Optimisation of ACC's fall prevention programmes for older people

FINAL REPORT
November 2008

M. Clare Robertson, PhD BSc(Hons) BCom A. John Campbell, MD FRACP

Dunedin School of Medicine, University of Otago Dunedin, New Zealand

Contact details: Associate Professor Clare Robertson Department of Medical and Surgical Sciences Dunedin School of Medicine, PO Box 913 Dunedin 9054, New Zealand Phone: 03 474 7007 ext 8508 Fax: 03 474 7641 Email: clare.robertson@stonebow.otago.ac.nz

CONTENTS
SUMMARY POINTS AND RECOMMENDATIONS PURPOSE OF THIS PROJECT PROJECT OBJECTIVES REPORT OVERVIEW SUMMARY OF FINDINGS FROM SYSTEMATIC REVIEWS Interventions for community dwelling older people Interventions in long term care and hospital settings Preventing injury from falls Relative cost effectiveness of interventions Rationale for choice of interventions to be modelled ESTIMATED IMPACT OF EFFECTIVE STRATEGIES General approach Methods and model assumptions Otago Exercise Programme Group exercise programmes Multicomponent group exercise Tai chi classes Home safety assessment and modification Cataract surgery Assessment and multifactorial intervention Multiple interventions Multifaceted small group learning Population approach Limitations of modelling methods Modelling falls strategies for the Australian population RECOMMENDED STRATEGIES REFERENCES 1 4 4 7 8 8 10 11 13 15 16 16 16 20 20 20 25 27 31 31 34 34 34 37 39 41 43

APPENDICES A Timeline for project B Methods used in systematic reviews C Interventions for community dwelling older people: systematic review D Interventions in long term care and hospital settings: systematic review E Preventing injury from falls: selected issues F Relative cost effectiveness of interventions: systematic review G Tables of ACC claims and costs

SUMMARY POINTS AND RECOMMENDATIONS

This project aimed to investigate the ways in which effective fall prevention strategies can be optimised for maximum cost effectiveness in preventing falls and injuries in older New Zealanders. Prevention of fall injuries includes 1) reducing the number of falls, 2) reducing the trauma associated with falls, and 3) maximising bone strength at all ages. The literature on maximising bone strength was outside the scope of this project. We used the current literature, meta-analyses, expert opinion, and economic modelling to estimate the potential impact of effective strategies on falls, fractures, and on healthcare costs. The main outcomes of the project are: 1) A comprehensive overview of pertinent information on the best strategies for targeting fall prevention programmes to specific groups of older people, and 2) A comparison of the efficiency (cost effectiveness) and potential for cost savings from the effective fall prevention strategies available internationally.

Falls prevention literature

The literature on falls prevention is large and complex, and requires expert knowledge and experience in the field to assimilate and interpret. Our three systematic reviews for this project included over 100 randomised controlled trials testing falls prevention interventions in community living older people, and 37 were in residential care facilities or hospitals. Given the number of good quality randomised controlled trials, there is no justification for using unsuccessful or untested programmes either for an individual or when choosing strategies for wider dissemination. Preventing falls prevents injuries resulting from falls, but the lack of information on injury outcomes in most of the randomised controlled trials makes estimation of the potential benefit in terms of injury prevention difficult. A systematic review of five controlled trials concluded that the population based approach to the prevention of fall related injury is effective and can form the basis of public health practice.

Community living older people

Exercise programmes were the most common single factor intervention evaluated (48 randomised controlled trials). Effective strategies included the Otago Exercise Programme, multicomponent group classes, and tai chi. Dynamic balance retraining may be the key component of the successful programmes. Exercise programmes that are individually tailored, progress in difficulty, and target carefully selected groups have resulted in the greatest reduction in falls and injuries. Other interventions effective in reducing falls include vitamin D supplementation in those with low vitamin D levels, cataract surgery, cardiac pacing, withdrawal of psychotropic medication, and a small group learning programme. Individual assessment and multifactorial interventions are effective in reducing falls, and this is always the method of choice for clinicians treating an individual patient. However resources are not available for every elderly New Zealand to receive this treatment.

Residential care facilities and hospitals

For those living in residential care, vitamin D supplementation is recommended. ACC has begun a phased national roll out of supplementing vitamin D in residential care facilities. In three individual trials, multifactorial interventions following individual assessment were also shown to be effective in reducing falls in residential care facilities. Effective delivery of these programmes depends on staff expertise, and in New Zealand this differs from the situation in Europe where the interventions were evaluated. There is an increased risk of falling for older hospital patients. Two multifactorial approaches have been successful in reducing falls in a mixture of acute and sub-acute wards. There are guidelines available for falls prevention programmes in both long term care and hospital settings. None of the evidence for successfully reducing falls has been from trials carried out in New Zealand.

Cost effectiveness of falls prevention strategies

Three randomised controlled trials tested interventions which were shown to be cost saving in subgroups in the community at high risk of falls. These were the Otago Exercise Programme, a home safety programme when delivered to those discharged from hospital who reported a previous fall, and a home based, multifactorial programme. Two controlled trials showed that a concerted population approach to fall prevention resulted in a significant reduction of medical costs, one study demonstrating a benefit to cost ratio (hospitalisations) of 26 to 1. We modelled the impact of delivering eight different effective strategies compared with no intervention to 1,000 older New Zealanders. The actual number of falls that would be prevented, and the cost of delivering the intervention were key to determining the potential value for money for each intervention. We recommend four effective strategies for implementation in New Zealand in the short term, Strategies 5 and 6 in the medium, and Strategies 7 and 8 in the longer term.

Recommended falls prevention interventions

1 2 3 4 5 6 7 8

Otago Exercise Programme 80 years, 75 with fall in previous year Tai chi classes (16 weekly classes) 60 years Home safety programme delivered by experienced occupational therapist 65 years, fall in previous year, on discharge from hospital Home safety programme delivered by experienced occupational therapist 75 years, severe visual impairment Multicomponent exercise classes (weekly all year) 65 years, physical performance impairment(s) Stepping On programme 70 years, fall in previous year Home based, individualised multifactorial programme 70 years, 4 of 8 targeted risk factors Cataract surgery 70 years, awaiting surgery

All eight recommended strategies are for community living older people. This reflects the strength of the current evidence and the potential to reduce falls in this setting. The results of our analyses endorse the fact that, to obtain maximum value for money for the population as a whole, effective strategies need to be targeted at those groups which have been shown to benefit most.

Further promising initiatives

ACC is currently funding a pilot research trial of a multifactorial and exercise programme in four long term care facilities in Auckland. A new technology, impact absorbing flooring, has been invented to reduce fractures and other injuries as a result of a fall. This flooring is estimated to be more cost effective than provision of hip protectors to those in long term care. A successful programme of medication review and advice to general practitioners by a pharmacist tested in the UK, would be appropriate for New Zealand residential care facilities.

Future directions

The current evidence indicating the potential for reducing falls in different settings justifies a weighting of ACC funding for falls prevention in older people of 75% in the community setting and 25% in residential care or hospitals. In funding research on injury prevention, ACC should look to answer questions where: a) there is evidence in the international literature but New Zealand conditions are likely to be different, or b) the answer cannot be provided from the literature by meta-analysis of existing studies, and primary or confirmatory research is needed. No information on falls prevention specific to Maori and Pacific populations has been identified in the literature. Research is needed to address this imbalance. Other gaps in the literature include the lack of comprehensive economic evaluations of the effective strategies in residential care or hospitals. Steps for the future include extending the economic models in this project over the longer term, such as the lifetime of people receiving the intervention. It will be important to test the predictive value of the models we developed by systematically collecting cost and effectiveness data for chosen interventions over a period of time. To aid planning and targeting of strategies for optimal value for money, the lack of information on the cost of falls to the New Zealand healthcare system, and to older people and their families, needs to be addressed.

This project was funded by ACC and the Ministry of Health (20072008)

PURPOSE OF THIS PROJECT

ACC and the Ministry of Health funded a comprehensive evaluation of the currently available information to determine 1) the best evidence on effective strategies to prevent falls and injuries in specific population groups of older people, 2) the relative cost effectiveness and the potential for cost savings for each effective prevention strategy identified from the literature, and 3) the best action to take for high claimants to ACC as a results of injuries from falls in order to reduce further claims by this group of older people. Reducing the risk of falls and the incidence of injuries resulting from falls in older people is a priority area for ACC. Given ACC's history of requiring the best available evidence as the basis for their prevention strategies, and given our publication record and international recognition in this area of research, the evidence based information provided to ACC from this project makes a substantial contribution to the advancement of injury prevention in New Zealand. The information produced is also be relevant to the implementation of the New Zealand Preventing Injury from Falls Strategy and to researchers worldwide. Reducing the incidence of injuries resulting from falls in older people is also a priority for the Ministry of Health. The Ministry of Health is the lead agency for the Health of Older People strategy, and supports ACC on the New Zealand Injury Prevention Strategy and the New Zealand Preventing Injury from Falls Strategy. The information produced will support the Ministry of Health to ensure that the most appropriate number and type of programmes are funded to reduce falls in older people. This project used the current literature, meta-analyses, expert opinion, claimant data from ACC, and economic modelling to investigate the ways in which effective fall prevention strategies can be optimised for maximum cost effectiveness in preventing falls and injuries in older New Zealanders.

PROJECT OBJECTIVES
The specific objectives of the project and the process taken to achieve these objectives are outlined below.

Objective 1
A. B.

To provide relevant information on the best strategies for targeting fall prevention programmes to specific groups of older people:

C. D. E.

Specific age groups Different living situations: i) Community dwelling ii) Rest homes iii) Long term hospital care iv) Acute hospital Maori and Pacific populations Older people with specific disabilities (for example, visual impairment) People presenting at emergency departments as a result of a fall.

The information needed for Objective 1 was gathered by undertaking two comprehensive systematic reviews and meta-analyses of the currently available scientific literature which were included in two reports from this project. The systematic review for strategies aimed at community living people was reported in our progress report dated 15 September 2007, and for rest homes residents and hospital inpatients in our progress report dated 15 December 2007. The information gathered in this first phase of the project was then used to develop summaries of the effectiveness of particular programmes and approaches to falls prevention in the subgroups of

older people listed in items A to E of Objective 1. The comprehensive search of the literature was to identify randomised controlled trials testing interventions to reduce falls and fall related injuries in older people. Up to date information about all potential falls prevention programmes is needed to ensure ACC and the Ministry of Health provide a cost effective approach to falls prevention, while optimising the benefits both to older people and to their own organisations. To assist with this part of the process and to ensure complete coverage of the literature, Clare worked with Lesley Gillespie in updating the Cochrane systematic review Interventions for preventing falls in older people. Because of the rapid growth of publications in this field, this review has been split into two, one addressing prevention strategies for community living older people led by Lesley Gillespie, and one led by Professor Ian Cameron, Sydney, for those in residential care and in hospital. Clare is a co-author for both these reviews. The protocols of these two reviews have been published (Gillespie 2008; Cameron 2005). The community review is currently going through the Cochrane peer review process and may be published in the Cochrane Library in February 2009, whereas the institutional reviews is still in a draft form. Our systematic reviews, and the two Cochrane reviews in preparation, report a considerable choice of options for ACC and New Zealand health system funders and providers for strategies that are effective in reducing falls. We consequently considered each of these strategies in terms of their potential impact in reducing the number of falls and injuries, cost effectiveness, and acceptability and feasibility for delivery in New Zealand.

