1
3
1
5
4
5
7
1
5
10
25
20
35
45
65
15
65
15
55
45
Age Age
Upper UI (excess deaths) Mean excess deaths Lower UI (excess deaths)
Upper UI (CoD counts) Mean CoD counts Lower UI (CoD counts)
all the excess deaths estimated by DisMod-MR will be depression is linked to higher rates of coronary heart
causally attributable to the disorder. A complex interplay disease (Charlson and others 2011). Lifestyle risk factors
of risk factors will typically contribute to the high rates and the use of medications in the treatment of some
of all-cause mortality in people with mental disorders. mental disorders contribute to higher morbidity and
Mental disorders can directly impact the risk of mortality rates through increased risk of obesity and
chronic disease through underlying biochemical mecha- metabolic dysfunction. Smoking rates are significantly
nisms (Stapelberg and others 2011). For example, major higher in people with mental disorders (Lasser and
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 47
others 2000); this group experiences disproportionate attributable to their condition: a strong and consistent
tobacco-related harm. relationship between schizophrenia and higher death
Despite their increased exposure to chronic disease rates has been shown; the onset of schizophrenia gen-
risk factors, people with mental disorders have inequita- erally precedes the physical health condition causally
ble access to health care, with less opportunity for met- associated with their death; and plausible biological
abolic risk factor screening (Crump and others 2013) pathways exist through the side effects of medication
and early cancer detection (Kisely, Campbell, and Wang and unhealthy behaviors directly related to the condition
2009) and lower rates of common prescriptions and (Laursen, Nordentoft, and Mortensen 2014). Although
procedures (Kisely and others 2007; Laursen and others poverty may be a confounding factor, with schizophre-
2009), even in HICs. nia more prevalent in low socioeconomic populations
that tend to experience poorer health outcomes, evi-
Schizophrenia. People with schizophrenia have well- dence indicates that people with schizophrenia move
documented premature mortality (Laursen 2011), but to these populations because of the impact of their
very few YLLs in GBD 2010. Although schizophrenia disorder, such as difficulty in securing education and
is one of the few mental disorders with cause-specific employment because of cognitive and social problems
deaths permissible by ICD, the number of cause-specific (Lambert, Velakoulis, and Pantelis 2003). Accordingly,
deaths globally (approximately 20,000) is noticeably schizophrenia can be the mediating factor for poorer
lower compared with the number of all-cause deaths socioeconomic and health outcomes.
(approximately 700,000) ascribed by the disorders
natural history. Bipolar Disorder. Approximately 1.3 million excess
Research from HICs suggests that men with schizo- deaths were estimated in the natural history model of
phrenia die about 15 years earlier than men without bipolar disorder. However, in contrast to schizophrenia,
schizophrenia; women with schizophrenia die, on aver- no cause-specific deaths are attributed to the disorder.
age, 12 years earlier than women without schizophrenia The natural history of the disease suggests that bipolar
(Crump and others 2013; Lawrence, Hancock, and Kisely disorder is associated with more excess deaths globally
2013). The majority of these deaths is due to chronic than schizophrenia. Research from the United Kingdom
disease; cardiovascular disease accounts for more than suggests that the excess mortality rates in schizophrenia
33 percent of all premature deaths in those with schizo- and bipolar disorder are comparable (Chang and others
phrenia (Crump and others 2013; Lawrence, Hancock, 2011); the higher number of deaths is likely explained
and Kisely 2013). Suicide, homicide, and accidents by the higher population prevalence of bipolar disorders
account for less than 15 percent of excess deaths (Crump (58.9 million cases in 2010, compared with 23.8 million
and others 2013; Lawrence, Hancock, and Kisely 2013). cases for schizophrenia) (Whiteford and others 2013).
The side effects of antipsychotic medications, par- An estimated 80 percent of premature deaths in peo-
ticularly weight gain and impaired glucose tolerance, ple with bipolar disorder is caused by physical disease,
increase the risk of excess mortality in people regularly almost 50 percent of which is cardiovascular disease
taking these medications. Despite concerns over the side (Westman and others 2013). Unnatural causes account
effects of antipsychotic medication, the lack of antipsy- for nearly 20 percent of premature deaths (Westman and
chotic treatment has been linked with higher all-cause others 2013).
mortality rates (hazard ratio [HR] 1.45; 95% confidence
interval [CI], 1.20-1.76), with the highest risks attrib- Autistic Disorder. GBD 2010 estimated that more than
uted to suicide (HR 2.07; 95% CI, 0.73-5.87) and cancer 100,000 excess deaths were caused by autistic disorder.
(HR 1.94; 95% CI, 1.13-3.32) (Crump and others 2013). There is clear evidence of premature mortality in
Research shows that although cancer-related death rates the natural history of autistic disorder, despite lack
are higher in this group, people with schizophrenia are of disorder-specific deaths registered using ICD codes.
at lower risk of developing cancer (Grinshpoon and People with developmental disorders are at twice the risk
others 2005). High mortality rates therefore likely reflect of premature death compared with the general population
inadequate and unequal access to health care and lower (Mouridsen and others 2008). There are several causes of
rates of diagnostic screening. Multiple medications and elevated death rates in autistic disorder, including acci-
discontinuation of medication also appear to increase dents, respiratory diseases, and seizures (Mouridsen and
the risk of all-cause death (Haukka and others 2008; others 2008; Shavelle, Strauss, and Pickett 2001). Autism
Joukamaa and others 2006). spectrum disorders are highly comorbid, with a range of
Research suggests that the majority of excess mor- potentially life-limiting physical conditions, including
tality in people with schizophrenia could be directly epilepsy and chromosomal disorders such as fragile X
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 49
Figure 3.6 Cause-Specific and Excess Deaths Attributed to Substance Use Disorders, by Age, with Uncertainty, 2010
0.5
0.20
Global deaths (millions)
0.05
0.1
0 0
4
4
+
4
9
4
9
4
4
4
4
4
4
+
1
7
75
1
5
1
1
2
3
4
5
6
7
75
10
15
20
25
35
45
55
65
10
15
20
25
35
45
55
65
Age Age
0.12 0.050
0.045
0.10 0.040
Global deaths (millions)
0.08 0.035
0.030
0.06 0.025
0.020
0.04 0.015
0.010
0.02
0.005
0 0
4
4
9
4
9
4
4
4
4
4
4
+
1
75
1
5
1
1
2
3
4
5
6
7
75
10
15
20
25
35
45
55
65
10
15
20
25
35
45
55
65
Age Age
Upper UI (excess deaths) Mean excess deaths Lower UI (excess deaths)
Upper UI (CoD counts) Mean CoD counts Lower UI (CoD counts)
direction of causality and the effects of confounding and intentional injuries, with strong evidence for
factors remain uncertain (Rao, Daley, and Hammen a dose-response relationship (Rehm, Baliunas, and
2000; Rohde and others 2001). others 2010).
The relationship between alcohol consumption and The risk of death through injuries and self-harm is
liver cirrhosis is well recognized, but alcohol use dis- elevated, accounting for approximately 30 million
orders appear to be more strongly related to cirrhosis YLLs globally.
mortality versus morbidity, as it negatively affects the
course of existing liver disease (Rehm, Baliunas, and The elevated risks in those with alcohol use disorders
others 2010). appear to be mediated by the quantity of alcohol con-
Heavy alcohol use is related to higher rates of infec- sumed and the drinking pattern (Rehm, Baliunas, and
tious diseases, such as tuberculosis, and unintentional others 2010).
Figure 3.7 Numbers of Cause-Specific and Excess Deaths Attributed to Neurological Disorders, by Age, with Uncertainty, 2010
a. Epilepsy b. Dementia
0.07 1.8
1.6
0.06
1.4
Global deaths (millions)
0.05
Global deaths (millions)
1.2
0.04 1.0
0.03 0.8
0.6
0.02
0.4
0.01
0.2
0 0
+
+
4
4
9
4
9
4
4
4
4
4
4
75
75
1
1
5
1
1
2
3
4
5
6
7
10
15
20
25
35
45
55
65
10
15
20
25
35
45
55
65
Age Age
Upper UI (excess deaths) Mean excess deaths Lower UI (excess deaths)
Upper UI (CoD counts) Mean CoD counts Lower UI (CoD counts)
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 51
Mu and others 2011). These differences could be due 2010); 55 percent of the deaths in the cohort were directly
partly to methodological differences or to genuine dif- related to epilepsy, including sudden, unexplained death
ferences caused by the etiology of the disease and envi- in 30 percent, definite or probable seizure in 15 percent,
ronmental risk factors. and accidental drowning in 10 percent.
The proportion of deaths attributable to epilepsy Another important risk factor for premature mortal-
differs by region. In GBD 2010, Sub-Saharan Africa ity is comorbid mental illness. Most studies of mortality
had the highest death rates caused by epilepsy (Murray risk in this population have been conducted in HICs,
and others 2012). Importantly, studies have shown that and the extent of this risk factor in resource-limited
a large proportion of these deathsthose attributable settings is largely unknown. In a Swedish retrospective
to falls, drowning, burns, and status epilepticusis study, 75 percent of epilepsy patients dying from an
preventable (Diop and others 2005; Jette and Trevathan external cause had comorbid psychiatric illness, most
2014). In a large cohort of people with active convulsive commonly depression and substance abuse (Fazel and
epilepsy in rural Kenya, 38 percent of epilepsy-related others 2013). In a population-based study in the United
deaths were caused by status epilepticus. Mortality in Kingdom, mortality among epilepsy patients was asso-
this cohort was more than six-fold greater than expected ciated with alcohol use and depression (Ridsdale and
and associated with nonadherence to (or unavailability others 2011). In a meta-analysis of studies on suicide in
of) anti-epileptic drugs, cognitive impairment, and age epilepsy patients, Pompili and others (2005) found that
(Ngugi and others 2014). the incidence of suicide was significantly higher among
Kamgno, Pion, and Boussinesq (2003) found simi- epilepsy patients than the general population. This strik-
larly high mortality rates in Cameroon, associated with ing mortality risk in epilepsy patients with mental disor-
poor access to or compliance with medical treatment. ders requires further study and intervention in LMICs,
In a study of 164 patients with epilepsy followed for 30 where the burden of epilepsy is highest.
years in Tanzania and treated with phenobarbital, 67.1
percent of the patients died, a mortality rate twice that Dementia. Our natural history model attributed more
of the rural Tanzanian population. The causes of death than two million excess deaths worldwide to dementia
were related to epilepsy in more than 50 percent of the in 2010, compared with 500,000 cause-specific deaths
patients and included status epilepticus, drowning, and derived from ICD records. Figure 3.7 shows that the
burns (Jilek-Aall and Rwiza 1992). majority of deaths caused by dementia, as expected,
In other LMICs outside Sub-Saharan Africa, the pre- occur in the elderly.
ventable causes of death in epilepsy patients are also a Excess mortality in dementia has been associated
significant factor. Drowning is the most common cause with functional disability leading to unhealthy lifestyle
of premature death in rural China (proportional mor- factors and comorbid physical conditions (Guehne,
tality ratio = 82.4 percent). This finding is attributed in Riedel-Heller, and Angermeyer 2005; Llibre and others
part to geographic and occupational risk hazards that 2008). Midlife cardiovascular risk factors have been
include living and working around ponds, paddy fields, associated with later mortality in patients who develop
cesspits, and wells (Mu and others 2011). dementia. In a Norwegian prospective study following
Epilepsy is associated with premature mortality, with patients for 35 years, dementia mortality was associated
the highest SMR in the first one to two years following with increased total cholesterol levels, diabetes mellitus,
diagnosis (Neligan and others 2010). Common causes and low body mass index in midlife (Strand and others
of premature mortality in epilepsy include acute symp- 2013). A study in seven countries found that smoking,
tomatic disorders, such as brain tumor or stroke; sudden hypercholesterolemia, high blood pressure, low forced
unexpected death in epilepsy; suicides; and accidents vital capacity, and previous history of cardiovascular
(Hitiris and others 2007). The epidemiology of premature disease at baseline were associated with a higher risk of
mortality is very relevant in LMICs, where 85 percent of death from dementia (Alonso and others 2009).
those with epilepsy live and where the risk of premature Dementia shows an increased mortality risk. In a
mortality is highest (Diop and others 2005; Jette and study of male civil servants who participated in the
Trevathan 2014; Newton and Garcia 2012). Particularly Israel Heart Disease study, patients with dementia had
concerning is the risk of premature mortality in childhood a hazard ratio for mortality of 2.27 compared with
onset epilepsy. In a prospective trial in Finland of patients patients without dementia (95% CI, 1.922.68) (Beeri
with childhood onset epilepsy followed for 40 years, and Goldbourt 2001).
24 percent of the patients died. This rate is three times The severity of disease is one of the most signifi-
higher than the expected age- and gender-adjusted mor- cant predictors of premature death in individuals with
tality in the general population (Sillanp and Shinnar dementia after controlling for other factors, with an HR
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 53
54
Table 3.2 Number of Cause-Specific and Excess Deaths, by Age, 2010
01 14 59 1014 1519 2024 2534 3544 4554 5564 6574 75+
Cause-specific deaths years years years years years years years years years years years years Total
Alzheimers disease and other dementias - - 869 605 578 642 1,259 2,302 4,575 12,559 41,622 420,710 485,721
Epilepsy 7,388 19,819 6,255 5,351 10,562 14,101 24,107 20,605 18,038 14,826 14,522 22,054 177,627
Schizophrenia - - - - - - 2,003 3,610 3,429 3,440 3,035 4,246 19,763
over and above cause-specific YLLs directly attributable Figure 3.8 Absolute YLLs Attributable to Mental, Neurological,
to these disorders. Variation in absolute YLLs among and Substance Use Disorders as Risk Factors for Other Health
regions is explained not only by population size, but Outcomes, 2010
also the distribution of the risk factors and outcomes 100
in each region. For example, YLLs attributable to alco-
hol use as a risk factor are greatest in Central Europe,
Eastern Europe, and Central Asiarather than South 80
Absolute YLLs (millions)
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 55
to all mental, neurological, and substance use disorders severe and persistent mental disorders may be less
combined. Nevertheless, presenting attributable YLLs likely to receive a timely diagnosis of physical illness
is another example of the deaths and YLLs caused by because of diagnostic overshadowing, that is, physical
these disorders, over and above the direct cause-specific complaints may be overlooked and attributed to psy-
deaths and YLLs allocated to each disorder in GBD chological and psychiatric factors (Bailey, Thorpe, and
2010. It is clear that the mortality-associated disease is Smith 2013). A review by Happell, Scott, and Platania-
significant. Phung (2012) found a reduced likelihood for people
with mental disorders to receive screening for breast,
cervical, and colorectal cancer or immunizations for
DISCUSSION AND IMPLICATIONS influenza and pneumonia, compared with the rest of
the population. Even in countries with well-established
Mental Disorders health care systems, people with mental disorders
The GBD findings of elevated rates of excess mortality receive lower-than-average prescriptions for medica-
across most mental and substance use disorders are sup- tion treating cardiovascular disease (Kisely, Campbell,
ported by the findings of a recent meta-analytic review and Wang 2009; Mitchell and Lord 2010) and are
(Walker, McGee, and Druss 2015). Moreover, recent less likely to receive coronary artery bypass grafting,
studies suggest that the majority of excess deaths are cardiac catheterization, or cerebrovascular arteriogra-
caused by preventable diseases, with a smaller propor- phy (Kisely, Campbell, and Wang 2009; Mitchell and
tion attributed to unnatural or unknown causes (Fekadu Lawrence 2011).
and others 2015; Lawrence, Hancock, and Kisely 2013). Strategies for reducing mortality associated with
The question remains as to what proportion of these mental and substance use disorders primarily target pre-
deaths can be directly attributed to mental disorders and venting onset, reducing case fatality, and preventing the
how much to subsequent confounding factors. development of fatal sequela. Growing evidence indi-
Despite the existence of complex relationships cates that excess mortality in people with these disorders
between mental disorders and premature mortality, can be reduced through established evidence-based
some relationships, such as that between mental dis- treatments and improved screening and treatment for
orders and suicide, are well-established (Li and others chronic disease.
2011). Mental disorders have also been linked to higher Psychiatric treatments, specifically pharmacother-
rates of death caused by cardiovascular disease, stroke, apies, may have some protective effect against excess
diabetes mellitus, respiratory diseases, and some cancers mortality (Weinmann, Read, and Aderhold 2009),
(Crump and others 2013; Hoyer, Mortensen, and Olesen although evidence suggests that this depends on the
2000). The relationship between mental disorders and use of medications according to best practice guidelines
a specific physical disease, leading to premature death, (Cullen and others 2013). However, some antidepres-
is also complex. People with major depression are more sants and second-generation antipsychotics may actually
likely to develop cardiovascular disease (Charlson and pose an elevated risk mediated by metabolic side effects
others 2011). Psychotropic medications can negatively (Newcomer 2005; Rummel-Kluge and others 2010;
impact cardiovascular and metabolic health (De Hert Smith and others 2008).
and others 2012). Obesity and metabolic disturbances Collaborative care by community-based health teams
are primary risk factors for cardiovascular disease and has the potential to reduce overall mortality, as well
type II diabetes, and these are two- to three-fold more as suicide deaths (Dieterich and others 2010; Malone
common in people with mental disorders, compared and others 2007). The use of collaborative care mod-
with the general population (Scott and Happell 2011). els to improve physical health in people with mental,
Major modifiable risk factors for chronic disease, such neurological, and substance use disorders is growing
as smoking (Lawrence, Mitrou, and Zubrick 2009), poor in HICs; these models have demonstrated a range of
diet, physical inactivity (Kilbourne and others 2007; positive health outcomes, including reduced cardiovas-
Shatenstein, Kergoat, and Reid 2007), and substance cular risk profiles (Druss and others 2010). The effec-
abuse (Scott and Happell 2011), are overrepresented in tiveness of these strategies in preventing premature
people with mental disorders. These risk factors may be mortality in LMICs has yet to be tested, but this may be a
the consequences of symptoms of mental, neurological, cost-effective approach to treatment in settings in which
and substance use disorders; medication effects; and trained mental health clinicians are scarce.
poor emotional regulation (Scott and others 2013). Known chronic disease risk factors, such as smok-
Mental disorders are associated with poorer clinical ing and obesity, are potentially modifiable. Lifestyle
management of comorbid conditions. People with interventions comprising a psycho-educational or
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 57
of stimulant dependence (Srisurapanont, Jarusuraisin, many LMICs and upper-middle-income countries (Jette
and Kittirattanapaiboon 2001). The RCTs of prescribed and Trevathan 2014). Legislation to ensure the avail-
psychostimulants in cocaine dependence have not found ability of affordable and efficacious anti-seizure medi-
that they lead to greater abstinence or retention in care cations, clinician education in prescribing anti-epileptic
(Castells and others 2010). medications, and patient education on the importance
In some regions, notably Asia, there is also wide- of medical adherence is critical to alleviate the epilepsy
spread delivery of non-evidence-based responses to treatment gap. Cost-effective epilepsy treatments are
psychostimulant dependence (Degenhardt and others available, and accurate diagnosis can be made without
2010, 2014). Illicit drug users may be detained in costly technical equipment. Targeting epilepsy risk fac-
closed settings, typically operated by military, gov- tors, including more common structural and metabolic
ernment security, or police for what is claimed to be causes of epilepsy, can decrease mortality risk. Education
treatment, most often for psychostimulant use (IHRD and information on safe lifestyle habits in epilepsy
2009; Pearson 2009; UNODC Regional Centre for patients will benefit populations in LMICs, as will edu-
East Asia and the Pacific 2006; WHO 2009). Detainees cation initiatives targeted to employers and teachers to
are often forced to comply with the interventions; dispel the myths associated with epilepsy.
evidence-based, effective drug treatment and HIV The mortality risk of dementia in many LMICs is
prevention are rarely delivered (General Department poorly known. Studies on the mortality rates due to
for Social Evils Prevention, Constella Group, and dementia and the incidence of preventable risk factors in
DFID 2008; IHRD 2009; UNODC Regional Centre for these regions are critical to develop strategies to alleviate
East Asia and the Pacific 2006; WHO 2009). External mortality in this fragile patient population. Mortality in
evaluations have concluded that there may be adverse dementia patients is commonly caused by preventable
impacts on drug use and HIV risk (Pearson 2009), in medical conditions. Caregiver education and support
addition to human rights violations (Human Rights services regarding proper care of patients with cognitive
Watch 2004; IHRD 2009; Pearson 2009; Rehm, Csete, decline will likely decrease infection rates and mortality.
and others 2010; WHO 2009). Government financial support for health care services
Although cannabis dependence had no YLLs, two and caregiver support would benefit this population.
million YLDs were attributed to the disorder. Behavioral Strategies to enhance nutrition, as well as monitoring
interventions are effective in the treatment of cannabis and treatment of vascular risk factors, are important
dependence (Denis and others 2013; Knapp and others measures. Raising awareness of the mortality risk among
2007); cognitive behavioral therapy and contingency the public, caregivers, and health workers can lead to
management show the greatest promise. Public health increased demand for services.
campaigns may be necessary to advise young people of
the risks of developing dependence on cannabis, because
many users fail to appreciate this risk. More research is
needed, however, into how to scale up these behavioral
CONCLUSIONS AND LIMITATIONS
approaches to reduce the population prevalence of these Quantifying mortality presents several challenges.
disorders (Knapp and others 2007). The cause-of-death data are affected by multiple fac-
tors, including certification skills among physicians,
diagnostic and other data available for completing
Neurological Disorders the death certificate, cultural variations in choosing
As the incidence of neurological disorders, including and prioritizing the cause of death, and institutional
epilepsy and dementia, grows in many resource-limited parameters governing mortality reporting (Lozano and
settings, strategies to decrease mortality rates in these others 2012). In LMICs, where many deaths are not
regions in particular must be addressed. Improvements medically certified, different data sources and diagnos-
in access to medical treatment, patient and clinician tic approaches are used to derive cause-of-death esti-
education, and a focus on preventable causes of death mates (Lozano and others 2012). Overall, improving
can substantially decrease mortality rates. and expanding sources of national mortality estimates
In resource-constrained settings, the mortality risk is imperative.
in epilepsy patients is up to six times higher than in Mortality directly related to mental, neurological, and
HICs and largely due to preventable causes (Kamgno, substance use disorders is particularly difficult to cap-
Pion, and Boussinesq 2003; Ngugi and others 2014). ture in cause-of-death data because of the complex web
The epilepsy treatment gap is more than 75 percent in of causality that links these disorders with other physical
low-income countries, and more than 50 percent in disorders. It is important to identify and quantify the
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 59
Beeri, M. S., and U. Goldbourt. 2001. Late-Life Dementia Schizophrenia Bulletin 39 (5): 115968. doi:10.1093/schbul
Predicts Mortality beyond Established Midlife Risk Factors. /sbs097.
American Journal of Geriatric Psychiatry 19 (1): 7987. Darke, S., L. Degenhardt, and R. P. Mattick, eds. 2006. Mortality
Bilder, D., E. L. Botts, K. R. Smith, R. Pimentel, M. Farley, amongst Illicit Drug Users. Cambridge, UK: Cambridge
and others. 2013. Excess Mortality and Causes of Death University Press.
in Autism Spectrum Disorders: A Follow Up of the 1980s Davoli, M., F. Forastiere, D. D. Abeni, E. Rapiti, and C. A. Perucci.
Utah/UCLA Autism Epidemiologic Study. Journal of 1994. Longitudinal and Cross-Sectional Mortality Studies
Autism and Developmental Disorders 43 (5): 1196204. in Injecting Drug Users. Journal of Epidemiology and
doi:10.1007/s10803-012-1664-z. Community Health 48 (1): 10102.
Brugal, M. T., A. Domingo-Salvany, R. Puig, G. Barrio, P. Garcia De Hert, M., J. Detraux, R. van Winkel, W. Yu, and C. U. Correll.
de Olalla, and L. de la Fuente. 2005. Evaluating the Impact 2012. Metabolic and Cardiovascular Adverse Effects
of Methadone Maintenance Programmes on Mortality Due Associated with Antipsychotic Drugs. Nature Reviews
to Overdose and AIDS in a Cohort of Heroin Users in Endocrinology 8: 11426. doi:10.1038/nrendo.2011.156.
Spain. Addiction 100: 98189. Degenhardt, L., C. Bucello, B. Mathers, C. Briegleb, H. Ali,
Caplehorn, J. R. M., and O. H. Drummer. 1999. Mortality M. Hickman, and J. McLaren. 2011. Mortality among
Associated with New South Wales Methadone Programs Regular or Dependent Users of Heroin and Other Opioids:
in 1994: Lives Lost and Saved. Medical Journal of Australia A Systematic Review and Meta-Analysis of Cohort Studies.
170: 10409. Addiction 106 (1): 3251.
Carpio, A., N. E. Bharucha, P. Jallon, E. Beghi, R. Campostrini, Degenhardt, L., A. J. Ferrari, B. Calabria, W. D. Hall,
and others. 2005. Mortality of Epilepsy in Developing R. E. Norman, and others. 2013. The Global Epidemiology
Countries. Epilepsia 46 (Suppl. 11): 2832. doi:10.1111 and Contribution of Cannabis Use and Dependence to
/j.1528-1167.2005.00404.x. the Global Burden of Disease: Results from the GBD 2010
Castells, X., M. Casas, C. Perez-Mana, C. Roncero, and Study. PLoS One 8 (10): e76635. doi:10.1371/journal
D. Capella. 2010. Efficacy of Psychostimulant Drugs for .pone.0076635.
Cocaine Dependence. Cochrane Database of Systematic Degenhardt, L., and W. Hall. 2012. Extent of Illicit Drug
Reviews (2). doi:10.1002/14651858.CD007380.pub3. Use and Dependence, and Their Contribution to the
Centre for Mental Health. 2010. The Economic and Social Global Burden of Disease. The Lancet 379 (9810): 5570.
Costs of Mental Health Problems in 2009/10. Centre for doi:http://dx.doi.org/10.1016/S0140-6736(11)61138-0.
Mental Health. http://www.centreformentalhealth.org.uk Degenhardt, L., W. D. Hall, M. Lynskey, J. McGrath, J. McLaren,
/pdfs/Economic_and_social_costs_2010.pdf. and others. 2009. Should Burden of Disease Estimates
Chang, C.-K., R. D. Hayes, G. Perera, M. T. M. Broadbent, Include Cannabis Use as a Risk Factor for Psychosis?
A. C. Fernandes, and others. 2011. Life Expectancy at Birth PLoS Medicine 6 (9): e1000133.
for People with Serious Mental Illness and Other Major Degenhardt, L., B. Mathers, P. Vickerman, T. Rhodes,
Disorders from a Secondary Mental Health Care Case C. Latkin, and others. 2010. Prevention of HIV Infection
Register in London. PLoS One 6 (5). doi:10.1371/journal for People Who Inject Drugs: Why Individual, Structural,
.pone.0019590. and Combination Approaches Are Needed. The Lancet
Charlson, F. J., A. E. Moran, G. Freedman, R. E. Norman, 376: 285301.
N. J. C. Stapelberg, and others. 2013. The Contribution of Degenhardt, L., B. M. Mathers, A. L. Wirtz, D. Wolfee,
Major Depression to the Global Burden of Ischaemic Heart A. Kamarulzamanf, and others. 2014. What Has Been
Disease: A Comparative Risk Assessment. BMC Medicine Achieved in HIV Prevention, Treatment and Care for
11 (250). doi:10.1186/1741-7015-11-250. People Who Inject Drugs, 20102012? A Review of the Six
Charlson, F. J., N. Stapelberg, A. J. Baxter, and H. Whiteford. Highest Burden Countries. International Journal of Drug
2011. Should Global Burden of Disease Estimates Include Policy 25 (1): 5360. doi:10.1016/j.drugpo.2013.08.004.
Depression as a Risk Factor for Coronary Heart Disease? Degenhardt, L., D. Randall, W. Hall, M. Law, T. Butler, and
BMC Medicine 9 (1): 47. others. 2009. Mortality among Clients of a State-Wide
Chesney, E., G. E. Goodwin, and S. Fazel. 2014. Risks of Opioid Pharmacotherapy Programme Over 20 Years: Risk
All-Cause and Suicide Mortality in Mental Disorders: A Factors and Lives Saved. Drug and Alcohol Dependence 105
Meta-Review. World Psychiatry 13 (2): 15360. doi:10.1002 (12): 915.
/wps.20128. Degenhardt, L., H. A. Whiteford, A. J. Ferrari, A. J. Baxter,
Crump, C., M. A. Winkleby, K. Sundquist, and J. Sundquist. F. J. Charlson, and others. 2013. Global Burden of Disease
2013. Comorbidities and Mortality in Persons with Attributable to Illicit Drug Use and Dependence: Findings
Schizophrenia: A Swedish National Cohort Study. American from the Global Burden of Disease Study 2010. The Lancet
Journal of Psychiatry 170 (3): 32433. doi:10.1176/appi 382 (9904): 156474.
.ajp.2012.12050599. Denis, C., E. Lavie, M. Fatseas, and M. Auriacombe. 2013.
Cullen, B. A., E. E. McGinty, Y. Zhang, S. C. Dosreis, Psychotherapeutic Interventions for Cannabis Abuse and/
D. M. Steinwachs, and others. 2013. Guideline-Concordant or Dependence in Outpatient Settings. Cochrane Database
Antipsychotic Use and Mortality in Schizophrenia. of Systematic Reviews. doi:10.1002/14651858.CD005336.pub3.
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 61
Jerome, L., A. Segal, and L. Habinski. 2006. What We Know Procedures, and Mortality from Heart Disease in Patients
about ADHD and Driving Risk: A Literature Review, Meta- with Severe Mental Disorder. Archives of General Psychiatry
Analysis and Critique. Journal of the Canadian Academy of 66 (7): 71320. doi:10.1001/archgenpsychiatry.2009.61.
Child and Adolescent Psychiatry 15 (3): 10525. Laursen, T. M., M. Nordentoft, and P. B. Mortensen. 2014.
Jette, N., and E. Trevathan. 2014. Saving Lives by Treating Excess Early Mortality in Schizophrenia. Annual Review
Epilepsy in Developing Countries. Neurology 82 (7): of Clinical Psychology 10: 42528. doi:10.1146/annurev
55253. -clinpsy-032813-153657.
Jilek-Aall, L., and H. T. Rwiza. 1992. Prognosis of Epilepsy Lawrence, D., K. J. Hancock, and S. Kisely. 2013. The Gap
in a Rural African Community: A 30-Year Follow-Up of in Life Expectancy from Preventable Physical Illness in
164 Patients in an Outpatient Clinic in Rural Tanzania. Psychiatric Patients in Western Australia: Retrospective
Epilepsia 33 (4): 64550. Analysis of Population Based Registers. British Medical
Joukamaa, M., M. Heliovaara, P. Knekt, A. Aromaa, R. Raitasalo, Journal (346): f2539. doi:10.1136/bmj.f2539.
and V. Lehtinen. 2006. Schizophrenia, Neuroleptic Lawrence, D., F. Mitrou, and S. Zubrick. 2009. Smoking and
Medication and Mortality. British Journal of Psychiatry 188 Mental Illness: Results from Population Surveys in Australia
(2): 12227. doi:10.1192/bjp.188.2.122. and the United States. BMC Public Health 9 (1): 285.
Kamgno, J., S. D. S. Pion, and M. Boussinesq. 2003.Demographic Lawrinson, P., R. Ali, A. Buavirat, S. Chiamwongpaet,
Impact of Epilepsy in Africa: Results of a 10-Year Cohort S. Dvoryak, and others. 2008. Key Findings from the WHO
Study in a Rural Area of Cameroon. Epilepsia 44 (7): Collaborative Study on Substitution Therapy for Opioid
95663. doi:10.1046/j.1528-1157.2003.59302.x. Dependence and HIV/AIDS. Addiction 103 (9): 148492.
Kilbourne, A. M., D. L. Rofey, J. F. McCarthy, E. P. Post, D. Welsh, Lee, L.-C., R. A. Harrington, J. J. Chang, and S. L. Connors. 2008.
and others. 2007. Nutrition and Exercise Behavior among Increased Risk of Injury in Children with Developmental
Patients with Bipolar Disorder 1. Bipolar Disorders 9 (5): Disabilities. Research in Developmental Disabilities 29:
44352. doi:10.1111/j.1399-5618.2007.00386.x. 24755.
Kisely, S., and L. A. Campbell. 2008. Use of Smoking Cessation Li, Z., A. Page, G. Martin, and R. Taylor. 2011. Attributable
Therapies in Individuals with Psychiatric Illness: An Update Risk of Psychiatric and Socio-Economic Factors for
for Prescribers. CNS Drugs 22 (4): 26373. Suicide from Individual-Level, Population-Based Studies:
Kisely, S., L. A. Campbell, and Y. Wang. 2009. Treatment of A Systematic Review. Social Science & Medicine 72 (4):
ISCHAEMIC Heart Disease and Stroke in Individuals 60816. doi:10.1016/j.socscimed.2010.11.008.
with Psychosis under Universal Healthcare. The British Liao, C. C., W. W. Shen, C. C. Chang, H. Chang, and
Journal of Psychiatry 195 (6): 54550. doi:10.1192/bjp T. L. Chen. 2013. Surgical Adverse Outcomes in Patients
.bp.109.067082. with Schizophrenia: A Population-Based Study. Annals of
Kisely, S., M. Smith, D. Lawrence, M. Cox, L. A. Campbell, and Surgery 257 (3): 43338.
others. 2007. Inequitable Access for Mentally Ill Patients Lim, S. S., T. Vos, A. D. Flaxman, G. Danaei, K. Shibuya, and
to Some Medically Necessary Procedures. Canadian others. 2012. A Comparative Risk Assessment of Burden of
Medical Association Journal 176 (6): 77984. doi:10.1503 Disease and Injury Attributable to 67 Risk Factors and Risk
/cmaj.060482. Factor Clusters in 21 Regions, 19902010: A Systematic
Knapp, P. W., B. Soares, M. Farrell, and M. Silva de Lima. Analysis for the Global Burden of Disease Study 2010.
