THE LANCET
HIV series
Paediatric HIV infection
Gabriella Scarlatti
By the year 2000 there will be six million pregnant women and five to ten million children infected with HIV-1.
Intervention strategies have been planned and in some instances already started. A timely and cost-effective strategy
needs to take into account that most HIV-1 infected individuals reside in developing countries. Further studies are
needed on immunological and virological factors affecting HIV-1 transmission from mother to child, on differential
disease progression in affected children, and on transient infection.
The first children with HIV-1 infection were described in
1983. As of late 1994, WHO estimated that about 15
million children had been infected world wide, with over
three-quarters of the cases in sub-Saharan Africa and
developing areas of the Americas. Moreover, nearly half
of all HIV-infected adults are women of childbearing age;
since most children who become infected acquire the
virus from their mother, the future impact of paediatric
HIV infection will be directly related to the increasing
incidence of infection among women. Interventions to
prevent further escalation in mother-to-child transmission
are urgently needed.
The rate of mother-to-child transmission of HIV-1 has
been evaluated in several epidemiological surveys, and
varies from 13% to 42%.1,2 Transmission rate is twice as
high in Africa as in Europe.3 This difference is still
unexplained, but contributory factors include concomitant
infections in the mother, differences in virulence of the
virus according to geographical origin, and frequency of
breastfeeding.
Timing of infection
An accurate understanding of the timing of HIV-1
transmission from mother to fetus is very important for
the design of intervention strategies. Evidence for both
early and late in-utero HIV-1 transmission is well
documented; there is probably more than one route of
transmission. Virus has been detected by polymerase
chain reaction (PCR) in fetal tissue specimens from the
first trimester of gestation,4 and by the presence of viral
p24 antigen in fetal serum samples taken during the
second trimester.5 Evidence for transmission comes from
the International Registry of HIV-exposed Twins: the risk
of HIV infection in the first-born twin is two-fold higher
than in the second-born twin.6 Thus, transmission seems
to occur during gestation from the first trimester, but
proportions of in-utero and intrapartum transmission
events are unclear.
Two approaches have been used to assess timing of
infection: (a) determination of antibody production to
viral proteins by the child itself; and (b) detection of
HIV-1 by PCR or virus isolation from the childs blood.
As evidence for infection late in pregnancy, De Rossia and
Lancet 1996; 348: 86368
Laboratory of Immunobiology, Centro San Luigi, San Raffaele
Scientific Institute, Via Stamira dAncona 20, 20 127 Milan, Italy
(G Scarlatti MD)
Vol 348 September 28, 1996
colleagues7 used the timing and pattern of seroconverion
to specific HIV-1 peptides in children at 46 weeks of
age. Comparison of the antibody pattern determined by
western blot in the mothers serum obtained during
gestation and at delivery with that of the newborn baby
suggested that at least 65% of children are infected during
the final 6 weeks of pregnancy and at delivery.8 According
to the working hypothesis of Bryson et al,9 detection of
virus by PCR within 48 hours of age indicates that
infection of the child took place during gestation, whereas
delayed appearance of virus implies intrapartum
transmission. This model does not exclude the possibility
that the virus resides in tissues other than blood, and may
become activated at birth.
Overall, therefore, about half of the transmission events
probably occur close to delivery.8,9 However, these figures
apply to live births. If early transmission events lead to
loss of the fetus, as suggested by Langston et al,10 the
actual frequency of in-utero transmissions may be higher.
Postnatal transmission can occur via breastfeeding.1 A
systematic review of published reports on the risk of
transmission via this route suggested an additional risk of
14% over transmission in utero or at delivery by mothers
who have been infected prenatally.11 Counselling mothers
about breastfeeding is especially difficult in developing
countries, where the practice has a protective effect
against infant death from other endemic diseases.
Accordingly, current recommendations about infant
feeding by HIV-1-infected mothers differ according to
geographical setting.12
Routes of transmission
The role of the placenta is puzzling. Does it function as a
barrier to the virus, or is it involved in cell-to-cell
transmission, leading to infection of the fetus by particular
virus variants? In-vivo infection of placental cells,
trophoblast, and villous Hofbauer cells has been shown as
early as 8 weeks gestation,13 and may therefore account
for early in-utero transmission. Differences in the ability
of placental cells to sustain virus replication might be
important for selection of virus variants during
transmission. Otherwise, intercurrent infections of the
placentaeg, chorioamnionitismay favour leakage of
the virus into the amniotic cavity. Whether placental
infection always results in infection of the fetus is
uncertain.