Objective 2
A. B.

Comparison of efficiency (cost effectiveness) and potential for cost savings from the effective falls prevention strategies available:

Firstly ACC will identify the amount and proportion of spending on fall injury claims from older people in different living situations and will provide this information in summary format to the researchers. The researchers will then undertake a formal comparison of the cost effectiveness of effective falls prevention strategies from both the societal and ACC's perspective with the view to optimising the combination and delivery of effective programmes. Should ACC have difficulty in providing the required information then the researchers will use the best available international evidence to inform the analysis.

Two further reviews of the literature were undertaken, and claims data broken down by community living and rest home claimants was received from ACC. We identified and extracted cost data from randomised controlled trials and controlled trials of falls prevention strategies that had included an economic evaluation in the study design or had reported cost outcomes. This review was included in a previous report from this project (progress report dated 15 March 2008). We also extracted data on fracture outcomes from these same trials and reported the results in a previous progress report (dated 15 June 2008). The information from all sources was used to develop the economic models addressing Objective 2 (see section of this report Estimated impact of effective strategies).

Objective 3

To work with ACC to develop a strategy to identify those who injure themselves frequently from falls and to decrease subsequent injuries and the high costs to ACC:

A. B. C.

Firstly ACC in conjunction with the researchers will investigate the information they have available on the characteristics of frequent claimants. The researchers will then work with ACC to design and implement a survey of frequent claimants in order to obtain a profile of relevant information. The researchers will analyse this profile and recommend strategies with the potential for reducing further fall injury claims in this particular group of older people.

Objective 3A

Identifying characteristics of high claimants

The collaboration with ACC to achieve this objective worked well and we are extremely grateful to Paula Eden and Lorna Bunt for their thoughtful input and practical contribution to this part of the project. Tables of data on ACC claims and payments gave information on those aged 65 and older who had a fall claim in the 2006/2007 financial year were received and attached to our progress report dated 15 March 2008. A summary of the information gathered from the tables was also provided. In general terms the tables showed that, for ACC, falls appear a high volume rather than a high cost per claimant issue. This was illustrated by the fact that 56% of the claims to date were for under $250, and 80% were under $700. Only 10% of claims were for over $1,600 and claims of over $10,000 made up 2.3% of the total. Even for people with a history of more than five claims, 80% of these in the 2006/2007 financial year were for $700 or less. In addition 82% of claims were short term, that is less than six months, with 57% less than eight weeks. When this was broken down by type of claim, 66% of claims for medical fees were for less than eight weeks and 38% of entitlement payments were over a period less than six months. The major component of the cost to ACC for fall claims for people with more than five fall claims was for medical fees (78%), with only 8% classed as entitlement payments. We also looked at further tables that gave similar types of data for those who had a fall at age 60 or older. One table of fall claims that were more than $10,000 provided a break down by diagnosis and site, and this showed that, as expected, the majority of these claims were for hip fractures, the most expensive and traumatic of injuries resulting from a fall. The claims and payments data provided valuable information for indicating the costs of fall injuries to ACC. In particular we noted the cost to ACC of falls in community living older people compared with those in rest homes. For all claims to ACC for falls for those aged 65 and older between July 2004 and June 2007, overall 12% were flagged as claims from rest home residents. For those aged 80 years and older, 24% of all falls claims were from rest home residents. The cost to ACC for all falls claims for those aged 80 years or older in 2006/07 was $8.7 million for rest home residents, and $30.1 million for those living in the community. Other claims (that is, claims excluding those for falls) made up around 24% and 36% respectively of the total incurred (other claims were $2.7 million for rest homes residents and $17.2 million for community living) people. Further ACC funding contributes to fall injuries through bulk funding to District Health Boards and to ambulance services, but as these are not broken down by type and cause of injury, the full cost of falls to ACC, and to the New Zealand health system, is not known.

Objective 3B

Survey high claimants

Analyse claimant data

Objective 3C

The initial intention was to carry out a detailed survey of ACC high claimants for falls as part of this project. It was thought that healthcare costs may be saved by providing interventions targeted specifically at those who could be categorised as high claimants. It would be necessary to determine the characteristics of this group of older people in order to choose appropriate prevention approaches. However the tables provided by ACC giving the numbers and payments for claims for fall injuries showed that fall events were a high volume rather than a high cost per claimant issue for ACC. Given these findings, ACC made the decision that the planned survey of high claimants (Objective 3B) would not go ahead. This was based on the view that, for ACC to benefit in terms of cost savings from claims, targeting falls prevention initiatives at high claimants may not be a particularly useful strategy. This decision fits with the findings in the literature which gives valuable information on particular subgroups who will benefit most from falls intervention programmes and in terms of reducing healthcare resource use.

Objective 3D

Recommend strategies

This aspect of the project forms the major new part of this final report. We have worked with ACC to ensure the information we provide is presented in a format that is useful to ACC and the Ministry of Health. To meet this final objective we have developed cost scenarios using the best available information:

We used the evidence base in the literature and our international networks to identify interventions that are effective in reducing falls and injuries in different settings and in different subgroups of older people. We have taken into account issues such as feasibility and acceptability of the interventions, and recommend only those likely to have the potential for support by ACC, the Ministry of Health, District Health Boards, and the primary care sector. We have estimated the cost of delivering each recommended intervention at 2008 prices to 1,000 older people in a particular subgroup, and the potential impact of these programmes on New Zealand healthcare costs using data reported from randomised controlled trials in New Zealand where available (Robertson 2001a, Garrett 2008). We have ranked the identified prevention strategies in terms of effectiveness and potential value for money in preventing falls and injuries from a societal perspective.

REPORT OVERVIEW
This final report from the project contains an overview of the findings presented in each of the progress reports. This is followed by the main body of the report, the section Estimated impact of effective strategies, which has a detailed description of the methodology and findings from the economic models developed in the last phase of this project to compare the potential impact of effective falls prevention strategies. For ease of reading we have presented our summary points and recommendations at the beginning of this document. We have summarised our findings and our recommendations for the next steps to be taken in New Zealand to reduce falls and related injuries in older New Zealanders. Appendices to the final report A separate document Optimisation of ACCs fall prevention programmes for older people: appendices to final report has been compiled to incorporate the results from all our progress reports for this project. The timeline for the project is provided in Appendix A. We describe the methods used in compiling the systematic reviews (Appendix B), and summarise the findings from the systematic review of interventions for community living older people (Appendix C), those in long term care or in hospital (Appendix D), relevant issues on preventing injury from falls (Appendix E), the systematic review on economic evaluation of interventions reported within randomised controlled trials (Appendix F), and in Appendix G the tables ACC generated showing ACC claims and costs.

SUMMARY OF FINDINGS FROM SYSTEMATIC REVIEWS

Prevention of fall injuries includes 1) reducing the number of falls, 2) reducing the trauma associated with falls, and 3) maximising bone strength at all ages. The literature on maximising bone strength was outside the scope of this project. Three systematic reviews, and a literature overview on selected topics concerning falls injury reduction, were completed as part of this project. A summary from each review is provided in this section, and the methodology used and the full report from each review as included in the progress reports are provided again in the Appendices of this final report. These reviews were essential to the project because the Cochrane review on falls prevention interventions is now out of date (Gillespie 2003). We divided the literature reviews into those addressing falls prevention in community living older people, and strategies aimed at people in long term care or in hospital. This is because the characteristics of people in terms of falls risk factors in these settings differ, and the types of effective intervention strategies and the health professionals who will deliver them may also differ. Terminology used In this project we divide interventions into single factor, multiple and multifactorial programmes. These terms are defined in the taxonomy for falls prevention interventions developed by ProFaNE (Prevention of Falls Network Europe), funded by the European Commission. Single interventions address one type of risk factor for falls whereas multiple and multifactorial programmes have multiple components and address more than one type of risk factor. Examples of single interventions are exercise programmes and vitamin D supplementation. Multifactorial interventions include a risk assessment with programme components to address the identified risks for each individual. In a multiple programme, the same components are delivered to each individual. In the falls prevention literature the lack of standard terminology when reporting results from clinical trials is problematic. The main difficulty is differentiating the results concerning the number of falls during the trial (falls), from the number of participants who had a fall (fallers). For this project we have chosen to use rate of falls and the term rate ratio for results comparing the number or rate of falls in the intervention group compared with the control group, and risk of falling and the term risk ratio for comparisons concerning number (proportion) of fallers. We did not pool these two types of results (falls, fallers) in our meta-analyses, as they each indicate quite different aspects concerning effectiveness of the intervention. A rate ratio is the preferred statistic because it answers the question Were the number of falls reduced by the intervention? A risk ratio, the cruder statistic of the two, addresses whether more people fell in the intervention group compared with the control group during the trial. Our own research has included the choice of appropriate statistical techniques for analysing the results of falls prevention trials (Robertson 2005).