2007. Psychosocial Interventions for Cocaine and Psycho- The Lancet 380 (9859): 222460.
stimulant Amphetamines Related Disorders. Cochrane Llibre R., J., A. Valhuerdi, I. I. Sanchez, C. Reyna, and
Database of Systematic Reviews 3. doi:10.1002/14651858 M. A. Guerra. 2008. The Prevalence, Correlates and Impact
.CD003023.pub2. of Dementia in Cuba: A 10/66 Group Population-Based
Lambert, T. J. R., D. Velakoulis, and C. Pantelis. 2003. Survey. Neuroepidemiology 31: 24351.
Medical Comorbidity in Schizophrenia. Supplement: Lozano, R., M. Naghavi, K. Foreman, S. Lim, K. Shibuya, and
Comprehensive Care for People with Schizophrenia Living others. 2012. Global and Regional Mortality from 235
in the Community. Medical Journal of Australia 178 (9): Causes of Death for 20 Age Groups in 1990 and 2010: A
6770. Systematic Analysis for the Global Burden of Disease Study
Lasser, K., J. Boyd, S. Woolhandler, D. U. Himmelstein, 2010. The Lancet 380 (9859): 2095128.
D. McCormick, and others. 2000. Smoking and Mental Malone, D., S. Marriott, G. Newton-Howes, S. Simmonds,
Illness: A Population-Based Prevalence Study. Journal and P. Tyrer. 2007. Community Mental Health Teams
of the American Medical Association 284 (20): 260610. (CMHTs) for People with Severe Mental Illnesses and
doi:10.1001/jama.284.20.2606. Disordered Personality. Cochrane Database of Systematic
Laursen, T. M. 2011. Life Expectancy among Persons Reviews 3. doi:10.1002/14651858.CD000270.pub2.
with Schizophrenia or Bipolar Affective Disorder. Mathers, B. M., L. Degenhardt, H. Ali, L. Wiessing, M. Hickman,
Schizophrenia Research 131 (13): 10104. doi:10.1016/j and others. 2010. HIV Prevention, Treatment, and Care
.schres.2011.06.008. Services for People Who Inject Drugs: A Systematic Review
Laursen, T., T. Munk-Olsen, E. Agerbo, C. Gasse, and P. Mortensen. of Global, Regional, and National Coverage. The Lancet
2009. Somatic Hospital Contacts, Invasive Cardiac 375 (9719): 101428.
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 63
Antipsychotics in the Treatment of Schizophrenia: A Coronary Heart Disease. Australian and New Zealand
Systematic Review and Meta-Analysis. Schizophrenia Journal of Psychiatry 45: 35169.
Research 123 (23): 22533. doi:http://dx.doi.org/10.1016/j Strand, B. H., E. M. Langballe, V. Hjellvik, M. Handal, O. Nss, G.
.schres.2010.07.012. P. Knudsen, and others. 2013. Midlife Vascular Risk Factors
Salomon, J. A., H. Wang, M. K. Freeman, T. Vos, and Their Association with Dementia Deaths: Results from
A. D. Flaxman, and others. 2012. Healthy Life Expectancy a Norwegian Prospective Study Followed Up for 35 Years.
for 187 Countries, 19902010: A Systematic Analysis for the Journal of Neurological Sciences 324 (1-2): 12430.
Global Burden Disease Study 2010. The Lancet 380 (9859): Swan, T. 2011. Hepatitis C Treatment Pipeline Report. New York,
214462. NY: Treatment Action Group.
Schneider, L.S., K. S. Dagerman, and P. Insel. 2005. Risk Thornicroft, G. 2013. Premature Death among People with
of Death with Atypical Antipsychotic Drug Treatment Mental Illness: At Best a Failure to Act on Evidence; at
for Dementia: Meta-Analysis of Randomized Placebo- Worst a Form of Lethal Discrimination. British Medical
Controlled Trials. Jama 294 (15): 193443. Journal 346. doi:10.1136/bmj.f2969.
Scott, D., and B. Happell. 2011. The High Prevalence of Poor Trinka, E., G. Bauer, W. Oberaigner, J. P. Ndayisaba, K. Seppi,
Physical Health and Unhealthy Lifestyle Behaviours in and others. 2013. Cause-Specific Mortality among Patients
Individuals with Severe Mental Illness. Issues in Mental with Epilepsy: Results from a 30-Year Cohort Study.
Health Nursing 32: 58997. Epilepsia 54 (3): 495501. doi:10.1111/epi.12014.
Scott, K. M., B. Wu, K. Saunders, C. Benjet, Y. He, and others. Turner, K. M. E., S. Hutchinson, P. Vickerman, V. Hope,
2013. Early-Onset Mental Disorders and Their Links to N. Craine, and others. 2011. The Impact of Needle and
Chronic Physical Conditions in Adulthood. In The Burdens Syringe Provision and Opiate Substitution Therapy on the
of Mental Disorders. Global Perspectives from the WHO World Incidence of Hepatitis C Virus in Injecting Drug Users:
Mental Health Surveys, 8796, edited by J. Alonso, S. Chatterji, Pooling of UK Evidence. Addiction 106 (11): 197888.
and Y. He. New York: Cambridge University Press. UNODC (United Nations Office on Drugs and Crime) Regional
Shatenstein, B., M.-J. Kergoat, and I. Reid. 2007. Poor Nutrient Centre for East Asia and the Pacific. 2006. HIV/AIDS
Intakes During 1-Year Follow-Up with Community and Custodial Settings in South East Asia: An Exploratory
Dwelling Older Adults with Early-Stage Alzheimer Dementia Review into the Issues of HIV/AIDS and Custodial Settings
Compared to Cognitively Intact Matched Controls. Journal in Cambodia, China, Lao PDR, Myanmar, Thailand and
of the American Dietetic Association 107 (12): 209199. Viet Nam. Bangkok: UNODC.
doi:http://dx.doi.org/10.1016/j.jada.2007.09.008. U.S. FDA (U.S. Food and Drug Administration). 2008.
Shavelle, R. M., D. J. Strauss, and J. Pickett. 2001. Causes of Information for Healthcare Professionals: Conventional
Death in Autism. Journal of Autism and Developmental Antipsychotics. http://www.fda.gov/Drugs/DrugSafety
Disorders 31 (6): 56976. /PostmarketDrugSafetyInformationforPatientsand
Sillanp, M., and S. Shinnar. 2010. Long-Term Mortality Providers/ucm124830.htm.
in Childhood-Onset Epilepsy. New England Journal of van Hasselt, F. M., P. F. Krabbe, D. G. van Ittersum,
Medicine 363 (26): 252229. M. J. Postma, and A. J. Loonen. 2013. Evaluating
Smith, M., D. Hopkins, R. C. Peveler, R. I. G. Holt, Interventions to Improve Somatic Health in Severe Mental
M. Woodward, and others. 2008. First- v. Second- Illness: A Systematic Review. Acta Psychiatrica Scandinavica
Generation Antipsychotics and Risk for Diabetes in 128: 25160. doi:10.1111/acps.12096.
Schizophrenia: Systematic Review and Meta-Analysis. Verhaeghe, N., J. De Maeseneer, L. Maes, C. Van Heeringen, and
British Journal of Psychiatry 192 (6): 40611. doi:10.1192 L. Annemans. 2011. Effectiveness and Cost-Effectiveness
/bjp.bp.107.037184. of Lifestyle Interventions on Physical Activity and Eating
Spencer, D. 2014. Understanding the Sources of Excess Habits in Persons with Severe Mental Disorders: A
Mortality in Epilepsy. Epilepsy Currents 14 (1): 3132. Systematic Review. International Journal of Behavioral
doi:10.5698/1535-7597-14.1.31. Nutrition and Physical Activity 8: 28.
Sporer, K. A. 2003. Strategies for Preventing Heroin Overdose. von Stumm, S., I. J. Deary, M. Kivimki, M. Jokela, H. Clark,
British Medical Journal 326 (7386). and others. 2011. Childhood Behavior Problems and
Sporer, K. A., and A. H. Kral. 2007. Prescription Naloxone: Health at Midlife: 35-Year Follow-Up of a Scottish Birth
A Novel Approach to Heroin Overdose Prevention. Annals Cohort. Journal of Child Psychology and Psychiatry 52 (9):
of Emergency Medicine 49 (2): 17277. 9921001. doi:10.1111/j.1469-7610.2011.02373.x.
Srisurapanont, M., N. Jarusuraisin, and P. Kittirattanapaiboon. Walker, E., R. E. McGee, and B. G. Druss. 2015. Mortality in
2001. Treatment for Amphetamine Dependence and Mental Disorders and Global Disease Burden Implications:
Abuse. Cochrane Database of Systematic Reviews (4). A Systematic Review and Meta-Analysis. JAMA Psychiatry
doi:10.1002/14651858.CD003022.pub2. 72 (4): 33441. doi:10.1001/jamapsychiatry.2014.2502.
Stapelberg, N. J. C., D. L. Neumann, D. H. K. Shum, Weinmann, S., J. Read, and V. Aderhold. 2009. Influence of
H. McConnell, and I. Hailton-Craig. 2011. A Topographical Antipsychotics on Mortality in Schizophrenia: Systematic
Map of the Causal Network of Mechanisms Underlying Review. Schizophrenia Research 113 (1): 111. doi:http://
the Relationship between Major Depressive Disorder and dx.doi.org/10.1016/j.schres.2009.05.018.
Excess Mortality from Mental, Neurological, and Substance Use Disorders in the Global Burden of Disease Study 2010 65
Chapter
4
Adult Mental Disorders
Steven Hyman, Rachana Parikh,
Pamela Y. Collins, and Vikram Patel
Corresponding author: Steven Hyman, Stanley Center for Psychiatric Research, Broad Institute of the Massachusetts Institute of Technology and Harvard
University; and Department of Stem Cell and Regenerative Biology, Harvard University, Cambridge, Massachusetts, United States of America, seh@harvard.edu.
67
Figure 4.1 Years Lived with Disability Caused by Unipolar Depression, Among the disorders discussed in this chapter,
Anxiety Disorders, Bipolar Disorder, and Schizophrenia Globally schizophrenia and bipolar disorder are the most highly
influenced by genes, with estimated heritabilities of 65
60 to 80 percent (Sullivan, Daly, and ODonovan 2012).
Genotyping of nearly 40,000 individuals with schizo-
50 phrenia and a larger number of healthy comparison
subjects has revealed 108 genomewide significant loci
that contribute to risk, with different combinations of
40
risk alleles acting in different individuals (Ripke and oth-
YLDs in millions
rs
rs
rs
rs
ys
ar
ea
ea
ea
ea
ea
nt
da
ye
9y
mo
+y
9y
9y
14
70
5
2
6
59
0
30
50
15
60
Cognitive behavioral therapy (Orgeta and others 2014; For refractory depression:
Wilson, Mottram, and Vassilas 2008)a Combined CBT and antidepressant (Wiles and
Behavioral therapies (Shinohara and others 2013)a others 2013)c
Psychodynamic therapies (Abbass and others 2014)a Electroconvulsive therapy (Martinez-Amoros and
others 2012; UK ECT Review Group 2003)a
Interpersonal psychotherapy (de Mello and others 2005)a
Transcranial magnetic stimulation (Gaynes and
others 2014)a
Notes Antidepressants are also effective for depression in people with physical illnesses (Rayner and others 2010).a
Antidepressants can be effectively prescribed in primary care settiings (Arroll and others 2009).a
Problem-solving therapy can be delivered by general practitioners (Huibers and others 2007).a
Group interpersonal therapy is effective in community-based, low-resource settings (Bass and others 2006).c
Older tricyclic antidepressants are similar in efficacy to newer drugs, but have greater side effects (Mottram, Wilson, and
Strobl 2006).a
Continuation of treatment with drugs for 912 months following response to medication reduces the risk of relapse
(Kaymaz and others 2008;b Wilkinson and Izmeth 2012a).
Evidence to suggest the superiority of one type of psychological intervention over another is limited (Cuijpers and others
2008;b Moradveisi and others 2013c).
Bipolar disorder Combination of second-generation antipsychotics and Psychotherapies like CBT, group psychoeducational
mood stabilizers for acute mania (Scherk, Pajonk, and therapy, and family therapy (Soares-Weiser and
Leucht 2007)a others 2007)a
Lithium, valproate, lamotrigine, and olanzapine for
maintenance therapy to prevent relapse (Soares-Weiser and
others 2007)a
Anxiety disorders
Anxiety disorders Antidepressants (Kapczinski and others 2003)a
Generalized anxiety disorder CBT-based psychotherapies (Hunot and others 2007)a
Panic disorder Combined therapy (CBT and antidepressants) or CBT alone
(Furukawa, Watanabe, and Churchill 2007)a
Post-traumatic stress No psychological intervention can be recommended routinely Non-trauma focused CBT and eye movement
disorder following traumatic events, and this may also have adverse desensitization and reprocessing (Bisson and others
effects on some individuals (Roberts and others 2009). 2013)a
SSRI antidepressants (Stein, Ipser, and Seedat 2006)a
CBT (particularly trauma-focused CBT) (Roberts and others 2010)
table continues next page
Persons with severe mental illnesses occasion- with severe mental illnesses (Kinoshita and others 2013).
ally require short periods of hospitalization and/or Systematic reviews have shown that life skills and social
longer-term supported housing because of the severity of skills training have moderate to strong effectiveness
their disorders and associated behaviors or abandonment to promote integration of persons with severe mental
by family. Systematic reviews have suggested that acute illnesses in communities where they live; and interven-
day hospitals can be as effective as inpatient care (Marshall tions with a greater client-centered approach have a
and others 2011) and that day hospitals may prevent the larger impact (Gibson and others 2011).
need for inpatient care (Shek and others 2009).
Packages of Care
Occupational Therapy. Occupational therapy inter- Promotion and Prevention. Indicated or targeted
ventions aim to support and improve skills for daily prevention of mental disorders is effective in the
living through life skills training, cognitive rehabilita- early and subclinical stages. A meta-analysis of 32
tion, supportive employment and education, and social studies (largely from Europe and the United States)
and interpersonal skills training. Occupational therapy concluded that preventive interventions could lower
is effective in rehabilitating persons with depression by the incidence of depression by 21 percent through
increasing productivity, reducing work-related stress, psychological interventions such as cognitive behav-
and helping in recovery (Hees and others 2013; Schene ioral therapy (CBT), interpersonal therapy, individual
and others 2007). Supported employment is effective in counseling, and group sessions (van Zoonen and
improving a number of vocational outcomes in persons others 2014). Psychological treatment of subclinical
Corresponding author: Tarun Dua, MD, MPH Programme for Neurological Diseases and Neuroscience Evidence, Research and Action on Mental and Brain Disorders,
Department of Mental Health and Substance Abuse, World Health Organization; duat@who.int.
87
EPILEPSY Epidemiology and Burden of Disease
Definitions A worldwide systematic review of prevalence has
not yet been published; in general, the prevalence
Epilepsy is a brain disorder traditionally defined as the in door-to-door studies has been reported to range
occurrence of two unprovoked seizures occurring more from 2.2 per 1,000 to 41.0 per 1,000 persons, often
than 24 hours apart with an enduring predisposition to with higher estimates in LMICs (Banerjee, Filippi, and
generate further seizures (Fisher and others 2014). In Allen Hauser 2009; Benamer and Grosset 2009; Burneo,
2014, the International League against Epilepsy provided Tellez-Zenteno, and Wiebe 2005; Forsgren and others
an enhanced definition of epilepsy (box 5.1). 2005; Mac and others 2007). The median incidence per
Epilepsy is considered to be resolved if a person has 100,000 per year is higher in LMICs at 81.7 (interquartile
an age-dependent syndrome that is now beyond the range (IQR) 28.0-239.5) compared with HICs at 45.0
expected age for this syndrome, or if the individual (IQR 30.3-66.7) (Ngugi and others 2011).
remained seizure free for the past 10 years and was off The higher estimates of prevalence or incidence
anti-epileptic drugs for at least the past five years (Fisher rates reported in many LMICs are thought to be
and others 2014). Those who continue to have seizures caused by the occurrence of endemic conditions, such
despite an adequate trial of a regimen of two tolerated as malaria or neurocysticercosis; the higher incidence
and appropriately chosen anti-epileptic drugs (AEDs), of road traffic injuries; birth-related injuries; and
whether in monotherapy or polytherapy, are considered variations in medical infrastructure, availability of
to be drug resistant. Epilepsy can be classified in three preventative health programs, and accessible care. In
categories: HICs, the prevalence of epilepsy is stable until after
age 50, when it increases; in contrast, the prevalence
Structural or metabolic epilepsies, for example, in LMICs tends to be stable in the third and fourth
epilepsy caused by a remote stroke decade of life, drops in the fifth decade, and, in some
Epilepsies of genetic or presumed genetic origin, for studies, increases again after age 60 (Banerjee, Filippi,
example, juvenile myoclonic epilepsy and Allen Hauser 2009).
Epilepsies of unknown causes (Berg and others 2010). Epilepsy is associated with premature mortality,
with the highest standardized mortality ratio encoun-
Examples of more common causes of epilepsy tered in the first year or two after diagnosis (Neligan
include brain tumors, infectious diseases, brain injury, and others 2010). In general, the standardized mor-
stroke, and hippocampal sclerosis. Less frequent causes tality ratio for epilepsy is approximately 3 (Hitiris
include genetic causes, autoimmune causes, and mal- and others 2007). The epidemiology of premature
formations of cortical development (Bhalla and others mortality is particularly relevant in LMICs, where 85
2011). Perinatal and infection-related etiologies often percent of those with epilepsy live and where the risk of
predominate in LMICs. premature mortality is highest (Diop and others 2005;
Jette and Trevathan 2014; Newton and Garcia 2012).
Most concerning is the fact that a greater proportion
of deaths in LMICs are potentially preventable, such as
Box 5.1 falls, drowning, burns, and status epilepticus (Diop and
others 2005; Jette and Trevathan 2014). For example, 38
percent of all epilepsy-related deaths in a large cohort
Definition of Epilepsy of people with convulsive epilepsy in rural Kenya
A person has epilepsy if he or she meets any of the follow- were caused by status epilepticus (Ngugi and others
ing criteria (Fisher and others 2014): 2014). Status epilepticus is defined as ongoing seizure
activity lasting five minutes or more, or two or more
seizures without recovery of consciousness in between
At least two unprovoked (or reflex) seizures occurring
(Lowenstein and others 2001). This is an important
more than 24 hours apart
definition, as evidence suggests that seizures lasting
One unprovoked (or reflex) seizure and a probability
more than five minutes are unlikely to self-terminate.
of further seizures similar to the general recurrence
Other common causes of premature mortality in those
risk (at least 60 percent) after two unprovoked seizures,
with epilepsy include acute symptomatic disorders (for
occurring over the next 10 years
example, brain tumor or stroke), sudden unexpected
Diagnosis of an epilepsy syndrome.
death in epilepsy, suicide, and accidents (Hitiris and
others 2007).
Neurological Disorders 89
sustainability and impact remain to be determined Pharmacological Interventions
(Fiest and others 2014). A broad approach is needed to The decision to initiate treatment with anti-epileptic
target stigma at the population level through legislation drugs can be challenging. Analysis of the Multicentre
and advocacy. In addition, education and information trial for Early Epilepsy and Single Seizures suggests
provision to dispel myths and enhance seizure manage- little benefit in initiating treatment for those who pres-
ment among employers and teachers should empower ent with a single seizure, with no known neurological
those with epilepsy to seek treatment and encourage disorder, and normal electroencephalograms (EEGs)
them to be more actively engaged in their communities. (Kim and others 2006). However, medical management
The cost-effectiveness of interventions to reduce stigma should be considered in those who are at moderate to
has not been formally assessed. high risk, defined as more than two to three seizures
at presentation, underlying neurological disorders, and
Legislation. One of the greatest contributors to the epi- abnormal EEGs (Kim and others 2006). More than 60
lepsy treatment gap in LMICs is the lack of availability randomized control trials (RCTs), mostly in HICs, have
of anti-epileptic drugs. The second-generation medica- examined the efficacy of anti-epileptic drugs, but there
tions are not available in the majority of countries, and continues to be a lack of well-designed RCTs examining
even the older anti-epileptic drugs are only available the efficacy of these medications for patients with gener-
sporadically. Investigators in Zambia who surveyed alized epilepsy syndromes and for children (Glauser and
111 pharmacies found that 49.1 percent did not carry others 2013). Newer AEDs tend to be better tolerated,
anti-epileptic drugs. Pediatric syrups that are extensively with fewer long-term side effects, but otherwise their
used in HICs were universally unavailable (Chomba and superiority has not been proven.
others 2010). Regrettably, personal communications Studies comparing the cost-effectiveness of anti-
with epilepsy care providers in other LMICs suggested epileptic drugs in new onset epilepsy have not been
that this problem may be widespread (Chomba and conducted. A recent systematic review summarizes the
others 2010). evidence regarding their efficacy as initial monotherapy
Clearly, policies are warranted to guarantee the in those with epilepsy. Monotherapy with any of the
ongoing availability of affordable and efficacious standard anti-epileptic drugs (carbamazepine, pheno-
anti-epileptic drugs to patients worldwide. Few coun- barbital, phenytoin, and valproic acid) should be offered
tries have a separate budget for epilepsy services, and to children and adults with convulsive epilepsy. Several
national funding support for epilepsy care is needed. lower-quality studies have demonstrated efficacy for
Out-of-pocket expenses are the primary source of phenobarbital in adults and children with partial onset
financing epilepsy care in 73 percent of low-income seizures and generalized onset tonic-clonic seizures
countries, including many countries in Africa, the (Glauser and others 2013). Given the acquisition costs,
Eastern Mediterranean, and South-East Asia, where the phenobarbital should be offered as a first option if
burden is highest (WHO 2011). Disability benefits do availability can be ensured. If available, carbamazepine
not exist in many regions, and patients are unable to should be offered to children and adults with partial
receive monetary support. onset seizures (WHO 2009b). Using the lowest possi-
ble dose should minimize side effects, improve seizure
Self-Management outcomes, and decrease the treatment gap. Valproic acid
Self-management is empowering patients to partici- and ethosuximide have been shown to be most effec-
pate more actively in managing their care. Patients are tive in the management of absence seizures, especially
likely to improve their understanding, adopt health- in children, although valproic acid is recommended,
ier lifestyles, and improve adherence to treatment as it is on the list of essential medicines. Ethosuximide
(Fitzsimons and others 2012). Self-management can is available as a complementary medication. However,
help those with epilepsy better identify and manage the medication should be avoided, when possible, in
their seizure triggers, which can reduce frequency and women of childbearing potential because of its higher
decrease health services utilization and health care association with major congenital malformations and
costs (Fitzsimons and others 2012). A few studies have poorer neurodevelopmental outcomes. Although newer
examined the effectiveness of self-management edu- therapeutic agents that are not metabolized by the liver
cation programs in adults and children and demon- are available, such as levetiracetam, the cost-effectiveness
strated some evidence of benefits; future research is of such therapies has not been studied in LMICs.
needed to examine the cost-effectiveness of such pro- Unfortunately, in LMICs, the availability and afford-
grams in LMICs (Bradley and Lindsay 2008; Lindsay ability of standard medications are poor and constitute
and Bradley 2010). barriers to treatment. One study found that the average
Neurological Disorders 91
Regrettably, those living in LMICs, where the bur- can be trained to provide basic treatment. Patient-
den of epilepsy is extensive, are the most affected related potential mechanisms for the treatment gap
by the epilepsy treatment gap (Jette and Trevathan include cultural beliefs, stigma, fear of side effects,
2014). The treatment gap is more than 75 percent in the hassle factor, and cost of treatment (Cameron and
low-income countries, more than 50 percent in many others 2012; Kale 2002; Mbuba and others 2008). All
LMICs and upper-middle-income countries, and less these reasons for the epilepsy treatment gap should
than 10 percent in most HICs (figure 5.1) (Meyer and be considered as potential targets for evaluation and
others 2010). action.
Proposed mechanisms for the epilepsy treatment One study examined the availability, price, and afford-
gap can be divided into two broad categories: health ability of anti-epileptic drugs in 46 countries (Cameron
care system and patient-related reasons (Cameron and others 2012). The study found that not only is
and others 2012; Kale 2002; Mbuba and others 2008). the availability of these medications lower in LMICs,
Health care system issues include lack of availability but their costs are highest where the treatment gap is
of anti-epileptic drugs, missed or delayed diagnosis, the greatest (Cameron and others 2012). This study
wrong treatment prescribed, treatment not offered supports the view that availability and affordability of
to patients, and lack of resources and personnel anti-epileptic drugs are likely major drivers in resource-
(Cameron and others 2012; Kale 2002; Mbuba and poor countries. Box 5.3 provides a summary of the
others 2008). Epilepsy diagnosis is predominantly potential targets for evaluation and action to improve
based on clinical history, and primary care physicians the epilepsy treatment gap.
Figure 5.1 Epilepsy Treatment Gap and Standard Errors Calculated from Lifetime Prevalence Estimates
100
80
Treatment gap (%)
60
40
20
0
a
sh
an
ia
bia
ica
ka
ina
on
la
ds
ain
e
ly
te
i
nc
or
liv
Ind
ma
Ita
de
an
lan
t
ro
ma
lom
Ch
Sp
ta
kis
ap
Fra
Bo
me
la
iL
te
dS
er
ng
Pa
Ja
ng
Co
a
Sr
th
Ca
Gu
Si
ite
Ba
Ne
Un
Neurological Disorders 93
and executive function that are at the late stages accom- of those were in LMICs. This figure will nearly double
panied by behavioral disturbances. Although age is the to 76 million in 2030 and to 145 million by 2050. The
most significant risk factor, dementia is not a normal majority (71 percent) of new cases will occur in LMICs
part of aging (Ganguli and others 2000; Kukull and oth- (figure 5.2) (Prince and others 2015; WHO 2015). The
ers 2002; Launer and others 1999). The clinical onset of steepest projected increases in numbers of people with
dementia is marked by the impact of cognitive decline dementia are expected in these settings because of rapid
in everyday activities, and diagnosis is often made by demographic changes. A new dementia case is diagnosed
physical and neurological examination with supporting every four seconds in the world, leading to 7.7 million
evidence from informant interviews. new cases per year; nearly 50 percent of new cases occur
Dementia is a syndrome that includes Alzheimers in Asia (WHO 2015).
disease; vascular dementia; frontotemporal dementia; In community-based samples, the prevalence of
Lewy body dementia; and reversible causes, for exam- dementia varies from 38 to 400 per 100,000 inhab-
ple, hypercalcemia, thyroid hormone abnormalities, itants, with an increasing incidence over 55 years.
vitamin B12 and folic acid deficiencies, HIV, sub- Frontotemporal dementia (9.7 percent), alcohol-related
dural hematoma, and normal pressure hydrocephalus. dementia (9.4 percent), traumatic brain injury (3.8
Alzheimers disease accounts for 5060 percent of all percent), and Huntingtons disease (3 percent) are more
late-life dementias, and vascular dementia accounts frequently present in early-onset dementia (EOD) com-
for up to 1520 percent. Although brain pathological pared with late-onset dementia (Picard and others
lesions differ across dementia subtypes, mixed forms 2011). Although dementia is more common in older age,
of dementia are common, and vascular brain damage some people develop symptoms at a younger age com-
often co-occurs. patible with EOD, a poorly understood and frequently
underdiagnosed condition.
Independent of the age at onset, most patients are
Epidemiology and Burden of Dementia cared for at home by close relatives. Need for one-on-
The most significant risk factor of dementia is increas- one care starts early, becomes increasingly intense, and
ing age; the incidence doubles with every five-year may change significantly throughout the natural history
increment after age 65 (WHO 2015). The graying of of the disease. Mood and behavioral changes, memory
societies in all global regions is expected to increase the impairment for recent events, and spatiotemporal disori-
number affected substantially. In 2015, approximately 47 entation, as well as problem-solving deficits that charac-
million people had some form of dementia; 63 percent terize the early stage, may expose people with dementia
Figure 5.2 Projected Growth in Number of People with Dementia in All Income Groups, 201050
120
Number of people with dementia (millions)
100
80
60
40
20
0
2000 2010 2020 2030 2040 2050 2060
High income Low and middle income
Figure 5.3 Distribution of the Total Societal Costs of Dementia Care, by World Bank Income Level
100
14.5
23.1
32.1 28.2
80
12.2
14.3
45.2
60
25.7
Percent
40
64.7
57.6
20 40.3 42.2
0
High-income Upper middle Lower middle Low-income
countries income countries income countries countries
Direct medical Direct social Informal care
Neurological Disorders 95
Interventions before diagnosis. Loss of body weight may increase mor-
Interventions need to address four key areas: bidity and mortality; yet, caregivers may be instructed
Timely diagnosis on simple practices and techniques to overcome prob-
Assessment and maintenance of physical health lems related to apathy and aversive feeding behaviors
Cognition, activity, and well-being; assessment and and may receive nutritional education to improve the
treatment of BPSD caloric and nutritional content of meals. Finally, moni-
Support for caregivers. toring and effective treatment of vascular risk factors
including high blood pressure, hypercholesterolemia,
smoking, obesity, and diabetesshould be encouraged
Detection and Diagnosis of Dementia to improve secondary prevention of cerebrovascular
The evidence does not support dementia screening events. Moreover, there is extensive and persuasive evi-
in the general population at present. Screening tools dence from mechanistic and well-designed prospective
in primary health services may be used for those who cohort studies that reducing the exposure to high blood
report initial concerns about their cognitive function. pressure and hypertension in mid-life, and to diabetes
Short versions of the Mini-Mental State Examination in mid- and late life, as well as the reduction in tobacco
(Folstein, Folstein, and McHugh 1973) take as little as use and increase in educational level of populations,
five minutes. However, unlike the Mini-Mental State can effectively reduce the dementia risk for populations
Examination, which has been validated in several set- (Prince and others 2014).
tings and languages, none of the short versions has been
validated in LMICs, and their use is not recommended Pharmacological Interventions
at present. Targets for pharmacological treatment include cognitive
Diagnosis requires a clinical and informant interview impairment; behavioral symptoms, such as agitation
and physical examination. Evidence from population- and aggression; and psychological symptoms, such as
based studies, for example, the 10/66 culture-fair diag- depression, anxiety, and psychosis. There is a large body
nostic algorithm (Prince and others 2003), suggests that of evidence for the efficacy of cholinesterase inhibitors
diagnosis can be achieved using highly structured inter- (ChEIs), such as donepezil, rivastigmine, and galantam-
views and examinations conducted by trained commu- ine, in the treatment of mild to moderate Alzheimers
nity health workers. Adaptations for use in clinical practice disease (Institute for Quality and Efficiency in Healthcare
are required, but the feasibility and cost-effectiveness of 2014). The use of each of these medications is associated
laboratory tests used in HICs to exclude treatable forms with modest and short-term comparable improvements
of dementia may limit their use in LMICs. Evidence from in cognitive function, global clinical state, and activities
HICs indicates that the good practice of disclosure of the of daily living. However, the evidence base for ChEIs in
dementia diagnosis allows better planning and may limit LMICs is limited. Moreover, the efficacy of this class of
distress; evidence from LMICs is lacking. drugs in severe dementia is unclear, although behavioral
Appropriate adaptation to local culture, language, symptom improvement was identified for galantamine
and beliefs should shape the design of programs and (Institute for Quality and Efficiency in Healthcare 2014).
activities planned and implemented, and involve stake- A fourth drug for the treatment of cognitive impair-
holders, policy makers, the media, and local health care ment, memantine, has a different mode of action and
services. Health and social services should be enhanced is well tolerated, but evidence for its efficacy is limited
to meet the projected increase in services. to people with moderate to severe dementia. ChEIs and
memantine are less efficacious in vascular dementia than
Physical and Care Needs Assessment other forms. Their efficacy in the treatment of behavioral
Information on care arrangements and resources disturbances is not established; manufacturer-sponsored
should be considered along with the evaluation of licensing trials and post hoc analyses indicate small
BPSD and the severity. A careful physical assessment is improvements.
very important to monitor hearing and visual impair- Use of haloperidol and atypical antipsychotic med-
ment, pain, constipation, urinary tract infections, and ications for the treatment of agitation and behavioral
bedsores that may explain exacerbation of psycholog- symptoms with BPSD indicate small treatment effects,
ical symptoms. Whether physical assessment improves most evident for aggression, although these must be
dementia prognosis, particularly the course of cognitive weighed against the associated mortality risk (Kales and
impairment, remains largely unknown. Nutritional sta- others 2012). Atypical antipsychotic drugs have been
tus should be carefully monitored during the course of widely prescribed for psychosis in dementia, but a meta-
the disease. Weight loss is common and may start even analysis of their efficacy indicated that only aripiprazole
Neurological Disorders 97
focused on simple curative interventions, and face high that the cost per quality-adjusted life year gained from
workloads. Given the frailty of many older people with early screening ranged from US$24,150 to US$35,661,
dementia, there is a need for outreach to assess and man- depending on the age group. The probability of screen-
age patients in their own homes. Dementia care should ing being cost-effective was highest in the group over age
be an essential component of any chronic disease care 75 years in a wide range of willingness to pay (WTP) (Yu
strategy. Training of nonspecialist health professionals and others 2015). The most cost-effective benefit of dis-
should focus on case-finding and conveying the diagno- ease modifying therapies has been seen in moderate to
sis to patients and caregivers together with information, severe dementia (Plosker and Lyseng-Williamson 2005).
needs assessment, and training and support. Training
can be service-based, as well as through changes to Pharmacotherapy
medical and nursing schools, public health, and rural Available pharmacoeconomic data from Europe and
health curricula. Medical and community care services the United States support the use of memantine as a
should be planned and coordinated to respond to the cost-effective treatment. Two cost-effectiveness analy-
increasing need for support as the disease progresses. ses of memantine in moderate-to-severe Alzheimers
disease have been conducted in Finland and the United
Community-Based Programs to Deliver Effective Kingdom; patient progression was simulated through
Treatments. Programs to support caregivers can be health states related to dependency, residential setting,
delivered individually or in groups by community and cognitive function (Francois and others 2004; Jones
health workers or experienced caregivers. Strain, possi- and others 2004). Memantine reduced total societal
bly associated with BPSD, should trigger more intensive costs by US$1,090 per patient per month, compared
interventions that include psychological assessment and with no pharmacological treatment, over 28 weeks in a
depression treatment for the caregiver, respite care, and resource utilization and cost analysis conducted along-
caregiver education and training. Such interventions side a pivotal trial in patients in the United States with
could be incorporated into horizontally constructed, moderate-to-severe Alzheimers disease (Wimo and oth-
community-based programs that address the generic ers 2003). Results were primarily driven by reductions
needs of frail, dependent, older people and their care- in total caregiver costs, which included the opportunity
givers, whether these needs arise from cognitive, mental, cost of time spent in caregiving tasks, and in direct
or physical disorders. Recent evidence has demonstrated nonmedical costs, which included the cost of care in a
the effectiveness of delivery of Internetbased caregiver nursing home or similar institution.
interventions (Czaja and Rubert 2002; Marziali and An analysis in Canada found that treatment with
Garcia 2011). rivastigmine yielded savings in the direct cost of caring
for patients with Alzheimers disease that exceed the cost
of the drug after two years of treatment (Hauber and
Dementia: Cost-Effectiveness of Interventions others 2000). In a 20-year Markov cohort model of dis-
The estimated worldwide societal cost of dementia ease modifying treatment in Alzheimers disease based
exceeded US$818 billion dollars in 2015 (Prince and on a Swedish population, the sensitivity analysis implied
others 2015). Direct costs include health service use no cost savings with disease modifying therapy, but most
and institutionalization; the indirect costs include those options indicated cost effectiveness verses the chosen
associated with inability to work and caregiver care. Both WTP (Skoldunger and others 2013). In another study
kinds of costs impose significant financial burdens on evaluating treatment with cholinesterase inhibitors or
individuals, families, and societies. Informal care costs memantine for those with mild to moderate vascular
are proportionally highest in LMICs, while the direct dementia, donepezil 10 mg orally daily was found to
costs for social care account for over half the costs in be the most cost-effective treatment (Wong and others
HICs (Prince and others 2015). Several studies, most in 2009).