Transmission during delivery may be due to mixing of
maternal and fetal blood during contractions, to
contamination through mucous membranes, or via
863
THE LANCET
swallowing infected maternal blood or cervico-vaginal
secretions when the fetus passes through the birth canal.
In addition, premature rupture of membranes could lead
to ascending transmission of HIV-1 through invasion of
the amniotic cavity; in this respect, vaginal secretions
seem to harbour virus,14 and p24 antigen has been
detected in amniotic fluid of seropositive women.5
Transmission through breastfeeding seems to be related
to the virus load in breast milk as well as to the length of
time the child was fed. Moreover, fetal gut cells are
susceptible to infection with HIV-1; such a route is
facilitated by the absence of an acid environment in the
newborn babys stomach that may allow HIV-1 to retain
its infectivity.15
Evidence for multiple mechanisms of transmission also
comes from analysis of GHHIV-1 sequences from
mother-child pairs.16 We observed that the HIV-1 variant
harboured by newborn babies is in some cases similar to
the variant detected in maternal peripheral blood
mononuclear cells (PBMC) and in other cases similar to
that detected in serum.16 This observation raises the
possibility that transmission may occur via either cell-
associated or cell-free virus.
Virological and immunological determinants of
transmission
Mothers with few or no symptoms can transmit infection
to their children, although they do so less frequently than
severely symptomatic women. There is a strong
correlation between low maternal CD4 T-lymphocyte
counts at delivery and increased risk of transmission.17
The precise CD4 T-cell value that correlates with
transmission is still unclear, and the CD4:CD8 ratio may
be a better marker of transmission. The predictive value
of maternal CD4 T-cell counts during early pregnancy
has not been established.
It is reasonable to assume that a high virus load in the
maternal circulation favours HIV-1 transmission to the
child. The increased transmission rate of mothers who
seroconverted late in pregnancy or after delivery during
breastfeeding18 accords with this view, and is presumably
due to the viraemic phase characteristic of primary HIV-1
infection. The European Collaborative Study showed
that, in a cohort of 330 mothers, those who were p24
antigen positive had a three-fold increase in risk of
transmission by comparison with antigen-negative
mothers.17 Furthermore, the recently concluded phase III
trial (ACTG076) of administration of zidovudine to a
selected cohort of women throughout pregnancy, labour,
and delivery, as well as to their babies for 6 weeks after
birth, suggests that virus load has a substantial impact on
the likelihood of transmission: by comparison with
placebo, rate of transmission was reduced by 675% in the
zidovudine-treated group.19 Direct evidence of the
importance of virus load in mother-to-child transmission
comes from the finding of a clear difference in
distribution of HIV-1 load, measured as virus RNA
concentrations, in plasma samples of transmitting and
non-transmitting mothers.20
Differences in virulence or cell tropism of virus strains
harboured by the mother may well contribute to the risk
of transmission. HIV-1 isolates can be classified according
to replicative capacity, cytopathogenicity, and tropism in
PBMC, macrophages, and tumour cell lines. Patients who
harbour viruses with high replicative and syncytium-
864
inducing capacity progress faster towards overt AIDS
than those who harbour slow-replicating, non-syncytium-
inducing virus variants. The replicative capacity of the
virus may likewise influence transmission to the child:
rapidly replicating syncytium-inducing viruses are isolated
more frequently from mothers who transmit virus to their
children than from those who do not.21,22
Quantitative and qualitative differences in immune
responses between transmitting and non-transmitting
mothers could provide valuable information about the
type of protective immunity involved and might reveal
risk factors for mother-to-child transmission. Several
studies suggest that the presence of maternal antibodies
that bind to viral envelope glycoprotein gp120 might alter
the risk of transmission, although conflicting results have
been reported.23,24 Furthermore, these studies were largely
based on laboratory strains or consensus sequences of
HIV-1, so the results cannot necessarily be extrapolated
to the in-vivo situation.
We have shown that a neutralising antibody response
towards the virus harboured by the mother at delivery
ie, autologous virusis associated with a reduced risk of
transmission.25 Moreover, the presence of these antibodies
seems to be linked to the capacity for cross-neutralising
several primary isolates. Two additional studies have
confirmed our results,21 and have also suggested that in
some cases transmission is favoured by the presence of
enhancing antibodies.