Interventions for community dwelling older people

A total of 103 randomised controlled trials testing a falls prevention programme in a community setting were identified and reviewed (41,348 participants) (see report dated 15 September 2007). Since then eight more randomised controlled trials have been identified. These will be included in the Cochrane systematic review due to be published next year, therefore we have not added them to the review in this project. In the 103 trials identified and meeting the inclusion criteria for the community setting, the interventions tested were: 81 single factor interventions 29 multifactorial programmes 11 multiple intervention programmes

 A total of 121 interventions were tested. The most common intervention was an exercise programme (48 were tested) and 12 used vitamin D supplementation. No information on falls specific to Maori and Pacific populations was identified in the literature. Interim results from a randomised controlled trial where monofocal glasses were provided to the intervention group were reported at the 3rd Australian and New Zealand Falls Prevention Society Conference in Melbourne in October 2008 (Haran 2008). Falls were reduced in those who tended to walk outside frequently, adding to the evidence that for older people who wear glasses, monofocal not bi-focal or multifocal glasses are recommended while walking. The pooled results from meta-analyses of trials with the same or similar interventions showed that the following interventions were effective in reducing falls in community living older people. Interventions successful in reducing falls in community living older people Number of randomised Pooled rate ratio controlled trials pooled (95% confidence interval) Otago Exercise Programme 3 0.66 (0.52 to 0.83) Tai chi classes 5 0.60 (0.50 to 0.73) Group exercise programmes 9 0.71 (0.62 to 0.81) Home safety programmes 3 0.83 (0.64 to 1.07) At high risk of falls 2 0.69 (0.56 to 0.84) Psychotropic medication withdrawal 1 0.34 (0.16 to 0.74) Cataract removal 2 0.67 (0.49 to 0.91) Cardiac pacing 1 0.42 (0.23 to 0.75) Multifactorial/multiple interventions 11 0.75 (0.65 to 0.88) Pooled risk ratio (95% confidence interval) 0.86 (0.75 to 1.00)

Vitamin D supplementation

In the pooled analysis, assessment and multifactorial interventions were effective in reducing falls, and this approach is recommended in falls prevention guidelines (NHS 2004, American Geriatrics Society 2001). However care is needed when planning the implementation of this strategy. In a recent randomised controlled trial in Upper Hutt, falls were not reduced when a falls nurse co-ordinator delivered this evidence based intervention (Elley 2008). It may be that this type of approach is more effective when referral to other health professionals is not the major mode of implementation. Vitamin D supplementation in our pooled analysis of falls risk did not quite meet significance (the upper limit of the 95% confidence interval included 1.00). It may be that this intervention is more effective in those with a low vitamin D level (preliminary results from the Cochrane community review update to be published in 2009). Although this medication is low cost, the cost of GP time and the blood test to determine vitamin D level would need to be considered in determining the cost effectiveness of this approach in the community. In one trial during winter months, an anti-slip device (Yaktrax Walker) reduced the rate of outdoor falls in community living people aged 65 or more who had fallen in the previous year, compared with the control group wearing usual winter footwear (McKiernan 2005). Interventions not successful in reducing falls in the community Currently there are several interventions trialed that were not effective in reducing falls. These include exercise programmes with only one type of exercise, for example resistance training only, walking groups, impact exercises, computerised balance exercises, agility training only, weight bearing exercises only, step up step down exercise only, and stretching and weight shifting exercises. Other interventions that have been tested and were not effective include

hormone replacement treatment, nutrient supplementation, falls education alone, and cognitive behavioural programmes. A recent randomised controlled trial has shown that an intervention aimed at reducing falls by improving vision actually significantly increased, rather than decreased, the rate of falls (Cumming 2007). Although vision should be optimal and cataract removal does prevent falls, older people should be advised to be extremely careful while adjusting to major changes in lens prescriptions.

Interventions in long term care and hospital settings

In addition to the randomised controlled trials in the community, a total of 37 randomised controlled trials meeting the inclusion criteria and testing programmes for those in rest homes and in hospital patients (both acute and long term acre) were identified (see report dated 15 December 2007). Eight of the trials were carried out in acute- or sub-acute wards in hospitals, or a combination of these wards (2,862 inpatients; 1,146 men; 1,716 women), and the remaining 29 trials were in long term care facilities (17,291 residents; 3,765 men; 13,481 women; 45 gender not specified). The interventions tested were: 26 single factor interventions (6 were tested in hospital wards) 11 multifactorial programmes (3 were tested in hospital wards) 4 multiple intervention programmes (none was tested in a hospital) A total of 41 interventions were tested. The most common single factor intervention was an exercise programme (11 were tested) and six trials used vitamin D supplementation. Three trials of multifactorial interventions showed good success in residential care homes in Europe (Becker 2003, Dyer 2004, Jensen 2002). Components of the interventions included educating staff on fall prevention, implementing exercise programmes, modifying the environment, supplying and repairing aids, reviewing drug regimens, providing free hip protectors, and having post-fall problem solving conferences. However, it may not be possible to deliver these interventions in a similar way in New Zealand given the different levels of staff expertise. It should also be noted that one falls prevention intervention in New Zealand rest homes significantly increased, rather than decreased, the rate of falls (Kerse 2004). This illustrates the fact that falls prevention is not an intuitive process, and only some interventions or combination of interventions will work in different settings and in different subgroups of older people. One promising intervention that would be appropriate for New Zealand was a medication review in rest home residents by a pharmacist (Zermansky 2006). After six months there was a reduction in the number of drug changes and in the number of falls. It was noted that general practitioners do not regularly review residents medications, but in this UK trial did accept 76% of the pharmacists recommendations. Meta-analyses pooling results from trials with the same or similar interventions showed that the following were effective in reducing falls in people in residential care facilities and in hospital. Key findings from the systematic review follow. Interventions successful in reducing falls in long term care facilities and hospitals Number of randomised Pooled rate ratio controlled trials pooled (95% confidence interval) Residential care facilities Vitamin D supplementation 5 0.65 (0.48 to 0.87) Hospital inpatients Multifactorial interventions 2 0.69 (0.54 to 0.88)

10

Summary of findings

Although falls are three times more common in long term care facilities than in community living older people, fewer randomised controlled trials testing effectiveness were identified in this setting (37 versus 111 now in the community). Vitamin D supplementation was the only intervention, when trials testing this approach were pooled, to show a significant benefit in terms of falls reduced in long term care facilities. In three individual trials in Europe, falls were reduced by using a multifactorial approach in a long term care setting. Effective delivery of these programmes depends on staff expertise, and in New Zealand this differs from the situation in Europe where the interventions were evaluated. Overall the current trials demonstrate no benefit on falls in using an exercise progamme alone. However, the types of exercise programmes tested have varied and do not appear to emulate successful community based exercise programmes. In a hospital setting there is an increased risk of falling for older people. Two multifactorial approaches have been successful in reducing falls in a mixture of acute and sub-acute wards. There are guidelines available for falls prevention programmes in both long term care and hospital settings. None of the evidence for successfully reducing falls has been from trials carried out in New Zealand.

Preventing injuries from falls

Extracting and pooling the information on injury as a result of a fall from the falls intervention trials identified proved problematic as the definitions of injuries differed markedly in the trials, or there was no definition of injury provided. A limited number of studies included in the systematic reviews reported the number of participants in each group who had a fracture as a result of a fall during the trial. This outcome is more robust as fractures can be verified by X-ray. We pooled fracture risk ratio data when appropriate, and when there were results from two or more studies available to pool. Interventions successful in reducing fractures in falls prevention trials Number of randomised Pooled risk ratio controlled trials pooled (95% confidence interval) Hip fractures Residential care facilities Multifactorial interventions 3 0.49 (0.24 to 0.98) Fracture (any site) Community living Group exercise programme

0.34 (0.16 to 0.72)

Summary and implications of findings from fall related injuries reported in the randomised controlled falls prevention trials

Although there has been a rapid growth in the number of randomised controlled trials published that tested falls prevention interventions in older people, only 27% (38 of 140) of the trials reported injury events or falls resulting in medical care being sought. Marked variations in the definition of a fall related injury event in the trials meant that these data, even if reported, could not be pooled.

11

Information gained from the falls prevention trials about the number of participants who sustained a fracture as a result of a fall is more robust since fractures can be verified by X-ray, but a limited number of trials reported this outcome. Three trials of multifactorial programmes in residential care reported the number of participants with a hip fracture as a result of a fall during the trial. The pooled risk ratio showed a significant reduction in fractures, but this finding needs to be viewed with caution due to the very small number of hip fracture events recorded. Pooling the risk ratio for sustaining a fracture from the three trials of exercise programmes in the community showed the number with fractures were significantly reduced. However, this result also must be viewed with caution as it represents only 6% (3 of 48) of the trials testing exercise programmes in the community. In addition, the exercise programmes differed in that two were group exercise programmes and one tested a home based programme for people with Parkinsons Disease. Given the small number of trials with fracture data available for pooling, this information did not provide definitive data we could confidently use in developing the models reported in this document.

Hip protectors and safety flooring

Currently there are several biomechanical means of reducing the trauma resulting from falls being developed and tested. Two that aim to prevent fractures from falls are hip protectors and safety flooring. Hip protectors Hip fractures, the most costly and traumatic injury from a fall, result almost exclusively from a simple fall and the impact of the greater trochanter with the floor (Dargent-Molina 1996, Norton 1997). Studies have addressed the problem by using hip protectors, but compliance has been estimated at only 25-30% (Chan 2000). Further problems with hip protectors result from displacement of the pads from the greater trochanter region prior to the fall. In the New Zealand study of the circumstances and consequences of falls in residential care, 35 falls occurred with hip protectors being worn at the time (Butler 2004). There were no hip fractures when residents fell wearing hip protectors, but over the 18 month period there were 12 hip fractures in the 917 falls without a hip protector in place. ACC has already investigated this approach by commissioning a review on the potential effectiveness of hip protectors in New Zealand residential care facilities, plus a cost effectiveness evaluation, both of which have informed this project. The evidence for effectiveness in preventing hip fractures indicates that hip protectors should be offered on an individual basis, and provision of hip protectors have formed part of the successful individualised multifactorial interventions in two trials in residential care facilities (Becker 2003, Jensen 2002). Safety flooring The researchers have been collaborating with the New Zealand inventor in the development and testing of an innovative new flooring material designed specifically to absorb energy on impact (Robertson 2006). This has the potential to reduce the number of fractures and other injuries as a result of falls in elderly people, especially those living in long term care (Cali 1995). Compared with known injury prevention programmes, this strategy has the advantage of being a long term approach plus eliminating the problems of individual compliance (Drahota 2007). A cost effectiveness evaluation of the new flooring has been carried out at the School of Population Health, University of Auckland by health economist Dr Paul Brown (Brown 2008). For residential care facilities this flooring material is more cost effective than supplying hip protectors, and has the potential to be cost saving from ACCs perspective.

12

We reported the results of our own study comparing gait patterns and balance of elderly women walking and standing on three different types of flooring at the 3rd Australian and New Zealand Falls Prevention Society Conference in Melbourne in October 2008 (Robertson 2008). There were no differences in balance meaures or in gait patterns between a standard noncompliant flooring (vinyl), a compliant flooring (carpet with good quality underlay) and Kradal, the new safety flooring manufactured by Acma Industries in Upper Hutt, Wellington. We concluded that Kradal had the potential to reduce fractures without affecting balance or walking patterns.