HICs, have evaluated the cost effectiveness of interven-
tions in dementia. Particular challenges in such studies Other Therapies
are the heterogeneity in etiology of dementia and the In terms of nonpharmacologic therapies, cognitive stim-
capture of cost-effectiveness in patients with milder ulation therapy has been shown to be cost-effective for
forms of cognitive impairment. people with mild-to moderate dementia when delivered
biweekly over 7 weeks though was found to have modest
Screening effects when continued for longer when added to admin-
A study in the Republic of Korea, where there is a nation- istration of acetylcholinesterase inhibitors (DAmico and
wide early detection program for dementia, showed others 2015). An exercise intervention was found to have
Neurological Disorders 99
Neurological Societies 2011). The cause is chronic exces- Self-Management
sive use of medications taken initially to treat episodic Stress is a common predisposing factor for migraine.
headache (Diener and Limmroth 2004). The overuse Improving the ability to cope is an alternative treat-
of all such medications is associated with this problem, ment approach, but the role of psychological therapies
although the mechanism through which it develops in migraine management is unclear. Most research has
undoubtedly varies among drug classes (Steiner and focused on high-end intensive treatment of individual
others 2007). cases of disabling and refractory headache, which has
limited relevance to public health. Yet there is potential
for low-cost delivery of group behavioral training, and
Epidemiology and Burden of Disease even some very limited evidence of benefit (Mrelle and
others 2008). This approach could be further explored
Estimating the global burden of headache disorders is
in LMICs.
a challenging task, given data paucity for many LMICs,
Obesity is a risk factor for migraine, especially for
variations in methodologies in epidemiological stud-
frequent migraine (Evans and others 2012). Regular
ies, and variation of cultural attitudes related to the
exercise and keeping fit can be beneficial. A study among
reporting of complaints. Much of the worlds popula-
obese adolescents with migraine found a significant
tion lives in countries where headache prevalence and
improvement in headache in those who participated in a
burden are incompletely known (Stovner and others
12-month weight-loss program (Evans and others 2012).
2007). Regardless, estimations have been done and
show that the global one-year prevalence of migraine
Pharmacological Interventions
constitutes 14.7 percent and TTH 20.8 percent of
Guidelines recommend a stepped-care approach com-
adults ages 1865 (Murray and others 2012). The
mencing with acute treatment using simple anal-
prevalence of all types of headache occurring on 15
gesics (aspirin or one of several other nonsteroidal
or more days per month (including chronic migraine,
anti-inflammatory drugs) (Steiner and others 2007).
chronic TTH, and MOH) is 3 percent (Stovner and
Good evidence demonstrates the efficacy and tolerability
others 2007). Although the prevalence of migraine is
of aspirin (Kirthi, Derry and Moore 2013), ibupro-
markedly lower in Asia (Stovner and others 2007) and
fen (Rabbie, Derry and Moore 2013), and diclofenac
was thought to be so in Africa, a study in Zambia has
potassium (Derry, Rabbie, and Moore 2013). The most
indicated a high one-year prevalence (22.9 percent),
desirable outcome of acute treatment is complete relief
coupled with very high prevalences of headache on
from pain within two hours, without recurrence or need
15 or more days a month (11.5 percent) and proba-
for further medication and without adverse events. This
ble MOH (7.1 percent), with considerable economic
outcome is not commonly experienced with simple anal-
impact (Mbewe and others 2015).
gesics alone.
The more easily achievable outcome referred to
as sustained headache relief (SHR) is defined as
Interventions reduction of pain to no worse than mild within two
Worldwide, at least 50 percent of headaches are self- hours of treatment, also without recurrence or need
treated, even in high-income countries (HICs) (WHO for further medication. Mild pain is assumed not
2011). Professional health care, when needed, should to be associated with disability, and SHR implies
be provided in primary care settings for the majority full functional recovery when functional impairment
of cases (WHO 2011), and guidelines for the man- was present initially. Aspirin alone provides SHR in
agement of headache disorders in these settings are an estimated 39 percent of users (Kirthi, Derry and
available (Steiner and others 2007). History and exami- Moore 2013); this is a modest effect in the sense that it
nation should take due note of warning features that leaves 61 percent without this benefit but at the same
might suggest an underlying condition (Steiner and time is among the most cost-efficient interventions to
others 2007). improve public health (Linde, Steiner, and Chisholm
Many instruments, including the HALT ques- 2015). Aspirin has the advantages of being universally
tionnaire, are available to assess the burden of head- available and on the WHO essential medicines list
ache symptoms on individual patients. (Steiner and (WHO 2013). Ibuprofen provides SHR in a somewhat
Martelletti 2007). Realistic goals of management include higher estimated proportion of users (45 percent)
understanding that primary headaches cannot be cured (Rabbie, Derry, and Moore 2013), at variable but
but can be managed effectively. We focus our further not always higher cost. Diclofenac is considerably
treatment discussions on migraine. more costly, without significantly greater efficacy
Corresponding author: Louisa Degenhardt, Ph.D., National Drug and Alcohol Research Centre, UNSW Australia, l.degenhardt@unsw.edu.au.
109
drug dependence if three or more indicators of depen- Family factors: poor quality of parent-child interac-
dence are present for at least one month within the past tion and relationships, parental conflict, and parental
year (WHO 1993). and sibling drug use
The Diagnostic and Statistical Manual of Mental Individual factors: male gender; having an external-
Disorders, 4th edition (DSM-4) used a similar classifica- izing disorder, such as attention-deficit hyperactivity
tion for substance abuse and substance dependence (APA disorder or conduct disorders in early childhood;
2000). However, the fifth edition (DSM-5) defines a sensation- and novelty-seeking personality traits; and
substance use disorder if two of 11 criteria grouped under low education levels
impaired control, social impairment, risky use, and Peer group factors: association with antisocial or
pharmacological dependence are present; it categorizes drug-dependent peers, which is one of the strongest
the severity along a continuum of mild, moderate, and risk factors for illicit drug dependence in adolescence
severe disorders, based on the number of criteria present and which operates independently of social, contex-
(APA 2013). tual, family, and individual factors.
Table 6.1 Summary of Population Platforms and Recommended Interventions for Illicit Drug Dependence
Universal prevention
and health promotion Evidence level CEA available? Notes
Legislation and regulation
Precursor chemical control May be effective No Some impact, short-term; some consequences difficult to predict
High-level law May be effective No Difficult to know if or when effect will occur; may be short-lived
enforcement
Street-level law Inconclusive No May have short-term, localized effect but leads to compensatory
enforcement increases elsewhere
Prescription monitoring May be effective No Poorly studied to date; may have some impact, although misuse of other
programs medications may occur
Information and awareness
Mass media campaigns Inconclusive No Limited research with inconsistent results, with some showing negative
and others positive impacts on drug attitudes and use
Intersector collaboration
Imprisonment Inconclusive No No evidence suggesting drug use is reduced on release, although
decreased use during imprisonment
Drug testing for offenders May be effective No Encouraging observational evidence from U.S. states where this has
been introduced
Court-mandated treatment Inconclusive No Includes mandated treatment and drug courts
Note: CEA = cost-effectiveness analysis.
Early Intervention with At-Risk Youth. There is lim- Overdose Prevention Education. Polydrug use
ited, low-quality, and inconsistent evidence about the increases the chances of fatal overdose, particularly the
effectiveness of school-based drug testing among high concurrent use of opioids and other drugs that depress
school students (Shek 2010). The evidence on the impact the central nervous system, like benzodiazepine and
of psychosocial interventions for young people using alcohol (Warner-Smith and others 2001). Educating
substances or at risk of doing so is limited and inconsis- people who use opioids, particularly by injection, about
tent (Strang and others 2012). these dangers and the risks of injecting alone or on the
streets, where assistance in case of overdose is limited, Although models differ, all SIFs provide sterile injecting
might reduce the risk of overdose (McGregor and others equipment and a hygienic environment where pre-
2001).5 However, the effectiveness of these strategies has obtained drugs can be injected.
not been rigorously evaluated. Observational evaluations in Vancouver and Sydney
have suggested that SIFs attract risky injectors, facilitate
Naloxone and Other Emergency Responses. Another safe-injection education, reduce syringe sharing, and
strategy is to improve bystander responses to opioid increase referral and entry into withdrawal management
overdoses by encouraging drug users who witness over- and drug treatment. Although reviews suggest that
doses to seek medical assistance and use simple but drug use does not change among clients or among drug
effective resuscitation techniques until help arrives injectors in the areas where SIFs are located (Kerr and
(Wagner and others 2010). This approach includes the others 2007; MSIC Evaluation Committee 2003), the
distribution of naloxone to opioid injectors and their evidence of their impact on HIV transmission is uncer-
peers. Naloxone is a narcotic antagonist that rapidly tain (Kimber and others 2010). However, reducing the
reverses the effects of acute narcosis, including respira- risk among the most vulnerable injecting drug users may
tory depression, sedation, and hypotension.6 An increas- increase the effectiveness of other interventions.
ing number of jurisdictions have been implementing
such programs, although evaluations have largely been Primary Health Care
observational (Tobin and others 2009). Screening and Brief Intervention. Some evidence sug-
gests that a single brief intervention in a clinical setting
Supervised Injecting Facilities. SIFs are located in areas can reduce illicit drug use (Baker and others 2005;
where injecting drug users are concentrated, typically Humeniuk and others 2012), although a recent system-
in areas with large, open drug markets. The goal is to atic review concluded that further studies were needed
reduce drug overdose deaths and BBV infections among (Young and others 2014). Brief interventions from
injectors who inject in public places. SIFs have poten- prescribers, such as tailored written letters to patients
tial community impact but exist in a limited number or consultations, reduced heavy benzodiazepine use up
of locations, only 61 cities in eight countries (Hedrich, to six months after intervention (Mugunthan, McGuire,
Kerr, and Dubois-Arber 2010; Kerr and others 2007). and Glasziou 2011).
Specialist Health Care users seek most often. It provides users with a respite
Detoxification and Withdrawal. Detoxification cen- from use, an occasion to reconsider their drug use,
ters provide supervised withdrawal from a drug of and a potential prelude to abstinence-based treat-
dependence with the aim of minimizing the severity ment. Detoxification has minimal, if any, enduring
of withdrawal symptoms. Detoxification is not a impact on dependence on its own (Mattick and
treatment, but it is the intervention that dependent Hall 1996).
NOTES REFERENCES
World Bank Income Classifications as of July 2014 are as fol- Allsop, D., J. Copeland, N. Lintzeris, A. Dunlop, M. Montebello,
lows, based on estimates of gross national income (GNI) per and others. 2014. A Randomized Controlled Trial of
capita for 2013: Nabiximols (Sativex) as an Agonist Replacement Therapy
during Cannabis Withdrawal. JAMA Psychiatry 71 (3):
Low-income countries (LICs) = US$1,045 or less 28191.
Middle-income countries (MICs) are subdivided: Amato, L., S. Minozzi, P. P. Pani, R. Solimini, S. Vecchi, and oth-
a) Lower-middle-income = US$1,046 to US$4,125 ers. 2011. Dopamine Agonists for the Treatment of Cocaine
b) Upper-middle-income (UMICs) = US$4,126 to Dependence. Cochrane Database of Systematic Reviews
US$12,745 12: CD003352. doi:http://dx.doi.org/10.1002/14651858
High-income countries (HICs) = US$12,746 or more. .CD003352.pub3.
Corresponding author: Mara Elena Medina-Mora, Instituto Nacional de Psiquiatra Ramn de la Fuente (Ramn de la Fuente National Institute of Psychiatry),
Mxico; medinam@imp.edu.mx.
127
(Babor, Robaina, and Jernigan 2014). These activities behavior and inattention of drinkers while intoxicated,
challenge the public health sector and governments to resulting in acts of violence, driving while impaired,
respond with public health strategies to minimize the inconsistent family environments affecting normal child
adverse health and societal consequences of the expand- development, and workplace absenteeism (WHO 2014a).
ing global markets in alcoholic beverages (Babor and An additional and increasingly significant conse-
others 2010). quence of maternal drinking during pregnancy is fetal
The high level of globalization has significant effects alcohol syndrome (FAS), a pattern of retarded growth
on markets. Transnational companies own the formulas and development, both neuropsychological and physical,
and grant licenses to local subsidiaries. Most product with typical facial dysmorphic features, that is found is
development targets external markets, and adver- some children exposed to alcohol in utero. A spectrum
tising is usually produced externally. Transnational of physical and neurodevelopmental abnormalities,
companies, supported by these economic advantages, which includes FAS, that is attributed to the effects of
are dynamic promoters of modifications in local alcohol on the fetus, is termed fetal alcohol syndrome
drinking practices, including the types and quantity of disorders (FASD). The level of maternal alcohol con-
beverages consumed (Room, Jernigan, Carlini, Gmel, sumption required to produce FASD, which has yet to be
and others 2013; Room, Jernigan, Carlini, Gureje, and established, is influenced by genetic and other maternal
others 2002). and fetal characteristics (Gray, Mukherjee, and Rutter
2009; May and Gossage 2011).
Alcohols impact on disease and injury is associated
Public Health Considerations with two dimensions: the overall volume consumed and
The substantial health and societal costs of alcohol con- the drinking patterns of how the volume is distributed
sumption outweigh its economic benefits and contribute by drinking. Heavy drinking episodes have particularly
to the view of public health professionals that alcohol damaging effects. The consequences associated with a
cannot be considered an ordinary commodity (Babor high volume of drinking or recurrent heavy drinking
and others 2010). Special policies are needed to curb the occur through three mechanisms: toxic and other effects
consequences of harmful use, especially in LMICs where on organs and tissues, behavior during intoxication, and
the burden is higher. alcohol dependence and other alcohol-induced mental
This public health perspective has received little disorders (APA 2013; WHO 1992, 2013a).
attention in international negotiations concerning trade
agreements and in resolutions of disputes under those
The Burden
agreements (Room, Jernigan, Carlini, Gmel, and others
2013; Room, Jernigan, Carlini, Gureje, and others 2002). Patterns of Alcohol Use and Trends
This lack of attention reduces the ability of LMICs to One of the most commonly used indicators of overall
ensure the internal regulation of markets (Grieshaber- alcohol consumption and comparison by location is
Otto, Schacter, and Sinclair 2006; Zhang and Monteiro per capita consumption. Although it is the best estimate
2013). Governments in LMICs are deterred or forced to available, it contains a substantial element of uncer-
abandon alcohol controls as a result of trade disputes; tainty, which increases where there are large proportions
for example, Thailand faces opposition from some of unrecorded production, which is more common in
World Trade Organization members to its proposed LMICs.
graphic warning labels on containers of alcohol sold Globally, per capita alcohol consumption in 2012 was
within its borders (OBrien 2013). an estimated 6.2 liters of pure alcohol by persons ages
15 years and older (WHO 2014a); 24.8 percent is con-
sumed as unrecorded alcohol (Lachenmeier, Sarash, and
ALCOHOL-RELATED DISORDERS Rehm 2009; WHO 2014a).
In general, HICs have the highest levels of per capita
Patterns of Alcohol Use, Alcohol Use Disorders, and consumption and often the highest prevalence of heavy
Fetal Alcohol Spectrum Disorders episodic drinking. The prevalence of heavy episodic
Alcohol is a major contributor to mortality, morbidity, drinking among adolescents ages 1519 years mirrors
and injuries. It is a causal factor in more than 60 diseases, that of the adult population, with the highest rates in the
including liver cirrhosis and cardiovascular disease, and World Health Organization (WHO) regions of Europe,
it is involved in the etiology of more than 200 other con- the Americas, and the Western Pacific (WHO 2014a).
ditions, such as neuropsychiatric conditions and diabetes Altogether, women drink less than men and have
mellitus. It also affects other people through the risky a lower prevalence of alcohol use disorders (AUDs);
(Diaz and others 2011; Nagel and others 2009). The Table 7.5 Screening and Brief Interventions in
mhGAP Intervention Guide (WHO 2011b) identifies Low- and Middle-Income Countries
three levels of interventions with individual problematic
Site Main results
drinkers:
Thailand (++) Motivational interviewing was tested in low-
Screening and brief interventions by trained primary resource settings (Noknoy and others 2010;
health care professionals Segatto and others 2011).
Early identification and treatment of AUDs in pri- China (+++) Counseling was supported with health promotion
mary health care booklets (Tsai and others 2009).
Referral and supervisory support by specialists. Brazil (+++) Brief advice was provided on cognitive
behavioral interventions (Marques and Formigoni
The WHO mhGAP action plan promotes scaling 2001).
up services for mental, neurological, and substance use Mexico (++) Motivation therapy showed a greater reduction
disorders, with more cases treated at the first level of of alcohol use compared with cognitive
care (WHO 2008). The program is based on a review behavioral therapy (Diaz and others 2011).
and evaluation of the strength of the evidence to submit Indigenous Brief psycho-educational intervention that
recommendations for action. Psychosocial support was communities/ included motivational care planning, problem
found to be more effective than no treatment, while Australia (++) solving, and impulse management showed
motivational interviewing and motivation enhancement significant benefits, compared with a control
were possibly more effective than standard psychoso- group (Nagel and others 2009).
cial treatment involving families and friends (mainly Kenya (++) Intervention was among people with HIV/AIDS
spouses), or no treatment, or individual counseling. (Papas and others 2010).
Evidence in LMICs is widespread and consistent, Note: ++ = good; +++ = very good; HIV/AIDS = human immunodeficiency virus and
showing positive results (table 7.5). acquired immune deficiency syndrome.
Medical and Social Detoxification, Treatment, treatment of comorbidity and possible referral to self-
Follow-Up, and Referral help groups (WHO 2008).
The recent evidence for LMICs is consistent with what
had previously been reported (Patel and others 2007) Self-Help and Support Groups
(table 7.6). Mutual help and self-help organizations for those inter-
The mhGAP recommends referral from first-level ested in reducing or ceasing drinking have been an impor-
care and supervisory support by specialists for patients tant part of the social response to alcohol in many societies.
with established alcohol dependence. The recommended Given that many religions forbid or discourage drinking,
actions include the planning of cessation of alcohol adherence to a religious congregation or group often car-
consumption and detoxification; if necessary, the treat- ries with it an expectation of mutual help to stop drinking.
ment of withdrawal symptoms with diazepam; the use In many social groups in Latin America and the Caribbean,
of medications to prevent relapses, such as naltrexone, joining a Protestant sect has often been a way out of socio-
acamprosate, or disulfuram; and the assessment and cultural expectations of heavy drinking, particularly for
COST-EFFECTIVENESS OF INTERVENTIONS
men (Butler 2006; Eber 2001). Many mutual help groups The addition of a cost component or economic dimen-
that are not affiliated with particular religions or that are sion to health impact assessment introduces the oppor-
entirely secular have formed in different countries (Room, tunity to identify alcohol prevention and control
Jernigan, Carlini, Gmel, and others 2013; Room, Jernigan, strategies that have better or worse value for money. For
Carlini, Gureje, and others 2002). The most well-known example, devoting scarce resources to interventions that
and widespread is Alcoholics Anonymous, which has do not discernibly reduce ill-health caused by the con-
proved adaptable to many cultural settings (Eisenbach- sumption of alcoholas is the case for information and
Stangl and Rosenqvist 1998; Mkel 1991). educationis a clear case of investing in interventions
Affiliation with Alcoholics Anonymous and similar that are not cost-effective. At the other end of the spec-
groups is not considered a form of formal treatment trum, in contrast, imposition and enforcement of taxa-
although some groups have affiliations with treatment tion policies offers an example of a highly cost-effective
institutionsand incorporating mutual help groups public health intervention that costs relatively little to
into a treatment system is likely to undercut their implement but reaps substantial health returns.
effectiveness. The principles of voluntary mutual help The available body of economic evidence to inform
organizations often do not allow random-assignment decisions around these alcohol control measures in
clinical trials to test their effectiveness; consequently, LMICs remains modest and is based on a modeling
not much research has been conducted on the impact approach that relies on data from higher-income settings
of these groups (Ferri, Amato, and Davoli 2006; Terra for some of its inputs. Rehm and others (2006) reported
and others 2007). However, survey results support on the comparative cost-effectiveness of a group of
the important role of these groups; 71 percent of the interventionsenactment of legislation on drinking
countries included in WHOs Atlas on Substance Use and driving, random breath testing, taxation of alcoholic
(2010) reported the presence of Alcoholics Anonymous. beverages, reduced hours of sale, and advertising bans
in East Asia and Pacific, Latin America and the Caribbean,
Latin America and the Caribbean. In Mexico accord- South Asia, and Sub-Saharan Africa. Increased taxation
ing to a National Household Survey (Medina-Mora and was the most cost-effective strategy, although it may have
others 2012), 44 percent of persons in treatment for a regressive impact on the incidence of alcohol consump-
alcohol problems reported being affiliated with self-help tion if accompanied by a rise in an already high level of
groups, while only 35 percent received professional treat- unrecorded consumption. The authors found reductions
ment. A study conducted in a nonprobabilistic sample of from 2 to 4 percent in the incidence of high-risk alco-
192 members of Alcoholics Anonymous found that the hol use, depending on regional drinking patterns. The
level of affiliation or involvement with the organization strategy of reducing the hours of sale produced a modest
was negatively related to relapse; with more involve- reduction of 1.53.0 percent in the incidence of high-risk
ment, mean participation time was higher, and activi- drinking, together with a 1.54.0 percent reduction in
ties related to service were more frequent. Most of the alcohol-related traffic fatalities.
Corresponding author: James G. Scott, The University of Queensland Centre for Clinical Research, Queensland; Royal Brisbane and Womens Hospital,
Queensland, Australia; james.scott@health.qld.gov.au.
145
of symptoms. The effectiveness of selected interven- suffer from anxiety disorders. Anxiety disorders and
tions is evidence based; these interventions have the ADHD were more common in adolescents compared
potential to be delivered in low- and middle-income with children.
countries (LMICs). The chapter does not discuss Most children and adolescents with mental and devel-
childhood depression, because of the overlap in inter- opmental disorders were in South Asia, reflecting the high
ventions with adult depression. population in this region and the reduction in mortality
The chapter considers interventions in terms of deliv- of infants and young children (Murray and others 2012).
ery platforms rather than specific disorders. This choice The populations of LMICs tend to have higher propor-
is because of the very high comorbidity between child- tions of children and adolescents than those of high-
hood mental and developmental disorders (Bakare 2012; income countries (HICs). For example, 40 percent of the
Rutter 2011). In addition, risk factors for childhood population in the least developed countries is younger
disorders are nonspecific and pluripotent. For example, than age 15 years, compared with 17 percent in more
children who are maltreated are at higher risk of a wide developed regions (United Nations 2011). Furthermore,
range of mental and developmental disorders (Benjet, population aging is occurring more slowly in LMICs, with
Borges, and Medina-Mora 2010). some low-income countries predicted to have the young-
est populations by 2050, given their high fertility rates
(United Nations 2011). These trends mean that childhood
NATURE OF CHILDHOOD MENTAL AND mental and developmental disorders will increase in sig-
DEVELOPMENTAL DISORDERS nificance in LMICs. Furthermore, the continuing reduc-
tions in infant mortality caused by infectious diseases
Childhood mental and developmental disorders are an mean more children will reach adolescence where the
emerging challenge to health care systems globally. Two prevalence of mental disorders increases and the onset of
contributing factors are the increases in the proportion adult mental disorders occurs. This will challenge already
of children and adolescents in the populations of LMICs, limited mental health services in these countries.
which is a result of reduced mortality of children under
age five years (Murray and others 2012), and the fact
that the onset of many adult mental and developmental Risk Factors for Childhood Mental and
disorders occurs in childhood and adolescence (Kessler Developmental Disorders
and others 2007). The risk factors for childhood mental and develop-
mental disorders shown in table 8.2 can be divided into
lifelong and age-specific risk factors (Kieling and others
Global Epidemiology and the Burden of Childhood 2011). The health of children is highly dependent on
Mental and Developmental Disorders the health and well-being of their caregivers; the envi-
Ascertaining the global epidemiology of mental disorders ronments in which the children live (including home
is a difficult task, given the significant paucity of data for and school); and, as they transition into adolescence,
many geographical regions, as well as the cultural varia- the influence of their peers. The relative importance of
tions in presentation and measurement. These issues are a particular risk factor should be considered in terms of
exacerbated when investigating mental disorders in chil- prevalence, strength of the association with an adverse
dren, particularly in LMICs where other health concerns, outcome, and potential to reduce exposure to that risk
such as infectious diseases, are priorities. The issue of data factor (Scott and others 2014). Using these criteria,
paucity was highlighted in the Global Burden of Disease efforts to address maternal mental health problems and
Study 2010 (GBD 2010) (Whiteford and others 2013). improve parenting skills have the greatest potential to
Epidemiologically, childhood mental disorders were reduce mental and developmental disorders in children.
relatively consistent across the 21 world regions defined
by GBD 2010. However, these prevalence estimates were
based on sparse data; some regions, such as Sub-Saharan Consequences of Childhood Mental and
Africa, have no data whatsoever for some disorders or Developmental Disorders
no data for specific disorders in childhood. Although The consequences of these disorders include the impact
regional differences may exist, the lack of data makes during childhood and the persistence of mental ill health
them difficult to ascertain. The 12-month global prev- into adult life. In childhood, the impact is broad, encom-
alence of childhood mental disorders in 2010 is shown passing the individual suffering of children, as well as the
in table 8.1. ADHD, conduct disorder, and autism were negative effects on their families and peers. This impact
more prevalent in males; females were more likely to may include aggression toward other children and
Illness or loss of Consanguinity In utero exposure to pes- Inadequate Inadequate Substance misuse
caregivers ticides and other toxins parenting parenting
Exposure to trauma, Birth hypoxia and other Developmental Inadequacies Early sexual activity
adversity, violence, obstetric complications and behavioral of schools or
or conflict problems teachers
Genetic background Maternal difficulties Developmental Risk-taking behaviors
adapting to pregnancy or and behavioral
arrival of newborn problems
Toxins Perinatal maternal Risk-taking School problems
mortality behaviors
Immigrant status
Source: Kieling and others 2011.
distraction of peers from learning. Children with mental in children. The World Health Organization (WHO)
and developmental disorders are at higher risk of mental has published a modular package for governments, pol-
and physical health problems in adulthood, as well as icy makers, and service planners, Child and Adolescent
increased likelihood of unemployment, contact with law Mental Health Policies and Plans, to address this need
enforcement agencies, and need for disability support. (WHO 2005b). The guidelines recommend attention to
a broad range of areas pertaining to childhood mental
and developmental disorders (box 8.1). The provision of
Trends in Childhood Mental and Developmental health services for children in isolation will not prevent
Disorders mental and developmental disorders or have significant
GBD 2010 estimated burden across five time points benefits for children with these disorders. Instead, an
(1990, 1995, 2000, 2005, and 2010) and found that the ecological approach that addresses problems in the sys-
prevalence and burden of childhood mental disorders tems around children (parents, family, and school) in
remained consistent between 1990 and 2010 (Erskine combination with targeted interventions for children is
and others 2015). Although the rates may not have necessary to make a meaningful difference (Kieling and
changed, population growth and aging have impacts on others 2011).
the burden of disease attributable to mental disorders
in childhood. As the population of children increases Child Protection Legislation
globally, the burden of disease attributable to mental Child maltreatment is a well-established risk factor
disorders in children will increase. for mental and developmental disorders in children
(Benjet, Borges, and Medina-Mora 2010). Child mal-
treatment is defined as any form of physical or emo-
INTERVENTIONS FOR CHILDHOOD MENTAL tional ill-treatment, sexual abuse, neglect or negligent
AND DEVELOPMENTAL DISORDERS treatment, or commercial or other exploitation that
results in actual or potential harm to a childs health,
Population Platform Interventions survival, development, or dignity in the context of a
Child and Adolescent Mental Health Policies and Plans relationship of responsibility, trust, or power (Krug and
Few countries have developed national policies and others 2002). Legislation to address child maltreatment
plans to address mental and developmental disorders requires the support of well-integrated systems that
Box 8.2
1. Compared with other children, did the child have 8. Does the child speak at all (can he/she make
any serious delay in sitting, standing, or walking? himself/herself understood in words, can he/she
2. Compared with other children, does the child say any recognizable words)?
have difficulty seeing, either in the daytime or 9. For children ages three to nine years, ask: Is the
night? childs speech in any way different from normal
3. Does the child appear to have difficulty hearing? (not clear enough to be understood by people
4. When you tell the child to do something, does other than his/her immediate family)?
he/she seem to understand what you are saying? For children age two years, ask: Can he/she name
5. Does the child have difficulty in walking or mov- at least one object (for example, an animal, a toy,
ing his/her arms, or does he/she have weakness a cup, a spoon)?
and/or stiffness in the arms or legs? 10. Compared with other children of his/her age,
6. Does the child sometimes have fits, become rigid, does the child appear in any way mentally
or lose consciousness? backward, dull, or slow?
7. Does the child learn to do things like other chil-
dren his/her age? Source: Zaman and others 1990.
Future implementation of programs to address childhood interventions to protect children is urgently required
mental and developmental disorders in LMICs should be in LMICs. Reducing bullying in schools may prevent
evaluated. Other evidence-based key recommendations mental disorders in childhood and later in life; however,
for interventions are summarized in table 8.3. there are no data to show effective programs in LMICs.
As the evidence presented in this chapter indicates, The widespread implementation and evaluation of
key interventions that have the potential to reduce parenting skills training, including psychosocial stim-
mental and developmental disorders in childhood are ulation and maternal mental health interventions, is
parenting skills training that includes psychosocial stim- recommended in all countries to achieve a meaningful
ulation, teacher training with The Incredible Years reduction in the global prevalence and burden of child-
program, and maternal mental health interventions. The hood mental and developmental disorders.
evidence suggests that these can be feasibly delivered in
LMICs, and that they have a strong efficacy in HICs.
CBT for anxiety disorders has a strong evidence base in NOTES
HICs, but much more work is needed to demonstrate
World Bank Income Classifications as of July 2014 are as
the feasible delivery of this intervention in LMICs. follows, based on estimates of gross national income (GNI)
Pharmacotherapy requires specialist care and assessment per capita for 2014:
that limits use in LMICs.
The screening of children for developmental dis- Low-income countries (LICs) = US$1,045 or less
orders is possible in LMICs; however, the evidence Middle-income countries (MICs) are subdivided:
for intervening once autism or intellectual disabil- a) lower-middle-income = US$1,046 to US$4,125
ity has been identified is limited. Similarly, child protec- b) upper-middle-income = US$4,126 to US$12,745
tion and reduction of bullying in schools are important High-income countries (HICs) = US$12,746 or more.
preventive strategies for childhood mental disorders.
The systems required for child protection are com- 1. This is a standard cutoff for cost-effectiveness used in the
plex and require collaboration across sectors and sig- United Kingdom, comparable to the US$50,000 threshold
nificant government investment. Further research on commonly used.
Corresponding author: Lakshmi Vijayakumar, Sneha, Voluntary Health Services, Chennai, India; Center for Youth Mental Health, University of Melbourne,
Australia; lakshmi@vijayakumars.com.