Protection from infection
Bryson et al26 described one child born to an HIV-1
infected mother in whom virus was no longer detectable
after 3 months of age. The child lost the maternal
antibody, never seroconverted, and remained well during
the 5 years of follow-up; these results suggest clearance of
infection. Similar cases of transient infections have been
reported subsequently. According to two major reports,
the frequency of such events may range from 27% to
64%.27,28 These studies raise several issues: (a) whether
the criteria for definition of infection are stringent enough
to exclude technical artifacts; (b) whether maternal
infected cells can reside in the childs blood, be detected
by sensitive assays such as PCR, and therefore give rise to
false-positive results; and (c) whether virus can reside in
tissues other than blood and under certain conditions
become activated.
If clearance of the virus can occur, study of the
underlying mechanisms would have important therapeutic
implications for interrupting transmission and progression
of the disease. One can reasonably assume that a strong
immune response might reduce virus load or even clear
the virus. In this regard, detection of HIV-1 envelope-
specific T-cell response in the cord blood of uninfected
children born to seropositive mothers indicates that
clearance may already occur in utero.29 Several groups of
investigators have described HIV-specific cytotoxic
lymphocytes (CTL),30 and also in-vitro antibody
production to viral gag and env proteins in apparently
uninfected infants.31 Roques and colleagues27 did not find
differences in neutralising antibody or antibody-
dependent cellular cytotoxicity (ADCC) titres in their
cohort of HIV-1 seronegative exposed children but the
children were tested only at birth. Differences in
susceptibility of host cells to the virus may also
contribute. Ometto et al32 showed that CD4 T cells and
Vol 348 September 28, 1996

Sensitivity (%)
At birth
Virus culture 1050
DNA PCR 1050
p24 antigen (CD) 1030
*To determine state of infection positive results are necessary on two separate
determinations from one or more of these tests (excluding cord-blood sample).
Table: Laboratory assays for HIV-1 detection recommended for
diagnosis in children (<18 months of age) born to HIV-
seropositive mothers*
also monocyte-derived macrophages of uninfected
newborn babies are less susceptible to in-vitro infection
with the mothers virus than with unrelated HIV-1
isolates.
Intervention strategies to reduce transmission
Transmission of HIV-1 from mother to child seems to be
a multifactorial event. Use of antiretroviral therapy with
zidovudine has shown that prevention of transmission is
achievable.19 However, the need for simpler strategies is
evident. Biggar et al33 found that cleansing of the birth
canal with chlorhexidine during labour was ineffective.
Another approach is to recommend breastfeeding only in
those areas where it is clearly necessary; two prospective
studies of breastfeeding versus bottle-feeding are planned
in Africa. Change in the mode of deliveryeg, opting for
surgical deliveryis another, albeit costly, strategy.
Several studies have shown a reduction (3051%) in risk
of transmission of HIV-1 to children delivered by
caesarean section.34 Statistically, this means that between
12 and 16 surgical deliveries are required to prevent
infection in one infant.34 To resolve this issue, a
multicentre trial has been started in Europe.
The ACTG076 trial, which included treatment from
the 14th week of gestation in women with CD4 counts of
over 200/L, prompts other considerations.19 Some
studies are now attempting to define the latest time in
pregnancy at which zidovudine treatment can begin. A
perplexing issue is the unnecessary treatment of mothers
and their uninfected babies. Although zidovudine
treatment has not shown severe side-effects, other than
anaemia, in newborn babies,19 the children born to treated
mothers must be followed to exclude any long-term
complications. Another issue is the consequence for the
child of the emergence of zidovudine-resistant virus
variants in the mother, which might possibly infect the
fetus.
Antibodies may neutralise or otherwise destroy the
virus, reducing viral load in the mother or acting directly
in the fetus/neonate. Immunological interventions, either
passive or active, are complex to test clinically but are
being evaluated. Trials of passive immunotherapy with
anti-HIV immunoglobulins given according to different
schedules to the mother and/or the neonate are planned
both in the USA and in Africa. Immunogenicity and
safety studies with viral envelope subunit vaccines (gp120
of MN or SF2) have also been started in pregnant women
and infants.
Diagnosis of HIV-1 infection in children
Early diagnosis of HIV-1 infection in children born to
seropositive mothers has been hampered because the
conventional antibody tests cannot be used owing to the
persistence of placentally transferred maternal IgG. These
antibodies can linger in the infants blood up to the age of
Vol 348 September 28, 1996
THE LANCET
18 months.35 During a workshop held in Siena, Italy, in
At >1 month 1992, virological, immunological, and molecular
>90
diagnostic tests were compared.36 There was general
>90 agreement that single tests are affected differently by
5080 factors such as age of the children, stage of the disease,
and antiviral therapy. Virus culture and PCR were judged
to be the most reliable techniques for determining
infection during the first 2 months of age in non breastfed
children, attaining a sensitivity of 90% compared with
5060% for other diagnostic tests (table).