Relative cost effectiveness of interventions

ACC and the Ministry of Health require not only evidence of effectiveness but also value for money when funding and supporting falls prevention strategies. Economic evaluation is the systematic, explicit analysis of alternative courses of action, assessed in terms of both costs and consequences. Economic evaluations provide a measure of efficiency for the purposes of decision making by comparing the costs and the consequences of alternative programmes (Drummond 2005). There are four main types of economic evaluations: 1) cost analysis, 2) cost effectiveness analysis, 3) cost utility analysis, and 4) cost benefit analysis. In a total of eight trials in a community setting, a comprehensive economic evaluation was reported for the seven different interventions tested (eight cost effectiveness analyses, one cost utility analysis, see report dated 15 March 2008). We found no comprehensive economic evaluations in the trials in residential care or in hospital. An article on the cost effectiveness of withdrawal of fall-risk-increasing medications in older people attending geriatric outpatients has recently been published (van der Velde 2008). When extracting data from the 103 randomised controlled trials in the community and 37 in an institutional setting identified in our systematic reviews, we noted the following economic evaluations had been carried out within these trials: For 8 trials testing an intervention in the community setting, the costs of delivering the intervention were reported For 1 intervention tested in a long term care setting, the costs of delivering the intervention were reported For 1 intervention in the community and 2 in a long term care setting, the costs of some healthcare use during the trial was reported, but no comprehensive economic evaluations were undertaken No information was reported on the costs or cost effectiveness of delivering a falls prevention programme in a hospital setting In a total of 8 trials in a community setting, a comprehensive economic evaluation was reported for the 7 different interventions tested (8 cost effectiveness analyses, one cost utility analysis). Investigation of the cost effectiveness of effective interventions has provided information on the likely value for money of several types of approach in the community setting, but none in long term care and hospitals. For our final recommendations we considered relevant information from peer reviewed articles of controlled (but not randomised) trials, and papers published on the web but not in peer reviewed journals. Examples of interventions evaluated and reported in this way include a medical vigilance system to reduce patient falls in hospital, cost effectiveness of methods to prevent hip fractures, a further economic evaluation of the Otago Exercise Programme tested in a controlled trial (Robertson 2001d), and the Queensland Stay on your feet programme (Beard 2006). The results from this systematic review were used to develop the models in the section of this report Estimated impact of effective strategies (see Tables 1, 6, 8, 10, 12, and 18).

13

Summary of findings from the costs, cost effectiveness and cost utility analyses in the randomised controlled trials

Although there has been a rapid growth in the number of randomised controlled trials published that tested falls prevention interventions in older people, analyses of the costs and cost effectiveness of the interventions within the trials to date is limited. Comprehensive cost effectiveness analyses were carried out in only eight of the 140 randomised controlled trials included in the systematic reviews of falls prevention interventions. All eight trials tested strategies that were effective in reducing falls in community living older people. For three interventions, the Otago Exercise Programme, psychotropic medication withdrawal, and a home safety programme effective in reducing falls in people with severe visual impairment, the costs of delivering the programmes and the cost effectiveness in New Zealand have been established. One trial in community living older people reported a cost utility analysis and extended the time period of the analysis from the trial duration to the participants remaining lifetime. There are limitations in using quality of life years gained (QALYs) for economic evaluations of complex interventions for older people since these interventions result in multiple benefits not captured by QALYs (Harwood 2008). We have not found quality of life measures sensitive to change in our falls prevention studies despite the beneficial outcomes of the trials. The negative effect on health related quality of life is larger for self reported fear of falling than for falls or a fracture (Iglesias 2008). The information provided by these comprehensive economic evaluations indicates there is some, although limited, evidence that falls prevention strategies can be cost saving during the trial period, and may also be cost effective over the participants remaining lifetime. The results of these analyses also endorse the fact that, to obtain maximum value for money, effective strategies need to be targeted at particular subgroups of older people. The cost of delivering the intervention and health service costs for participants during the trial were available for only one of the trials in residential care. No cost effectiveness analyses in this setting were reported. For the randomised controlled trials in a hospital setting, no cost data or cost effectiveness analyses were reported.

Other findings ACC estimated that supplementation with vitamin D in residential care could lead to a reduction of over 5,000 falls, and 330 claims to ACC, potentially saving an estimated $NZ 2.3 million in the 2008/2009 financial year (Williams 2008). A paper presented at the 3rd Australian and New Zealand Falls Prevention Society Conference in Melbourne in October 2008 investigated the incremental cost effectiveness of fall risk screening in hospitals (Haines 2008). The conclusion was that preventing inhospital falls with a targeted falls prevention intervention approach using physiotherapist clinical judgement, was more cost effective than a no intervention approach. Two controlled trials showed that a concerted population approach to fall prevention resulted in a significant reduction of medical costs, one study demonstrating a benefit to cost ratio (hospitalisations) of 26 to 1 (Beard 2006), Tinetti 2008).

14

Rationale for choice of interventions to be modelled

ACC and the Ministry of Health require information to use for supporting strategies which will give the most benefit in terms of effectiveness and cost effectiveness in reducing falls and injuries. The approaches also need to be acceptable and feasible for the New Zealand healthcare system. In choosing approaches to recommend, we took into account the type of intervention, the health professionals delivering the intervention, the different settings, and characteristics of potential participants including age group. We considered the evidence carefully for all these factors and selected the following effective interventions to investigate further: the Otago Exercise Programme, tai chi and multicomponent exercise classes, home safety assessment and modification programmes delivered to selected subgroups of older people, cataract surgery, multifactorial programmes aimed at the persons individual fall risk factors, and two multiple approaches the multifaceted group learning programme Stepping On, and a population approach. We modelled the likely impact of these interventions when delivered to 1,000 older New Zealanders. The methods used and the results of the modelling are reported in the next section of this report. Several further strategies that were shown to be effective in reducing falls in the randomised controlled trials were considered but not investigated further. These include cardiac pacing, an effective treatment for a very select subgroup of older people, and the gradual withdrawal of psychotropic medications, which our own research has shown was successful in reducing falls in people regularly taking these medications (Campbell 1999, Robertson 2001d). While withdrawal of sleeping medication is recommended for a motivated individual, there are several problems to be overcome for widespread dissemination. We consider the resources needed to implement this approach would be considerable. GPs would need training and incentives to prescribe withdrawal to their patients. This would involve outreach visits (oneon-one visits to GPs by a pharmacist or other health professional) and primary care professional meetings led by a respected expert. In our trial we found people reluctant to withdraw from their medication, and one month after the end of the trial, eight of the 17 who had withdrawn successfully, had restarted their medication. We consider that a psychologist would need to be involved in order to improve uptake and adherence, and provide support and advice on alternative methods for sleep, in order to prevent people restarting their psychotropic medication. We have not modelled the potential impact of the two types of successful interventions in residential care and hospitals. Our meta-analyses and those of others conclude that vitamin D supplementation is an effective strategy to reduce falls, particularly for those in residential care facilities. It is accepted therefore that this strategy should be used for falls prevention in this setting. As ACC has recently commenced implementation of this strategy nationally (Williams 2008), we have not investigated this approach further in this project. Although three trials of multifactorial interventions have shown good success in rest homes in Europe (Becker 2003, Dyer 2004, Jensen 2002), the New Zealand situation means that fewer staff have been trained and are available to implement the components of the strategies. ACC is currently funding a pilot trial of an exercise programme and a multifactorial approach based on the Green Box in four Auckland rest homes. We are co-investigators for this research which is led by Associate Professor Ngaire Kerse at the School of Population Health, University of Auckland. A multifactorial approach was effective in longer stay hospital patients in Australia (Haines 2004), but no data were available on the costs of this particular intervention. District Health Boards are required to have an in-hospital falls prevention policy, and should use the published guidelines until more research is available.

15

ESTIMATED IMPACT OF EFFECTIVE STRATEGIES General approach

In this section of the project we have estimated the likely impact on healthcare costs of the interventions shown in the systematic reviews to be effective in reducing falls and with good rationale for dissemination in New Zealand. The economic models we have developed are aimed at assisting in clinical and health policy decision making regarding falls prevention strategies for older people. They involve assumptions and are intended to give an indication only of the relative impact of each intervention. Sources of data The systematic reviews in this project provided information on which interventions are effective in reducing falls in each of the settings and for each subgroup outlined in Objective 1. The systematic review of cost outcomes and economic evaluations reported in the randomised controlled trials provided information on the incremental cost effectiveness compared with usual care for several, but not all, of the effective interventions. Cost effectiveness and cost utility analyses do not attempt to put a monetary value on the fall events that were prevented by the intervention. The measure of relative cost effectiveness and cost utility in these studies is expressed as a ratio of the dollar amount for healthcare as a result of delivering the programme, compared with usual care per fall event prevented, or per QALY gained. For the models in this section we have combined information on effectiveness of falls prevention strategies obtained from the literature, with information on the costs of programme delivery and the consequences of programme delivery in terms of health care costs incurred or averted. In addition to the randomised controlled trials in the systematic review, we have used published information on effectiveness and cost effectiveness from controlled trials, that is, trials in which participants were not randomised to the groups used for comparison purposes. If the cost per person of delivering the intervention in the trial was not available we have calculated this to use as an estimate. The models We have intentionally kept the models as simple and transparent as possible. In each model we estimated the likely total costs incurred, or cost savings, if the intervention is delivered to 1,000 participants. First we estimated the number of falls per person receiving the intervention that were prevented over a one year period in the relevant clinical trial. We then used the incremental cost per fall prevented reported in the trial, or if this was not available, an estimate of the mean cost per person for delivering the intervention and a mean cost of a potential fall event, and extrapolated the likely total costs or cost savings after one year for 1,000 participants.

Methods and model assumptions

The main assumption in the models developed in this section of the report is that the results from the trials (effectiveness in terms of number of falls per person prevented, mean costs of intervention delivery and other healthcare use, incremental cost effectiveness ratios) can be extrapolated for 1,000 participants simply by multiplying the per person and per fall event trial results by 1,000. This assumption ignores the importance of issues such as economies of scale, and the likely variations in costs and effectiveness within a group of real world participants due to the delivery scenario, the particular setting, the type of intervention, or related to the fall risk mix of the group themselves. It should be noted that the economic evaluation techniques we have incorporated assume that any cost savings identified will result in this same amount being spent by ACC or other health agencies on alternative worthwhile programmes, whether for falls prevention or other injury prevention or healthcare activities. However, if the freed or saved amount is in fact wasted by being spent on ineffective or untested programmes, costs to ACC and the health system may actually increase, rather than decrease.