163
Suicide Mortality The mean age of suicide in HICs is higher than in
WHO reports that 804,000 suicide deaths occurred LMICs, 50.4 versus 42.0 years, respectively, a difference
globally in 2012. The demographic characteristics and largely accounted for by the difference in the median
regional distribution of suicides, and the changes in sui- ages of the populations. Despite the higher rates of
cide rates between 2000 and 2012, are shown in table 9.1 suicide in the elderly, for males and females in LMICs,
and figure 9.1. Substantial differences exist in the rates over 63 percent of all suicides occur in individuals ages
and characteristics of suicide between LMICs and 1549 years.
high-income countries (HICs) as well as among LMICs
in the six WHO regions. To facilitate the comparison of Relative Importance of Suicide as a Cause of Death
rates between regions and countries, the rates reported Suicide accounted for 1.7 percent of all deaths in HICs
here per 100,000 population are all standardized to the and 1.4 percent in LMICs in 2012, making suicide the
age distribution of the global population in 2012. 11th most important cause of death in HICs and the
17th most important cause in LMICs. Among ages
1529 years in LMICs, suicide accounts for 7.9 percent
Overall Suicide Rates
of all deaths and is the third most important cause of
The 2012 age-adjusted suicide rate in HICs (12.7) was
death; among persons ages 3049 years, suicide accounts
slightly higher than that in LMICs (11.2); over 75 percent
for 3.4 percent of all deaths and is the seventh most
of all global suicides occur in LMICs, given their larger
important cause of death. Another measure of the public
proportion of the global population. Among LMICs, the
heath importance of suicide is that it is the most impor-
region-specific suicide rate in the six regions varies over
tant type of intentional violent death (which includes
a threefold range (from 6.1 to 17.7); the country-specific
suicides, murders, and war-related deaths): in LMICs,
rate varies over a 100-fold range, from 0.44 in the Syrian
suicide accounts for 44 percent of all violent deaths in
Arab Republic to 44.2 in Guyana.
males and 70 percent of all violent deaths in females.
Suicide
165
Figure 9.1 Percent Change in Age-Adjusted Suicide Rate in Different There are two sources of data for suicide attempts:
Regions of the World from 2000 to 2012 Based on WHO Global Health self-reports from community surveys and reports from
Estimates emergency departments of general hospitals (where
most suicide attempts that receive medical care are
5
0 treated). For the majority of the survey data and emer-
5 gency department data about suicide attempts available
200012 percent change
10
15 from LMICs, the lack of standardized methods for iden-
in suicide rate
ca
as
an
ia
c
al
rie
cifi
op
As
tri
fri
ric
ne
ob
nt
ur
Pa
un
me
rra
Gl
t
ou
nE
s
n
co
rn
ite
i
A
ec
-E
si
Cs
te
me
he
th
ed
IC
es
I
LM
ou
nt
M
co
LM
W
inc
nS
si
-in
rn
he
le-
ste
gh
si
nt
M
idd
IC
Hi
Ea
si
L
LM
dm
he
IC
si
w-
males and females in HICs, 0.03 for males and 0.06 for
IC
Lo
LM
Figure 9.2 Suicide Rates by Gender and Age for Selected Regions, Based on WHO Global Health Estimates, 2012
a. Global (197 countries) b. 56 High-income countries (HICs) c. 141 Low- and middle-income
countries (LMICs)
100 100 100
90 90 90
80 80 80
70
Suicide rate (per 100,000)
70 70
60 60 60
50 50 50
40 40 40
30 30 30
20 20 20
10 10 10
0 0 0
10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80
Age (years) Age (years) Age (years)
Male Female
d. 46 LMICs in WHO Africa region e. 26 LMICs in WHO Americas region f. 16 LMICs in WHO Eastern
Mediterranean region
100 100 100
90 90 90
80 80 80
Suicide rate (per 100,000)
70 70 70
60 60 60
50 50 50
40 40 40
30 30 30
20 20 20
10 10 10
0 0 0
10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80
Age (years) Age (years) Age (years)
g. 20 LMICs in WHO Europe region h. 11 LMICs in WHO South-East Asia region i. 21 LMICs in WHO Western Pacific region
100 100 100
90 90 90
80 80 80
Suicide rate (per 100,000)
70 70 70
60 60 60
50 50 50
40 40 40
30 30 30
20 20 20
10 10 10
0 0 0
10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80
Age (years) Age (years) Age (years)
Male Female
Note: The countries included in each region are listed in annex 2 of WHO 2014a. HICs = high-income countries; LMICs = low- and middle-income countries; WHO = World Health Organization.
Suicide 167
persons ages 18 years or older (15.4), globally there are geographic regions of large LMICs, like China (Phillips
about 20 self-reported suicide attempts for each death and others 2002) and India (Patel and others 2012).
by suicide in persons ages 18 or older; this amounts to Some of these differences can be attributed to limita-
15 million suicide attempts worldwide each year. tions or biases in the reporting of suicides, but most of
The limited nationally representative data avail- the reported differences reflect real differences in suicide
able from HICs suggest that the case-fatality of medi- rates. Given the magnitude of these differences, policy
cally treated suicide attempts is greater for males than makers and planners should be cautious when transpos-
females for all methods and increases with age, but it ing a prevention strategy from HICs to LMICs, from one
is unknown whether this pattern is also true in LMICs nation to another, or even from one region to another
(WHO 2014a). in a country. Development and ongoing quality control
of registry systems that monitor the changing rates,
demographic profile, and methods of fatal and nonfatal
Methods of Suicide and Suicide Attempts suicidal behavior in the country or region is essential for
Collecting information about the methods used in fatal planning and implementing interventions.
and nonfatal suicidal behavior, the demographic pro-
file of individuals who use different methods, and the
case-fatality of the different methods is an impor- RISKS AND PROTECTIVE FACTORS IN LMICs
tant component of a comprehensive suicide prevention
plan. Unfortunately, only a minority of countries pro- The identification of risk and protective factors is a key
vides method-specific data when reporting mortality component of any prevention strategy and guides the
data to WHO, although International Classification of development of appropriate interventions. Risk fac-
Diseases-10 (ICD-10) codes exist for all methods of sui- tors can be present in different categoriesindividual,
cide. Of the 140 LMICs, only 36 provided data on suicide relationships, community, society, and health system
methods at any time after 2005, and these countries only that can have multiple points of overlap (WHO 2014a).
accounted for 11 percent of all suicides in LMICs in 2012. There are several theoretical ways to conceptualize how
In the absence of national-level data from WHO, it risk factors influence suicidal behaviors. One approach
is necessary to consider reviews of subnational data. A to conceptualize risk factors is to view their influence as
systematic review (Gunnell, Eddleston, and others 2007) being proximal versus distal. Proximal risk factors include
of the global literature from 1990 to 2007 estimated psychiatric disorder, physical disorder, psychosocial life
that about 30 percent of all suicides worldwide are crisis, availability of means, and exposure to models of
caused by pesticide self-poisoning, most of which occur suicide. Distal risk factors include genetic susceptibility/
in LMICs, particularly in rural areas where residents loading, personality characteristics such as impulsivity or
practice small-scale agriculture and have easy access to aggression, early traumatic events, and neurobiological
pesticides. Based on this result, pesticide ingestion is the disturbances such as serotonin dysfunction (Hawton and
most common method of suicide globally. However, it van Heeringen 2009).
is probable that the choice of method varies greatly by There are also different patterns of risk across the life-
region, gender, age, urban versus rural residence, and span. For example, risk factors for the elderly differ from
over time, so each nation must develop standardized those for adolescents and young adults. What is universal
methods for routinely obtaining this information to help is that the greater the number of risk factors present, the
inform country-specific and community-specific means greater is the likelihood of a range of suicidal behaviors
restriction strategies. For countries that already provide (Phillips and others 2002).
ICD-10 cause of death mortality data to WHO, this
could be accomplished relatively easily by mandating
that all reports of accidental deaths include the corre- Risk Factors
sponding X-code. The relative importance of certain risk factors differs by
country and region, such as age of onset of a psychiatric
disorder, religious orientation and practice, geographical
Role of Surveillance in Suicide Prevention in LMICs location, age ranges, and gender distribution. Even within
The available evidence suggests that substantial a region, national and intranational differences exist in
cross-national variation in the rates, demographic pro- the prevalence of risk factors; any listing of risk factors
file, and methods of suicide and attempted suicide is the may not apply to all LMICs, even in the same region.
rule rather than the exception. Other reports also indi- Risk factors are variable over time and may be
cate large differences in suicide rates between different influenced by the rapidity of change occurring within
Suicide 169
these factors are also associated with suicide. In Turkey, Exposure to Models
from 1990 to 2010, economic problems, relationship Risk of suicidal behavior can be influenced by exposure
problems, and educational failure were the most com- to similar behavior by other people.
mon reasons for suicide (Oner, Yenilmez, and Ozdamar A substantial body of evidence indicates that certain
2015). In Brazil, from 1980 to 2006, the most dominant types of media reporting and portrayal of suicidal
sociodemographic characteristics of those who died behavior can influence suicide and self-harm in the
by suicide were low educational level and single status general population (Pirkis and Blood 2010). Newspaper
(Lovisi and others 2009). Another study from Brazil reporting of suicides can be particularly influential if it
found that income inequality represents a community- is sensational, if it includes dramatic headlines and pic-
level risk factor for suicide rates (Machado, Rasella, and tures, if it reports methods of suicide in detail, and if the
Dos Santos 2015). subject is a celebrity (Stack 2003).
One of the most distressing features of suicide in
Urban versus Rural Locations LMICs is the frequent occurrence of suicide pacts and
Globally, suicide rates are higher in urban than in rural family suicides, which constitute an estimated 1 percent
areas, but these can vary across countries by age and of suicides. Family suicides are often a suicide-homicide,
gender. In LMICs, living in a rural area increases risk. In in which the adults murder their children prior to their
China, the suicide rates are three times higher in rural own suicide. These suicides are frequently driven by
areas than urban areas (Cao and others 2000; Phillips and debt, poverty, and other social issues rather than by
others 2002); in Sri Lanka, the rural suicide rate is twice depression or mental disorders (Gupta and Gambhir
that of urban areas (Jayasinghe and de Silva 2003); and in Singh 2008; Vijayakumar and Thilothammal 1993).
India, about 90 percent of the suicides occur in rural areas
(Gajalakshmi and Peto 2007; Joseph and others 2003).
Distal Risk Factors
Availability of Means and Methods Several biological systems might be involved in suicidal
When a person is contemplating suicide, access to spe- behavior, particularly with regard to the serotonin,
cific methods might be the factor that leads from suicidal noradrenalin, and hypo-thalamic-pituitary-adrenal axis
thoughts and plans to action. systems (Mann 2003).
The easy availability of highly lethal methods is a Family history of suicide increases the risk at least
significant factor in suicides in LMICs. As many as twofold, particularly in girls and women, independent of
30 percent of global suicide deaths might involve inges- family psychiatric history (Qin, Agerbo, and Mortensen
tion of pesticides (Gunnell, Eddleston, and others 2007). 2003). Studies from India (OR = 1.33; confidence inter-
This situation is compounded by the limited availability val (CI) = 0.593.09) (Vijayakumar and Rajkumar 1999)
of appropriate health care services and professionals, and China (OR = 3.9; CI = 2.46.3) (Phillips and others
and by the complexity of managing pesticide overdoses 2002) corroborate these findings.
that lead to increased fatalities.
In Turkey, from 1990 to 2010, the most common History of Suicide Attempts
suicide method was hanging, and men used firearms A history of self-harm or suicide attempts is seen as
more frequently than women did (Oner, Yenilmez, and a very strong risk factor. Studies from China, India,
Ozdamar 2015). In Brazil, the most common methods and Sri Lanka reveal that around one-third of those
were hanging, firearms, and poisoning (Lovisi and who died by suicide had made a prior suicide attempt
others 2009). In Africa, the most frequently used meth- (Abeysinghe and Gunnell 2008; Phillips and others 2002;
ods of suicide were hanging and pesticide poisoning Vijayakumar and Rajkumar 1999).
(Mars and others 2014).
In a systematic review and meta-analysis of the Early Traumatic Events
most common methods of suicide in the Eastern Childhood adversities, including physical, emotional,
Mediterranean region, the pooled proportions of and sexual abuse, have been associated with higher risk
hanging, self-immolation, and poisoning were 39.7, for suicide. A highly significant relationship between
17.4, and 20.3 percent, respectively (Morovatdar and domestic violence and suicidal ideations has been found
others 2013). More females died by self-immolation in many LMICs, with 48 percent of women in Brazil,
than males (29.4 percent versus 11.3 percent); more 61 percent in the Arab Republic of Egypt, 64 percent
males died by hanging than females (38.8 percent versus in India, 11 percent in Indonesia, and 28 percent in the
26.3 percent); and more females died by poisoning than Philippines reporting suicidal ideations and domestic
males (32.0 percent versus 19.0 percent). violence (WHO 2001).
Suicide 171
the worlds internally displaced and refugee populations. Religious and Spiritual Beliefs
Accordingly, it is essential to take steps to provide appro- Religious and strong cultural beliefs that discourage sui-
priate interventions (Reed and others 2012). cide are seen as major protective factors. The protective
value of religion and spirituality probably arises in part
Sexual Minorities from providing access to a socially cohesive and support-
In many LMICs, discrimination against sexual minori- ive community. Islam and Christianity, and specifically
ties, such as lesbians, gays, bisexuals, and transgenders, Catholicism, prohibit the taking of ones own life, and
is ongoing, endemic, and systemic. This problem can this prohibition can have a strong inhibitory effect on
lead to the continued experience of stressful life events, suicidal behavior. Data from Islamic countries and from
such as loss of freedom, rejection, stigmatization, and countries in Latin America and the Caribbean that are
violence that can lead to suicidal behaviors (Haas and predominantly Catholic bear this out; however, the
others 2011). There have been no studies that have com- strong stigma associated with suicide in these cultures
pared suicide rates among sexual minorities in countries may mean that underreporting is likely. The rates of sui-
with or without social acceptance of alternative lifestyles. cide in Islamic countries are very low; for example, Saudi
Arabia and Syria have a similar rate of 0.4 per 100,000
Survivors of Suicide Loss (WHO 2014a). Islam also prohibits alcohol consump-
People bereaved by the suicide of loved ones or a close tion, a known risk factor for suicide.
contact often experience significant emotional distress as A survey of young people from nine Latin American
a result of their loss. These feelings are often accompanied and Caribbean countries reported that attendance at
by feelings of stigma, loss of trust, and social isolation. religious services and connectedness with parents and
Many survivors experience suicidal thoughts themselves. school reduced risk behaviors (Blum and others 2003).
Every year, an estimated four million people may be A study from India revealed that religiosity acted as a
actively experiencing the aftermath of a suicide, many strong protective factor against suicide (Vijayakumar
of them children, due to the high proportion of young 2002). Due to the lack of reliable data, the debate
married women in China and India who die by suicide. remains open as to whether it is the religious beliefs per
Many LMICs do not provide programs for survivors se or the social connectedness that occurs in the context
in any systematic way. Families in which suicide has of religious involvement that is protective.
occurred may be ostracized and isolated, and the mar-
riage prospects of sisters and daughters of people who Positive Coping Strategies and Well-Being
die by suicide may be marred (Khan and Prince 2003). Subjective personal well-being and effective positive
These attitudes may affect the ways in which people coping strategies seem to be protective against suicide
respond to survivors and may reduce the likelihood that (Sisask and others 2008). However, ample debate remains
survivors seek what limited services might be available. regarding the international measures of national and
individual well-being, making the relationship between
well-being and suicide less than simple.
Protective Factors
Use of upstream approaches, such as addressing risk
The role of protective factors, such as resiliency, social and protective factors early in the life course, has the
support, self-esteem, problem-solving skills, and religious potential to shift the odds in favor of more adaptive
affiliation have not been as well studied as risk factors. outcomes. Moreover, upstream approaches may simul-
taneously impact a wide range of health and societal
Strong Personal Relationships outcomes, such as suicide, substance abuse, violence, and
The promotion and maintenance of healthy close rela- crime (Jan-Llopis and others 2005).
tionships can increase resilience and act as a protective Figure 9.3 provides a list of key risk factors for suicide
factor against the risk of suicide. In a study in Brazil, aligned with their possible interventions.
the protective factors for boys and girls included having
good family relationships and feeling liked by friends and
teachers, and these factors seemed beneficial (Anteghini
SUICIDE PREVENTION IN LMICs
and others 2001). Similarly, a survey of adolescents from
nine Caribbean countries reported that strong connec- This section summarizes the evidence for suicide pre-
tions with family and school provided the best protective vention in LMICs. It provides an overview of poten-
factors (Blum and others 2003). Relationships are espe- tial populationwide, community-based, and health and
cially protective for adolescents and elderly persons, who social care interventions and describes the development
have higher levels of dependency. of national suicide prevention strategies.
Barriers to accessing
Health systems
health care
Access to means
Follow-up and
Harmful use of alcohol
commumity support
Assessment and
Job or financial loss
management of
Individual suicidal behaviors Indicated
Hopelessness
Behaviors assessment and
management of mental
Chronic pain and substance
use disorders
Suicide was once commonly viewed as a mental mental disorders. Moreover, in LMICs, the availability
health problem that needed to be addressed primarily of mental health professionals needed to deliver mental
by clinical intervention, especially by the treatment of health interventions is often limited.
depression. Suicide is now recognized as a public health WHO has produced several documents on suicide pre-
issue that should be addressed by social and public vention. Based on these documents and recent literature,
health programs, as well as clinical activities targeting table 9.2 highlights potential interventions in LMICs;
Suicide 173
Table 9.2 Potential Interventions for Suicide in LMICs
Population platform interventions
Universal prevention and health promotion
Restrict the availability of toxic pesticides and other commonly used methods
Decriminalize suicide
Reduce the availability and excessive use of alcohol and illicit drugs
Work with national and local media organizations to limit inappropriate reporting of suicides
Conduct campaigns to reduce the stigma associated with suicide and mental disorders and to encourage help-seeking behavior
Provide adequate economic and welfare support to individuals who are unemployed, disabled, or destitute
the relevance of these to a particular LMIC depends on of suicide. In this section, we consider interventions
its epidemiology of suicide, key risk factors, and social specific to suicidal behavior, such as restricting access
context, as well as the available resources in the country. to commonly used methods of suicide, and those to
The evidence is of mixed quality; in some cases, improve the mental health of the population in general,
it extrapolates from research in HICs. Furthermore, where an impact on suicide seems probable.
because of the low incidence of suicide, the evidence for
several of the interventions comes from trials that have
used suicide attempts, rather than suicide, as the primary Population Platform Interventions
outcome measure. Restricting Access to Lethal Means
Some of the interventions highlighted in other chap- Research has demonstrated that one of the most effec-
ters, such as those to reduce the incidence of alcohol mis- tive approaches to reducing suicide is restricting access
use and depression, will help to decrease the incidence to highly lethal and commonly used methods (Mann
Media Reporting
Improving the portrayal of suicide in the media is an Community Platform Interventions
important component of suicide prevention. Sensational Services of Nongovernmental Organizations
reporting can raise awareness (cognitive availability) of Most LMICs do not have the financial or person-
high-lethality suicide methods that, if popularized, may nel resources to support suicide prevention programs,
Suicide 175
especially health care systemdriven models. It has schoolteachers, people caring for refugees and victims
become imperative to develop low-cost interventions of disaster, hospital emergency department staff, prac-
that can be delivered by lay volunteers or community titioners of traditional and alternative medicine, police,
health workers. prison staff, and youth leaders. Training gives these
This enormous gap in mental health services has individuals the skills to identify and respond to at-risk
been the catalyst for the emergence of nongovernmental individuals (WHO 2012, 2014a).
mental health organizations. Many African and South- Although research evidence to support this activity
East Asian countries have such organizations, often is limited to institutional settings (Mann and others
taking the form of suicide prevention centers, staffed 2005), it appears to be intuitively sensible and is valued
largely by volunteers and operating as crisis centers or by front-line personnel and communities.
hotlines, providing free service in many LMICs. For
example, the Beijing Suicide Research and Prevention Other Community Platform Interventions
Center in China established a national hotline and Recently, there has been interest in multifaceted,
provides standardized training to other hotline services community-based approaches to improving the iden-
around the country. tification and treatment of depression and reducing
The primary goal of these prevention centers is to suicide. Hungary participated in the European Alliance
provide emotional support to suicidal persons through against Depression Programme. The program includes
befriending and counseling in person or by telephone. four levels of intervention: general practitioner training
In many countries, as the primary or sole agency for sui- workshops, a public information campaign, training
cide prevention, they have enlarged their perspectives by community facilitators (gatekeepers), and interventions
being proactive in rural and remote areas and in special targeted at high-risk groups. Szekely and others (2013)
populations. Although many innovative programs for report data from the intervention (population 77,000)
raising awareness and increasing help-seeking behavior and control (population 163,000) regions of Hungary;
have been developed, most have not been evaluated they find evidence of a significantly greater reduction
(Vijayakumar and Armson 2005). in suicide in the intervention region compared with the
control area.
School-Based Interventions A multifaceted suicide prevention program in a
There is mixed evidence concerning the effectiveness Brazilian municipality, the Program for Promotion of
of school-based interventions for preventing suicide. In Life and Suicide Prevention, was designed to reduce sui-
the largest randomized control trial (RCT) carried out cide rates in the general population (Conte and others
to datethe Saving and Empowering Young Lives in 2012). The components of the program included trying
Europe trialmental health awareness and skills training to break taboos and talking about death, improving and
reduced the incidence of suicidal thoughts and attempts streamlining the process of care, and reorganizing work
among secondary school children (Wassermann and processes in the basic network. Although suicide rates
others 2015). More research is needed in this area in fell in the municipality, the lack of comparison informa-
LMICs. tion from control areas means it is not possible to deter-
mine whether the reduction was due to the program or
Safe Storage of Pesticides other influences.
Multiple projects have investigated approaches to Campaigns to reduce stigma associated with suicide
restricting access to pesticides in farming communities and encourage help-seeking have been suggested as
in rural Asia. These include studies of lockable safe a population-level intervention; such campaigns may
storage boxes in Sri Lanka (Hawton and van Heeringen also be appropriately carried out by local communities.
2009; Konradsen and others 2007) and a centralized Activity might also focus on groups identified as being at
community pesticide storage facility in southern India high risk in the particular community, such as victims of
(Vijayakumar and others 2013). These approaches show domestic abuse, people who abuse alcohol, or those who
some promise, although the possibility of adverse effects engage in gambling.
has been raised. A randomized trial of locked storage An unusual intervention in the Islamic Republic of Iran
devices that is enrolling 200,000 people is underway in used videos documenting the stories of self-immolation
Sri Lanka (Pearson and others 2011). victims (Ahmadi and Ytterstad 2007). Young women
from socioeconomically deprived groups who were iden-
Gatekeeper Training tified as at high risk were targeted. There was some evi-
A gatekeeper is anyone in a position to identify whether dence of a beneficial effect on self-immolation and overall
someone may be at risk of suicide. Gatekeepers include suicide attempts compared with a nonintervention city.
Suicide 177
is no longer operational. In India, suicide prevention A substantial minority of individuals who attempt
is included in the countrys national mental health suicide or die by suicide in these settings does not have a
program. mental disorder. Psychosocial and economic risk factors
Although many risk factors for suicide are shared by need to be acknowledged, and interventions need to be
all countries, their relative importance in determining developed that target these factors. In LMICs, suicide
the local incidence of suicide varies. The first step in prevention is more of a social and public health objective
informing priority areas for suicide prevention is to than a traditional mental health sector objective.
collect good quality, nationally representative data on Before intervening, information about the prevalence,
the age- and gender-specific incidence of suicide, the demographic patterns, and methods of suicide in the
methods used by those who take their lives, and the key country or community is needed. Data from represen-
risk factors. Guidelines by WHO to set up a surveillance tative locations on the pattern of deaths is particularly
system and the process to be followed can be accessed important in countries without effective registry systems.
from the STEPwise approach to surveillance at http:// Several evidence gaps exist. A more refined estimate of the
www.who.int/chp/steps/en. burden and modeling that focuses on risk factor abate-
ment, resilience enhancement, and intervention effects
will effectively direct future suicide prevention activities.
COST-EFFECTIVENESS OF PREVENTION
EFFORTS
NOTE
The cost of treating suicide attempts, particularly self-
poisoning by pesticides in LMICs, is high (Sgobin and Portions of this chapter are based on work that will appear in
the International Handbook of Suicide Prevention, 2nd edition,
others 2015; Wickramasinghe and others 2009). Suicide
forthcoming from Wiley.
prevention control measures may need to be tailored to The authors are very grateful to Mr. Sujit John, Senior
the context of a specific country, taking into consideration Research Coordinator, Schizophrenia Research Foundation,
the epidemiological, geographic, and gender distribution for his technical assistance in the preparation of the chapter.
of suicide, political will, perceptions of stigma, legisla- World Bank Income Classifications as of July 2014 are as
tion, and resource availability to deliver appropriately follows, based on estimates of gross national income (GNI)
designed prevention programs. As such programs are per capita for 2013:
developed, there will be a need to generate cost and cost-
effectiveness information. Although there have been some Low-income countries (LICs) = US$1,045 or less
promising interventions in LMICs, the evidence of cost- Middle-income countries (MICs) are subdivided:
effectiveness remains sparse, and evidence on costs and a) lower-middle-income = US$1,046 to US$4,125
cost-effectiveness from HICs may not be relevant (WHO b) upper-middle-income (UMICs) = US$4,126 to
US$12,745
2010b). No economic evaluation was conducted for the
High-income countries (HICs) = US$12,746 or more.
multicountry RCT of BIC (Fleishmann and others 2008),
but the clinical costs were equal to treatment as usual.
Chapter 12 in this volume (Levin and others 2015) pro-
REFERENCES
vides a review of costs and cost-effectiveness for mental
health interventions more broadly. Abeyasinghe, R., and D. Gunnell. 2008. Psychological Autopsy
Study of Suicide in Three Rural and Semi-Rural Districts
of Sri Lanka. Social Psychiatry and Psychiatric Epidemiology
43 (4): 28085.
CONCLUSIONS Ahmadi, A., R. Mohammadi, D. C. Schwebel, N. Yeganeh,
Suicide is a major public health problem in LMICs. The A. Soroush, and S. Bazargan-Hejazi. 2009. Familiar Risk
magnitude of the problem and the paucity of resources Factors for Self Immolation: A Case Control Study. Journal
in these countries necessitate a need for collaboration of Womens Health 18 (7): 102531.
and cooperation across a variety of stakeholders to Ahmadi, A., and B. Ytterstad. 2007. Prevention of
Self-Immolation by Community-Based Intervention.
implement strategies that are culturally relevant and
Burns 33 (8): 103240.
cost-effective. The huge variability in the prevalence, Anteghini, M., H. Fonseca, M. Ireland, and R. W. Blum. 2001.
demographic patterns, and methods of suicide should Health Risk Behaviors and Associated Risk and Protective
be considered when making global cross-national Factors among Brazilian Adolescents in Santos, Brazil.
prevention recommendations. LMICs need to adopt a Journal of Adolescent Health 28 (4): 295302.
process whereby they can decide on the interventions Barbosa, L. P., L. Quevedo, G. da Silva Gdel, K. Jansen,
that are appropriate for their cultures and populations. R. T. Pinheiro, and others. 2014. Childhood Trauma and
Suicide 179
Khan, M. M., and M. Prince. 2003. Beyond Rates: The Tragedy Review and Meta-Analysis. Archives of Suicide Research
of Suicide in Pakistan. Tropical Doctor 33 (2): 6769. 17 (4): 33544.
Knipe, D. W., R. Carroll, K. H. Thomas, A. Pease, D. Gunnell, NCRB (National Crime Research Bureau). 2014. Accidental
and C. Metcalfe. 2015. Association of Socio-Economic Deaths and Suicide in India. Ministry of Home Affairs,
Position and Suicide/Attempted Suicide in Low and Government of India, New Delhi.
Middle Income Countries in South and South-East Oner, S., C. Yenilmez, and K. Ozdamar. 2015. Sex-Related
AsiaA Systematic Review. BMC Public Health 15: 1055. Differences in Methods of and Reasons for Suicide in
doi:10.1186/s12889-015-2301-5. Turkey between 1990 and 2010. Journal of International
Konradsen, F., R. Pieris, M. Weerasinghe, W. Van der Hoek, Medical Research 43 (4): 48393.
M. Eddleston, and A. H. Dawson. 2007. Community Otzen, T., A. Sanhueza, C. Manterola, and J. A. Escamilla-Cejudo.
Uptake of Safe Storage Boxes to Reduce Self-Poisoning from 2014. Mortalidadporsuicidioen Chile: Tendenciasenlosanos
Pesticides in Rural Sri Lanka. BMC Public Health 7 (1): 13. 19982011 [Trends in Suicide Mortality in Chile from 1998
Krawczyk, N., A. Meyer, M. Fonseca, and J. Lima. 2014. Suicide to 2011]. Revista Medica de Chile 142: 30513.
Mortality among Agricultural Workers in a Region with Patel, V. 2007. Mental Health in Low- and Middle-Income
Intensive Tobacco Farming and Use of Pesticides in Brazil. Countries. British Medical Bulletin 81 (1): 8196.
Journal of Occupational and Environmental Medicine 56 (9): Patel, V., C. Ramasundarahettige, L. Vijayakumar, J. S. Thakur,
9931000. V. Gajalakshmi, and others. 2012. Suicide Mortality in
Kumar, S., J. Lakshmanan, S. Saradha, and R. C. Ahuja. 2005. India: A Nationally Representative Survey. The Lancet
Domestic Violence and Its Mental Health Correlates 379 (9834): 234351.
in Indian Women. British Journal of Psychiatry 187 (1): Pearson, M., F. Konradsen, D. Gunnell, A. H. Dawson,
6267. R. Pieris, and others. 2011. A Community-Based Cluster
Lovisi, G. M., S. A. Santos, L. Legay, L. Abelha, and E. Valencia. Randomised Trial of Safe Storage to Reduce Pesticide Self-
2009. Epidemiological Analysis of Suicide in Brazil Poisoning in Rural Sri Lanka: Study Protocol. BMC Public
from 1980 to 2006. Brazilian Journal of Psychiatry 31 (2): Health 21 (11): 879.
S8694. Phillips, M. R., G. Yang, Y. Zhang, L. Wang, H. Ji, and M. Zhou.
Machado, D. B., D. Rasella, and D. N. Dos Santos. 2015. Impact 2002. Risk Factors for Suicide in China: A National Case-
of Income Inequality and Other Social Determinants on Control Psychological Autopsy Study. The Lancet 360 (9347):
Suicide Rate in Brazil. PLoS One 10 (4): e0124934. 172836.
Malakouti, S. K., F. Davoudi, S. Khalid, M. A. Asl, M. M. Khan, Pirkis, J., and R. W. Blood. 2010. Suicide and the News
and others. 2015. The Epidemiology of Suicide Behaviors and Information Media: A Critical Review. http://www
among the Countries of the Eastern Mediterranean Region .mindframe-media.info/__data/assets/pdf_file/0016
of WHO: A Systematic Review. Acta Medica Iranica 53 (5): /5164/Pirkis-and-Blood-2010,-Suicide-and-the-news-and
25765. -information-media.pdf.
Mann, J. J. 2003. Neurobiology of Suicidal Behaviour. Nature Pridemore, W. A., M. B. Chamlin, and E. Andreev. 2013.
Review Neuroscience 4: 81928. Reduction in Male Suicide Mortality Following the
Mann, J. J., A. Apter, J. Bertolote, A. Beautrais, D. Currier, and 2006 Russian Alcohol Policy: An Interrupted Time-Series
others. 2005. Suicide Prevention Strategies: A Systematic Analysis. American Journal of Public Health 103: 202126.
Review. Journal of the American Medical Association 294 Qin, P., E. Agerbo, and P. B. Mortensen. 2003 Suicide Risk
(16): 206474. in Relation to Family History of Completed Suicide and
Mars, B., S. Burrows, H. Hjelmeland, and D. Gunnell. 2014. Psychiatric Disorders. A Nested Case-Control Study. Ugeskr
Suicidal Behaviour across the African Continent: A Laeger 165 (25): 257377 [original article in Danish].
Review of the Literature. BMC Public Health 14 (14): 606. Rahman, A., and A. Hafeez. 2003. Suicidal Feelings Run High
doi:10.1186/1471-2458-14-606. among Mothers in Refugee Camps: A Cross-Sectional
McColl, H., K. McKenzie, and K. Bhui. 2008. Mental Survey. Acta Psychiatrica Scandinavica 108 (5): 39293.
Healthcare of Asylum-Seekers and Refugees. Advances in Reed, R. V., M. Fazel, L. Jones, C. Panter-Brick, and A. Stein.
Psychiatric Treatment 14 (6): 45259. 2012. Mental Health of Displaced and Refugee Children
Milner, A. J., G. Carter, J. Pirkis, J. Robinson, and M. J. Spittal. Resettled in Low-Income and Middle-Income Countries:
2015. Letters, Green Cards, Telephone Calls and Postcards: Risk and Protective Factors. The Lancet 379 (9812): 25065.
Systematic Review and Meta-Analytic Review of Brief Schlebusch, L. 2005. Suicidal Behaviour in South Africa.
Contact Interventions for Reducing Self-Harm, Suicide Pietermaritzburg, University of KwaZulu-Natal Press.
Attempts and Suicide. British Journal of Psychiatry 206: Sgobin, S. M., A. L. Traballi, N. J. Botega, and O. R. Coelho. 2015.
18490. Direct and Indirect Cost of Attempted Suicide in a General
Mishara, B. L., and D. N. Weisstub. 2014. Suicide Laws: An Hospital: Cost-of-Illness Study. Sao Paolo Medical Journal
International Review. International Journal of Law and 133 (3): 21826. doi:10.1590/1516-3180.2014.8491808.
Psychiatry (in press). Sisask, M., A. Vrnik, K. Kolves, K. Konstabel, and
Morovatdar, N., M. Moradi-Lakeh, S. K. Malakouti, and D. Wasserman. 2008. Subjective Psychological Well-Being
M. Nojomi. 2013. Most Common Methods of Suicide (Who-5) in Assessment of the Severity of Suicide Attempt.
in Eastern Mediterranean Region of WHO: A Systematic Nordic Journal of Psychiatry 62 (6): 43135.
Suicide 181
Chapter
10
Population and Community Platform Interventions
Inge Petersen, Sara Evans-Lacko, Maya Semrau,
Margaret Barry, Dan Chisholm, Petra Gronholm,
Catherine O. Egbe, and Graham Thornicroft
Corresponding author: Inge Petersen, University of KwaZulu Natal, Durban, Psychology, School of Applied Human Sciences, South Africa, peterseni@ukzn.ac.za.