Detection of p24 antigen in the serum or plasma from
newborn babies is less sensitive than virus culture or
PCR, but can achieve similar sensitivities by 6 months of
age.37 Even higher sensitivity can be achieved by acid
treatment of sera, which disrupts the immune
complexes.38 With this approach, several groups correctly
diagnosed HIV-1 infection of babies tested within the first
few months of life.
IgA and IgM do not cross the placenta, so detection of
HIV-specific antibodies of such classes in a child indicates
infection. However, IgA is detected in only half of the
infected infants by the age of 36 months,39 and IgG
antibodies of maternal origin have to be removed before
testing owing to competition with the antigen. Detection
of IgA seems to be a more sensitive assay than detection
of HIV-specific IgM, probably because IgM production is
transient.39
In-vitro antibody production assays detect HIV-specific
antibody-producing B-cells.41 False-positives have been
observed during the first 23 months of life, probably
from carry over of maternal antibodies attached to the
infants B cells.40 Consequently, this assay is suitable only
for older children. However, sensitivity of the assay also
depends on several other factorseg, B-cell immaturity of
the neonate, state of the immune system, and antiviral
therapy.
Overall, diagnosis of HIV-1 infection at birth is possible
but unreliable, presumably because of the small number
of infected cells in the blood. Although the number of
different samples and assays required for diagnosis is still
controversial, an accurate diagnosis can be achieved in
non-breastfed children born to seropositive mothers by
the age of 3 months.36
Clinical features of paediatric infection
Patterns of disease expression and progression differ
among HIV-1 infected children.41,42 2326% have a
rapidly downhill course, developing features characteristic
of AIDS within the first year of life.1,42,43 Others develop
AIDS slowly over several years. Some children do not
have any signs or symptoms of disease by the age of 810
years. Most children with HIV-1 infection acquired from
the mother will display features of HIV-1 infection within
6 months of life.1,43 At birth, only a very low proportion of
children have signs of infection, but by 1 year 8090%
will have manifested some features of infection. Age of
onset of any sign of HIV-1 infection predicts lengths of
survival.44 The mortality rate for children who develop
features of AIDS early in life is substantially higher than
for those who become symptomatic later during
childhood.41
Onset of HIV-1 infection in children has a wide
spectrum of clinical manifestations.43,44 Some infants
present with features of severe immunodeficiency,
whereas others have non-specific findings such as
865
THE LANCET

Panel 1: Opportunistic infections in HIV-1 infected children
Pneumocystis carinii pneumonia
Candidosis, oesophageal or pulmonary
Cytomegalovirus infection
Cryptosporidiosis
Non-tuberculous mycobacteria
Herpes zoster infection
Tuberculosis
Toxoplasmosis
hepatosplenomegaly, failure to thrive, unexplained
persistent fever, parotitis, and recurrent gastroenteritis.
During the first year, lymphadenopathy, splenomegaly,
and hepatomegaly, singularly or combined, have been
observed in more than 50% of children; other signs,
including failure to thrive, fever, diarrhoea, and AIDS-
defining secondary infections, are frequently observed,
but can also present at a later age.
About 3050% of children present with an early onset
of opportunistic infections. Pneumocystis carinii
pneumonia (PCP), although it can occur at any age, is
most commonly diagnosed in young infants.42 Half of the
children with AIDS diagnosed before 1 year of age have
PCP. Children are more likely than adults to develop
PCP, with a peak incidence between 3 and 6 months of
age,43 and less likely to develop other opportunistic
infections such as toxoplasmosis, tuberculosis,
cryptococcosis, and histoplasmosis (panel 1). The survival
rate of these children is worse than that of children
without early onset of opportunistic infections.1,43 At 3
years of age, 48% of children with early onset of
opportunistic infections have died, compared with 97%
without such infections. The prevalence and severity of
infectious complications can be altered by supportive
care.45 The recommendations of the Centers for Disease
Control for the prophylaxis of PCP were initially based on
absolute CD4 counts. In view of the variability of CD4
counts in young children, and because PCP can occur in
children with counts above the cut-off level, the revised
guidelines (panel 2) propose prophylaxis in all children
born to HIV-1 positive mothers by 46 weeks of age.45
Serious bacterial infections, cytomegalovirus infection,
lymphoid interstitial pneumonitis (LIP), and
encephalopathy are more common in infected children
than in adults. Recurrent and prolonged bacterial
infections are common in children after the first year of
life.42 However, the frequency of bacterial infection is
falling, perhaps as the result of prophylactic antibiotics
(especially co-trimoxazole) and intravenous
immunoglobulin.46 Immunisations against Haemophilus
influenzae and Streptococcus pneumoniae are recommended;
and HIV-1 infected children should receive all routine
immunisations, except BCG, which should be withheld
from children with symptomatic disease in regions where
tuberculosis prevalence is low. Although there have been
no reports of an increased number of adverse reactions, it
is not clear whether immunisations affect HIV replication.