16

Viewpoint of the models We have stated the perspective taken for estimating healthcare use and costs in each particular model. Wherever possible we have used a societal perspective the broadest perspective because of the broad nature of the problems caused by falls (Drumond 2005). This perspective considers not only costs to the health system, but also to the fallers themselves, their carers, and their families. The models could be used by ACC to estimate the impact from their own perspective by substituting the price ACC pays, or is willing to pay, per person for delivery of a particular intervention, and the average cost to ACC of fall events in the subsample and setting of the older people being targeted. Timeframe For simplicity and ease of comparison, all our models are based on estimates of the number of falls prevented during one year only from the start of the intervention. One exception is the model for 16 weeks of tai chi classes, where falls were monitored for only six months in the trial with cost effectiveness information available (Voukelatos 2007). The benefit of an intervention will not always cease after programme delivery ceases. For example, if people who receive an exercise programme keep exercising, the benefits will continue for no further costs. Similarly for equipment provided and behaviour change sustained long term following a home safety programme. Therefore the timeframe of one year taken is likely to underestimate the cost effectiveness of the interventions. Target group In each of the models we have assumed that the group being targeted and choosing to participate in the real world setting would be similar to the group who took part in the trial (the one being used as the basis for the particular model), in terms of age group and presence of other fall risk factors. Measure of effectiveness The interventions chosen for this section were those shown in randomised controlled trials to be effective in significantly reducing the rate of falls. There is no justification for using unsuccessful or untested programmes either for an individual or when choosing strategies for wider dissemination. In clinical trials, the effectiveness of a falls prevention intervention is expressed as an incidence rate ratio the rate of falls during the trial for the intervention group compared with (divided by) the rate of falls in the control group. Since the control group is arbitrarily assigned an incidence rate of 1.00, an incidence rate ratio of 0.60 indicates there was a 40% reduction in the rate of falls in the intervention compared with the control group during the trial. When the upper limit of the 95% confidence interval for the incidence rate ratio is below 1.00, the interpretation of the trial results is that fall events were significantly reduced by the intervention. However the impact of the intervention in terms of fall events prevented, and therefore the potential for cost savings in healthcare, depends on the actual number of falls and injuries prevented, not just the relative rate reduction between the control and intervention groups. A baseline fall rate of 2.00 falls per person year which is reduced by 40% results in 0.8 falls being prevented per person after one year, whereas a baseline fall rate of 1.00 per person year in those receiving the intervention means 0.4 falls per person would be prevented. Although both these scenarios result in a 40% reduction in falls, the first example will result in double the number of fall events prevented during one year for the same number of people receiving the intervention. Therefore we have chosen the number of falls prevented per person receiving the intervention in the particular trial forming the basis of the model as our estimation of effectiveness in each model. Specific assumptions are that:

The number of falls per person in the control group during the trial is generalisable to New Zealanders in the same age group and with similar characteristics in terms of fall risk who are not offered the intervention, or who choose not to participate.

17

The number of falls per person in the intervention group during the trial, and the reduction in falls calculated as the difference between falls per person in the control and intervention groups, is generalisable to New Zealanders in the same age group and with similar characteristics in terms of fall risk who would accept the intervention if offered. Effectiveness calculated as the number of falls prevented per person gives a better estimation of likely impact in the real world than the number of falls prevented per person year. In the trials, falls were generally monitored for one year, but in real life, the number of falls prevented per person taking part in the programme would be the measure of success.

Adherence to the intervention The use of clinical trial results for developing our models means that the levels of adherence to the intervention and the drop out rate are assumed to be similar to those in the relevant trial. In some cases adherence rates may be greater, in some cases lower, particularly if less experienced or motivated staff deliver the programme or a different subgroup of older people are offered the same programme. Particular subgroups (for example frail or unfit individuals taking part in an exercise programme; fitter people in good health) may respond differently to varying levels and amounts of professional assistance and support (King 2006). Predictors of adherence are specific to each intervention and cover a wide range of physiological, demographic, psychosocial, health related, and environmental factors. Estimation of costs in 2008 The trials reported cost outcomes in different currencies and from different years. To enable comparison of these costs we present all monetary values in the reported currency and also in 2008 New Zealand dollars. We first used Purchasing Power Parity values in the year of the prices to convert the currency to New Zealand dollars, and then inflated this value to 2008 prices using a Reserve Bank of New Zealand web calculator based on the New Zealand consumer price index (http://www.rbnz.govt.nz/inflationcalculator/calculate.do). Costs of implementing the intervention An assumption in each of the models is that the average cost for delivering the intervention to 1,000 participants will be similar to the average cost as reported in the relevant trial. That is: 1) a similar recruitment strategy will be used and the uptake rate will be similar, 2) costs of training staff to deliver the programme will be similar, 3) there will be a similar input in terms of time by those delivering the programme and the same salary levels, and 4) administration and the costs of equipment and other resources will be similar. When the cost of delivering the intervention was not reported in the trial, we estimated the average cost per person by identifying the cost items that were required and valuing these items at 2008 New Zealand prices. We used 25% of resource use to estimate the cost of overheads (office accommodation, administration, financial services, computer use, and depreciation of equipment used for running the programme). Healthcare costs of a fall event We have used two published estimates of the cost of a fall in community living older New Zealanders as appropriate in our models. One trial reported the hospital and all other healthcare costs related to falls from a societal perspective for 233 women aged 80 years and older taking part in a randomised controlled trial of the Otago Exercise Programme (Robertson 2001a). During this trial 27% of total hospital costs for these participants were related to falls. The total cost of all fall related healthcare for the first year of the trial was $120,401 at 1995 prices, and 240 falls were reported. Therefore the average cost of a fall to New Zealand society was $501.67 ($668.43 at 2008 prices). Fall related costs were defined as those incurred from the day of the fall until the end of the following month. This underestimates the cost of a fall event. If a person received follow up home services or was admitted to a rest home as a result of the fall, substantial costs would continue to be incurred longer term.

18

In the two years of this trial, 22% (77 of 358) of the falls resulted in healthcare services being used (Robertson 2001a). Of note is the fact that hospital admission costs made up 90% of the total fall related costs. This means that fall related hospital admission costs averted are a good indication of the potential healthcare cost savings as a result of preventing fall injuries. In a randomised controlled trial of women and men aged 75 years and older with a fall in the previous year carried out in the Hutt Valley in 2005, the healthcare costs related to falls were monitored for six months for the first 202 participants (Garrett 2008). Of the 97 falls reported during this time period, 18 (19%) required medical care but there were no hospital admissions, and the median cost of a fall (societal cost) was $153.53 ($169.38 at 2008 prices). Cost effectiveness of the intervention The incremental cost effectiveness ratios reported in the clinical trials were of three main types: 1) the ratio was the incremental cost of delivering the intervention compared with usual care per fall event prevented during the trial (the difference between the costs of delivery of the intervention and the costs of the control group activity if any, divided by the difference in the number of falls in the control group and the intervention group), 2) the incremental costs used to calculate the ratio included not only the cost of delivering the intervention, but also all fall related healthcare costs during the trial, and 3) the incremental costs used to calculate the ratio included all healthcare costs during the trial, whether or not related to falls. An assumption of our models is that the incremental cost per fall prevented reported for a particular intervention compared with usual care in the relevant clinical trial, is generalisable to the New Zealanders in the same age group and with similar characteristics in terms of fall risk who would accept the same intervention if offered. Costs or cost savings from delivering each intervention We have used two ways of calculating the total costs or cost savings from delivering the intervention to 1,000 participants. When an incremental cost effectiveness ratio was available from the trial, we arrived at our estimate by multiplying the incremental cost per fall prevented by the likely number of falls that would be prevented in one year. When a cost effectiveness ratio was not available, we first estimated the likely savings in healthcare costs from the falls that would be prevented in one year, and took the total costs or cost savings to be the difference between the cost of programme delivery to the 1,000 participants and these potential healthcare cost savings. The assumptions here are that all the data used in these calculations are accurate, appropriate, and generalisable. An outline of each intervention, and tables showing the sources of data used in the models and the models themselves, are provided in the next section.

19

Otago Exercise Programme

The strong research base for this home based, muscle strengthening and balance retraining programme comes from our own research group at the Dunedin School of Medicine, University of Otago. This is the only falls prevention programme with more than one trial demonstrating effectiveness in reducing falls and injuries (Campbell 1997, Robertson 2001b, Robertson 2001c). Overall in four trials with 1,016 people aged 65 years to 97 years in nine cities and towns in New Zealand, the Otago Exercise Programme reduced the rate of both falls and fall injuries by 35% (Robertson 2002). Subgroups benefiting most were those aged 80 and older, and those with a previous fall. The Programme was effective when delivered by a research physiotherapist and by nurses, trained and supervised by the physiotherapist. ACC provided the research funding for the first two trials, and has supported nationwide dissemination of the Programme for some time. Training of physiotherapists as exercise instructors or to train and supervise nurses to deliver the Programme is through a website coordinated by the Auckland University of Technology. ACC published a manual for instructors and provides the manual free of charge to all New Zealand health professionals (ACC 2003). The Otago Exercise Programme is now used routinely throughout the world. The evidence for effectiveness and cost effectiveness reported in the literature is summarised in Table 1. Results from the West Auckland trial showed that the Otago Exercise Programme was cost saving for those aged 80 years and older (see also Table 18). A further randomised controlled trial of the programme is in progress in Vancouver, Canada. Interim results show that the Programme reduced falls in the first 74 people in the trial. Further exciting results show that executive functioning was significantly improved in those receiving the exercise programme compared with the control group (Liu-Ambrose 2008). This may help explain the mechanism whereby exercise reduces falls, that is, not only by improving balance and muscle strength, but also by improving cognitive performance. Table 2 shows the results of modelling the impact of the Otago Exercise Programme based on three separate controlled trials, and in different age groups (Robertson 2001a, Robertson 2001b, Robertson 2001c). Recruitment of clients could be through general practitioners, A & E departments for those presenting with a fall, and other health professionals. There is good evidence for ACC, the Ministry of Health, and District Health Boards to continue supporting the Otago Exercise Programme in the home based format. A current pilot trial in Dunedin led by Dr Stephanie Woodley is testing whether group classes will be as effective in improving strength and balance measures, but is not powered for falls. Meantime the home based approach is endorsed by a recent finding from a survey of 5,440 people aged 54 and older in the UK (Yardley 2008). Over 60% of respondents said they would consider doing a home based strength and balance training programme at home (36.4% said they would definitely do it), whereas only 22.6% would definitely attend group exercise sessions.

Group exercise programmes

Multicomponent group exercise Exercise classes, typically with strength, balance, agility and flexibility exercises included and taught by a trained instructor, were the most common single falls prevention intervention in the literature (23 trials). Our meta-analysis showed there was significant reduction in the rate of falls by 29% in the 9 trials using this combination of exercises and with data on the rate of falls available for pooling. A systematic review published in the Journal of the American Geriatrics Society last month also showed that group exercise programmes are effective in reducing falls (Sherrington 2008). From this review it appears that balance retraining exercises are the crucial component of an effective group exercise programme. We chose a multicomponent group exercise programme for people with one or more physical impairments to model in this section (Barnett 2003). Physical impairments that are important risk factors for falls lower limb weakness, poor balance or reaction time were assessed using a standard assessment screen.