183
Table 10.1 Matrix of Best-Practice and Good-Practice Interventions
Identification and case Treatment, care,
Delivery platform Promotion and primary prevention detection and rehabilitation
Population
Legislation and regulation Laws and regulations to reduce demand for Mental health
alcohol use: taxes laws and
Laws and regulations to reduce demand regulations that
for alcohol use; enforcement of BAC limits, are in line with
advertising bans, and minimum ages the best practice
and human rights
Laws and regulations to restrict access to standards
means of self-harm and suicide
Child protection laws
Laws and regulations promoting healthy
lifestyles, for example, tobacco control
Laws and regulations to promote improved
control of neurocysticercosis
Information and awareness Mass public awareness campaigns
Community
Workplace Integrating mental health promotion
strategies, such as stress reduction and
awareness of alcohol and drug misuse, into
occupational health and safety policies
Schools Universal and targeted SEL programs for Identification and case
vulnerable children detection in schools of
Awareness programs children with MNS disorders
line with human rights standards would be protective. Legislation and Regulations for Promotion and
Additional interventions were identified on the assump- Primary Prevention
tion that addressing the known determinants of MNS Reducing Harmful Alcohol Use
disorders should promote mental health and lead to a The prevention of harmful alcohol use in adults provides
reduction in MNS disorders, but these interventions were benefits across diseases. It can help prevent the develop-
not recommended as good practice, given the lack of ment of alcohol use disorder and unipolar depression, as
evidence of their effectiveness. The thorough review in well as other chronic diseases, such as cardiovascular dis-
this volume of the available evidence of the most effec- ease, diabetes, and cirrhosis of the liver, and it can reduce
tive and cost-effective interventions for the respective the risk of contracting human immunodeficiency virus
disorders was used as the evidence base, supplemented (HIV). It can also help with the prevention of accidental
by a desk review of the best evidence where necessary and intentional injuries or death (Rehm and others 2006).
(see online annex table 10A.) For further information on Evidence from HICs and LMICs indicates that the
the cost-effectiveness of the mental health interventions most cost-effective strategy for reducing alcohol con-
referenced in this chapter, see chapter 12 in this volume sumption is increased taxation or pricing of alcohol
(Levin and Chisholm 2015). products, followed by bans on alcohol advertising,
Box 10.1
Self-poisoning with pesticide is the most common Organization toxicity Class 1 pesticides in 1995 and
method of suicide in Sri Lanka, accounting for two- the banning of endosulfan, a Class II toxicity pesti-
thirds of suicide deaths. The suicide rate in Sri Lanka cide, in 1998, the suicide rate halved from 1996 to
reached a peak in 1995 at 47 deaths per 100,000 2005, with a reduction of 19,769 suicides, compared
population. With the banning of all World Health with 198695 (Gunnel and others 2007).
Best-practice interventions were identified on the basis Identification and Case Detection
of two criteria: Evidence on the identification and case detection of MNS
disorders in the workplace could only be sourced from
Evidence of their effectiveness based on sufficient HICs. An evaluation of the APPRAND program in France
evidence from LMICs, using the ACE framework, as provided evidence on individuals on sick leave who were
well as their cost-effectiveness in HICs. screened by company health physicians and identified as
Evidence of their feasibility in relation to cultural having anxiety and depressive disorders and who received
acceptability and capacity for scale-up in resource- an awareness-raising and referral intervention. Those
constrained settings in LMICs. individuals displayed higher remission and recovery
rates, compared with individuals in other centers who
Good-practice interventions were identified on the were not screened and who did not receive the inter-
basis of sufficient evidence of their effectiveness in HICs vention (Godard and others 2006). Positive effects have
and/or promising evidence of their effectiveness in also been reported for a mental health first aid course in
LMICs, using the ACE framework. Australia that included training in screening for mental
disorders (Kitchener and Jorm 2004).
For neurological disorders, positive outcomes have
Workplaces been reported in the United States for migraine and
Promotion and Primary Prevention headache management programs that have included
Workplace settings provide an ideal delivery channel screening questionnaires and educational initiatives.
for promotion and prevention interventions for adults. These interventions resulted in an increase in the number
Evidence from HICs indicates that individual- and of participants seeking help from physicians, an improve-
organization-level interventions improve and maintain ment in headache symptoms, a reduction in absenteeism
mental health in the workplace. These interventions include among those affected, and a reduction in the cost burden
screening and cognitive behavioral therapy (CBT) for pre- to employers (Page and others 2009; Schneider and oth-
clinical symptoms of depression and anxiety to prevent the ers 1999). No evidence for screening for MNS disorders
onset of these disorders (Nytro and others 2000; WHO in the workplace could be sourced from LMICs, and these
2000). However, the evidence base from LMICs is sparse. interventions are not yet recommended.
Box 10.2
Classroom, Community, and CampBased Intervention in the West Bank and Gaza
Classroom, community, and campbased inter- and girls (ages 611 years), as well as in adolescent
vention provides structured expressive-behavioral girls (ages 1216 years), enabling them to func-
group activities over 15 sessions to reduce traumatic tion as other children would in relation to family,
stress reactions and strengthen childrens resiliency school, and peers. However, this effect was not the
to cope with the stress of ongoing violence and case with adolescent boys (ages 1216 years), who
trauma. The program was delivered by trained demonstrated an increased tendency to use avoid-
school counselors and other social workers to more ance of cognitions and feelings as a defense mecha-
than 100,000 children in the West Bank and Gaza. nism, which may relate to their greater exposure to
A randomized control trial involving 664 children violence.
ages 616 years found that the program improved
psychological functioning and coping in young boys Source: Khamis, Macy, and Coignez 2004.
Box 10.4
An exploratory study in So Paulo, Brazil, tested not initially respond correctly to the vignettes,
the effectiveness of an educational strategy to build researchers found at least 50 percent had learned
teachers capacity to identify students with possible to identify and make referrals of problematic cases
mental health problems and subsequently make following the training, and 60 percent learned to
appropriate referrals. Teacher training involved identify normal adolescent behaviors. The study
two two-hour sessions that included a lecture suggests that brief training can increase teachers
followed by theoretical and practical exercises. capacity to identify mental health problems and
Teachers were evaluated on their ability to identify make appropriate referrals, especially among those
and refer students with mental health problems in who initially struggled to do so.
a hypothetical vignette scenario. When assessing
responses specifically among teachers who did Source: Vieira and others 2014.
appropriate action. RCTs from HICs provide evidence The intervention was associated with improvements in
for training in indicated screening of developmental knowledge, attitudes, and recognition of MNS disor-
and behavioral disorders in schools. Programs such as ders (Eustache, Becker, and Wozo 2014). In Chandigarh
Mental Health First Aid for High School Teachers have city, India, a one-off educational intervention package
been tested using a cluster RCT (Jorm and others 2010). improved teachers knowledge, attitudes, and skills
Data from LMICs are limited. However, evidence regarding epilepsy immediately after the intervention,
supports the feasibility and reliability of identify- and at the three-month follow-up. However, it was
ing and assessing MNS disorders in primary and noted that further workshops would likely be required
secondary school students (Becker and others 2010a, for long-term benefit (Goel and others 2014).
2010b; Opoliner and others 2013; Vieira and others Given sufficient evidence from HICs, as well as
2014) (box 10.4). In Haiti, a 2.5 day training program emerging promising evidence from LMICs, the identi-
for secondary school teachers focused on recogniz- fication and case detection in schools of children with
ing, responding to, and referring students at risk for MNS disorders are recommended as good practice.
MNS disorders following the earthquake in 2010. Further research adapting and developing, validating,
Corresponding author: Rahul Shidhaye, Centre for Chronic Conditions and Injuries, Public Health Foundation of India, rahul.shidhaye@phfi.org.
201
These three key delivery channels map well onto the health care services. Health promotion interventions
commonly cited Service Organization Pyramid for an delivered at the population level can be important in
Optimal Mix of Services for Mental Health supported by improving mental health literacy by helping people
the World Health Organization (WHO) (figure 11.1) to recognize problems or illnesses, increasing their
(WHO 2003a). At each subsequent level of the pyramid, knowledge about the causes of disorders and options
the mental health needs of individuals become greater for treatment, and informing them about where to go
and require more intensive professional assistance, usu- to get help (see chapter 10 in this volume, Petersen and
ally resulting in higher costs of care. In certain settings others 2015).
beset with conflict, natural disaster, or other emer- Informal health care comprises service providers
gencies, a further channel for delivering much-needed who are not part of the formal health care system, such
mental health care is humanitarian aid and emergency as traditional healers, village elders, faith-based organi-
response. zations, peers, user and family associations, and lay
people (WHO 2003a). Traditional and religious healers
are of particular significance, as populations through-
Self-Care and Informal Health Care out East Asia and Pacific, South Asia, Latin America
The foundation of the health care delivery platform rests and the Caribbean, and Sub-Saharan Africa often use
on self-care and emphasizes health workerpatient part- traditional medicine to meet their health needs (WHO
nerships. Persons with MNS disorders and their family 2002). In many parts of the world, making contact with
and friends play a central role in the management of men- such informal providers represents the initial pathway
tal health problems. The role of individuals may range to care (Bekele and others 2009); these service provid-
from collaborative decision-making concerning their ers are typically very accessible and more acceptable
treatment, to actively adhering to prescribed medication, because they are integral members of the local com-
to changing health-related behaviors, such as drug and munity. Given the widespread presence of traditional
alcohol use, stress management, and identification of and religious healers and the shortage of human
seizure triggers and avoiding them for seizure control. resources in mainstream biomedical services, it is
Self-care is important for MNS disorders, but it imperative that primary health and other formal care
is also important for the prevention and treatment services establish strong links with informal health care
of physical health problems (WHO 2003a). Self-care providers, especially traditional healers (Patel 2011). It
is most effective when it is supported by popula- is also critical to note that the evidence base regarding
tionwide health promotion programs and formal the effectiveness of services provided by traditional
and religious healers is limited. Nevertheless, it is
essential to engage with them, as they provide acces-
Figure 11.1 World Health Organization Service Organization Pyramid sible, acceptable, and affordable care, and efforts need
for an Optimal Mix of Services for Mental Health to be made to ensure that their practices do not harm
the patients.
Low High
Peers are another key human resource at this level
of health care. Peer-led education and behavioral inter-
Long-stay ventions have been effective with target populations
facilities and
specialist services and health issues in low- and middle-income countries
Psychiatric Community (LMICs) (Manandhar and others 2004; Medley and
services in mental health
general services others 2009; Tripathy and others 2010). Peers are more
hospitals
numerous, may be perceived as more approachable, and
Costs
Frequency of need
mental health
lf-c
Informal community care health conditions with members of the target population
Informal services
Relationships among Different Delivery Channels Table 11.1 summarizes the evidence base for
No single service delivery channel can meet all mental interventions by various delivery channels. The inter-
health needs. For example, on the one hand, primary ventions are intended as examples rather than as
mental health care must be complemented by special- recommendations.
ist care services that primary health workers can use
for referrals, support, and supervision; on the other
hand, primary mental health care needs to promote SYSTEM-STRENGTHENING STRATEGIES FOR
and support self-care and informal community care INTEGRATED HEALTH CARE DELIVERY
that encourages the involvement of people in their
own recovery. Support of self-care and management The availability of evidence-based interventions does not
can be provided via routine primary care visits or via ensure their translation into practice. In this section, we
group sessions led by health or lay workers in health address the question of how to integrate evidence-based
care settings or community venues. Another increas- mental health care interventions into primary care and
ingly accessible option for the effective support of self-care delivery channels and how to link this integra-
self-care and management is telephone- or Internet- tion to specialist care.
based programs. In short, the potential of the health A comprehensive and multifaceted approach that
care system as a delivery platform for enhanced men- contains the following elements is essential for the
tal health and well-being can only be fully realized if successful integration of mental health into health care
genuine continuity and collaboration of care occur systems:
across the three service delivery channels; continuity
and collaboration, in turn, rely on an appropriate A whole-of-government approach involves the pro-
flow of support, supervision, information-sharing, motion, pursuit, and protection of health through
and education. concerted action by many sectors of government.
Table 11.2 Key Characteristics of a Public Health Approach to MNS Disorder Prevention and Management
Prevention essentials Management essentials
Promotion of healthy behaviors Person-centered care and support
Prevention of exposure to adverse events and risks Family and community support
Early detection Coordinated, holistic care
Intersectoral collaboration Continuity of care and proactive follow-up
Life course approach
Source: WHO and Calouste Gulbenkian Foundation 2014.
Note: MNS = mental, neurological, and substance use.
Recovery of adults suffering from depression or Improvement in the mental well-being, burden,
anxiety, or both, at 712 months following the and distress of caregivers of people with dementia
intervention Decrease in the amount of alcohol consumed by
Reduction in symptoms for mothers with perina- people with alcohol-use disorders
tal depression symptoms Reduction in functional impairment of children
Reduction in the prevalence and the symptoms affected by post-traumatic stress disorder at six
of adults with post-traumatic stress disorder over and 12 months following the intervention.
six months
Improvement in symptoms of people with Sources: Clarke, King, and Prost 2013; van Ginneken and others 2013.
dementia
might need to help ensure the effective delivery of ser- mental health specialists, increases the skills of primary
vices for depression, psychosis, epilepsy, and alcohol use care workers and builds mental health networks (WHO
disorders in Sub-Saharan Africa (IOM 2013). Table 11.3 and WONCA 2008).
lists the steps to strengthen human resource compe-
tencies for MNS disorders; the core competencies for Specialist Transitioning
all service providers across MNS disorders are listed in Specialists, especially in LMICs, are usually engaged in
table 11.4. In addition to the common competencies for service delivery. It is imperative to make a transition from
all service providers, the IOM framework also focuses on providing clinical services to training and supervising
a diverse range of cadre-specific competencies. the primary health care staff and providing direct clin-
Pre-service and in-service training of primary care ical interventions judiciously and sparingly. In separate
workers on mental health issues is an essential prereq- projects focusing on integrated primary care for mental
uisite for the integration of mental health into primary health in the city of Sobral, Brazil, and the Sembabule dis-
care platforms. The training, to the extent possible, trict of Uganda, specialists together with medical officers
should happen in primary care or community mental in primary care visited primary care settings and assessed
health care facilities, to ensure that practical experience patients. Over time, psychiatrists started taking less active
is gained and that ongoing training and support are roles, while general practitioners assumed added respon-
facilitated (WHO and WONCA 2008). The effects of sibilities, under the supervision of the psychiatrists.
training are nearly always short lived if health workers Specialists can interact with primary care staff via referral
do not practice newly learned skills and receive ongoing and back-referral (WHO and WONCA 2008).
specialist supervision. A trial from Kenya did not find
any impact of the training program of medical officers Planning and Consultation
on improvement in diagnostic rates of mental disor- Involving primary health care staff in the overall program
ders (Jenkins and others 2013). A quasi-experimental planning and rollout process enhances ownership and
study from Brazil had similar findings and noted that commitment to achieve the planned outcomes within
wider changes in the system of care may be required agreed timelines (Patel and others 2013). Consultations
to augment training and encourage reliable changes in with general practitioners have been demonstrated to be
clinical practice (Goncalves and others 2013). Ongoing one of the key factors in the success of the new mental
support and supervision from mental health specialists health services in Australia (WHO and WONCA 2008).
are essential. Case studies from Australia, Brazil, and Decisions need to be made after careful consideration
South Africa have demonstrated that a collaborative of local circumstances; this requires consultation with
stepped care approach, in which joint consultations and policy makers as well as users of mental health services
interventions occur between primary care workers and and their families and the primary care staff.
Table 11.4 Core Competencies for All Service Providers across MNS Disorders
Competency
Screening and identification
Demonstrate awareness of common signs and symptoms of MNS disorders
Recognize the potential for risk to self and others
Demonstrate basic knowledge of causes
Provide the patient and community with awareness and education
Demonstrate cultural competence
Demonstrate knowledge of other MNS disorders
Corresponding author: Carol Levin, Department of Global Health, University of Washington, Seattle, WA, United States; clevin@uw.edu.
219
the World Health Organizations (WHO) CHOosing since 2000 in English. The search combined terms
Interventions that are Cost-Effective (CHOICE) project. for specific mental health interventions with eco-
Such model-based studies rely on existing data, as well nomic terms such as cost, cost-effectiveness, or
as several analytical assumptions; these studies have quality-adjusted life year (QALY), as well as the
adopted an epidemiological, population-based approach names of all LMICs and their respective regions (see
that identifies the expected costs and health impacts of annex 12A for a list of search terms used to identify
delivering evidence-based interventions at scale in the relevant literature). Where little or no literature was
population as a whole, whether a specific country or found for LMICs on interventions of potential impor-
an entire region. We also review this form of economic tance, this systematic search was augmented by selec-
evidence and comment on important gaps in the current tive searches of the literature available since 1995 for
evidence base, as well as the relative strengths and limita- high-income countries (HICs); however, these results
tions of this approach. are not included in the figures or tables. Annex 12B
One important limitation of conventional cost- provides the search statistics.
effectiveness analysiswhether garnered through Articles included in the review were graded using the
trial-based or model-based approachesis that it is checklist of Drummond and others (2005) to generate a
restricted to consideration of the specific implementa- quality score for each article, with most studies graded
tion costs and health-related outcomes of an interven- between 7 and 10. Annex 12C provides a list of studies
tion; it does not typically extend to the nonhealth or that were used to generate the tables and figures pre-
wider economic or social value of investing in mental sented in this chapter. It presents detailed information
health innovation and service scale-up. In particu- on the intervention characteristics and comparators,
lar, cost-effectiveness analysis in its conventional form target population group, geographic location, method-
has little to say about the equitable distribution of ology, results, and quality scores. All cost-effectiveness
costs and health gains across different groups of the results are presented in 2012 US$ except where noted
target population. Incorporation of such concerns into otherwise. Consistent with earlier iterations of DCP,
economic evaluation represents a major objective of reported regional estimates refer to the World Banks
extended cost-effectiveness analysis, which is explored categorization of countries by income.
and addressed specifically in chapter 13 in this volume
(Chisholm, Johansson, and others 2015).
In this chapter, we review the available cost- COST-EFFECTIVENESS OF MENTAL HEALTH
effectiveness evidence for the different levels and under- PROMOTION AND PROTECTION MEASURES
pinning strategies of the mental health care system, AT THE POPULATION AND COMMUNITY
with a focus on information generated in or for LMICs.
Based on the overall analytical framework and priority
LEVELS
intervention matrices developed for this volume, the Economic evaluation has yet to be extensively applied to
remainder of the chapter is presented as follows. First, mental health promotion, largely because of the chal-
we consider the economic evidence for mental health lenges associated with using conventional methods and
prevention and protection at the population and principles of cost-effectiveness analysis in the context
community levels of the health and welfare system, of such programs, in particular, the limitations of exper-
including legislative, regulatory, and informational mea- imental study design; the multifaceted, complex, and
sures at the public policy level (population platform), long-term nature of anticipated program benefits; and
as well as school-, workplace-, and community-based the shortage of sensitive or suitable outcome measures
programs (community platform). We then examine the (Petticrew and others 2005). Moreover, many of the
economic evidence relating to the identification and determinants of poor mental health and mental health
treatment of MNS disorders (health care platform), inequalities lie outside the health sector, thereby requir-
focusing on the relative cost-effectiveness or efficiency ing an evaluation of intersectoral action. Certain mental
of treatment programs implemented in nonspecialized health promotion strategies are not amenable to con-
versus more specialized health care settings. Finally, we trolled studies, because it is not feasible or ethical to
assess the financial costs and budgetary implications of exclude a segment of the target population from exposure
implementing or scaling up a set of prioritized, cost- to the intervention in question. Since cost-effectiveness
effective interventions. is by definition a relative concept, this limitation makes
Our review is based on available, published litera- estimation of the relative or comparative efficiency of one
ture. A systematic search of the literature for LMICs strategy over another problematic. Where such compar-
was undertaken in PubMed to find articles published isons are not possible, prospective observational studies,
1,511
(Thailand; Prukkanone and others 2012)
236
(Nigeria; Gureje and others 2007)
Note: ** = effects measured in quality-adjusted life years gained; all other effect estimates are measured as disability-adjusted life years averted; MNS = mental, neurological, and substance
use; SSRI = selective serotonin reuptake inhibitor; TCA = tricyclic antidepressant. All reported cost-effectiveness estimates have been converted to 2012 US$.
A WHO-CHOICE (World Health Organization life year gained ranged from less than US$100 for
CHOosing Interventions that are Cost-Effective) acute management with simple analgesics to thou-
analysis was conducted for a selected core set of sands or even tens of thousands of US$ for treat-
interventions for migraine in four countries: China, ment of analgesic nonresponders with triptans.
India, the Russian Federation, and Zambia. The
analysis included first-line analgesics, such as acetyl- The most cost-effective strategy by far is acute man-
salicylic acid 1,000 milligrams (mg), and second-line agement with simple analgesics; it was less than
medications, such as sumatriptan 50 mg, for acute US$100 per disability-adjusted life year averted and
treatment of attacks. It was assumed that the latter therefore represents a highly cost-effective use of
would be used only by nonresponders to first-line resources for health. Adding consumer education
medications (a stepped care treatment paradigm). and improving adherence has a small upward influ-
The analysis included prophylactic drugs, such as ence on cost-effectiveness. Compared with no treat-
amitriptyline 100 mg daily. The expected conse- ment at all, this strategy is less than US$150 per
quences of adding consumer education, in the form healthy life year gained; compared with use of simple
of posters and leaflets in pharmacies explaining how analgesics without consumer education, the incre-
to acquire and use these medications, and train- mental cost to be paid to obtain one extra healthy life
ing for health care providers were also modeled. year rises to US$600.
Compared with no treatment, the cost per healthy Source: Linde, Chisholm, and Steiner 2015.
Specialist Health Care the basis of expected cost savings; inadequate expenditure
Specialized mental health care covers hospital-based on community-based care is quite likely to result in poor
outpatient and inpatient care for acute and severe outcomes for the individuals and families concerned
episodes or cases of mental disorder. In many LMICs, (Knapp and others 2011).
mental hospitals absorb a disproportionate share of the Detailed analysis of this kind has not been conducted
government mental health budgetover 70 percent in the context of ongoing efforts to relocate services in
in many casesyet such institutions are commonly LMICs. However, a simple comparison of the cost of a
associated with isolation, human rights violations, and community-based versus hospital-based service model
poor outcomes. Such expenditure patterns also curb has been carried out as part of the WHO-CHOICE
the development of more equitable and cost-effective analysis for schizophrenia and bipolar affective disorder.
community-based services. For schizophrenia, the costs of the hospital-based ser-
The dramatic deinstitutionalization observed in vice model exceeded those of the community-based
most HICs in recent decades has been accompanied service model by 3350 percent, reflecting greater use of
by a certain amount of economic research into the resource-intensive services, such as acute and long-term
costs, needs, and outcomes of persons relocated into psychiatric inpatient care (Chisholm 2005; Chisholm
community-based care. Such research has shown that and others 2008). Even if one assumes no improved out-
community-based care is certainly associated with better comes for persons treated under the community-based
health and social outcomes, and it is not inherently more service model, there is a clear difference in terms of
costly than institutions, once account is taken of indi- cost-effectiveness; the costs of the community-based
viduals needs and the quality of care (Knapp and others service model are 2540 percent lower.
2011). New community-based care arrangements could Relocating services and resources away from long-
be more expensive than long-stay hospital care, but they stay mental hospitals toward nonspecialized health set-
may still be seen as more cost-effective because, when tings is a key financing issue for mental health systems.
appropriately set up and managed, they deliver better Efforts to change the balance of mental health care are
health and economic outcomes. Accordingly, such a pro- often hindered by a lack of appropriate transitional
cess of deinstitutionalization should not be predicated on funding. Transitional or dual funding is required over a
Table 12.2 Cost of Implementing Resourceful Adolescent ProgrammeAdolescent Version in Four Countries
Cost item Ethiopia India Mexico Mauritius
Total population age 12 years (%) 1.4 1.1 1.0 0.8
Health educators needed per 1 million population 2.8 2.3 2.1 1.7
(at 100% coverage)
Cost per head of population at 100% coverage (US$) 0.03 0.03 0.11 0.24
Source: World Health Organization, CHOICE (database), http://www.who.int/choice/costs.
2.50
Cost per head of population (2012 US$)
2.00
1.50
1.00
0.50
0
Current Target Current Target Current Target Current Target Current Target
Sodo district Sehore district Chitwan district Kenneth Kaunda Kamuli district
(Ethiopia) (India) (Nepal) district (Uganda)
(South Africa)
Alcohol use disorders $0.006 $0.052 $0.023 $0.057 $0.002 $0.035 $0.061 $0.477 $0.003 $0.054
Epilepsy $0.036 $0.210 $0 $0 $0.028 $0.136 $0 $0 $0.062 $0.184
Psychosis $0.017 $0.144 $0.030 $0.098 $0.007 $0.317 $1.081 $1.505 $0.011 $0.072
Depression $0.024 $0.155 $0.014 $0.097 $0.006 $0.178 $0.284 $0.530 $0.007 $0.073
modeling studies have been conducted for a range Africa, the other in South-East Asia). This wide range of
of disorders, permitting comparison of relative cost- cost-effectiveness points to the importance of carefully
effectiveness with other DCPs. Arguably, there is now evaluating and choosing an appropriate set of inter-
sufficient evidence to counteract or debunk the overgen- ventions for scaled-up investment and implementation;
eralized claim that treatment of mental disorders is not a selecting an inefficient set will waste money and limit
cost-effective use of scarce health care resources. potential health gains. Unfortunately, however, a high
As with any other area of health, the reality is that proportion of mental health budgets is being used in the
the range of possible interventions varies a great deal provision of the least cost-effective interventions, such
with respect to their cost-effectiveness. An analysis of as long-term inpatient treatment of severe mental dis-
500 single and combined interventions assessed by the orders in mental hospitals. Very little is invested in more
WHO-CHOICE project for the prevention and control cost-effective strategies, including the community-based
of noncommunicable diseases and injuries in two LMIC provision of adjuvant psychosocial treatment for severe
regions found that costs differed by at least three orders mental disorders, and measures to reduce access to or
of magnitude (from a few cents to more than US$10 per marketing of alcohol.
capita), as did cost-effectiveness (from US$10 to more Ultimately, policies are enacted and resources
than US$100,000 per healthy life year gained) (Chisholm allocated at the level of individual countries. It is
and others 2012). important that more economic evidence be generated
In the economic analysis for MNS disorders in this alongside clinical trials or other evaluations at the
series, Chisholm and Saxena (2012) found a very sub- national level, rather than relying on international
stantial range of cost-effectiveness, with alcohol control estimates that may lack sensitivity to local priorities
measures, drug treatment for epilepsy, and depression or health system characteristics. Our review high-
treatment identified as offering the best value for money lighted several cost-effectiveness studies from high
in the two WHO subregions assessed (one in Sub-Saharan as well as lower-income country settings to show
Corresponding author: Dan Chisholm, Department of Health System Financing, World Health Organization; chisholmd@who.int.
237
by private providers of variable and sometimes poor of MNS disorders in two distinct geographical and
qualityoften by cutting other household spending and health system contexts: India and Ethiopia. India is a
investment, or by liquidating assets or savingsor go very large, lower-middle-income country in South Asia;
without treatment altogether. Ethiopia is a large, low-income country in East Africa.
Either way, MNS disorders pose a direct threat to We selected these two countries for in-depth analysis
the well-being of households. In India, for example, because both have recently articulated ambitious plans
the National Sample Survey Organization found that to enhance mental health service quality and coverage,
in 2004, national OOP expenditures for treatment of as well as to extend financial protection or health insur-
psychiatric disorders amounted to nearly Rs 7 billion ance for their citizens.
(US$280 million in 2012 US$), half of which was bor-
rowed, and a further 40 percent drawn from household
income or savings (Mahal, Karan, and Engelgau 2010). Extended Cost-Effectiveness Analysis: Principles
Another study, conducted in the Indian state of Goa, and Practice
found that 15 percent of women with common mental Objectives and Components
disorders, such as depression or anxiety, spent more than In addition to health gains, a potential nonhealth ben-
10 percent of household income on health-related care efit of specific interventions or policies, such as public
(Patel and others 2007). financing, is the value that some form of health insur-
The high, potentially catastrophic cost to households ance bestows on households that would otherwise pay
of securing needed health services and goods is a funda- privately for health services and goods. Because OOP
mental concern underlying the drive toward universal spending for the care and treatment of MNS disorders
health coverage (UHC). Direct OOP payments represent can be considerable and enduring, the reduction or
a regressive form of health financingpenalizing those elimination of such expenditures can represent major
least able to afford careand are an obvious channel savings or even financial salvation for affected house-
through which impoverishment may occur or deepen. holds. Public financing of health service costs can also
Prepayment mechanisms, such as national or social increase the use of services, especially for those whose
insurance, more equitably safeguard at-risk populations incomes are so low that they do not access services in
from the adverse financial consequences of mental disor- the first place.
ders. Accordingly, ongoing efforts to move toward UHC Our application of ECEA to MNS disorders focuses
focus on increasing (1) the proportion of the population on public financing as an instrument for financial risk
covered by some form of financial protection; (2) the protection (FRP). Public financing provides FRP ben-
proportion of total costs covered by some form of pre- efits to households by shielding them from the OOP
payment, such as health insurance; and (3) the depth of costs and impoverishment-related consequences of the
coverage (the range of services or interventions available covered health care services (Verguet and others 2015).
to insured persons) (WHO 2010a). Our approach to the measurement of FRP is described
Current coverage of essential health care and treat- in box 13.1.
ment services for MNS disorders is limited, in terms Another essential component of ECEA is its examina-
of access and financial protection or benefit inclusion. tion of the distribution of health and economic benefits
Efforts to scale up community-based public health by population subgroup, for example, by geographical
services for these conditions can contribute strongly to location, care setting, or income quintile. Such an anal-
greater equality of access, because such services will serve ysis enables policy makers to understand how an inter-
more people in need, with less reliance on direct OOP vention or a policy such as public financing would affect
spending. This chapter explores the veracity of this claim different segments of the population, particularly those
through an innovative approach to economic evalua- with low incomes or high vulnerability.
tion called extended cost-effectiveness analysis (ECEA) In short, ECEA provides a tool to amplify under-
(Verguet, Laxminarayan, and Jamison 2015; Verguet and standing of the extent and distribution of health
others 2015). and financial benefits associated with health policies
ECEA goes beyond conventional cost-effectiveness and interventions. Elucidation and enumeration of
analysis (CEA) not only by considering the distribution these benefits provides a more holistic assessment of
of costs and outcomes across different socioeconomic the expected returns on health service investments
groups in the population, but also by explicitly examin- while providing new, evidence-based insights to the
ing the extent to which interventions or policies protect national policy makers responsible for setting prior-
households against the financial risk of medical impov- ities and allocating resources within and beyond the
erishment. We apply this ECEA approach to a range health sector.
Several metrics can be used to quantify the finan- To estimate the FRP, we first estimated the individu-
cial risk protection (FRP) benefits of health pol- als expected income before public financing, which
icies. One approach is to estimate the amount of depends on treatment coverage and associated OOP
households private out-of-pocket (OOP) expen- costs. We then estimated the individuals certainty
ditures averted by the policy; another is to esti- equivalent by assigning individuals a utility function
mate the number of cases of poverty averted by that specifies their risk aversion, which is equivalent
counting the number of individuals no longer to calculating their willingness to pay for insurance
falling under a poverty line/threshold because against the risk of medical expenditures. Finally, we
of substantial OOP medical expenditures. In derived a money-metric value of the insurance pro-
this study, we used as FRP metric the money- vided by public financing (risk premium) as the dif-
metric value of insurance provided by public ference between the expected value of income and the
financing (Verguet, Laxminarayan, and Jamison certainty equivalent (Verguet, Laxminarayan, and
2015), which quantifies insurance risk premiums; Jamison 2015). Aggregating the money-metric value
it reflects risk aversion, in which individuals of insurance with the income distribution of the
would prefer the certainty of insurance over the populationwith a proxy based on the countrys
uncertainty/risk of possible OOP expenditures, gross domestic product per capita and Gini
and hence are willing to pay a certain amount of coefficientyielded a dollar value of FRP at the soci-
money to avoid that risk. etal level.
Application to Mental, Neurological, and Substance health worker care package, and the prevention of com-
Use Disorders mon mental disorders and substance use disorders as
ECEA is applicable to many interventions to prevent part of a school-based intervention package.
or treat MNS disorders, whether considered separately These analyses focus on establishing the distribu-
or in combination. However, since this approach to tional consequences and the value of FRP resulting
economic analysis is new and yet to be tried in the from increased levels of publicly financed interventions.
context of MNS disorders, our first goal was to test its Because the availability and use of mental health services
applicability and assess its internal validity. We accom- in most low- and middle-income countries is very low,
plished this by constructing a series of equation-based however, the economic benefit associated with a switch
ECEA models that employed the same epidemio- from private to public payment for services would
logical and treatment cost-outcome input data used be correspondingly small. Accordingly, we assess the
in previous CEA studies, such as the treatment of impacts of increased FRP and increased service coverage.
psychosis, bipolar disorder, and depression with psy-
chosocial treatment and psychotropic medication,
which Chisholm and Saxena (2012) already examined TOWARD UNIVERSAL HEALTH COVERAGE:
in the contexts of Sub-Saharan Africa and South-East
Asia. Additional information output from the ECEA
TWO COUNTRY ANALYSES
modelparticularly the estimated value of FRP aris- Although analysis has only been conducted for the
ing from public financing of health care costscould two countries presented, the insights and lessons
then be readily interpreted with reference to this ear- from it have a far broader applicability that can be
lier published work. confirmed through further country-based work using
We combined the results of these intervention-specific the methods and models developed for this chapter.
analyses to evaluate the impact of defined packages of Analysis of this kind can be of particular informa-
care. Future applications of the ECEA approach could tional value to other countries planning to reform
focus more on prevention, including the prevention of their mental health programming and public health
childhood behavioral disorders as part of a community financing policies.