By contrast, children rarely develop Kaposis sarcoma
and other HIV-associated tumours. Progressive
encephalopathy has been reported as the first
manifestation of HIV-1 infection in around 1015% of
children.42 In some children, however, signs of central
nervous system impairment are not manifest until many
years later. The development of encephalopathy is
associated with a poor prognosis.42
LIP is a common chronic pulmonary disease of HIV-1-
infected children. It results from the proliferation of
lymphoid tissue in the interstitium of the lung
parenchyma and produces tachypnoea, hypoxia, and
clubbing of the digits. The chest radiograph shows diffuse
reticulonodular infiltrates, associated with hilar
lymphadenopathy. By comparison with PCP, LIP has a
later onset of symptoms and a much less severe impact.41
Median survival is about five times shorter for children
diagnosed with PCP than for those diagnosed with LIP.
Prognosis for children with LIP is substantially better
than those with any other AIDS-defining condition.42
Consequently, in the CDC revised classification of HIV-1
infection for children (1994), LIP is no longer included
as an AIDS-defining illness, although it has to be
registered.47 Thus, early diagnosis has important
implications for health planning and care provision.

Underlying mechanisms of disease progression

Panel 2: Guidelines for PCP prophylaxis for HIV-exposed Knowledge of the mechanisms underlying different
infants and HIV-infected children (modified from Centers for disease progressions in HIV-1-infected children is
Disease Control and Prevention45) important for planning of interventions. One hypothetical
Age, status of HIV-infection PCP prophylaxis CD4 cell count explanation for such differences is the timing of
monitoring transmission from mother to child. Although not proven,
it is likely that children infected in utero have a faster
Birth4/6 weeks; No 1 month
HIV exposed
disease course than those infected during delivery. If so,
46 weeks to 4 months; Yes 3 months this could be related to the different state of development
HIV exposed of the immune and nervous systems, and to the possibility
412 months; of earlier infection of thymic cells. However, children in
HIV infected or Yes 6, 9, and 12 the European Collaborative Study who developed AIDS
undetermined months in the first year of life could not be distinguished from
HIV infection excluded No None infected children who did not, in terms of mode of
15 years; HIV infected Yes, if At least every 34 delivery, birth weight, gestational age, or perinatal
CD4 count months
<500 cells/L or
findings.43
CD4 percentage Emerging results suggest that overt disease in children
<15% correlates with presence of rapidly replicating syncytium-
612 years; HIV infected Yes, if At least every 34 inducing viruses,48 as also observed in adults. Neonates
CD4 count months can harbour virus variants with both slow and rapid
<200 cells/L or replicating capacity.22,32,48 Recovery of a rapid replicating,
CD4 percentage syncytium-inducing virus from a neonate predicts fast
<15% progression of disease.48 However, as in HIV-1 infected

866 Vol 348 September 28, 1996


adults, fast progression can also occur in the individuals
with slow virus isolates. A high virus burden, evaluated as
virus RNA load in plasma or provirus load in PBMC,
during the first year of life is associated with fast
progression of disease.49
Differences in host immune control of virus replication
by CTL, neutralising antibody response, or ADCC may
be important for disease outcome. We have shown that
children with slow progression of the disease develop a
persistent neutralising antibody response towards their
own virus,48 whereas those with fast progression do not. In
an earlier study, a neutralising antibody response and
ADCC against laboratory strains of HIV-1 in infected
babies correlated with a better clinical status and delayed
disease progression, independently of the virus strain
tested.50 However, a rapid decline in the antibody titre
during the first few months of life suggested that the
antibody was of maternal origin. Thus, transplacental
passage of maternal neutralising antibodies may also play a
part in disease outcome of the infected child.
Conclusion
For diagnosis of HIV-1 infection in children, it is no
longer necessary to wait for clinical signs of AIDS to
appear or for the child to reach 18 months of age, when
conventional serological tests can be used. With
appropriate techniques, early diagnosis is now possible by
3 months of age.