20

Table 1. Effectiveness and cost effectiveness of delivering the Otago Exercise Programme compared with no exercise programme for one year in three clinical trials Study Currency Target group ICER for Number of falls Incremental Relative delivery of prevented per cost per reduction person for intervention in rate of person receiving $NZ 2008 the intervention* delivering falls intervention ($NZ year (95% CI of prices) $NZ 2008 for ($NZ year incidence of prices) rate ratio) 32% (0.52 to 0.90) 0.540 (152/11788/116) $213 ($173) $418 ($314) ICER for delivery and fall related hospital admissions averted $NZ 2008 ($NZ year of prices)

Robertson 2001a $NZ 1995 Women 80 years Robertson 2001b $NZ 1998 Women and men 75 years Women and men 80 years Robertson 2001c $NZ 1998 Women and men 80 years

46% (0.32 to 0.90) 61% (0.14 to 1.05) 30% (0.59 to 0.84)

0.255 (109/11980/121) 0.611 (81/6143/60)

$549 ($432)

$2,291 ($1,803) $867 ($682)

$197 ($155)

$549 ($432)

$732 ( $576)

0.275 $531 ($418) (105/120198/330)

$1,930 ($1,519)

CI denotes confidence interval. ICER denotes incremental cost effectiveness ratio (incremental cost per fall prevented). $NZ denotes New Zealand dollars. *Numbers in brackets are raw data used for this calculation (number of falls in control group/number in control group minus number of falls in intervention group/number in intervention group). Main outcome results are reported in Campbell 1997. Negative values indicate cost savings.

21

Table 2. Estimated impact of delivering the Otago Exercise Programme for one year to 1,000 participants Study Scenario Target group Cost for delivery to 1,000 participants $NZ 2008 Number of falls prevented for 1,000 participants Cost of falls prevented* for 1,000 participants $NZ 2008 Total cost/cost saving* for 1,000 participants $NZ 2008 $148,303

Robertson 2001a Delivered by research physiotherapist Women 80 years Robertson 2001b Delivered by nurse Women and men 75 years Women and men 80 years Robertson 2001c Delivered by nurse Women and men 80 years

$213,000

541 (2,703)

$361,303 ($1,806,514)

$549,000 $549,000

255 611

$50,197 $447,400

$531,000

275

$201,300

$NZ denotes New Zealand dollars. *Negative values indicate cost savings. Estimated using average cost of a fall $NZ668.43 (at 2008 prices, societal perspective) from Robertson 2001a (women 80 years). Estimated from cost per fall prevented ($NZ197 per fall prevented, 2008 prices) reported in Robertson 2001b (75 years see Table 1). Estimated from cost saving per fall prevented ($NZ732 per fall prevented, 2008 prices) reported in Robertson 2001b (80 years see Table 1).

22

Table 3. Effectiveness and cost of weekly multicomponent group exercise classes compared with no exercise classes for one year Study Currency Target group Relative reduction in rate of falls (95% CI for incidence rate ratio) 40% (0.36 to 0.99) Number of falls prevented per person receiving the intervention* Incremental cost per person for delivering intervention $NZ 2008 $547

Barnett 2003 65 years, 1 physical performance impairment

0.341 (0.9460.605)

$NZ denotes New Zealand dollars. CI denotes confidence interval. *Numbers in brackets are data used for this calculation (rate of falls in control group minus rate of falls in intervention group, no raw data reported). No cost outcomes reported. Cost per person for one year of classes, based on estimations in Table 4.

Table 4. Estimated cost of exercise classes for 83 participants for one year as in Barnett 2003 Unit cost $NZ 2008 52.00 30.00 95.00 250.00 24.00 52.00 30.00 15.00 5.00 1.00 Total cost $NZ 2008 468 90 285 750 1992 20202 7770 1245 415 3071 9072 $45,360 $547

Cost item Training costs Exercise instructors (n=3) Materials Supervisor Recruitment Advertisements Staff time* Programme delivery Exercise instructor Venue hire Ankle cuff weights Manual for home exercise, diary to record home exercise Tea/coffee Overhead costs Total cost Average cost per participant

Resource use 3 hours each Manuals, photocopying 3 hours 3 newspaper advertisements 30 minutes per person recruited 1.5 hours per class, 37 classes for 7 streams, 12 people per class 1 hour per class, total 37 classes for 7 streams 1 per person 1 per person Per person for each class

$NZ denotes New Zealand dollars. *Includes assessment for inclusion criteria. Office accommodation, financial and administration services, depreciation on equipment for staff completing the assessments and exercise instructors calculated as 25% of resource use.

23

Table 5. Estimated impact of delivering multicomponent exercise classes to 1,000 participants for one year Study Target group Cost for delivery to 1,000 participants $NZ 2008 Number of falls prevented for 1,000 participants Cost of falls prevented* for 1,000 participants $NZ 2008 Total cost/cost saving* for 1,000 participants $NZ 2008 $319,065 $489,241

Barnett 2003 65 years, 1 physical performance impairment

$547,000

341

$227,935 $57,759

$NZ denotes New Zealand dollars. *Negative values indicate cost savings. Cost per person for one year of classes based on estimations in Table 4. Estimated using average cost of a fall $NZ668.43 (at 2008 prices, societal perspective) from Robertson 2001a (women 80 years). Estimated using median cost of a fall $NZ169.38 (at 2008 prices, societal perspective) from Garrett 2008 (women and men 75 years).

24

The screening tool included an inability to stand from a 45 cm high chair in less than two seconds, a need to step to maintain balance when performing a near-tandem balance test, and an inability to catch a rod dropped from above the hand within 300 milliseconds. The structured exercise classes were held for an hour each week over four terms for one year (37 classes in all). Accredited exercise instructors taught the classes. Table 3 shows the evidence for effectiveness of the intervention; there were no cost outcomes in the trial. We estimated the cost of delivering the classes in 2008 New Zealand dollars (Table 4). Table 5 shows the estimated impact if 1,000 people were recruited and took part in the classes for a year. There have been other effective group exercise programmes in terms of relative rate reduction published, but with insufficient data on falls and intervention details to estimate the costs and impact (Lord 2003, Skelton 2005). A recent systematic review of exercise programmes to prevent falls concludes with the following key messages (Robertson 2007): Many different risk factors contribute to falls but muscle weakness and poor balance underlie most falls. The most common components of successful exercise programmes to reduce falls are moderate intensity strength training against resistance and dynamic balance retraining, but there are insufficient good quality negative studies to determine whether some types of exercises are ineffective. Exercise programmes that are individually tailored, progress in difficulty, and target carefully selected groups at high risk, have so far resulted in the greatest absolute reduction in falls and injuries. Researchers should make programme details available so that those running falls prevention programmes in particular settings or subgroups of older people can use effective, evidence based interventions. Tai chi classes Tai chi has been shown to improve balance and has been suggested as a form of exercise to reduce falls since the successful trial by Wolf and colleagues (Wolf 1996). In this trial 72 participants aged 70 and older attended tai chi classes twice a week for 15 weeks. Falls were a secondary outcome of the study and were monitored for seven to 20 months. Compared with the 138 participants receiving computerised balance training or health education classes, the rate of falls was reduced by 47.5%. A further larger trial by Wolf, this time in a frailer group of older people and with a more intense version of tai chi, after 48 weeks showed a non-significant reduction in the rate of falls (Wolf 2003). There are now three other trials with data available for pooling. Participants in the trial with the largest reduction in the rate of falls (55%) were physically inactive 70 to 92 year olds in the US who were monitored for six months (Li 2005). The pooled rate ratio from our meta-analysis of the five published trials (n = 1,629 participants) shows that tai chi reduced the rate of falls by 40% in a combination of healthy people aged 60 and older plus older, frailer people. The Cochrane systematic review recently submitted for peer review, will endorse our finding and report that tai chi is effective in reducing both the rate of falling and the risk of falls. ACC has funded a randomised controlled trial of tai chi in people aged 65 years and older with a fall in the previous year, and this pragmatic trial will be completed next year. Of the tai chi trials currently published, the Central Sydney Tai Chi Trial best emulates the conditions expected in the New Zealand community (Voukelatos 2007). The 702 relatively healthy participants were aged 60 and older (mean 69, standard deviation 6.5, range 60 to 96 years). Those randomised to receive tai chi attended classes for one hour once a week for 16 weeks. The style of tai chi taught varied (majority Sun style, two classes Yang style) but all instructors had at least five years experience or had completed an accredited tai chi trainers course. Falls were monitored for only 24 weeks (at least 12 months is considered optimal for a falls prevention trial) when the rate of falls had been reduced by 33%.

25

Table 6. Effectiveness and cost effectiveness of 16 weekly tai chi classes compared with no classes after 24 weeks Study Currency Target group Relative reduction in rate of falls (95% CI for incidence rate ratio) 13% (0.46 to 0.96 Number of falls prevented per person receiving the intervention* Incremental cost per person for delivering intervention $NZ 2008 (reported currency) $303 ($A234) $373 ($A298)
CI denotes confidence interval. ICER denotes incremental cost effectiveness ratio (incremental cost per fall prevented). $NZ denotes New Zealand dollars, $A Australian dollars. *Numbers in brackets are data used for this calculation (number of falls in control group / number in control group minus number of falls in intervention group / number in intervention group). Main outcomes are reported in Voukelatos 2007. Calculated from raw data in Haas 2006 (total cost of tai chi classes $A81,232/347 in intervention group = $A234). Estimated using cost for 15 weeks tai chi classes twice weekly ($A470 at 2008 prices) calculated pro rata (classes once a week for 16 weeks) from Day 2008 (exchange rate $A1 = $NZ1.249 in mid-June 2008, Purchasing Power Parity not yet available for 2008).

ICER for delivery of intervention $NZ 2008 (reported currency)

ICER for delivery and total healthcare use $NZ 2008 (reported currency) $2,176 ($A1,683)

Haas 2006 $A 2002 60 years

0.126 (126/337 86/347)

$2,442 ($A1,889)

Table 7. Estimated impact of 16 weekly tai chi classes to 1,000 participants after 24 weeks Study Target group Cost for delivery to 1,000 participants $NZ 2008 $303,000 $373,000 Number of falls prevented for 1,000 participants 126 (630) 126 $21,342 Cost of falls prevented* for 1,000 participants $NZ 2008 Total cost/cost saving for 1,000 participants $NZ 2008 $274,176* $351,658

Haas 2006 60 years

$NZ denotes New Zealand dollars. *Includes costs of delivery and all healthcare use (Haas 2006, see Table 6). Estimated using median cost of a fall $NZ169.38 (at 2008 prices, societal perspective) from Garrett 2008 (women and men 75 years with a previous fall).

26

The evidence of effectiveness and cost effectiveness from the Central Sydney trial is summarised in Table 6. The Centre for Health Economics Research and Evaluation published a separate report on the cost effectiveness of tai chi in this trial (Haas 2006). In Table 7 we show the estimated impact after 24 weeks of 16 weekly classes with 1,000 participants. The modified version of tai chi used in most of the classes in this trial was designed by Dr Paul Lam especially for older people with arthritis (www.taichiproductions.com). Tai Chi for Arthritis has 12 movements based on the Sun style. In 2005 ACC invited Dr Lam to help design safety measures and set up training courses for class teachers to be used throughout New Zealand.