Universal Health Coverage for Mental, Neurological, and Substance Use Disorders: An Extended Cost-Effectiveness Analysis 239
India make several simplifying assumptions so that the results
Indias health sector is undergoing a rapid and stark are comparable to the ECEAs presented for schizophre-
transition, not only in epidemiological terms as the nia and depression treatment. For example, treatment-
deaths and disabilities from chronic diseases and injuries seeking costs, such as travel expenses, were omitted. The
take an ever-higher toll, but also in systemic terms as analysis by Megiddo and others (2016) also employs
efforts to improve service quality and expand financial differing government and consumer costs, but here we
protection take effect (Patel and others 2011). In par- assume the costs of a given service to be equal, regardless
ticular, there is a strong push to move toward universal of the purchaser.
public finance (UPF)the government finances an Prevalence and other epidemiological parameters
intervention irrespective of who is delivering or receiv- came from the Global Burden of Disease (GBD) 2010
ing itto reverse decades of high, often impoverishing study estimates for South Asia (Whiteford and others
OOP health care expenditures and to allocate resources 2013). For calculation of healthy life-years, we applied
more equitably. the following disability weights: 0.072 for seizure-free
This subsection estimates the expected health and patients, 0.319 for patients with seizures, and 0.420 for
economic benefits of scaling up services for the treat- untreated individuals with epilepsy (IHME 2012). For
ment of three prominent contributors to the burden of each scenario, we estimated the policys impact on pop-
MNS disorders: epilepsy, schizophrenia, and depression. ulation health (healthy life-years gained), direct govern-
All monetary values are expressed in 2012 US$. ment expenditures, OOP expenditures averted, and the
FRP provided.
Enhanced Financial and Service Coverage of Epilepsy The results, presented in table 13.1, relate to a popu-
Treatment lation of one million persons in the general population,
Fewer than half of the estimated 6 million to 10 million divided into equal household income quintiles of 200,000
individuals with epilepsy in India receive any treatment persons. The model is dynamic, and the values change
(Meyer and others 2010). To counter this health and over time (meaning that the data for each point in time
financial burden, the Ministry of Health is considering a are needed to replicate the results exactly): here we pre-
national epilepsy program that could increase access to, sent the results for the average year. The estimated disease
and utilization of, treatment through three interventions burden associated with epilepsy amounts to 2,200 lost
(Tripathi and others 2012): public awareness campaigns, years of healthy life per one million population. Current
better training of health workers, and UPF for first- and intervention efforts lead to 503 healthy life-years gained
second-line anti-epilepsy drugs (AEDs) and epilepsy (23 percent of the total estimated disease burden); the
surgery. The ECEA that follows examines UPFa policy three enhanced-coverage intervention scenarios result
intervention that would also address the financial risk in gains of between 1,118 and 1,251 healthy life-years,
posed by OOP spending on epilepsy treatment. The equivalent to more than 50 percent of the measured
incremental impacts of three UPF interventions were disease burden. Public financing of second-line AEDs
assessed: UPF for first-line AEDs (intervention 1); UPF as well as first-line AEDs to 80 percent of those in
for first- and second-line AEDs (intervention 2); and need (intervention 2) generates 90 more healthy life-
UPF for first- and second-line AEDs and epilepsy sur- years than intervention 1 alone; the addition of surgery
gery (intervention 3). (intervention 3) adds a further 44 healthy life-years per
First-line AEDs include carbamazepine, phenytoin, and one million population. Intervention health benefits are
valproate, as well as phenobarbital; the second-line AED distributed equitably across income quintiles.
is lamotrigine. Seventy percent of patients are expected The total cost of implementing intervention 1 is
to respond to first-line AEDs; the remaining 30 percent US$0.16 per capita, rising to US$0.30 for intervention
are allocated equally to three groups: those receiving 3 (table 13.1). Compared with no intervention, the cost
second-line AED treatment, those receiving surgery, and per healthy life-year gained for all three intervention
refractory cases who do not respond to any treatment. scenarios falls below US$200 (range: US$112US$181).
Each intervention increases access to the treatment Relative to the current situation, the incremental cost-
provided by UPF to 80 percent (from less than 50 percent effectiveness of intervention 1 is US$70 per healthy life-
without UPF). We estimate that 70 percent of all treat- year gained; intervention 3 is the next most cost-effective
ment costsincluding outpatient visits, inpatient visits, (incremental cost-effectiveness ratio US$850).
and drugsare paid OOP in the baseline and that the UPF coverage would avert more than US$100,000
interventions reduce OOP expenditures for the covered in OOP expenditures per one million population
services to zero. Relative to the full model and detailed under intervention 1, and US$190,000 and US$208,000
results presented by Megiddo and others (2016), we under interventions 2 and 3, respectively. Finally, the
monetized value of insurance was found to amount to Enhanced Financial and Service Coverage of
US$11,000 per one million population for interventions Schizophrenia Treatment
2 and 3, with evidence of a clear trend for it to decrease Schizophrenia poses a considerable public health
with wealth. For example, the poorest quintile derives and social policy challenge because of its severity, its
37 percent of the total insurance value, compared with often catastrophic effect on the welfare and income of
8 percent for the wealthiest. family members, and the significant risk that patients
The primary conclusion from this analysis is that will suffer severe human rights violations. Here we
intervention 1 is the most cost-effective and least costly analyze the impact of enhanced public financing and
strategy to implement from a public payer perspec- provision of schizophrenia treatment on health and
tive, but intervention 3increased service and finan- financial outcomes, including increased uptake of
cial coverage of first- and second-line AEDs, as well as treatment (leading to more health gains), reduced
surgerywould generate the greatest level of health gain OOP treatment costs, and greater insurance against
and offer the greatest level of financial protection at the catastrophic health expenses (Raykar, Nigam, and
population level. Chisholm 2015).
Universal Health Coverage for Mental, Neurological, and Substance Use Disorders: An Extended Cost-Effectiveness Analysis 241
In this model, all persons treated for schizophrenia 30 percent in the poorest income group to 50 percent
in nonspecialized health care settings receive a combi- in the richest). Target coverage for all income groups
nation of first-generation antipsychotic drugs, such as was set at 80 percent, meaning that 80 percent of those
haloperidol or chlorpromazine, as well as basicor, for needing treatment would receive publicly financed care.
a small proportion, intensivepsychosocial treatment. Schizophrenia prevalence rates for South Asia were
Fifteen percent of cases are expected to require short- taken from the GBD 2010 study (Whiteford and others
term inpatient psychiatric care; 2 percent are assumed to 2013), stratified by region, age, and gender, but not by
be long-term residential patients in community-based income. To derive prevalence rates by income group,
facilities; and 50 percent receive hospital outpatient care these estimates were applied to the household survey in
(Chisholm and others 2008). India (District Level Household and Facility Survey-3);
The resulting cost per treated case is US$177 per this showed a higher prevalence among higher-income
year. Given that OOP spending as a share of total health groups, which could reflect better detection, greater
expenditure amounts to at least 70 percent for noncom- health service uptake, or both. Disability weights, which
municable diseases in India (Mahal, Karan, and Engelgau are necessary for the calculation of healthy life-years lost
2010), we estimate that the annual expected cost to or gained, are 0.576 and 0.756 for residual and acute
households would be US$124. Treatment improves the cases, respectively (IHME 2012). A composite disability
average level of functioning or disability by an estimated weight of 0.612 was used, based on a weighted average of
24 percent (Chisholm and others 2008); adherence to acute (20 percent) and residual (80 percent) cases.
treatment was set at 76 percent (Chatterjee and others The results, displayed in table 13.2, indicate that the
2014). The estimated proportion of total cases currently current public health burden of schizophrenia amounts
receiving treatment in India is 40 percent (Murthy to 1,700 lost healthy life-years per one million popula-
2011), to which we applied a socioeconomic gradient tion. Treatment of schizophrenia with a combination
to account for increased detection and health care uti- of psychosocial treatment and antipsychotic medica-
lization rates among wealthier groups (ranging from tion generates 126 healthy life-years at current levels of
Table 13.2 Extended Cost-Effectiveness Analysis of Publicly Financed Schizophrenia Treatment in India
Income quintile
Total (per one
Outcome I II III IV V million persons)
Averted disease burdena
Current burden (healthy life-years lost) 307 316 333 354 394 1,704
Current-coverage averted burden (healthy life-years 17 20 24 29 36 126
gained)
Target-coverage averted burden (healthy life-years 45 46 49 52 57 249
gained)
Cost of care ($)b
Current-coverage total costs 26,721 32,042 38,666 46,156 57,059 200,644
Current-coverage private expenditures averted 18,705 22,429 27,066 32,309 39,942 140,451
(under UPF)
Target-coverage total costs 71,257 73,238 77,331 82,055 91,295 395,176
Target-coverage private expenditures averted 49,880 51,267 54,132 57,439 63,906 276,623
(under UPF)
Insurance value ($)c 7,282 5,587 4,972 4,302 2,439 24,582
Source: Raykar, Nigam, and Chisholm 2015.
Note: UPF = universal public financing for 80 percent of the population in need. Results are based on a population of one million people, with intervention benefits equally divided
among income quintiles of 200,000 persons each (quintile I having the lowest household income and quintile V the highest). Target coverage of UPF for schizophrenia treatment for
all income groups was set at 80 percent. All monetary values are expressed in 2012 US$.
a. The estimated disease burden, expressed as healthy life-years lost or gained, is drawn from the Global Burden of Disease 2010 study for South Asia (Whiteford and others 2013).
b. Total costs = (direct government expenditures) + (private expenditures, including out-of-pocket costs).
c. Insurance value = financial risk protection provided, based on current coverage.
Figure 13.1 Distribution of Public Spending and Insurance Value of UPF for Schizophrenia Treatment in India, by Income Quintile
7,000
25
6,000
Public health spending (%)
20
5,000
15 4,000
3,000
10
2,000
5
1,000
0 0
I II III IV V I II III IV V
Income quintile Income quintile
Current distribution (without UPF)
Revised distribution (with UPF)
Universal Health Coverage for Mental, Neurological, and Substance Use Disorders: An Extended Cost-Effectiveness Analysis 243
be close to US$35 (Chisholm and Saxena 2012; Patel and coverage, which is skewed in favor of the richer quintiles.
others 2011), of which 70 percent (US$25) is projected The total cost of providing this elevated level of service
to be paid by households. Treatment affects the dura- coverage approaches US$700,000 per one million pop-
tion of a depressive episode and is expressed here as an ulation per year, or US$0.70 per head of population,
improvement in the remission rate by 35 percent, sub- compared with US$0.28 now. Publicly financing this
sequently adjusted downward to reflect expected rates scaled-up treatment will avert more than US$477,000
of nonadherence of 70 percent (Chisholm and Saxena of OOP spending per one million population, shared
2012). We modeled the impact of moving from current fairly equally among income quintiles. The overall
coverage (ranging from an estimated 10 percent for the insurance value is approximately US$5,400, much lower
lowest-income quintile to 30 percent for the highest) to than that of schizophrenia treatment because of the
a target coverage of 50 percent for all income groups. lower coverage rate and cost of treatment, and also
As shown in table 13.3, the public health burden of much flatter (there is no clear income gradient between
depression is considerable (more than 14,000 healthy quintiles IIV).
life-years lost per one million population). At current
coverage rates in the population, treatment is estimated Combination Package
to generate 729 healthy life-years (equivalent to only Combining the results of these analyses of UPF for the
5 percent of current disease burden) per million pop- treatment of epilepsy, schizophrenia, and depression,
ulation. With coverage scaled up to 50 percent, close to several findings become apparent. First, over 90 percent
1,800 healthy life-years would be gained, equivalent to of the total avertable burden of disease, in healthy life-
12 percent of the current disease burden; as a proportion years gained per one million population, is attributable
of current burden, the impact is similar to that of schizo- to UPF of treatment for depression and epilepsy; UPF of
phrenia treatment, but because of the higher prevalence treatment for schizophrenia accounts for only 7 percent
of depression, the absolute amount of avertable health of the 3,683 healthy life-years. Second, UPF for treat-
gain in the population is at least five times greater. ment of depression also accounts for the greatest share
As in the case of schizophrenia treatment, health ben- of averted OOP spending at specified target-coverage
efits are distributed much more evenly across income levelshalf in this instance (US$477,000 of a total of
groups at the assumed scaled-up coverage level of US$962,000 per one million population). Both of these
50 percent among all income groups than under current findings reflect the larger number of prevalent cases
Table 13.3 Extended Cost-Effectiveness Analysis of Publicly Financed Depression Treatment in India
Income quintile Total
(per one
million
Outcome I II III IV V persons)
a
Averted disease burden
Current-coverage burden (healthy life-years lost) 2,754 2,817 2,914 2,996 3,153 14,633
Current-coverage averted burden (healthy life-years gained) 67 104 143 184 232 729
Target-coverage averted burden (healthy life-years gained) 337 345 357 367 386 1,793
Cost of care ($) b
Current-coverage total costs 25,669 39,385 54,318 69,821 88,178 277,371
Current-coverage private expenditures averted (under UPF) 17,968 27,569 38,023 48,875 61,725 194,160
Target-coverage total costs 128,346 131,282 135,795 139,642 146,964 682,028
Target-coverage private expenditures averted (under UPF) 89,842 91,897 95,056 97,750 102,875 477,420
c
Insurance value ($) 1,101 1,167 1,232 1,183 717 5,400
Note: UPF = universal public financing for 50 percent of the population in need. Results are based on a population of one million people, with intervention benefits equally divided
among income quintiles of 200,000 persons each (quintile I having the lowest household income and quintile V the highest). Target coverage of UPF for depression treatment for all
income groups was set at 80 percent. All monetary values are expressed in 2012 US$.
a. The estimated disease burden, expressed as healthy life-years lost or gained, is drawn from the Global Burden of Disease 2010 study for South Asia (Whiteford and others 2013).
b. Total costs = (direct government expenditures) + (private expenditures, including out-of-pocket costs).
c. Insurance value = financial risk protection provided, based on current coverage.
Figure 13.2 Composite Value of Insurance through UPF for Treatment of MNS Disorders in India, by Illness and Income Quintile
10,000
9,000
Value of insurance per 1 million
8,000
population (2012 US$)
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
I II III IV V
Income quintile
Note: MNS = mental, neurological, and substance use; UPF = universal public finance. Value of insurance = financial risk protection provided at current coverage. Results are based on a population
of one million people, equally divided into income quintiles of 200,000 persons each (quintile I having the lowest household income and quintile V the highest). Results assume target coverage
levels of 80 percent for all income groups.
Universal Health Coverage for Mental, Neurological, and Substance Use Disorders: An Extended Cost-Effectiveness Analysis 245
CEA of the National Mental Health Strategy (Bjerkreim depression and epilepsy. The costs and health benefits
Strand and others 2015). The selected interventions of the intervention package are estimated to be higher
include phenobarbital for epilepsy, fluoxetine combined for the lowest-income groups (table 13.4) based on
with cognitive therapy and proactive case management the higher prevalence and treatment gap among those
for depression, valproate combined with psychosocial groups. Similarly, the measured value of insurance
therapy for bipolar affective disorder, and first-line anti- is highest among the lowest-income group. Although
psychotic medication (haloperidol or chlorpromazine) UPF would reduce household private expenditures for
plus psychosocial treatment for schizophrenia. those with current access to care, the averted OOP
As with the Indian analyses, the ECEA splits the pop- expenditures would be extremely low, given the very low
ulation into five income quintiles and runs the analyti- current access to and coverage of treatment services (less
cal model for each income group with quintile-specific than 5 percent), particularly among the lower-income
prevalence rates. The average age-specific disease preva- quintiles (Bjerkreim Strand and others 2015). In other
lence rates used in the standard CEA (Bjerkreim Strand words, the FRP of UPF is extremely low because of the
and others 2015) were distributed into income-quintile- low current level of private spending on mental health
specific prevalence rates, using a population-based care in Ethiopia, a direct consequence of the very low
prevalence study conducted in Ethiopia (n = 1,497) coverage of services.
(Fekadu and others 2014).1 Disease-specific mortality, Findings from this ECEA indicate that investing
intervention coverage, and intervention effectiveness in UPF of public mental health will create substantial
were held constant in each income group. Estimates of health benefits, but it will most likely produce a low
the efficacy of interventions were drawn from system- degree of FRP. Accordingly, while the ECEA approach
atic reviews, meta-analyses, and randomized controlled captures FRP and equity in the economic evaluation of
trials (full details can be found in Bjerkreim Strand and mental health policy, the FRP benefits are less relevant
others 2015). when the current utilization and spending on care is
Current treatment coverage for all disorders is less low, as they are in Ethiopia. Nevertheless, we expect that
than 5 percent (Bjerkreim Strand and others 2015). many families experience impoverishing loss of income
Following the introduction of UPF, and in line with the because of mental disorders.
National Mental Health Strategy, coverage for all income
groups is modeled to reach 75 percent for treatment of Productivity Impact of Scaled-Up Depression
schizophrenia and epilepsy, 50 percent for treatment of Treatment
bipolar disorder, and 30 percent for treatment of depres- Owing to low levels of current investment, OOP spend-
sion (Federal Democratic Republic of Ethiopia 2012). ing averted and FRP conferred as a result of switching
Target coverage for depression is lower than the other to a publicly financed model of mental health care
disorders because of its higher prevalence and lower are modest. However, implementation of the National
detectability. Mental Health Strategy can lead to other important wel-
A significant proportion of total health spending in fare gains, in particular, productivity at the household
Ethiopia is from OOP expenditures, varying between 30 and societal levels.
and 40 percent of the total over the past 10 years (World Therefore, we also explored the expected productiv-
Bank 2014). This analysis assumes a current household ity gains from scaling up the provision of depression
contribution of 34 percent toward the cost of treatment; care and treatment. We focused only on depression
the government covers the remaining 66 percent. To because the disease burden of depression is high,
estimate the amount of household OOP expenditures and evidence indicates that depression has a substan-
averted by UPF, we quantified what households would tial impact on productivity (Clark and others 2009;
pay for illness-related treatment cost at current service Goetzel and others 2004). Between 1 and 3 percent of
delivery levels. the adult Ethiopian population is estimated to have a
For the country as a whole, which had a population of depressive episode at any given time, with an average
94.6 million in 2012 (United Nations 2015), the expected duration of 8.4 months (Bjerkreim Strand and others
annual cost of implementing the defined mental and 2015). Productivity is lost during such episodes because
neurological health care package at specified target cov- of increased absence from work (absenteeism) and
erage levels is approximately US$153 million, equivalent decreased work performance when present at work
to a little more than US$1.60 per capita (Johansson and (presenteeism). Depression treatment programs have
others 2015). The return on this investment, in total been shown to improve rates of employment by up
population health gain, exceeds 155,000 healthy life- to 5 percent in the United Kingdom (Clark and oth-
years, the majority of which derives from treatment of ers 2009); in the United States, costs associated with
presenteeism have been estimated to be higher than the Disability days (per month) because of depression are
costs of treatment (Goetzel and others 2004). estimated to be 2.9 in low-income settings (Alonso
To estimate the productivity impact across income and others 2011). Hence, we assumed treatment would
groups from scaling up treatment of depression in reduce the number of disability days by 8.7 days in total
Ethiopia, we first adapted the Goetzel and others (2004) (2.9 days * 2.9 months). Subsequently, the population
approach to presenteeism to the context of Ethiopia. We with depression, target coverage (30 percent), and aver-
used epidemiological, demographic, efficacy, and cost age daily income (per wealth quintile in the productive
data from the contextualized CEA of mental health care age groups [ages 1560 years]) were multiplied by this
in Ethiopia by Bjerkreim Strand and others (2015). It change in absenteeism (8.7 days) to derive an estimate
was estimated that treatment led to an average reduction of the potential productivity gains in Ethiopia. In addi-
in the duration of a depressive episode of 2.9 months tion, persons with depression have been found to have
(8.4 months * efficacy of 0.35). Second, this reduction 3.7 days with partial disability per month in low-income
in duration was converted to reduction in absenteeism. countries (Bruffaerts and others 2012). Partial disability
Universal Health Coverage for Mental, Neurological, and Substance Use Disorders: An Extended Cost-Effectiveness Analysis 247
means that on-the-job productivity is reduced because by income group, for example, provides an important
of disease; it was estimated that patients with depression equity dimension that has so far been largely absent from
had 1.2 full days lost per month because of presentee- conventional economic evaluation methods (including
ism, based on the assumption that each partial day is the WHOs CHOICE [CHOosing Interventions that are
equivalent to one-third of a full lost day. Subsequently, Cost-Effective] project and earlier editions of Disease
the associated productivity gain was estimated using the Control Priorities). Identification of the averted OOP
same method as for absenteeism.2 spending associated with a move to UPF usefully com-
The results shown in table 13.5 indicate that scaled-up plements other research related to UHC, such as estima-
depression treatment at 30 percent coverage could lead tion of the costs of scaling up services.
to total productivity gains of close to US$40 million We found ECEA to be a feasible approach and a use-
per year. The largest benefits accrue to the wealthier ful addition to the methodological toolbox available to
quintiles because of their higher average income level analysts, particularly since it can be incorporated into
(Johansson and others 2015). Our estimates indicate existing cost-effectiveness modeling frameworks. The
that the expected productivity gain from scaled-up treat- main additional data requirement is to be able to break
ment of depression is likely to reduce the expected gov- down epidemiological and other key input parameters
ernmental cost of the treatment program by 71 percent. by income group, the source of which would typically
We acknowledge that it is problematic to apply a be nationally representative demographic and health
high-income country method to an agrarian economy surveys. Static and more dynamic approaches to ECEA
like Ethiopia to estimate productivity losses. Nevertheless, modeling have been developed and employed; for MNS
calculations of productivity impact, based on presentee- disorders with long-term impacts, or for other inter-
ism and absenteeism, are applied to illustrate how such ventions, a dynamic, agent-based approach to modeling
information may be an important supplement to infor- can be used that requires more data as well as analytical
mation on the expected FRP of mental health care in a expertise, but may be better able to capture socio-
low-income context. Appropriate measures of presen- demographic changes and disease interactions over time.
teeism and absenteeism need to be contextualized and Whichever approach is used, both are subject to
found for each particular setting. More conceptual and the inherent uncertainty surrounding population-level
empirical work on this issue is needed. projections of intervention costs, impacts, and con-
sequences, consideration of which is contained in the
primary analyses underlying the base case findings
CONCLUSIONS AND RECOMMENDATIONS reported in this chapter (Johansson and others 2015;
This chapter employed a novel approach to the eco- Megiddo and others 2016; Raykar, Nigam, and Chisholm
nomic analysis of mental health care interventions, with 2015). These uncertainty analyses indicate that results
a view to gaining insights into intervention or policy for FRPas well as overall costs and health effectsare
impacts other than health gain itself. Assessment of the sensitive to assumptions around target coverage rates to
health and nonhealth impacts of scaled-up treatment be achieved in the population, the proportion of total
Income quintile
Total
Cost/outcome I II III IV V population
Government cost of depression treatment program ($, millions) 15.1 13.2 11.2 9.3 7.3 56.1
a
Productivity gain from scaled-up depression treatment ($, million)
Caused by absenteeism 3.0 4.9 5.9 6.6 7.9 28.3
Caused by presenteeism 1.2 2.0 2.4 2.7 3.3 11.6
b
Net societal cost of depression treatment program ($, million) 10.9 6.3 2.9 0.0 3.9 16.2
Source: Johansson and others 2015.
Note: Results are based on the total Ethiopian population, with intervention costs equally divided among income quintiles of the population (quintile I having the lowest household
income and quintile V the highest). All monetary values are expressed in 2012 US$.
a. Total societal income/wealth in productive ages (1560 years) (2012) in Ethiopia is US$879: by quintile (Q), US$281 for QI, US$536 for QII, US$772 for QIII, US$1,072 for QIV, and
US$1,732 for QV.
b. Net societal cost = (governmental cost) (productivity gain).
Universal Health Coverage for Mental, Neurological, and Substance Use Disorders: An Extended Cost-Effectiveness Analysis 249
(US$1.21 per capita in Ethiopia and US$1.37 in India). Low-income countries (LICs) = US$1,045 or less
Furthermore, a UPF policy can lead to a more equita- Middle-income countries (MICs) are subdivided:
ble allocation of public health resources across income a) Lower-middle-income = US$1,046US$4,125
groups, and benefit the lowest-income groups most in b) Upper-middle-income (UMICs) = US$4,126US$12,745
High-income countries (HICs) = US$12,746 or more.
terms of the value of insurance, used here as a measure
of financial protection: the poorest 40 percent of house- 1. For each disorder, based on data extracted from Fekadu
holds receive over 50 percent of the combined value of and others (2014), we extract a prevalence ratio between
insurance in India, and 76 percent in Ethiopia. income quintiles using a risk index by income quintile (Q)
It should be pointed out, however, that because exist- (QI, 1.4; QII, 1.2; QIII, 1; QIV, 0.8; and QV, 0.6) applied
ing treatment coverage is low (especially in Ethiopia, to the mean prevalence of each disorder (Johansson and
where it is 5 percent or less), averted OOP expenditures others 2015).
arising from a switch to public finance of treatment 2. The total gain in productivity by wealth quintile i due to
costs will be correspondingly low (table 13.6). This again absenteeism averted is given by: Prod_Ai = AP * Incomei *
points to the substantial shortage of appropriate mental Durdis * Eff * Popi * Cov, where AP is the number of days of
absenteeism prevented (8.7 days); Incomei is the average
health services in Ethiopia. It should also be noted that
daily income in each wealth quintile i; Durdis is the average
private expenditures on complementary or traditional
duration of a depressive episode (8.4 months); Eff is the
remedies would not be covered by such public financing, efficacy of the intervention (SSRI + cognitive therapy +
and this might continue to be a significant drain on the proactive case management = 0.35); Popi is the number
income or resources of some household groups. of people with depression in each wealth quintile i; and
Only when a substantial increase in service cov- Cov is the target coverage of treatment (0.30). The total
erage is modeled does one see the true scale of the gain in productivity by wealth quintile i due to presentee-
private expenditures that would pertain in the absence ism averted is given by: Prod_Pi = PP * Incomei * Durdis
of UPF. It is vital that increased financial protection * Eff * Popi * Cov, where PP is the number of full days of
goes hand in hand with enhanced coverage of an presenteeism prevented by going from depressed to non-
essential package of care. Improved service access depressed (1.2); and the other variables are identical to
those in Prod_Ai. The estimated annual number of people
without commensurate financial protection will lead
with depression (ages 1560 years) per quintile (Q) is QI,
to inequitable rates of service uptake and outcomes,
900,000; QII, 771,000; QIII, 641,000; QIV, 511,000; and
but improved financial protection without appropriate QV, 381,000.
service scale-up will bring little improvement at all. In
short, a concerted, multidimensional effort is needed
if the much-needed move toward UHC for MNS REFERENCES
disorders is to be realized.
Alonso, J., M. Petukhova, G. Vilagut, S. Chatterji, S. Heeringa,
and others. 2011. Days Out of Role Due to Common
Physical and Mental Conditions: Results from the WHO
ACKNOWLEDGMENT World Mental Health Surveys. Molecular Psychiatry 16:
The authors thank Ramanan Laxminarayan for his comments 123446.
on an earlier draft of this chapter. Bjerkreim Strand, K. B., D. Chisholm, A. Fekadu, and
K. A. Johansson. 2015. Scaling-Up Essential Neuropsychiatric
Services in Ethiopia: A Cost-Effectiveness Analysis. Health
Policy and Planning doi:10.1093/heapol/czv093.
NOTES Bruffaerts, R., G. Vilagut, K. Demyttenaere, J. Alonso,
Disclaimer: Dan Chisholm is a staff member of the World A. Alhamzawi, and others. 2012. Role of Common Mental
Health Organization. The authors alone are responsible for the and Physical Disorders in Partial Disability around the
views expressed in this publication, and they do not necessarily World. British Journal of Psychiatry 200: 45461.
represent the decisions, policy, or views of the World Health Chatterjee, S., S. Naik, S. John, H. Dabholkar, M. Balaji,
Organization. and others. 2014. Effectiveness of a Community-Based
This chapter was previously published as Health and Intervention for People with Schizophrenia and Their
Economic Benefits of Public Financing of Epilepsy Treatment Caregivers in India (COPSI): A Randomized Controlled
in India: An Agent-Based Simulation Model. I. Megiddo, A. Trial. The Lancet 383 (9926): 138594.
Colson, D. Chisholm, T. Dua, A. Nandi, and R. Laxminarayan. Chisholm, D., O. Gureje, S. Saldivia, M. V. Calderon,
Epilepsia. 2016. Epub January 14. doi:10.1111/epi.13294. R. Wickremasinghe, and others. 2008. Schizophrenia
World Bank Income Classifications as of July 2014 are as Treatment in the Developing World: An Interregional and
follows, based on estimates of gross national income (GNI) Multinational Cost-Effectiveness Analysis. Bulletin of the
per capita for 2013: World Health Organization 86 (7): 497576.
Universal Health Coverage for Mental, Neurological, and Substance Use Disorders: An Extended Cost-Effectiveness Analysis 251
DCP3 Series Acknowledgments
Disease Control Priorities, third edition (DCP3) We thank the many contractors and consultants
compiles the global health knowledge of institu- who provided support to specific volumes in the form
tions and experts from around the world, a task that of economic analytical work, volume coordination,
required the efforts of over 500 individuals, includ- chapter drafting, and meeting organization: the Center
ing volume editors, chapter authors, peer reviewers, for Disease Dynamics, Economics & Policy; Centre
advisory committee members, and research and staff for Chronic Disease Control; Centre for Global Health
assistants. For each of these contributions, we convey Research; Emory University; Evidence to Policy Initiative;
our acknowledgment and appreciation. First and Public Health Foundation of India; QURE Healthcare;
foremost, we would like to thank our 33 volume University of California, San Francisco; University of
editors who provided the intellectual vision for their Waterloo; University of Queensland; and the World
volumes based on years of professional work in their Health Organization.
respective fields, and then dedicated long hours to We are tremendously grateful for the wisdom and
reviewing each chapter, providing leadership and guidance provided by our advisory committee to the
guidance to authors, and framing and writing the editors. Steered by Chair Anne Mills, the advisory com-
summary chapters. We also thank our chapter authors mittee assures quality and intellectual rigor of the high-
who collectively volunteered their time and expertise est order for DCP3.
to writing over 160 comprehensive, evidence-based The National Academy of Medicine, in collaboration
chapters. with the InterAcademy Medical Panel, coordinated the
We owe immense gratitude to the institutional spon- peer-review process for all DCP3 chapters. Patrick Kelley,
sor of this effort: The Bill & Melinda Gates Foundation. Gillian Buckley, Megan Ginivan, and Rachel Pittluck
The Foundation provided sole financial support of managed this effort and provided critical and substan-
the Disease Control Priorities Network. Many thanks tive input.
to Program Officers Kathy Cahill, Philip Setel, Carol The World Bank External and Corporate Relations
Medlin, and (currently) Damian Walker for their Publishing and Knowledge division provided excep-
thoughtful interactions, guidance, and encouragement tional guidance and support throughout the demanding
over the life of the project. We also wish to thank production and design process. We would particularly
Jaime Seplveda for his longstanding support, including like to thank Carlos Rossel, the publisher; Mary Fisk,
chairing the Advisory Committee for the second edition Nancy Lammers, Rumit Pancholi, and Deborah Naylor
and, more recently, demonstrating his vision for DCP3 for their diligence and expertise. Additionally, we thank
while he was a special advisor to the Gates Foundation. Jose de Buerba, Mario Trubiano, Yulia Ivanova, and
We are also grateful to the University of Washingtons Chiamaka Osuagwu of the World Bank for providing
Department of Global Health and successive chairs King professional counsel on communications and marketing
Holmes and Judy Wasserheit for providing a home base strategies.
for the DCP3 Secretariat, which included intellectual Several U.S. and international institutions contrib-
collaboration, logistical coordination, and administra- uted to the organization and execution of meetings that
tive support. supported the preparation and dissemination of DCP3.
253
We would like to express our appreciation to the reproductive and maternal health volume consulta-
following institutions: tion, November 2013)
National Cancer Institute, cancer consultation
University of Bergen, consultation on equity (June (November 2013)
2011) Union for International Cancer Control, cancer con-
University of California, San Francisco, surgery sultation (November 2013, December 2014)
volume consultations (April 2012, October 2013,
February 2014) Carol Levin provided outstanding governance for cost
Institute of Medicine, first meeting of the Advisory and cost-effectiveness analysis. Stphane Verguet added
Committee to the Editors (March 2013) invaluable guidance in applying and improving the
Harvard Global Health Institute, consultation on extended cost-effectiveness analysis method. Shane Murphy,
policy measures to reduce incidence of noncommu- Zachary Olson, Elizabeth Brouwer, Kristen Danforth, and
nicable diseases (July 2013) David Watkins provided exceptional research assistance
Institute of Medicine, systems strengthening meeting and analytic assistance. Brianne Adderley ably managed
(September 2013) the budget and project processes. The efforts of these indi-
Center for Disease Dynamics, Economics & Policy viduals were absolutely critical to producing this series, and
(Quality and Uptake meeting, September 2013; we are thankful for their commitment.
Tarun Dua
SERIES EDITORS
Tarun Dua is a Medical Officer working in the Evidence,
Research and Action on Mental and Brain Disorders Dean T. Jamison
unit in the Department of Mental Health and Substance Dean T. Jamison is a Senior Fellow in Global Health
Abuse at the World Health Organization. Dr. Dua serves Sciences at the University of California, San Francisco,
as the focal point for neurological disorders in the and an Emeritus Professor of Global Health at the
organization. University of Washington. He previously held academic
255
appointments at Harvard University and the University and the International Development Research Centre,
of California, Los Angeles; he was an economist on the among others, in work carried out in over 20 low- and
staff of the World Bank, where he was lead author of the middle-income countries. She led the work on nutrition
World Banks World Development Report 1993: Investing for the Copenhagen Consensus in 2008, when micronu-
in Health. He was lead editor of DCP2. He holds a PhD trients were ranked as the top development priority. She
in economics from Harvard University and is an elected has served as Associate Provost of Graduate Studies at
member of the Institute of Medicine of the National the University of Waterloo, Vice-President Academic at
Academy of Sciences. He recently served as Co-Chair Wilfrid Laurier University in Waterloo, and interim dean
and Study Director of The Lancets Commission on at the University of Toronto Scarborough.