Multivariate analysis of virological and immunological
variables soon after birth should provide a reliable picture
of markers of disease progression. The evidence of
abundant virus replication at an early age has important
implications for clinical management, and for initiation of
medical therapies. The discovery of a class of HIV co-
receptors gives a new insight into virus-host interactions;
control of viral load and cell tropism via different use of
co-receptors has become important to the understanding
of disease progression and transmission.
I thank Eva Maria Feny, Joakim Coster, Emily Carrow, Laura Rancillio,
and Anna Bucceri for their help with the manuscript.
References
1 European Collaborative Study. Children born to women with HIV-1
infection: natural history and risk of transmission. Lancet 1991; 337:
25360.
2 Simonon A, Lepage P, Karita E, et al. An assessment of the timing of
mother-to-child transmission of human immunodeficiency virus type 1
by means of polymerase chain reaction. J AIDS 1994; 7: 95257.
3 Dabis F, Msellati P, Dunn D, et al. Estimating the rate of mother-to-
child transmission of HIV: report of a workshop on methodological
issues, Ghent (Belgium). AIDS 1993; 7: 113948.
4 Courgnaud V, Laure F, Broussard A, et al. Frequent and early in utero
HIV-1 infection. AIDS Res Hum Retroviruses 1991; 7: 33741.
5 Viscarello RR, Cullen MT, DeGennaro NJ, et al. Fetal blood sampling
in human immunodeficiency virus-seropositive women before elective
midtrimester termination of pregnancy. Am J Obstet Gynecol 1992;
167: 107579.
6 Goedert J, Dulliege A-M, Amos C, et al. High risk of HIV-1 infection
for first-born twins. Lancet 1991; 338: 147175.
7 De Rossi A, Zanotto C, Mammano F, et al. Patterns of antibody
response against the V3 loop in children with vertically acquired
human immunodeficiency virus type 1 (HIV-1) infection. AIDS Res
Hum Retroviruses 1993; 8: 22128.
8 Rouzioux C, Costagliola D, Burgard M, et al. Timing of mother-to-
child transmission depends on maternal status. AIDS 1993; 7: S4952.
9 Bryson YJ, Luzuriaga K, Sullivan JL, et al. Proposed definitions for in
utero versus intrapartum transmission of HIV-1. N Engl J Med 1992;
327: 124647.
10 Langston C, Lewis DE, Hammill HA, et al. Excess intrauterine fetal
demise associated with maternal human immunodeficiency virus
infection. J Infect Dis 1995; 172: 145160.
Vol 348 September 28, 1996
THE LANCET
11 Dunn DT, Newell ML, Ades AE, et al. Risk of human
immunodeficiency virus type 1 transmission through breastfeeding.
Lancet 1992; 340: 58588.
12 World Health Organisation. Global Programme on AIDS: consensus
statement from WHO/UNICEF consultation on HIV transmission and
breast-feeding. Wkly Epidemiol Rec 1992; 67: 17779.
13 Lewis SH, Reynolds-Kohler C, Fox HE, Newton JA. HIV-1 in
trophoblastic and villous Hofbauer cells, and haematological
precursors in eight-week fetuses. Lancet 1990; 335: 56568.
14 Henin Y, Mandelbrot L, Henrion R, et al. HIV in the cervicovaginal
secretions of pregnant and non-pregnant women. J AIDS 1993; 6:
7275.
15 Batman PA, Fleming SC, Sedgwick PM, et al. HIV infection of human
fetal intestinal explant cultures induces epithelial cell proliferation.
AIDS 1994; 8: 16167.
16 Scarlatti G, Leitner T, Halapi E, et al. Comparison of variable region 3
sequences of human immunodeficiency virus type 1 from infected
children with the RNA and DNA sequences of the virus populations of
their mothers. Proc Natl Acad Sci USA 1993; 90: 172125.
17 European Collaborative Study. Risk factors for mother-to-child
transmission of HIV-1. Lancet 1992; 339: 100712.
18 Tovo P-A, Palomba E, Gabbiano C, et al. Human immunodeficiency
virus type 1 (HIV-1) seroconversion during pregnancy does not
increase the risk of perinatal transmission. Br J Obstet Gynaecol 1991;
98: 94042.
19 Centers for Disease Control. Zidovudine for the prevention of HIV
transmission from mother to infant. Morbid Mortal Wkly Rep 1994; 43:
28587.
20 Fang G, Burger H, Grimson R, et al. Maternal plasma human
immunodeficiency virus type 1 RNA levels: a determinant and
projected threshold for mother-to-child transmission. Proc Natl Acad
Sci USA 1995; 92: 1210004.