Home safety assessment and modification

Modifications in the home aimed at reducing home hazards and risky behaviour associated with falls are a component of many of the multifactorial falls prevention programmes. However, effectiveness of this approach was not known until trials testing home assessment and modification alone were completed. We found three trials testing a home safety programme as a single intervention and with the rate of falls available for pooling. Two of these trials showed a significant reduction in the rate of falls (Campbell 2005, Cumming 1999). A systematic review published this month (Clemson 2008) endorses the effectiveness of home safety assessment and modification programmes delivered by an experienced occupational therapist for preventing falls in community living older people. We would recommend targeting subgroups at high risk of falls, as this approach was not effective in reducing falls in a general population (Stevens 2001). People recently discharged from hospital In the trial by Cumming where the majority of participants were recruited on discharge from hospital, the subgroup who benefited most from the home safety programme were those reporting a fall in the previous year (hazard ratio 0.75, 95% confidence interval 0.58 to 0.96). A comprehensive economic evaluation of the home safety programme in Cummings trial confirms that the programme is more likely to be cost effective in older people who have a history of falls (Salkeld 2000). People with severe visual impairment Older people with visual impairment are at double the risk of falling compared with their normally sighted peers. A home safety programme modified for use in people with severe visual impairment was tested in the VIP trial (Campbell 2005). It was delivered by experienced occupational therapists during one or two home visits. The occupational therapists facilitated provision of any equipment such as grab rails, mobility aids and improvements in lighting. After one year falls were reduced by 41%. This trial was carried out in Auckland and Dunedin, New Zealand with participants recruited through low vision clinics and the Royal New Zealand Foundation of the Blind. The cost effectiveness of the home safety programme in people with visual impairment was also established (Campbell 2005). Estimated impact Table 8 shows the information on effectiveness and cost effectiveness available from these two trials, and also for completeness, from another economic evaluation a decision analytic model (Smith & Widiatmoko 1998). However this article reports the results from a hypothetical intervention programme and uses effectiveness data from previous literature. We have therefore modelled the impact of delivering a home safety programme to 1,000 participants using the approaches in the two clinical trials only (Table 9). Delivery of home safety programmes We noted in the VIP trial that falls were reduced inside the home, which was the focus of the intervention, and also away from home. This was also a finding in the home safety trial in Sydney (Cumming 1999). This suggests that the removal of hazards in the home is not the only

27

reason this intervention works in reducing falls. Raising awareness regarding safety in the public environment and mobility training may also have a part to play. Dr Lindy Clemson, a senior researcher in the field of occupational therapy from Monash University, Melbourne, Australia provided us with the following points, and some notes regarding best practice for environmental fall prevention home visits by an occupational therapist (see also Peterson & Clemson 2008). 1. The occupational therapy approach is very different from the pre-determined hazard list with a safety brochure provided. It is clear from the literature that this will not make a significant difference to the number of falls sustained. An occupational therapy home visit where reducing falls risk is the focus, differs from other occupational therapy visits where improving access and making things easier to accomplish are the expected outcomes.

2.

Take into account the person-environment fit. Ladder safety or alternatives to climbing and community safety may be an appropriate focus for an active person, whereas tripping hazards and safe reaching are likely more important for a frailer and more sedentary lifestyle. Understand the perceptions of the older person of their experience of falls and their beliefs about causes of falls. Challenge the individual to appraise their risk and to explore understanding of potential causes of falls. We need to listen to what people are saying and be observant of what they are doing. Use a problem solving approach. Audit the home collaboratively with the client to address behavioural and environmental concerns and develop priorities. Understand the meaning of home, activities and roles, and sense of control. These provide the context for working to make changes in the environment or to working with the person to change habits that might be risky behaviours. We know these factors will directly influence whether or not people will follow through with recommendations and decisions. Consider risk taking behaviours and encourage protective adaptations. Use situational cues and target the behaviour to be changed. Solutions are very often aimed at changing habitual behaviours. Have up-to-date knowledge on options for important hazards that are frequently involved in falls. For example, be knowledgeable about slip resistant products, strategies to fix loose floor coverings and new home safety products. Use environmental re-design strategies. For example, remove or re-locating furniture to reduce clutter and allow turning space. Encourage awareness raising and generalisation to other situations. Encourage awareness and safe behaviours in a wide range of situations. Assess and practice safe mobility strategies at home and in the community. Strategies such as scanning ahead and heel-toe walking, combined with improved balance, can give people confidence getting out and about. Encourage the right kind of exercise and physical activity. Integrate exercises and activity into daily life routines and in ways that are safe. Adequate follow up. Follow up and assist with practical solutions and modifications where necessary.

28

Table 8. Effectiveness and cost effectiveness at one year for delivering a home safety assessment and modification programme compared with no home safety programme from two clinical trials and one analytic model Study Currency Target group Relative reduction in rate of falls (95% CI for incidence rate ratio) 41% (0.42 to 0.83) Number of falls prevented per person receiving the intervention* Incremental cost per person for delivering intervention $NZ 2008 (reported currency) $369 ($NZ325) ICER for delivery of intervention $NZ 2008 (reported currency) ICER for delivery and healthcare costs $NZ 2008 (reported currency)

Campbell 2005 $NZ 2004 75 years visual acuity 6/24 or worse Salkeld 2000 $A 1998 65 years recently discharged from hospital 65 years recently discharged from hospital, fall in previous year Smith & Widiatmoko 1998 $A 1996 75 years

0.535 (271/193 172/198)

$738 ($NZ650)

14% (0.71 to 1.03) 25% (0.58 to 0.96)

0.362 (324/266 226/264) 0.913 (193/103 99/103)

$7,075 ($A4,986)

$5,648 ($A3,980)

NA

NA

$251 ($172)

$2,504 ($A1,721)

CI denotes confidence interval. ICER denotes incremental cost effectiveness ratio (incremental cost per fall prevented). $NZ denotes New Zealand dollars, $A Australian dollars. NA denotes these data are not available (not reported). *Numbers in brackets are raw data used for this calculation (number of falls in control group/number in control group minus number of falls in intervention group/number in intervention group). Main outcome results are reported in Cumming 1999. Note that a sensitivity analysis which excluded 15 outliers (if total cost was >3 standard deviations above the group mean and/or they had reported >50 falls in the past year) indicated cost savings (monetary value of cost savings not reported). Estimation of a hypothetical intervention programme.

29

Table 9. Estimated impact after one year from delivering a home safety assessment and modification programme to 1,000 participants Study Target group Cost for delivery to 1,000 participants $NZ 2008 Total number of falls prevented for 1,000 participants 535 Total cost of falls prevented* for 1,000 participants $NZ 2008 $357,610 Total cost/cost saving* for 1,000 participants $NZ 2008 $11,390

Campbell 2005 75 years, visual acuity 6/24 or worse Salkeld 2000 (1) 65 years recently discharged from hospital 65 years recently discharged from hospital, fall in previous year Salkeld 2000 (2) 65 years recently discharged from hospital 65 years recently discharged from hospital, fall in previous year

$369,000

362

$2,561,150

913

$5,156,624

$369,000

362

$241,972

$127,208

$369,000

913

$610,277

$241,277

$NZ denotes New Zealand dollars. *Negative values indicate cost savings. Estimated using average cost of a fall $NZ668.43 (at 2008 prices, societal perspective) from Robertson 2001a. Calculated using cost per fall prevented reported in Salkeld 2000 (includes all healthcare use for one year). Cost of delivery assumed the same as in VIP trial (Campbell 2005).

30

Cataract surgery
As noted above, older people with visual impairment are at double the risk of falling compared with their normally sighted peers. Three trials have tested an intervention aimed at reducing falls by improving vision. Two trials tested whether expedited cataract removal would reduce falls in women over 70 who had been referred to a hospital ophthalmology department (Harwood 2005; Foss 2006). Patients on the waiting list were randomised to expedited surgery (approximately four weeks) or to remain on the waiting list (12 months wait). The rate of falls was significantly reduced following first eye cataract surgery compared with those on the waiting list (Harwood 2005), but not following second eye surgery (Foss 2006). However, pooling of the rate ratios from the two trials showed a significant 33% reduction in falls. A comprehensive economic evaluation of the cost effectiveness of first eye cataract surgery in the trial has been reported (Sach 2007) and Table 10 summarises the evidence on effectiveness and cost effectiveness of this intervention. We have modelled the impact of this intervention using the reported data, and the cost of eye cataract surgery in New Zealand in a private health setting (Table 11). This trial also reported a cost utility analysis and extended the time period of the analysis from the trial duration to the participants remaining lifetime. The cost-utility ratios from the National Health Service perspective, and from the personal social services perspective for one year, were above the current accepted willingness to pay value of UK30,000. However, when the costs and QALYs were modelled over the participants expected lifetime, the incremental cost per QALY was within this limit at UK13,172 (at 2005 prices). The authors also provided cost effectiveness acceptability curves. The QALY provides a generic measure to enable comparison of cost effectiveness across different types of health services and treatments. However there are limitations in using QALYs for economic evaluations of complex interventions for older people since these interventions result in multiple benefits not captured by health related quality of life measures (Harwood 2008). We have not found quality of life measures sensitive to change in our falls prevention studies despite the beneficial outcomes of the trials. The negative effect on health related quality of life is larger for self reported fear of falling than for falls or a fracture (Iglesias 2008).

Assessment and multifactorial intervention

There is strong evidence that multifactorial interventions are effective in reducing falls in older people in the community, and good evidence for their effectiveness in residential care facilities and hospitals. Multifactorial programmes address the individuals risk factors for falls and this is always the approach taken by a clinician treating an individual patient. The first intervention tested in a randomised controlled trial that was successful in reducing falls was a multifactorial programme addressing individual risk factors for falls delivered at home by a nurse and physiotherapist (Tinetti 1994). Participants were aged 70 or over and had one or more of eight specified risk factors for falling. The intervention was cost saving in those with four or more of the eight risk factors at baseline (Rizzo 1996). The evidence for effectiveness and cost effectiveness from this trial is summarised in Table 12, and we have estimated the impact if the intervention were to be delivered to 1,000 participants (Table 13). We have chosen this particular intervention to recommend because it is feasible to run in New Zealand, and further evidence suggests that this type of approach results in overall reduction in hospitalisations from falls when implemented widely (Tinetti 2008). Other multifactorial interventions worthy of consideration include treatment by a geriatrician, and home visits by an occupational therapist to those who visited an A & E department as a result of a fall (Close 1999), and medical treatment, physiotherapy, and occupational therapy to this same subgroup of older people (Davison 2005). Although a multifactorial approach is the ideal for the individual patient, for a public health approach, multifactorial programmes may hold no advantages over single factor approaches (Campbell 2007). In a meta-analysis of good quality randomised controlled trials, we showed that falls were reduced by a similar amount by the single factor interventions and those with multiple components.