Investing in Health.
Prabhat Jha
Rachel Nugent Prabhat Jha is the Founding Director of the Centre for
Rachel Nugent is a Research Associate Professor in Global Health Research at St. Michaels Hospital and
the Department of Global Health at the University holds Endowed and Canada Research Chairs in Global
of Washington. She was formerly Deputy Director of Health in the Dalla Lana School of Public Health at
Global Health at the Center for Global Development, the University of Toronto. He is Lead Investigator of
Director of Health and Economics at the Population the Million Death Study in India, which quantifies the
Reference Bureau, Program Director of Health and causes of death and key risk factors in over two mil-
Economics Programs at the Fogarty International lion homes over a 14-year period. He is also Scientific
Center of the National Institutes of Health, and senior Director of the Statistical Alliance for Vital Events, which
economist at the Food and Agriculture Organization of aims to expand reliable measurement of causes of death
the United Nations. From 199197, she was Associate worldwide. His research includes the epidemiology and
Professor and Department Chair in Economics at Pacific economics of tobacco control worldwide.
Lutheran University. She has advised the World Health
Organization, the U.S. government, and nonprofit orga-
nizations on the economics and policy environment of Ramanan Laxminarayan
noncommunicable diseases. See the list of Volume Editors.
257
Petra Gronholm Cathrine Mihalopoulos
Centre for Global Mental Health, Institute of Psychiatry, Deakin University, Melbourne, Victoria, Australia
Psychology, and Neuroscience, Kings College London,
London, United Kingdom Maristela Monteiro
Pan American Health Organization, Washington DC,
David Gunnell United States
University of Bristol, Bristol, United Kingdom
Aditi Nigam
Wayne D. Hall Center for Disease Dynamics, Economics & Policy,
Centre for Youth Substance Abuse Research, University Washington, DC, United States
of Queensland, Brisbane, Queensland, Australia
Rachana Parikh
Steven Hyman
Public Health Foundation of India, New Delhi,
Stanley Center for Psychiatric Research, Broad Institute
India
of MIT and Harvard and Department of Stem Cell and
Regenerative Biology, Harvard University, Cambridge, Inge Petersen
Massachusetts, United States University of KwaZulu-Natal, Durban, South Africa
David Jernigan
Michael R. Phillips
Johns Hopkins Bloomberg School of Public Health,
Shanghai Mental Health Center, Shanghai Jiao Tong
Johns Hopkins University, Baltimore, Maryland,
University School of Medicine, Shanghai, China;
United States
Departments of Psychiatry and Global Health,
Nathalie Jette Emory University, Atlanta, Georgia, United States
University of Calgary, Calgary, Alberta, Canada
Martin J. Prince
Kjell Arne Johansson Institute of Psychiatry, Psychology, and Neuroscience,
University of Bergen, Bergen, Norway Kings College London, London, United Kingdom
258 Contributors
James G. Scott John Strang
University of Queensland Centre for Clinical Research, National Addiction Centre, Kings College London,
Brisbane, Queensland, Australia; Metro North Mental London, United Kingdom
Health, Royal Brisbane and Womens Hospital,
Kiran T. Thakur
Brisbane, Queensland, Australia
Columbia University College of Physicians and
Maya Semrau Surgeons, New York, New York, United States
Centre for Global Mental Health, Institute of Psychiatry, Graham Thornicroft
Psychology, and Neuroscience, Kings College London, Centre for Global Mental Health, Institute of Psychiatry,
London, United Kingdom Psychology, and Neuroscience, Kings College London,
Rahul Shidhaye United Kingdom
Public Health Foundation of India, New Delhi, India; Stphane Verguet
CAPHRI School for Public Health and Primary Care, Department of Global Health and Population,
Maastricht University, Maastricht, the Netherlands Harvard T. H. Chan School of Public Health, Boston,
Morton M. Silverman Massachusetts, United States
Suicide Prevention Resource Center, Education Lakshmi Vijayakumar
Development Center, Waltham, Massachusetts, United SNEHA, Voluntary Health Services, Chennai, India;
States, The University of Colorado Denver School of Centre for Youth Mental Health, University of
Medicine, Aurora, Colorado, United States; The Jed Melbourne, Melbourne, Victoria, Australia
Foundation, New York, New York, United States
Theo Vos
Timothy J. Steiner
Institute for Health Metrics and Evaluation, University
Norwegian University of Science and Technology,
of Washington, Seattle, Washington, United States
Trondheim, Norway; Imperial College London,
London, United Kingdom Harvey A. Whiteford
School of Public Health, University of Queensland,
Emily Stockings
Herston, Queensland, Australia; Queensland Centre
National Drug and Alcohol Research Centre, University
for Mental Health Research, Wacol, Queensland,
of New South Wales, Sydney, Australia
Australia; Institute for Health Metrics and Evaluation,
Kirsten Bjerkreim Strand University of Washington, Seattle, Washington,
University of Bergen, Bergen, Norway United States
Contributors 259
Advisory Committee to the Editors
261
Carol Medlin Jaime Seplveda
Senior Health and Nutrition Specialist, Executive Director, Global Health Sciences, University
Health, Nutrition, and Population Global Practice, of California, San Francisco, San Francisco, California,
World Bank, Washington, DC, United States United States
Alvaro Moncayo Richard Skolnik
Researcher, Universidad de los Andes, Bogot, Lecturer, Health Policy Department, Yale School of
Colombia Public Health, New Haven, Connecticut, United States
Jaime Montoya Stephen Tollman
Executive Director, Philippine Council for Health Professor, University of the Witwatersrand,
Research and Development, Taguig City, the Johannesburg, South Africa
Philippines Jrgen Unutzer
Ole Norheim Professor, Department of Psychiatry, University of
Professor, University of Bergen, Bergen, Norway Washington, Seattle, Washington, United States
Folashade Omokhodion Damian Walker
Professor, University College Hospital, Ibadan, Senior Program Officer, Bill & Melinda Gates
Nigeria Foundation, Seattle, Washington, United States
Toby Ord Ngaire Woods
President, Giving What We Can, Oxford, Director, Global Economic Governance Programme,
United Kingdom Oxford University, Oxford, United Kingdom
K. Srinath Reddy Nopadol Wora-Urai
President, Public Health Foundation of India, Professor, Department of Surgery, Phramongkutklao
New Delhi, India Hospital, Bangkok, Thailand
Sevkat Ruacan Kun Zhao
Dean, Ko University School of Medicine, Istanbul, Researcher, China National Health Development
Turkey Research Center, Beijing, China
263
Gregory Simon Steven D. Vannoy
Group Health Research Institute, Seattle, Washington, University of Massachusetts, Boston, Boston,
United States Massachusetts, United States
264 Reviewers
Index
265
childhood disorders resulting in adult disorders, pricing and market regulation, 138
146, 194 primary health care, 1314t, 15
for dementia, 8 prohibition and partial bans, 131, 132t
of suicide and self-harm, 164, 165t, 16667f, 166f quasi-experimental studies, 130
Alcoholics Anonymous, 137 reducing availability of alcohol, 131
alcohol use disorders, 12743 school-based, 19, 139
age of death attributable to, 53, 54t screening and brief interventions, 13536, 136t,
binge drinking, 49, 128 139, 140
burden of disease, 3234t, 12830 self-help and support groups, 13637
challenges for LMICs, 13839 sobriety checkpoints, 134
classification of beverages, 127 specialist health care delivery, 15
consequences, 129 suspension of drivers license, 135
co-occurring disorders and, 4950 taxation, 131t, 13233, 138, 221, 229
cost-effectiveness of interventions for, 19, 20f, warning labels, 135, 139
13738, 13940, 22122, 226, 227t liver cirrhosis and, 50
DALYs and, 129 patterns of, 128
deaths associated with, 7t, 44, 129 prohibition and partial bans, 131, 132t
fetal alcohol syndrome disorders (FASD), 128, public health considerations, 128
135, 139 quasi-experimental studies, 130
GBD findings of excess mortality for recommendations for LMICs, 139
estimated number of cause-specific and excess societal response, 130
deaths, 46t, 49 suicide and, 50, 129, 169, 175
gender differences, 45, 46f unintentional injuries and violence, 50, 129
gender differences, 12829 YLLs and, 4345, 45f, 53, 55, 55t
globalization of alcohol beverages industry, 128 Alzheimers disease. See also dementia
indigenous communities and burden of disease, 3234t, 94
community platform interventions, 135, 136t DALYs and, 95
prohibition, 131, 131t deaths associated with, 7t, 37, 53, 54t
industry role, 12728 goal to identify cure by 2025, 99
interventions for, 13037 pharmacological interventions, 96, 98
advertising bans, 139 amphetamine dependence. See also illicit drug
availability control and licensing of sellers, 138 dependence
blood alcohol concentration (BAC) testing of amphetamine-type stimulants, 109
drivers, 134 burden of disease, 3234t
breath testing of drivers, 134 consumption trends, 111
community platform interventions, 13031, deaths associated with, 7t, 110
135, 136t age of death, 54t
control of unrecorded market, 133, 13839 estimated number of cause-specific and excess
cross-border shopping, 133 deaths for, 46t, 50f
delivery platforms, 14t prevalence in Southeast Asia and Australasia, 37
demand reduction strategies, 229, 229f rates of dependence, 110
driving countermeasures, 13435, 139 YLLs and, 43
education campaigns, 139 Anderson, P., 19
family-based interventions, 135, 139 anorexia nervosa, 3. See also eating disorders
health care platform interventions, 13537 antidepressants. See depression
individual-based, 130 anti-epileptic drugs (AEDs), 88, 90, 91, 240
law enforcement measures to reduce driving antipsychotics
while impaired, 132t for dementia patients, 53, 9697
mass media campaigns, 135 for schizophrenia, 48
medical and social detoxification, follow-up, anxiety disorders, 7071
and referral, 136, 137t, 139 age of onset, 7071
population platform interventions, 130, burden of disease, 3234t, 36, 7071
13135, 131t childhood, 145
pregnant women and, 135, 139 clinical features and course, 70
266 Index
DALYs associated with, 36 planning and consultation with primary health care
deaths associated with, 7t staff in, 210
epidemiological surveys on, 38 Positive Parenting Program (Triple P), 223
epidemiology, 7071 autistic spectrum disorders
estimated number of cause-specific and excess age of cause-specific and excess deaths
deaths for, 46t, 49 attributed to, 47f
gender differences, 68 burden of disease, 3234t, 36
generalized anxiety disorder, 70 co-occurring disorders with, 4849
interventions for, 9t. See also adult mental disorders deaths associated with, 7t
limited access to, 12 estimated number of cause-specific and excess
pharmacologic and psychological treatment, deaths for, 46t, 47f, 4849
7476, 75t interventions for, 9t
primary health care, 1314t, 15
self-care, 15 B
obsessive-compulsive disorder (OCD), 70 Babor, T. F., 135
panic disorder, 70 Baker-Henningham, H., 150
simple phobias, 70 Balanced Care Model, 207
social anxiety disorder, 70 Beijing Suicide Research and Prevention Center, 176
YLDs and, 68, 68f best practice interventions, 4b, 12, 15, 22, 56, 183,
Asia. See also specific countries and regions 184t, 188
amphetamine dependence in, 37 binge drinking, 49, 128
drug users, detention and treatment of, 58 bipolar disorder, 2
illicit substance use in, 111 ADHD and, 49
Aspergers syndrome, 36. See also autistic spectrum adult bipolar disorder, 7172
disorders age of cause-specific and excess deaths attributed
aspirin, 100101 to, 47f
Assessing Cost-Effectiveness (ACE) prevention burden of disease, 32t, 34t, 7172
framework, 188 clinical features and course, 71
Atkins diet, 91 cost-effectiveness of interventions for, 19, 227t
Atlas on Substance Use (WHO), 137 community-based vs. hospital-based services, 228
attention-deficit hyperactivity disorder (ADHD) deaths associated with, 7t
age of occurrence, 146 epidemiology, 7172
bipolar disorder and, 49 estimated number of cause-specific and excess
burden of disease, 3234t, 36 deaths for, 46t, 47f, 48
cost-effectiveness of pharmacological gender differences, 68
interventions, 155 interventions for. See also adult mental disorders
defined, 145 pharmacologic and psychological treatment,
estimated number of cause-specific and excess 7476, 75t
deaths for, 46t, 49 specialist health care delivery, 15
interventions for, 9t, 12 YLDs and, 68, 68f
medications, 154 birth trauma, 185
school-based, 192 blood alcohol concentration (BAC) testing of
Australia drivers, 134
collaborative stepped care approach in, 210 Boussinesq, M., 52
drug dependence in Brazil
court-mandated treatment, 113 childhood mental and developmental disorders in
economic costs, 111 CBT for children with anxiety disorders, 154
interventions, cost of, 118, 119 community-based interventions, 149
indigenous communities and alcohol consumption collaborative stepped care approach in, 210
in, 131 cost-effectiveness of drug therapy for schizophrenia
mental health first aid course in, 188 and depression in, 225
methadone maintenance and buprenorphine primary care interventions in, 208
maintenance in, 226 training of primary care workers, 210
parenting interventions in, 155, 223 specialists training primary health care staff in, 210
Index 267
suicide in, 169, 170, 171, 172 Cause of Death Ensemble Modeling (CODEm), 42
Program for Promotion of Life and Suicide CBT. See cognitive behavioral therapy
Prevention, 176 Central/Eastern Europe and Central Asia
teacher training program to identify and assess alcohol-related deaths in, 44, 55
mental health problems in, 191b cost-effectiveness of interventions in, 221, 227t
breath testing of drivers, 134 suicide of women in, 171
brief psychological intervention YLL rates in, 44, 4546f, 55
for alcohol use disorders, 13536, 136t, 139, 140 child abuse, 68, 146, 148, 152
for drug dependence, 115, 116t, 117 Child and Adolescent Mental Health Policies and Plans
for suicide, 177 (WHO), 148
Building Back Better (WHO), 204 childhood mental and developmental disorders,
bullying, 150, 156, 189 14561. See also attention-deficit hyperactivity
buprenorphine maintenance, 57, 117, 118, 119, 226 disorder (ADHD); autistic spectrum disorders
burden of MNS disorders, 45b, 58, 22, 2940. See also anxiety disorders, 145, 146
mortality rates bullying, 150, 156, 189
adult mental disorders, 6768 burden of disease, 146
alcohol use disorders, 3234t, 12830 consequences of, 14647
childhood mental and developmental disorders, 146 cost-effectiveness of interventions, 15556
depression, 6970 epidemiology, 146
Global Burden of Disease Study 2010 (GBD 2010), gender differences in, 5, 33, 146, 147t
2930. See also Global Burden of Disease interventions for, 910t, 12, 14955, 149b, 156t
Study 2010 child and adolescent mental health policies and
illicit drug dependence, 32, 3234t, 34, 111, 118 plans, 148
implications of study findings, 3637 child protection legislation, 14849
limitations of study and directions for future cognitive behavioral therapy (CBT), 15355, 156
research, 3738 community platform interventions, 14951
methodology of study, 3031 delivery platforms, 13t
neurological disorders, 87 early child development, 14950
overview, 2930 early intervention strategies, importance of, 36
Burundi, integration of mental health care into health care platform interventions, 15154
primary care program, 17b maternal mental health interventions,
15253, 212
C medications for ADHD, 154
Canada medications for conduct disorder, 154
cost of Alzheimers disease treatment in, 98 multisystem therapy, 155
fetal alcohol syndrome (FAS) warning labels in, 135 parenting skills training, 152, 21314
TEAMcare Canada, 213 population platform interventions, 14849
cancer, 41, 48, 49, 56 problem-solving skills therapy (PSST), 155
cannabis dependence. See also illicit drug dependence psychosocial treatments for conduct disorder,
burden of disease, 3234t 15455
cannabis products, 109 school-based interventions, 150. See also
deaths associated with, 7t, 110 education and schools
estimated number of cause-specific and excess screening and community rehabilitation for
deaths for, 46t developmental disorders, 15152, 156
medications for, 118 specialist health care, 15455
rates of dependence, 110 Ten Questions screen, 151, 151b
schizophrenia and, 53, 55t voluntary sector programs, 150
cardiovascular disease, 48, 51, 52, 56, 213 nature of, 146
Carroll, A. E., 150 risk factors for, 14648, 148t
case studies. See also Ethiopia; India trends, 148
scaling up interventions for MNS disorders, types of, 145
1617b, 24 Chile
catastrophic financial effects, 2b CBT depression program in, 222
268 Index
National Depression Detection and Treatment communicable compared to noncommunicable
Program, 1617b, 213 diseases in global burden of disease, 30, 36, 41
postpartum depression interventions in, 153, 2078 community-based care
school-based interventions in, 192 for childhood mental and developmental disorders,
suicide rates in, 169 15152, 156
China compared to hospital level of care, 228
alcohol consumption in, 129 for illicit drug dependence, 11415, 116t
taxation, 133 residential facilities, 1516, 203
violence associated with, 129 community outreach teams, 16, 56, 7778, 79
Central Government Support for the Local community platform interventions, 1314t, 15,
Management and Treatment of Severe 18792, 194
Mental Illnesses Project, 16b adult mental health, 7374
depression, treatment of, 70 alcohol use disorders, 13031, 135, 136t
drowning as premature cause of death in, 52 childhood mental and developmental disorders,
headache interventions in, 102, 208, 228b 14951
suicide in, 164, 168, 170, 172, 175, 177 gender equity and economic empowerment
survey of mental disorders in, 38 interventions, 193
Chisholm, D., 19, 232 identification and case detection, 193
CHOosing Interventions that are Cost-Effective illicit drug dependence, 11314, 115t
(CHOICE) project (WHO), 220, 226, 228, neighborhood groups, 19293, 194
228b, 232 parenting. See parenting interventions
chronic or relapsing course, 1 in schools. See education and schools
cocaine dependence. See also illicit drug dependence suicide and, 17576
age of death attributable to, 54t treatment, care, and rehabilitation, 193
burden of disease, 3234t workplace. See workplace
consumption trends, 111 comparative risk assessments (CRAs), 6, 37, 42,
deaths associated with, 7t 55t, 59
estimated number of cause-specific and excess counterfactual burden and, 44, 5356
deaths for, 46t, 50f competency-based education, 20910, 209t
pharmacotherapies conduct disorders
effectiveness for, 58 burden of disease, 3234t, 36
for psychostimulant dependence, 118 childhood, 145
prevalence in North American and Latin deaths associated with, 7t
America, 37 defined, 145
rates of dependence, 110 estimated number of cause-specific and excess
YLLs and, 43 deaths for, 46t, 49
Cochrane Collaboration review, 207 interventions for, 9t
CODEm (Cause of Death Ensemble Modeling), 42 contingency management approach to drug
cognitive behavioral therapy (CBT) dependence, 117
for adult mental disorders, 76 continuous quality improvement, 214
for childhood mental and developmental disorders, Convention on the Rights of Persons with
15355, 156 Disabilities, 215
for depression, 225 co-occurring disorders, 1, 8, 12, 47
in adolescents, 222 alcohol use disorders and, 49
for illicit drug dependence, 117 autistic spectrum disorders and, 4849
maternal and child health programs, 212 bipolar disorders and, 48
school-based, 192 dementia and, 5253
workplace, 188, 189 epilepsy and, 52
cognitive rehabilitation for dementia, 97, 213 integrating mental health into health programs
collaborative care models, 56, 77, 79, 207 for, 21214
collaborative stepped care, 15, 77, 2079, 213 schizophrenia and, 48
Colombia, cost-effectiveness analysis of coping strategies and well-being, 172
antidepressants in, 225 Corrigan, P. W., 187
Index 269
cost-effectiveness and affordability of interventions, DART-AD (dementia antipsychotic withdrawal trial)
12, 1819, 20f, 21936 trial, 53
for adult mental disorders, 78 decriminalization of suicide, 175
affordability, 22931 Degenhardt, L., 31
alcohol use, demand reduction strategies for, dementia, 9399. See also Alzheimers disease
229, 229f age of death attributable to, 53, 54t
costs of scaling up, 232f premature death, 41
school-based social and emotional learning age of onset, 8
interventions, 22930 burden of disease, 30, 3234t, 87, 9495, 94f, 95
of alcohol-related legislation, 22122 caregiver stress, 95, 97
for alcohol use disorders, 19, 20f, 13738, 13940, coping strategy program, cost-effectiveness
22122 of, 225
for childhood mental and developmental disorders, community health workers detection abilities, 193
15556 co-occurring disorders and, 5253
CHOosing Interventions that are Cost-Effective cost-effectiveness of interventions for, 9899
(CHOICE) project (WHO), 220, 226, 228, cost of, 8
228b, 232 definitions of, 31, 9394
collaborative care models, 56 detection and diagnosis, 96
community-based parenting programs, 22223 early-onset dementia, 94
for dementia, 9899 epidemiology, 9495
economic evaluation of treatment and GBD findings of excess mortality for
prevention, 18b estimated number of cause-specific and excess
for epilepsy, 19, 20f, 58, 90, 93 deaths, 46t, 51f, 5253
extended cost-effectiveness analysis (ECEA), implications, 58
19, 21f, 238 interventions for, 10t, 9698, 186
financial risk protection, 19 capacity of health care teams, 9798
for headache disorders, 102 caregivers, 97
for illicit drug dependence, 118, 120 community-based programs, 98
lack of evidence, 1215, 21 health care delivery interventions, 97
limitation of conventional cost-effectiveness integration into health care, 213
analysis, 220 nonpharmacological interventions, 97
for MNS disorders, 22329 other interventions, 9899
by country, 224f pharmacological interventions, 9697, 98
international studies, 226 specialist health care delivery, 15
national studies, 22426 recommendations for, 103
primary health care, 22427 YLLs and, 43, 45f
nonspecialized treatment settings, 23031 dementia antipsychotic withdrawal trial (DART-AD)
overview, 21920 trial, 53
paucity of trials, 219, 222, 230 Dementia Society of Goa, 207
population and community levels, 22023 demographic factors, 3b
school-based social and emotional learning depression, 6970. See also bipolar disorder;
interventions, 222, 22930 postpartum depression
specialist health care delivery, 22829 age of cause-specific and excess deaths attributed
costs of mental health care, 23738 to, 47f
counseling sessions, 76 alcohol use disorders and, 4950
court-mandated treatment for drug dependence, 112t, 113 antidepressants, 56, 75t, 225, 226
criminal activity related to illicit drugs, 111, 119 burden of disease, 32, 3234t, 34, 36
criminal justice platforms, 112t, 113 chronic illnesses associated with, 68
cross-border shopping for alcohol, 133 clinical features and course, 69
co-occurring disorders with, 47, 56, 69
D cost-effectiveness of interventions for, 19, 20f,
DALYs. See disability-adjusted life years 227t, 230
DARE (Drug Abuse Resistance Education) antidepressants and CBT, 225, 226
program (US), 114 enhanced financial and service coverage, 24344
270 Index
DALYs associated with, 36 driving impaired and traffic accidents, 49, 129, 137
deaths associated with, 6, 7t cost-effectiveness of countermeasures, 221
epidemiological surveys on, 38 countermeasures for, 13435, 139
epidemiology and burden of disease, 6970 helmet laws, 186
estimated number of cause-specific and excess Drug Abuse Resistance Education (DARE) program
deaths for, 46t, 47f, 49 (US), 114
gender differences, 68 drug dependence. See illicit drug dependence
interventions for, 8, 9t, 12, 75t drug education, 114, 115t
collaborative care, 207 drug testing
electroconvulsive therapy (ECT), 74 of offenders, 112t, 113
European Alliance against Depression in workplace, 11314, 115t
Programme, 176 Drummond, M. F., 220
limited access to, 12
primary health care, 1314t, 15, 2078 E
psychosocial interventions for adolescents, 190 early child development, 14950, 19293
self-care, 15 early intervention
specialist health care delivery, 15 drug dependence of at-risk youth, 114
transcranial magnetic stimulation as treatment for psychosis treatment, 77
for, 74 East Asia and Pacific
serotonin-norepinephrine reuptake inhibitors alcohol consumption in
(SNRIs) and, 225 cost-effectiveness of interventions, 137, 221
suicide and, 69, 176 driver testing and arrest, 135
YLDs and, 68, 68f taxation, 133, 137
YLLs and, 49 cost-effectiveness of interventions in, 227t, 239
detoxification suicide prevention organizations in, 176
alcohol use disorders, 136, 137t, 139 suicide rates in, 164
substance abuse, 116, 120 traditional medicine in, 202
developed countries. See high-income countries (HICs) YLLs in, 44, 4546f
developing countries. See low- and middle-income Eastern Europe. See Central/Eastern Europe and
countries (LMICs) Central Asia
developmental disorders, children with. See childhood Eastern Mediterranean Region
mental and developmental disorders suicide in, 170
developmental disorders, people with, 48. See also WHO proposed regional framework in, 2324b
childhood mental and developmental disorders eating disorders, 3, 3234t
diabetes, 41, 56, 128, 129, 184, 213 ECEA. See extended cost-effectiveness analysis
3 Dimensions of Care for Diabetes (UK), 213 economic effects
Diagnostic and Statistical Manual of Mental Disorders of illicit drug dependence, 111
(DSM), 31 of mental, neurological, and substance use (MNS)
DSM-4, 110, 208 disorders, 8
DSM-5, 7172, 110 economic evaluation of treatment and prevention, 18b.