21 Kliks SC, Wara DW, Landers DV, et al. Features of HIV-1 that could
influence maternal-child transmission. JAMA 1994; 272: 46774.
22 Scarlatti G, Hodara V, Rossi P, et al. Transmission of human
immunodeficiency virus type 1 (HIV-1) from mother-to-child
correlates with viral phenotype. Virology 1993; 197: 62429.
23 Rossi P, Moschese V, Broliden PA, et al. Presence of antibodies to
HIV-1 gp120 env epitopes correlates with uninfected status of the
children born to seropositive mothers. Proc Natl Acad Sci USA 1989;
86: 805558.
24 Parekh BS, Shaffer N, Chou-Pong P, et al. Lack of correlation
between maternal antibodies to V3 loop peptides of gp120 and
perinatal HIV-1 transmission. AIDS 1991; 5: 117984.
25 Scarlatti G, Albert J, Rossi P, et al. Mother-to-child transmission of
human immunodeficiency virus type 1: correlation with neutralizing
antibodies against primary isolates. J Infect Dis 1993; 168: 20710.
26 Bryson YJ, Pang S, Wei LS, et al. Clearance of HIV infection in a
perinatally infected infant. N Engl J Med 1995; 332: 83338.
27 Roques PA, Gras G, Parnet-Mathieu F, et al. Clearance of HIV
infection in 12 perinatally infected children: clinical, virological and
immunological data. AIDS 1995; 9: F1926.
28 Newell M-L, Dunn D, De Maria A, et al. Detection of virus in
vertically exposed HIV-antibody-negative children. Lancet 1996; 347:
21315.
29 Clerici M, Sison AV, Berzoisky JA, et al. Cellular immune factors
associated with mother-to-infant transmission of HIV. AIDS 1993; 7:
142733.
30 Rowland-Jones SL, Nixon DF, Aldhous MC, et al. HIV-specific
cytotoxic T-cell activity in an HIV-exposed but uninfected infant.
Lancet 1993; 341: 86061.
31 Borkowsky W, Krasinski K, Moore T, et al. Lymphocyte proliferative
responses to HIV-1 envelope and core antigens by infected and
uninfected adults and children. AIDS Res Hum Retroviruses 1990; 6:
67378.
32 Ometto L, Zanotto C, Maccabruni A, et al. Viral phenotype and host-
cell susceptibility to HIV-1 infection as risk factors for mother-to-child
HIV-1 transmission. AIDS 1995; 9: 42734.
33 Biggar RJ, Miotti PG, Taka TE, et al. Perinatal intervention trial in
Africa: effect of a birth canal cleansing intervention to prevent HIV
transmission. Lancet 1996; 343: 164750.
34 European Collaborative Study. Caesarean section and risk of vertical
transmission of HIV-1 infection. Lancet 1994; 343: 146467.
35 Rogers MF, Ou C, Kilbourne B, et al. Advances and problems in the
diagnosis of human immunodeficiency virus infection infants. Pediatr
Infect Dis J 1991; 10: 52331.
36 Report of a Consensus Workshop, S. Italy. Early diagnosis of HIV
infection in infants. J AIDS 1992; 5: 116978.
37 Borkowsky W, Krasinski K, Pollack H, et al. Early diagnosis of human
immunodeficiency virus infection in children <6 months of age:
comparison of polymerase chain reaction, culture, and plasma antigen
capture techniques. J Infect Dis 1992; 166: 61619.
38 Miles SA, Balden E, Magpantay L, et al. Rapid serologic testing with
867
THE LANCET
immune-complex-dissociated HIV p24 antigen for early detection of
HIV in neonates. N Engl J Med 1993; 328: 297302.
39 Schpbach J, Wunderli W, Kind C, et al. Frequent detection of HIV-
and IgG-specific IgM and IgA antibodies in HIV-positive cord-blood
sera: fine analysis by western blot. AIDS 1989; 3: 58389.
40 Amadori A, De Rossi A, Giaquinto C, Faulkner-Valle G, Zocchello F,
Chiecto-Brianchi L. In-vitro production of HIV-specific antibodies in
children at risk of AIDS. Lancet 1988; i: 85254.
41 Scott GB, Hutto C, Makuch RW, et al. Survival in children with
perinatally acquired human immunodeficiency virus type 1 infection.
N Engl J Med 1989; 321: 179196.