disability-adjusted life years (DALYs) See also cost-effectiveness and affordability of
alcohol use disorders and, 129 interventions
caused by MNS disorders, 5, 3035, 31f, 32t, 34t ECT (electroconvulsive therapy), 74
cost per DALY averted, 18, 19, 20f education and schools
gender differences, 3233, 32t, 3433, 34f, 34t alcohol education campaigns, 139
illicit drug dependence and, 36, 111 early childhood enrichment programs, 19293
disasters and refugees, 177 epilepsy education, 58
Disease Control Priorities in Developing Countries, 2b illicit drug dependence
DisMod-MR, 43, 47, 49, 59 drug education, 114, 115t
disruptive behavioral disorders. See attention-deficit skills training, 114, 115t
hyperactivity disorder (ADHD); conduct mental health awareness, 7374
disorders overdose prevention education, 11415
domestic violence legislation, 186 peer-led education, 202
Dretzke, J., 155 preschool educational programs, 19293
Index 271
school-based interventions, 18992, 194 gender differences in, 5, 33
alcohol use, 19, 139 interventions for, 8, 10t, 8993
childhood mental and developmental alternative therapies, 91
disorders, 150 anti-stigma interventions, 8990
emergency response, 190, 191b helmet laws, 186
HealthWise program (South Africa), 190, 190b legislation, 90
identification and case detection, 19091 management of infectious etiologies, 91
illicit drug dependence, 114, 115t optimizing health care delivery, 9192
information and awareness, 189 pharmacological interventions, 58, 9091
Mental Health First Aid for High School population platform interventions, 89
Teachers, 191 primary health care, 1314t, 15
social and emotional learning interventions, self-management, 90, 203
18990, 222, 22930 surgical management, 91, 93
suicide and self-harm, 176 treatment gap, 58, 92f, 93b
teacher training program, 191b mental illness and, 52
treatment, care, and rehabilitation, 192 recommendations for, 103
for vulnerable children, 190 status epilepticus, 88
whole-of-school approaches, 150 suicide and, 52
Egypt YLDs and, 90
childhood mental and developmental disorders, YLLs and, 4345, 45f
community-based interventions in, 149 Ethiopia
suicide in, 170 alcohol use, demand reduction strategies for, 229
Eickmann, S. H., 149 cost-effectiveness of interventions in, 18b
elderly persons. See also Alzheimers disease depression interventions in, 248t
Home Care Program for (Goa), 207 productivity impact of scaled-up treatment,
neurological disorders in, 36 24647
suicide rates of, 164 extended cost-effectiveness analysis of publicly
electroconvulsive therapy (ECT), 74 financed mental and neurological health care
emergency response package in, 24546, 247t
drug-related interventions, 115 comparison with India, 249t
humanitarian aid, 16 parenting skills training in, 152
mental health care, 204 school-based social and emotional learning
school-based interventions, 190, 191b intervention in, 230
environmental events, 3b European Alliance against Depression Programme, 176
epilepsy, 8893 European Headache Federation, 208
alcohol use disorders and, 49 evidence-based interventions for health care delivery,
anti-epileptic drugs (AEDs), 88, 90, 91, 240 204, 205t
autistic spectrum disorders and, 48 extended cost-effectiveness analysis (ECEA), 19,
birth trauma and, 185 21f, 238
burden of disease, 30, 3234t, 87, 8889 application to MNS disorders, 238, 239
co-occurring disorders and, 52 comparison of India and Ethiopia, 249t
cost-effectiveness of interventions for, 19, 20f, 58, Ethiopia analyses, 24548
90, 93, 227t Indian analyses, 24045
extended cost-effectiveness analysis, 241t principles and practice, 238
DALY ranking of, 90
deaths associated with, 7t, 41 F
age of death, 53, 54t faith-based organizations, 202
definition of, 31, 88 family impacts and involvement, 1
epidemiology, 8889 alcoholics, family-based interventions for, 135, 139
GBD findings of excess mortality for family history of suicide, 170
estimated number of cause-specific and excess illicit drug dependence, 110
deaths, 46t, 5152, 51f in treatment, 74, 79
implications, 58 farmers, suicides of, 171
272 Index
Farrington, D. P., 150 Global Health Estimates of disease burden, 5
fetal alcohol syndrome disorders (FASD), 128, 135, 139 globalization of alcohol beverages industry, 128
financial risk protection (FRP), 19, 203, 238, 239b Gmel, G., 133
Finland Good Behavior Game (US), 114
Alzheimers disease, pharmacological interventions good practice interventions, 12, 15, 183, 184t, 188,
for, 98 193, 214
epilepsy-related deaths in, 52 Grading of Recommendations Assessment,
illicit drug use in, 111 Development and Evaluation (GRADE)
workplace treatment, care, and rehabilitation in, 189 guidelines, 8, 13031
fluoxetine, 226 Gunnell, D., 17b
folic acid
deficiency, 94 H
food fortification, 185 Handwerk, M., 155
forensic psychiatry, 203 Happell, B., 56
fragile X syndrome, 4849 HCV (hepatitis C), 51, 57, 111, 119
France headache disorders, 99102. See also migraine
APPRAND program, 188 burden of disease, 100
fetal alcohol syndrome (FAS) warning labels in, 135 cost-effectiveness of interventions for, 102
FRP (financial risk protection), 19, 203, 238, 239b epidemiology, 100
Fuhr, D., 19 interventions for, 100101, 188
alternative therapies, 101
G optimizing health care delivery, 102, 208
gatekeeper training, 176, 223 pharmacological interventions, 100101
GBD. See Global Burden of Disease Study 2010 (GBD public education programs, 1012
2010) self-management, 100
gender differences in burden of MNS disorders, 5, training health care providers, 102
3234, 32t, 34f, 34t, 68 medication-overuse headache, 99100
alcoholic consumption, 12829 recommendations for, 103
childhood mental and developmental disorders, 5, tension-type headache, 99
33, 146, 147t Headache Management Trial, 208
illicit drug dependence, 5, 45, 46f health care platform interventions, 4b, 1314t, 1516,
suicide and, 164, 165t, 16667f 20118
YLLs and, 45, 46f for alcohol use disorders, 13537. See also alcohol
generic drugs. See medications use disorders
genotyping, 68 for childhood mental and developmental disorders,
Global Burden of Disease Study 2010 (GBD 2010), 3, 15154. See also childhood mental and
2930. See also burden of MNS disorders developmental disorders
comparative risk assessments. See comparative risk collaborative stepped care. See collaborative
assessments (CRAs) stepped care
excess mortality from MNS disorders, 4165 for depression, 70
assessment as risk factors for other health elements of, 2014
outcomes, 44. See also co-occurring disorders emergency mental health care, 204. See also
cause-specific death estimates, 42, 4453. See also emergency response
specific MNS disorders evidence-based, 204
implications, 5658 hospital level of care, 1314t, 15, 203
methodology of study, 4244. See also years of life for illicit drug dependence, 11418, 116t. See also
lost (YLLs) illicit drug dependence
transition from communicable to integrating mental health into existing health
noncommunicable diseases, 30, 36, 41 programs, 21214
Global Burden of Disease Study 2013 (GBD 2013), 38 for mood and psychotic disorders, 7478
Global Campaign against Headache, 101 primary level. See primary health care level
Global Campaign against Headache for Europe, PRogramme for Improving Mental health carE
102, 208 (PRIME), 209, 231
Index 273
for psychiatric services, 203 illicit drug dependence, 10925. See also amphetamine
quality of care, 15, 21415 dependence; cannabis dependence; cocaine
relationships among difference delivery channels, 204 dependence; opioid dependence
self-care and informal health care, 2023. See also age of death attributable to, 54t
self-care burden of disease, 32, 3234t, 34, 111, 118
specialists. See also specialist health care delivery consequences, 11011
training primary health care staff by, 210 consumption trends, 111
for suicide, 17778 cost-effectiveness of interventions, 118, 120
system-strengthening strategies for, 2049 criminal activity, 111, 119
task-sharing approach, 20912, 210b DALYs associated with, 36, 111
Healthnet Transcultural Psychosocial Organization definition of, 31, 10910, 120n1
(TPO), 17b delivery platforms for, 14t
HealthWise program (South Africa), 190, 190b economic costs of, 111
hepatitis B, 51, 111 externalizing disorders, 114, 120n4
hepatitis C (HCV), 51, 57, 111, 119 family factors, 110
heroin. See illicit drug dependence; opioid dependence GBD findings of excess mortality for
high-income countries (HICs). See also specific estimated number of cause-specific and excess
countries deaths, 49, 50f, 51
alcoholic consumption in, 128, 138 implications of, 5758
burden of MNS disorders in, 5, 29 gender differences in, 5, 45, 46f
cost-effectiveness of interventions in, 19 global trends, 111
drug dependence treatments and implications for low- and middle-income countries,
interventions, 118 11819
dementia care costs in, 95f individual factors, 110
epilepsy-related deaths in, 51 injecting drug risks, 51, 53, 55t, 57, 226
intervention delivery platforms in, 12 internalizing disorders, 114, 120n4
screening children for developmental disorders, interventions and policies, 11t, 11118
15152 access to treatment, 114
specialist services, 2 brief psychological intervention, 115, 116t, 117
Hip Hop Stroke (awareness program for children), 189 cognitive behavioral therapy (CBT), 117
HIV/AIDS community-based care, 11415, 116t
alcohol use and, 184 community platform interventions, 11314, 115t
anti-epileptic drugs for people with, 91 contingency management approach, 117
burden of disease, 57 control of supply, 112
cost-effective prevention strategy, 226 court-mandated treatment, 112t, 113
dementia and, 94 criminal justice platforms, 113
illicit drug use and, 51, 58, 11011, 115, 119 delivery platforms for, 14t
integrating mental health into existing programs detoxification and withdrawal, 116, 120
for, 212 drug education, 114, 115t
mental health needs of persons with, 21314 drug testing of offenders, 112t, 113
methadone maintenance and, 117, 118 early intervention with at-risk youth, 114
suicide and, 169 health care platform interventions, 11418, 116t
HIV antiretroviral therapy, 57 imprisonment, 112t, 113, 120
home care programs, 207 law enforcement, 112, 112t, 11920
Honduras, epilepsy treatment in, 89, 186 legislation, 186
hospital level of care, 1314t, 15, 203 medication for cannabis dependence, 118
cost-effectiveness of, 22829 medication for heroin and opioid dependence,
humanitarian aid and emergency response, 16, 204 116t, 117
human rights violations, 58, 67, 214, 215 naloxone and other emergency responses, 115
overdose prevention education, 11415
I population platform interventions, 11213, 112t
ibuprofen, 100 prescription monitoring programs, 11213, 112t
ICD-10. See International Classification of Diseases primary health care, 1314t, 15, 115, 116t
274 Index
psychosocial interventions, 117 maternal and infant health programs in, 153
public awareness campaigns, 112t, 113 National Sample Survey Organization, 238
residential rehabilitation, 117 pesticide ban in, 185
school-based prevention programs, 114, 115t school-based social and emotional learning
self-help and mutual aid groups, 114, 115t intervention in, 230
skills training in schools, 114, 115t suicide in, 168, 170, 178
specialist health care delivery, 15, 11617, 116t religious and spiritual beliefs, 172
supervised injecting facilities, 57, 115 safe storage of pesticides, 176
therapeutic community (TC) model, 117 universal health coverage, 240, 245f
workplace drug testing, 11314, 115t indigenous communities and alcohol consumption
medical vs. moral models of addiction, 12, 119 community platform interventions, 135, 136t
mortality rates, 11011 prohibition, 131, 131t
narcotic antagonists, 119, 120, 120n6. See also individual factors
buprenorphine maintenance; naltrexone alcohol use disorders, 130
maintenance illicit drug dependence, 110
natural history of dependence, 110 infants
novel psychoactive substances, 109, 120n2 immunization programs for, 185
peer group factors, 110 maternal and infant health programs, 153
pharmacotherapies, effectiveness of, 5758 psychosocial interventions for, 149, 150
polydrug use, 114, 120n5 salt iodization programs for, 185
precursor chemicals, 112, 112t, 120n3 informal health care. See self-care
recommendations, 11920 information and communication packages, 78
research needs, 119, 120 injecting drug risks, 51, 53, 55t, 57, 226
risk factors, 110 Institute for Health Metrics and Evaluation at
social and contextual factors, 110 University of Washington, 38
suicide and, 51, 55t, 110, 175 Institute of Medicines Forum on Neuroscience and
YLLs and, 45f, 53, 110 Nervous System Disorders, 209
gender differences and, 45, 46f intellectual disability
imprisonment for drug offenses, 112t, 113, 120 burden of disease, 3234t
India childhood, 145
alcohol consumption in, 129 defined, 145
demand reduction strategies for, 229 effective interventions for, 10t
taxation on, 133 Inter-Agency Standing Committees Guidelines on
unrecorded production and consumption, 221 Mental Health and Psychosocial Support in
collaborative care for mental illnesses in, 77 Emergency Settings, 204
community-based rehabilitation in, 78 International Classification of Diseases (ICD-10), 5, 31,
cost-effectiveness of interventions in, 18b 35, 37, 42, 46, 49, 72, 109, 168
depression, 230, 24344, 244t International Convention on the Rights of Persons with
extended analysis for schizophrenia, 19, 21f Disabilities, 186
Dementia Society of Goa, 207 International Labour Organization, 188
District Mental Health Programme, 208 International League against Epilepsy, 88
education and schools in interventions for MNS disorders, 812
examination stress, 175 for adult mental disorders, 7378. See also adult
teacher training for youth health promotion mental disorders
program, 222 for alcohol use disorders, 13037. See also alcohol
teacher training to improve epilepsy use disorders
knowledge, 191 case studies, 1617b
epilepsy interventions in, 240, 241t for childhood mental and developmental disorders,
extended cost-effectiveness analysis (ECEA), 24045 14955. See also childhood mental and
comparison with Ethiopia, 249t developmental disorders
headache interventions in, 228b collaborative care models, 56
MANAS (MANashanti Sudhar Shodh, or project to community-based. See community platform
promote mental health), 78, 207, 22425 interventions
Index 275
costs. See cost-effectiveness and affordability of traditional medicine in, 202
interventions YLLs in, 44, 4546f
delivery platforms of, 1217, 1314t law enforcement
for dementia, 9698 alcohol use disorders and, 132t
effective essential interventions, 812, 911t illicit drug dependence and, 112, 112t, 11920
health care. See health care platform interventions legislation
for illicit drug dependence, 11t, 11118. See also on alcoholic beverages, 22122
illicit drug dependence child protection legislation, 14849
limited access to, 12 on epilepsy, 90
population-based. See population platform illicit drugs legislation, 186
interventions mental health legislation, 73
quality of care, 15, 21415 restricting access to lethal means of suicide, 17b,
iodine deficiency, 185 176, 185, 194, 222
Iran, suicide in, 176, 177 licensing of alcoholic beverages sellers, 138
Israel life expectancy gap in people with mental disorders, 41,
community-based interventions for childhood 42, 57. See also years of life lost (YLLs)
mental and developmental disorders, 149 lifestyle risk factors, 47, 5657, 58, 101
Heart Disease study, 52 Lim, S. S., 44, 129
List of Essential Medicines (WHO), 57
J liver cirrhosis, 50, 184
Jamaica low- and middle-income countries (LMICs). See also
childhood emotional and behavioral problems specific countries
in, 150 alcohol consumption in, 128
psychosocial interventions for malnourished infants challenges for, 13839
in, 149, 150 cost-effective interventions, 221
mortality rates associated with, 130
K recommendations for, 139
Kamgno, J., 52 burden of MNS disorders in, 29
Kenya cause-of-death data from, 58
epilepsy-related deaths in, 52, 88 childhood mental and developmental disorders,
epilepsy treatment in, 89 community-based interventions in, 149
training of primary care workers in, 210 dementia care costs in, 95f
ketogenic diet, 91 epilepsy-related deaths in, 51
key messages, 4b epilepsy treatment gap in, 58, 92f
Kilian, R., 155 illegal substance dependence in
knowledge gaps, effect on scaling up, 2122 assessment issues, 118
Korea, Republic of burden of disease, 118
dementia detection program in, 98 cost-effectiveness of interventions, 118
suicide in, 175 health care infrastructure and capacity, 119
implications, 11819
L medical vs. moral models of addiction, 119
Lachenmeier, D., 13334 opioid substitution treatment (OST), 57, 119
Latin America and the Caribbean potential new treatments, 119
alcohol consumption in, 129 research needs, 119
cost-effectiveness of interventions, 137, 221 intervention delivery platforms in, 12, 29
partial bans on, 132 MNS disorders in, 5
self-help and support groups, 136, 137 mood and anxiety disorders in, 69
taxation on, 13233 neurological disorders in, 87
cost-effectiveness of interventions in, 221, 226, 227t suicide surveillance in, 168
substance use disorders in, 44 survey of mental disorders in, 38
suicide in transition from communicable to
religious and spiritual beliefs, 172 noncommunicable diseases in, 41
risk factors, 169 vital registration systems, lack of, 163
survivors of suicide loss, 172 lung cancer, 42
276 Index
M mortality rates associated with, 6. See also mortality
major depressive disorder. See depression rates
Malaysia, suicide in, 175, 177 need for action to address, 2223
MANAS (MANashanti Sudhar Shodh, mental health significance for global health, 58
project in India), 78, 207, 22425 substance abuse. See illicit drug dependence
mania, 69, 71. See also bipolar disorder years lived with disability (YLDs) and, 5, 6f. See also
mass media campaigns. See public awareness years lived with disability
campaigns years of life lost (YLLs) and, 5, 6, 6f. See also years of
maternal depression. See postpartum depression life lost
maternal mental health interventions, 15253, 212 mental disorders. See adult mental disorders; childhood
Mauritius mental and developmental disorders; mental,
preschool program in, 150 neurological, and substance use (MNS) disorders;
school-based prevention program for adolescent specific disorders (e.g., anxiety, depression)
depression in, 222, 229 mental health awareness campaigns, 73, 18687. See
school-based social and emotional learning also public awareness campaigns
intervention in, 230 mental health first aid training, 188, 191, 193, 223
Maximizing Independence at Home project, 97 Mental Health Gap Action Programme (mhGAP).
media reporting of suicide and self-harm, 175 See World Health Organization (WHO)
medical marijuana, 91 mental health legislation, 73. See also legislation
medical vs. moral models of addiction, 12, 119 mental health workers
medications health centers or home visitation programs
access of people with mental disorders to, 48 using, 193
for ADHD, 154 human resource competencies for MNS disorders
antipsychotics for dementia patients, 53, 9697 in, 210
for conduct disorder, 154 low availability of, 12
cost-effectiveness of, 226 pre-service and in-service training of primary care
for epilepsy. See anti-epileptic drugs (AEDs) workers, 210
low-cost generics, 22, 226 methadone maintenance, 57, 111, 113, 117, 118, 226
morbidity and mortality rates related to treatment Mexico
with, 47 alcoholic beverages in
pharmacotherapies cost-effectiveness of interventions, 138
for dementia, 9697, 98 demand reduction strategies for, 229
for epilepsy, 9091 unrecorded production of, 133
for headache disorders, 100101 illicit substance use in, 111
for heroin and opioid dependence, 116t, 117 school-based social and emotional learning
for mood and psychotic disorders, 74, 7576t, 79 intervention in, 230
for substance use disorders, 48, 5758, 118 mhGAP. See World Health Organization
prescription monitoring programs, 11213, 112t microfinance, 193
psychotropic medications Middle East and North Africa
effects of, 48, 56 cost-effectiveness of interventions in, 227t
primary care staff prescribing, 212 illicit drug dependence in, 44
Megiddo, I., 240 suicide of women in, 171
memantine, 96, 98 midwives, role of, 212
men. See gender differences migraine. See also headache disorders
mental, neurological, and substance use (MNS) burden of disease, 3234t, 87
disorders, 1, 2. See also neurological disorders; cost-effectiveness of interventions for, 228b
specific types of disorders DALYs associated with, 36
adults. See adult mental disorders deaths associated with, 7t
alcohol abuse. See alcohol use disorders definition of, 99
children. See childhood mental and developmental estimated number of cause-specific and excess
disorders deaths for, 46t
disability-adjusted life years (DALYs) due to, 5. See interventions for, 10t, 188
also disability-adjusted life years primary health care, 1314t, 15
economic output lost due to, 8 self-care, 15
Index 277
Mihalopoulos, C., 155, 223 YLDs and, 87
Millennium Development Goals, 214 YLLs and
Mini-Mental State Examination, 96 gender differences and, 45, 46f
MNS. See mental, neurological, and substance use regional differences and, 55
(MNS) disorders New Zealand
monitoring and evaluation of interventions, 22 fetal alcohol syndrome (FAS) warning labels in, 135
monitoring and reporting systems indigenous communities, alcohol consumption by, 135
dementia, 213 Nigeria
suicide and self-harm, 177 alcohol use, demand reduction strategies for, 229
mood disorders, 6870, 75t. See also anxiety disorders; community-based awareness in, 73
depression depression in, 69, 70, 208, 225
moral vs. medical model of addiction, 12, 119 epilepsy in, 230
morphine maintenance, 117 schizophrenia in, 225, 230
mortality rates. See also Global Burden of suicide in, 169
Disease Study 2010 (GBD 2010); years of life noncommunicable diseases
lost (YLLs) compared to communicable diseases in global
for alcohol use disorders, 7t, 44, 129 burden of disease, 30, 36, 41
cause-of-death data, difficulty in capturing, 5859 integrating mental health into primary care for, 213
illicit drug dependence, 11011 nonspecialist human resource cadres, 15, 22
MNS disorders associated with, 67, 7t, 22, 41 Norwegian dementia mortality study, 52
models used in estimating, 43, 43f novel psychoactive substances, 109, 120n2
suicide mortality rates, 164, 165t
multiple sclerosis, 3, 29, 3234t, 87 O
multisystem therapy, 155 obesity, 49, 5657, 100
music therapy, 74 obsessive-compulsive disorder (OCD), 70
mutual aid groups. See support groups occupational therapy, 76, 79
Open the Doors program, 187
N opioid dependence. See also illicit drug dependence
naloxone and other emergency responses, 115 burden of disease, 3234t
naltrexone maintenance, 57, 117, 119 consumption trends, 111
narcotics. See illicit drug dependence DALYs associated with, 36
National Institute for Health and Care Excellence deaths associated with, 7t, 37, 41
(NICE), 99 age of death, 54t
natural history models, 43, 4546, 48, 49, 59 GBD findings of excess mortality for
needle programs, 57 estimated number of cause-specific and excess
neighborhood factors, 3b deaths, 46t, 50f, 51
neighborhood groups, 19293, 194. See also self-help implications, 5758
programs; support groups illicit opioids, 109
neurocysticercosis, 89b, 91, 186 naltrexone maintenance, 117
neurological disorders, 87108. See also epilepsy; opioid substitution treatment (OST), 57, 113,
headache disorders 114, 118, 119, 120. See also methadone
burden of disease, 30, 3234t, 87 maintenance
community health workers detection abilities, 193 prevalence in Australasia and Western Europe, 37
cost-effectiveness of interventions for, 226 rates of dependence, 110
in elderly persons, 36 substance use disorders and, 51
GBD findings of excess mortality for supervised injectable heroin maintenance, 57, 117
estimated number of cause-specific and excess YLLs and, 37, 43
deaths, 46t, 51, 51f overdose prevention education, 11415
implications, 58 oxycodone, 111. See also opioid dependence
gender differences in, 45, 46f
interventions for, 10t P
delivery platforms, 14t Pakistan
school-based interventions, 189, 192 depressive disorder related to suicide in, 169
278 Index
mental health awareness among school post-traumatic stress disorder (PTSD), 49, 68, 70, 75t,
children in, 74 77, 192
preventive maternal and child health care in, 153 poverty. See also financial risk protection (FRP)
rural secondary schools in, 189 microfinance schemes and, 193, 221
suicide in, 175 schizophrenia and, 48
depressive disorder related to, 169 suicide and, 169, 170, 175
women, 171 Powell, C., 149
Thinking Healthy Programme, 212 precursor chemicals, 112, 112t, 120n3
Palestine, school-based intervention in, 191b pregnancy
panic disorder, 70, 75t alcohol use in, 135, 139
parenting interventions, 193 vulnerability for MNS disorders, 185
community-based program for, cost-effectiveness premature mortality, 5. See also years of life
of, 22223 lost (YLLs)
skills training, 152, 21314 Preventing Suicide: A Global Imperative (WHO), 177
Parkinsons disease prevention of MNS disorders, 4b, 812, 911t. See also
burden of disease, 30, 3234t, 87 interventions for MNS disorders
gender differences in, 5, 33 adult mental disorders, 7677
peer-led interventions cost-effectiveness of. See cost-effectiveness and
education, 202 affordability of interventions
illicit drug dependence, 110 primary health care level, 1314t, 15, 203
self-help groups and peer support, 203 for alcohol use disorders, 1314t, 15
Perinatal Mental Health Project (South Africa), 212 competency-based education, 20910, 209t
pesticides cost-effectiveness of, 22427
regulation to restrict access to, 17b, 176, 185 international studies, 226
safe storage of, 176 national studies, 22426
self-poisoning, 175, 177 evidence-based, 205t
Phanthunane, P., T. Vos, 225 for illicit drug dependence, 1314t, 15, 115
pharmacologic treatment. See medications planning and consultation, 210
phobias, 70 pre-service and in-service training of
Pion, S. D. S., 52 workers, 210
Plan Do Study Act, 214 psychotropic medications, prescription authority
Platania-Phung, C., 56 for, 212
pneumonia, 53, 56 Prince, M., 44
political will, effect on scaling up, 21 problem-solving skills therapy (PSST), 155
polydrug use, 114, 120n5 PRogramme for Improving Mental health carE
Pompili, P., 52 (PRIME), 209, 231
population platform interventions, 1314t, 15, 18387, psychiatric services, 203. See also specialist health
19394 care delivery
for adult mental health, 73 psychosis, 2
for alcohol use disorders, 130, 13135, 131t extended cost-effectiveness analysis for, 19
for childhood mental and developmental disorders, interventions for, 8
14849 medications for, 226
for epilepsy, 89 primary health care for, 1314t, 15
for illicit drug dependence, 11213, 112t specialist health care delivery for, 15
information and awareness campaigns, 18687 psychosocial interventions
key findings, 183 for conduct disorder, 15455
legislation and regulations, 18485. See also for illicit drug dependence, 117
legislation for malnourished infants, 149
protecting persons with MNS disorders, 186 psychosocial life crises and suicide, 16970
restricting access to means of suicide, 185 psychotherapy for mood and psychotic disorders, 74,
for suicide, 17475 7576t
postpartum depression, 75t, 77, 15253, 212 psychotropic medications, effects of, 48, 56
womens support groups for, 153 PTSD. See post-traumatic stress disorder
Index 279
public awareness campaigns, 18687, 194 extended cost-effectiveness analysis for, 19, 21f
alcohol consumption, 135 health system barriers and opportunities for, 2122
headaches, 101, 102 knowledge gaps as factors, 2122
illicit drug dependence, 112t, 113 political will as factor, 21
mental health, 73 proposed regional framework in WHO Eastern
Public Health Action for the Prevention of Suicide Mediterranean Region, 2324b
(WHO), 177 strategies for strengthening health system, 22
public health considerations schizophrenia, 7273
alcohol use disorders, 128 age of cause-specific and excess deaths attributed
health platform related, 208 to, 47f
integrating mental health into existing programs, 21214 burden of disease, 3234t
maternal mental health, 153 cannabis dependence and, 53, 55t
suicide, 173 clinical features and course, 72
co-occurring disorders with, 48
Q cost-effectiveness of interventions for, 19, 2021f,
quality-adjusted life years (QALYs), 155, 225 225, 227t, 230
quality of care, 15, 21415. See also health care platform community-based vs. hospital-based
interventions services, 228
extended cost-effectiveness analysis, 242t
R DALYs associated with, 36
RAP-A program, 230, 230t deaths associated with, 7t
refugees, suicides of, 17172, 177 age of death, 54t
rehabilitation environmental factors associated with, 68
cognitive rehabilitation for dementia, 97 epidemiology and burden of disease, 7273
illicit drug dependence, 117 estimated number of cause-specific and excess
mental disorders, 78 deaths for, 4648, 46t, 47f
Rehm, J., 13334, 137, 221 extended cost-effectiveness analysis for, 19, 21f,
relaxation techniques, 74 24145
religious beliefs and suicide, 172 gender differences in, 5, 33, 68
religious healers, 202 genotyping of individuals with, 68
research and development initiatives, 22 interventions for. See also adult mental disorders
for illicit drug dependence, 119, 120 enhanced financial and service coverage, 24143
residential facilities, 1516, 203 pharmacologic and psychological treatment, 76t
for illicit drug dependence interventions, 117 side effects of antipsychotic medications for, 48
Resourceful Adolescent Programme-Adolescent version suicide and self-harm and, 48
(RAP-A) program, 22930, 230t YLDs and, 68, 68f
respiratory diseases, 41, 56 YLLs and, 43, 45f, 48
risk factors schools. See education and schools
for childhood mental and developmental disorders, Scott, D., 56
14648, 148t screenings
for illicit drug dependence, 110 for alcohol use disorders, 13536, 136t, 139
for suicide, 37, 16871, 173f, 178 for childhood mental and developmental disorders,
risperidone, 226 15152, 156
Russian Federation, headache interventions in, 102, for comorbid health issues, 5657
208, 228b for dementia, 98
for illicit drug dependence, 115, 117
S for mental health disorders, 77
Saxena, S., 232 SDG (sustainable development goal), 2b
scaling up, 45b self-care, 1314t, 15, 2023
affordability and. See cost-effectiveness and for epilepsy, 90
affordability of interventions evidence-based, 205t
case studies of interventions for MNS disorders, for headache disorders, 100
1617b for mood and psychotic disorders, 74
280 Index
self-harm. See suicide and self-harm specialist health care delivery, 15, 7476, 203
self-help programs, 57, 2023 for alcohol use disorders, 15
for alcohol use disorders, 13637 for childhood mental and developmental
for illicit drug dependence, 114, 115t disorders, 15455
self-immolation. See suicide and self-harm cost-effectiveness of, 22829
serotonin-norepinephrine reuptake inhibitors evidence-based, 205t
(SNRIs), 225 for extended-stay facilities, 203
Service Organization Pyramid for an Optimal for illicit drug dependence, 15, 11617, 116t
Mix of Services for Mental Health (WHO), Sri Lanka
202, 202f children with developmental delays in, 151
sexual minorities, suicide of, 172 suicide in, 170, 171, 175, 17778
shame and fear, 67 prevention through pesticide regulation,
SHR (sustained headache relief), 100 17b, 185, 185b
simple phobias, 70 safe storage of pesticides, 176
Single Convention on Narcotic Drugs, 120n1 START (STrAtegies for RelaTives) study, 99
Six Sigma, 214 Statistical Process Control, 214
Skeen, S., 194 stigma and discrimination, 1, 5b, 22, 67
smoking, 4748, 49, 52, 5657 anti-stigma interventions, 8990, 187
Sneha (suicide prevention organization), 175 limiting access to interventions, 12
sobriety checkpoints, 134 limiting access to screenings, 56
social anxiety disorder, 70 quality of care and, 214
social causation pathway, 3b self-care and, 203
social change, 3b suicide and, 175, 176
social determinants, 1, 3b Strang, J., 8
for illicit drug dependence, 110 stroke, 41, 56
social drift pathway, 3b Sub-Saharan Africa
societal response to alcohol use disorders, 130 alcohol consumption in, 129
socioeconomic status, 3b. See also poverty cost-effectiveness of interventions, 137, 221
drinking and, 129 mortality associated with, 129
schizophrenia and, 48 taxation, 133, 137
SOLVE training package, 188 childhood mental disabilities, lack of data on, 146
Sornpaisarn, B., 133 cost-effectiveness of interventions in, 221, 226, 227t,
South Africa 231, 239
alcohol consumption of pregnant women in, 135 epilepsy in, 44, 45, 52, 89, 91
collaborative stepped care approach in, 210 human resource competencies for MNS disorders
epilepsy treatment in, 89 in, 210
HealthWise program in, 190, 190b illicit drug dependence in, 44
HIV/AIDS treatment integrated with mental cost-effectiveness of interventions, 118
health in, 214 microfinance in, 193
parenting skills training in, 152 traditional medicine in, 202
Perinatal Mental Health Project, 212 YLL rates in, 44, 4546f, 55
Primary Care 101 (PC101), 214 gender differences and, 45, 46f
primary care practitioners in, 208 substance use disorders. See illicit drug dependence
workplace interventions in, 189 suicide and self-harm, 42, 16381
South Asia of adolescents, 171
alcohol consumption in age pattern of, 164, 165t, 16667f
cost-effectiveness of interventions, 137, 221 alcohol consumption and, 50, 129, 169, 175
taxation, 133, 137 as cause of death, 41, 164
cost-effectiveness of interventions in, 226, 227t, changes in rates (2000-12), 164, 165t
231, 239 coping strategies and well-being, 172
suicide prevention organizations in, 176 cost-effectiveness of prevention efforts, 178
suicide rates in, 164 decriminalization of, 175
traditional medicine in, 202 definition of, 163
Index 281
depression and, 69, 176 suicide mortality rates, 164, 165t
drug misuse and, 175 surveillance in LMICs, 168
early traumatic events associated with, 17071 survivors of suicide loss, 172
economic issues and, 175 urban vs. rural locations, 170
effective interventions for, 11t WHO prevention guidelines, 57, 163, 173
epidemiology, 163 YLLs and, 53, 55
epilepsy and, 52 supervised injecting facilities, 57, 115
exposure to models, 170 support groups, 19293, 194. See also self-help
family history of suicide, 170 programs
of farmers, 171 for alcohol use disorders, 13637
gender differences, 164, 165t, 16667f, 171 for illicit drug dependence, 114, 115t
interventions for, 11t, 17478 suspension of drivers license, 135
brief intervention and contact, 177 sustainable development goal (SDG), 2b
community platform interventions, 17576 sustained headache relief (SHR), 100
delivery platforms, 14t Sweden
disasters and refugees, 177 cost of Alzheimers disease treatment in, 98
examination stress, 175 dementia-related deaths in, 53
gatekeeper training, 176 epilepsy-related deaths in, 52
health care platform interventions, 17778 Szekely, A., 176
medical management of poisoning with
pesticides, 177 T
monitoring and reporting systems, 177 Tanzania
national suicide prevention strategies, 17778 epilepsy-related deaths in, 52
nongovernmental organization services, 17576 epilepsy treatment in, 89
population platform interventions, 17475, 194 task-sharing approach, 20912, 210b
restricting access to lethal means, 17b, 17475, taxation of alcoholic beverages, 131t, 13233, 138,
176, 185, 194, 222 221, 229
safe storage of pesticides, 176 Taylor, B. J., 13334
school-based interventions, 176 teacher training program to identify and assess mental
stigma and discrimination, 175 health problems, 191, 191b
media reporting of, 175 TEAMcare USA and TEAMcare Canada, 213
mental disorders and alcohol misuse associated telemedicine, 78
with, 169 Ten Questions screen, 151, 151b
methods, availability of, 168, 170 testing
pesticide self-poisoning, 175, 177 blood alcohol concentration (BAC) testing of
Sri Lanka suicide prevention through pesticide drivers, 134
regulation, 17b, 185b breath testing of drivers, 134
physical disorders and, 169 Thailand
prevention in LMICs, 17274 alcoholic beverages, regulation of, 128
prior suicide attempts, 170 antidepressants and CBT as cost-effective
protective factors, 172 interventions for depression in, 225
psychosocial life crises and, 16970 therapeutic community (TC) model, 117
as public health issue, 173 Thinking Healthy Programme (Pakistan), 212
of refugees and internally displaced persons, 3 Dimensions of Care for Diabetes (UK), 213
17172, 177 Total Quality Management, 214
religious and spiritual beliefs, 172 TPO (Healthnet Transcultural Psychosocial
risk factors for, 37, 16871, 173f, 178 Organization), 17b
schizophrenia and, 48 traditional healers, 202
of sexual minorities, 172 traffic accidents. See driving impaired and traffic
stigma of, 176 accidents
strong personal relationships and, 172 transcranial magnetic stimulation, 74
substance use disorders and, 51, 55t, 110 traumatic brain injury, measures to protect
suicide attempt rates, 16468 against, 186
282 Index
treatment gap, 214 unrecorded market of alcohol production and sales,
in epilepsy, 58, 92f 133, 13839
treatment of MNS disorders, 4b, 812, 911t. See also urinary tract infections, 53
interventions for MNS disorders
cost-effectiveness of. See cost-effectiveness and V
affordability of interventions vicious cycle of social determinants, 3b
Ttofi, M. M., 150 violence
tuberculosis, 50, 129, 212, 21314 alcohol-attributable, 50, 129
Turkey domestic violence legislation, 186
early childhood enrichment project in, 192 Vreeman, R. C., 150
suicide in, 170, 171
W
U warning labels on alcoholic beverages, 135, 139
Uganda web-based psychological therapy, 15, 78
epilepsy treatment in, 89 Whiteford, H. A., 31
specialists training primary health care WHO. See World Health Organization
staff in, 210 whole-of-government approach, 205
United Kingdom whole-of-school approach, 150
Alzheimers disease, pharmacological interventions women. See also gender differences; pregnancy
for, 98 gender equity and economic empowerment
bipolar disorder research in, 48 interventions, 193
cognitive rehabilitation for dementia in, 97 postnatal depression. See postpartum depression
community-based awareness in, 73 suicide rates of, 171
community-based vs. hospital-based programs workplace interventions, 194
in, 73 drug testing, 11314, 115t
coping strategy program for mental health of epilepsy, anti-stigma interventions for, 89
dementia caregivers in, 225 identification and case detection, 188
crisis intervention teams in, 77 mood and anxiety disorders intervention, 73
epilepsy-related deaths in, 52 promotion and primary prevention, 188
parenting programs in, 223 World Development Report (1993), 2b
3 Dimensions of Care for Diabetes, 213 World Health Organization (WHO)
United Nations Assessment Instrument for Mental Health Systems
Childrens Fund Multiple Indicator Cluster (WHO-AIMS) survey, 73
Survey, 151 Atlas on Substance Use, 137
Convention on the Rights of Persons with Building Back Better, 204
Disabilities, 215 Child and Adolescent Mental Health Policies and
United States Plans, 148
court-mandated drug treatment in, 113 CHOosing Interventions that are Cost-Effective
depression in, 69 (CHOICE) project, 220, 226, 228, 228b, 232
Drug Abuse Resistance Education (DARE) Comprehensive Mental Health Action Plan, 22, 24
program, 114 cost-effectiveness analysis, 18
fetal alcohol syndrome (FAS) warning labels in, 135 detection of mental disorders, system for, 77
Good Behavior Game for classroom behavior Global Health Estimates, 5, 163, 166f
management with young children, 114 High-Level Meeting on Non-communicable
stroke awareness program for children in, 189 Diseases (2011), 229
TEAMcare USA, 213 Integrated Management of Adult and Adolescent
universal health coverage, 5b, 18b, 19, 23751. See also Illness (IMAI), 213
extended cost-effectiveness analysis (ECEA) List of Essential Medicines, 57
health system goals, 237 Mental Health Gap Action Programme (mhGAP), 1,
MNS coverage, 23738 5, 8, 17b, 22, 74, 77, 204, 245
pay out-of-pocket (OOP) for treatment, 23738 alcohol consumption and, 135, 136
University of Washingtons Institute for Health Metrics depression and, 208
and Evaluation, 38 Mental Health Global Action Program, 177
Index 283
Ministerial Conference on Global Action Against population attributable fractions (PAFs) and, 44
Dementia (2015), 99 schizophrenia, 68, 68f
noncommunicable disease interventions years of life lost (YLLs), 5, 6, 6f, 3031, 35, 35f, 37, 41
package, 213 alcohol use disorders, 4345, 45f, 53, 55, 55t
Preventing Suicide: A Global Imperative, 177 amphetamine dependence, 43
proposed regional framework in Eastern attributions needed for more accurate
Mediterranean Region, 2324b representation of MNS disorders, 55
Public Health Action for the Prevention of cause of death and, 4243, 46, 5355
Suicide, 177 cocaine dependence, 43
on quality improvement (QI) mechanisms, 214 dementia, 43, 45f
QualityRights Project, 186 depression, 49
QualityRights Toolkit, 215 differences in patterns of MNS prevalence and,
recommended method for integrating hospital- 4445, 45f
based and community-based mental health gender differences, 45, 46f
services, 16b illicit drug dependence, 45f, 53, 110
Service Organization Pyramid for an Optimal Mix of gender differences and, 45, 46f
Services for Mental Health, 202, 202f natural history models and, 4546
on suicide and suicide prevention, 57, 163, 173, 177, neurological disorders
178 gender differences and, 45, 46f
World Mental Health Action Plan (2013-2020), 177 regional differences and, 55
World Mental Health (WMH) surveys, 6970, 71, opioid dependence, 37, 43
102, 166 population attributable fractions (PAFs) and, 44
YLDs assigned to MNS disorders, 30 regional differences and, 5354, 55f
World Psychiatric Association, 187 schizophrenia, 43, 45f, 48
suicide and, 53, 55
Y
years lived with disability (YLDs), 5, 6f, 3031, Z
35, 35f, 37 Zambia
adult mental disorders, 6768, 68f epilepsy treatment in, 90
epilepsy, 90 headache disorders and treatment in, 100,
from natural history models, 43 102, 228b
neurological disorders, 87 trauma-focused CBT for children in, 154
284 Index
ECO-AUDIT
Environmental Benefits Statement
The World Bank Group is committed to reducing its environmental footprint.
In support of this commitment, the Publishing and Knowledge Division leverages
electronic publishing options and print-on-demand technology, which is located
in regional hubs worldwide. Together, these initiatives enable print runs to be
lowered and shipping distances decreased, resulting in reduced paper consumption,
chemical use, greenhouse gas emissions, and waste.
The Publishing and Knowledge Division follows the recommended standards
for paper use set by the Green Press Initiative. The majority of our books are
printed on Forest Stewardship Council (FSC)certified paper, with nearly all
containing 50100 percent recycled content. The recycled fiber in our book paper
is either unbleached or bleached using totally chlorine-free (TCF), processed
chlorine-free (PCF), or enhanced elemental chlorine-free (EECF) processes.
More information about the Banks environmental philosophy can be found
at http://www.worldbank.org/corporateresponsibility.