42 Tovo PA, DeMartino M, Gabiano C, et al. Prognostic factors and
survival in children with perinatal HIV-1 infection. Lancet 1992; 339:
124953.
43 European Collaborative Study. Natural history of vertically
acquired human immunodeficiency virus. Pediatrics 1994; 94:
81519.
44 Galli L, de Martino M, Tovo P-A, et al. Onset of clinical signs in
children with HIV-1 perinatal infection. AIDS 1995; 9: 44561.
45 Centers for Disease Control and Prevention. Revised guidelines for
prophylaxis against Pneumocystis carinii pneumonia for children
Public health
Targeting the vulnerable in emergency situations: who is
vulnerable?
Austen P Davis
Summary
Background Emergencies such as wars and natural
disasters increase the vulnerability of affected populations
and expose these populations to risks such as disease,
violence, and hunger. Emergency public health
interventions aim to mitigate these effects by providing
basic minimum requirements, reducing vulnerability, and
reducing exposure to risk. Targeted services are generally
aimed at children under 5. Mortality rates among young
children are higher than the crude mortality rate (CMR)
among the whole population in emergency settings, so
attention is focused on this age group. However, even
under normal conditions mortality is higher in young
children. This analysis compared the relative risk of death
for young children with that for older children and adults
under normal conditions and in emergency settings.
Methods Mortality data from refugee camps set up in
response to three different emergencies were ex amined.
Baseline mortality rates in the refugees countries of origin
were calculated from published data. Relative risks
between normal and emergency conditions were calculated
and compared.
Findings Mortality rates were higher among children under
5 than among older children and adults both under normal
circumstances and in the emergency setting in camps in
Tanzania, Uganda, and Z aire. However, the relative risk for
under-5 versus over-5 mortality was smaller under
emergency conditions than under normal circumstances.
Mdecins Sans Frontires Holland, Nairobi, Kenya
(A P Davis MSc )
Correspondence to: 7 Bishops Road, London N6 4HP, UK
868
infected with human immunodeficiency virus. Morbid Mortal Wkly Rep
1995; 44: 111.
46 Working Group on Antiretroviral Therapy, National Pediatric HIV
Resource Center. Antiretroviral therapy and medical management of
the human immunodeficiency virus-infected child. Pediatr Infect Dis J
1993; 12: 1322.
47 Centers for Disease Control. 1994 Revised Classification System for
Human Immunodeficiency Virus Infection in Children Less Than 13
years of Age. Morbid Mortal Wkly Rep 1994; 43: 110.
48 De Rossi A, Giaquinto C, Ometto L, et al. Replication and tropism of
human immunodeficiency virus type 1 as predictors of disease
outcome in infants with vertically acquired infection. J Pediatr 1993;
123: 92936.
49 Palumbo PE, Kwok S, Waters S, et al. Viral measurement by
polymerase chain reaction-based assays in human immunodeficiency
virus-infected infants. J Pediatr 1995; 126: 59295.
50 Robert-Guroff M, Oleske JM, OConnor EM, Epstein LG,
Minnefor AB, Gallo RC. Relationship between HTLV-III neutralising
antibodies and clinical status of pediatric acquired immunodeficiency
syndrome (AIDS) and AIDS-related complex cases. Pediatr Res 1987;
21: 54750.
Thus, children over 5 and adults are disproportionately
more affected by ex posure to emergency risks than are
younger children.
Interpretation If the objective of intervention, to reduce
mortality, is to be achieved, the population over the age of
5 cannot be ignored. Emergency public health needs to
develop specific tools to investigate risk in other age
groups (as well as children under 5), to identify causes,
and to design programmes to address such needs.
Lancet 1996; 348: 86871
See Commentary page 840
Introduction
Emergency public health interventions aim to mitigate the
adverse health effects of natural and man-made disasters
by providing the basic minimum requirements for healthy
life (food, shelter, water, a sanitary environment, and
access to health care), by trying to reduce exposure to
health threats, and by treating the sick.
Because of the sudden onset of emergency situations
and the scale and urgency, intervention within a very
short time is often essential.1 Information is generally
scarce, and planning and intervention capacity is limited.
The initial intervention decisions must rely partly on
experience from previous disasters. The health problems
created are fairly consistent, differing only in terms of the
severity of the event and the coping capacity of the
affected population.2 A survey of senior relief agency staff 3
indicated that the commonest causes of death in children
in emergency settings were malnutrition, diarrhoeal
diseases, measles, and acute respiratory-tract infections.
Emergency public health interventions generally take
one of two formsnon-targeted programmes in which
Vol 348 September 28, 1996