31

Table 10. Effectiveness and cost effectiveness of expedited first eye cataract surgery compared with no surgery after one year Study Currency Target group Relative reduction in rate of falls (95% CI for incidence rate ratio) 34% (0.45 to 0.96) Number of falls prevented per person receiving the intervention* ICER for Incremental ICER for cost per delivery of delivery and all healthcare person for intervention costs plus carer delivering and all costs intervention healthcare $NZ 2008 $NZ 2008 costs (reported $NZ 2008 currency) (reported currency) $3,700 $11,901 (UK4,390) $10,801 (UK3,983)

Sach 2007 UK 2005 Women >70 years

0.456 (1.3900.934)

$NZ denotes New Zealand dollars, UK pounds sterling. *Numbers in brackets are raw data used for this calculation (reported mean number of falls per person in control group minus mean number of falls per person in intervention group). Main outcomes are reported in Harwood 2005. Cost of pre-surgery visit to private ophthalmologist, cataract surgery including hospital and anaesthetist costs, 2 follow up visits (personal communication Marinoto Clinic, October 2008).

Table 11. Estimated impact after one year from expedited first cataract surgery for 1,000 participants Study Target group Cost for delivery to 1,000 participants $NZ 2008 Total number of falls prevented for 1,000 participants 456 456 $3,700,000 456 $304,804|| Total cost of falls prevented* for 1,000 participants $NZ 2008 Total cost/cost saving* for 1,000 participants $NZ 2008 $5,426,856 $4,925,256 $3,395,196

Sach 2007 Women >70 years

$NZ denotes New Zealand dollars. *Negative values indicate cost savings. Calculated using cost per fall prevented reported in Sach 2007 (includes all healthcare use for one year). Calculated using cost per fall prevented reported in Sach 2007 (includes all healthcare use and carer costs for one year). Cost of private cataract surgery (see Table 10). ||Estimated using average cost of a fall $NZ668.43 (at 2008 prices, societal perspective) from Robertson 2001a.

32

Table 12. Effectiveness and cost effectiveness of individualised multifactorial intervention compared with no intervention after one year Study Currency Target group Relative reduction in rate of falls (95% CI for incidence rate ratio) 31% (0.52 to 0.90) Number of falls prevented per person receiving the intervention* Incremental cost per person for delivering intervention $NZ 2008 (reported currency) $1,870 ($US905) ICER for delivery of intervention $NZ 2008 (reported currency) ICER for delivery and total healthcare use $NZ 2008 (reported currency) <$0 (<$US0)

Rizzo 1996 $US 1993 70 years, 1 of 8 targeted risk factors for falls

0.499 (164/14494/147)

$3,670 ($US1,772)

$NZ denotes New Zealand dollars, $US United States dollars. CI denotes confidence interval. ICER denotes incremental cost effectiveness ratio (incremental cost per fall prevented). *Numbers in brackets are data used for this calculation (number of falls in control group/number in control group minus number of falls in intervention group/number in intervention group). Main outcomes are reported in Tinetti 1994. Calculated using mean cost of intervention delivery per fall prevented. Calculated using mean cost of total healthcare use per fall prevented. <$0 indicates cost saving (monetary value of cost savings not reported).

Table 13. Estimated impact after one year from individualised multifactorial intervention delivered to 1,000 participants Study Target group Cost for delivery to 1,000 participants $NZ 2008 Number of falls prevented for 1,000 participants Cost of falls prevented* for 1,000 participants $NZ 2008 Total cost/cost saving for 1,000 participants $NZ 2008 Cost saving*

Rizzo 1996 70 years, 1 of 8 targeted risk factors for falls

$1,870,000

499

$1,870,000

499

$333,547

$1,536,453

$NZ denotes New Zealand dollars. *No data reported to estimate the amount of cost saving. Estimated using average cost of a fall $NZ668.43 (at 2008 prices, societal perspective) from Robertson 2001a (women 80 years).

33

Multiple interventions
Multifaceted group learning A programme that may well provide a possible multiple intervention option for ACC and the Ministry of Health to promote is Stepping On (Clemson 2004). In a 14 month randomised controlled trial in Sydney, Australia, small group learning (2-hour weekly sessions for seven weeks) led to a significant 31% reduction in falls in community living people who had had a fall in the previous 12 months (Table 14). Falls prevention programmes in New Zealand often take the form of group learning sessions run by community organisations. An ad hoc approach may not be effective, therefore not cost effective in preventing falls, so that these organisations should consider using the Stepping On programme instead. The emphasis of this programme is on behaviour change to avoid falls. The programme includes sessions on fall risk appraisal, exercise, home hazards, strategies to get around the local community, safe footwear, vision as a risk factor for falls, vitamin D, hip protectors, medication management, mastering safe mobility, and a home visit to follow through the falls prevention strategies and activities, and to assist with home adaptations and modifications if required. A booster session was held after three months. We estimated the cost of delivering this intervention in 2008 New Zealand dollars since no cost data were reported from this trial (Table 15). A manual has been published that provides all the information an experienced occupational therapist would require to run the sessions. The estimated impact of the programme after one year if 1,000 participants with a previous fall were recruited, is shown in Table 16. Another successful multiple intervention, this one with a focus on home safety, was a comprehensive geriatric assessment (all participants) followed by a home safety visit (intervention participants only) (Nikolaus 2003). The subgroup who reported having had two or more falls in the previous year showed the greatest benefit. Population approach Despite methodological limitations of the evaluation studies reviewed (five controlled but not randomised trials), the consistency of findings in a Cochrane systematic review led to the conclusion that a population based approach to the prevention of fall related injury is effective and can form the basis of public health practice (McClure 2005). Two controlled studies also provide good evidence that a concerted population approach to falls prevention will result in a reduction in healthcare use and costs. The results of a cost benefit study of the Stay on Your Feet programme in Queensland, Australia is summarised in Table 17 (Kempton 2000, Beard 2006). The cost benefit evaluation compared hospital records (admissions) between a matched sample and the intervention areas. The programme activities were community based and aimed at older adults (55 years). They included selective advertising using specialty products, mass media campaigns, distribution of educational material (pamphlets, manuals, booklets), partnerships with GPs, and other health professionals and workers, partnerships with local government, workshops and training sessions regarding home modifications and appropriate exercises. The risk factors addressed were balance and gait problems, insufficient exercise, inappropriate footwear, poor vision, medication use, underlying medical conditions, and environmental hazards. A controlled study by Tinetti and colleagues has recently been published (Tinetti 2008). The authors compared rates of injuries from falls in a region in the US exposed to interventions to change clinical practice (522 primary care clinicians, 133 outpatient rehabilitation facilities, 26 home care agencies, 7 acute care hospitals with emergency departments, and 41 senior centres), and a region where clinicians had not been exposed to such interventions (460 clinicians, 146 outpatient rehabilitation facilities, 7 acute hospitals and emergency departments, and 43 senior centres). A multidisciplinary team used the media, web sites, posters, brochures, educational materials, opinion leaders, advertising, and visits (outreach) to everyone in the main group of clinicians and facilities being targeted.

34

Table 14. Effectiveness and cost of a multifaceted group learning intervention Stepping On compared with no intervention after 14 months Study Target group Relative reduction in rate of falls (95% CI for incidence rate ratio) 31% (0.50 to 0.96) Number of falls prevented per person receiving the intervention* Incremental cost per person for delivering intervention $NZ 2008 $885

Clemson 2004 70 years, fall in previous year

0.527 (255/153179/157)

CI denotes confidence interval. $NZ denotes New Zealand dollars. *Numbers in brackets are data used for this calculation (number of falls in control group /number in control group minus number of falls in intervention group/number in intervention group). No cost outcomes reported. Cost per person for the intervention, based on estimations in Table 15.

Table 15. Estimated cost of 'Stepping On' intervention for 157 participants as in Clemson 2004 Unit cost Total cost ($NZ 2008) ($NZ 2008) 45.00 106.00 63.96 45.00 52.00 45.00 65.00 55.00 30.00 15.00 5.00 1.00 24.00 45.00 10.00 60.00 135 106 10042 11700 2028 585 845 715 6240 2355 785 1256 4992 14130 1570 1260 14686 $73,430 $885

Cost item Training costs Class facilitator Materials

Resource use Occupational therapist, 3 hours Manual $A85 Mail outs, advertisements, promotional material, health professional referral

Recruitment* Programme delivery Class facilitator 2.5 hours per class, 8 classes for 13 streams Physiotherapist 1 hour in weeks 1 and 6, 40 minutes in week 2 Speaker on pedestrian safety 40 minutes in week 4 (+travel time) Low vision expert 40 minutes in week 5 (+travel time) Pharmacist 40 minutes in week 6 (+travel time) Venue hire 2 hours per class, 8 classes for 13 streams Ankle cuff weights 1 per person Handout material Photocopied material for each person Tea/coffee Per person for each class Transport costs Taxi for 2 per class Occupational therapist (home visit) 2 hours per visit (includes travel time) Increasing lighting levels, applying non-slip Home modifications tape, fixing pathways Visit to optician 21 initiated a vision assessment Overhead costs Total cost Average cost per participant
*As reported for the trial in Clemson 2007.

Office accommodation, financial and administration services, depreciation on equipment for staff completing the assessments and exercise instructors calculated as 25% of resource use.

35

Table 16. Estimated impact of delivering multifaceted group learning classes Stepping On to 1,000 participants after one year Study Target group Cost for delivery to 1,000 participants $NZ 2008 Number of falls prevented for 1,000 participants Cost of falls prevented* for 1,000 participants $NZ 2008 Total cost/cost saving* for 1,000 participants $NZ 2008 $532,737 $795,737

Clemson 2004 70 years, fall in previous year

$885,000

527

$352,263 $89,263

$NZ denotes New Zealand dollars. *Negative values indicate cost savings. Cost per person for the Stepping On intervention (8 classes and one home safety visit by occupational therapist) based on estimations in Table 15. Estimated using average cost of a fall $NZ668.43 (at 2008 prices, societal perspective) from Robertson 2001a (women 80 years). Estimated using median cost of a fall ($NZ169.38 (at 2008 prices, societal perspective) from Garrett 2008 (women and men 75 years).

Table 17. Estimated impact from Stay on Your Feet delivered 1992/93 to 1995/96 in Queensland Study Currency Target group Cost for delivery of the programme $NZ 2008 (reported currency) $1,187,206 ($A805,579) Total healthcare costs averted* $NZ 2008 (reported currency) Total healthcare costs averted $NZ 2008 (reported currency) Average benefit to cost ratio

Beard 2006 $A 1995/96 >60 years

$16,730,439 ($A11,352,453)

$14,772,036 ($A10,023,577)

20.6:1

$NZ denotes New Zealand dollars. $A Australian dollars. *Calculated using method 1: Comparison based on hospital admission rates for 60 years in interstate control region and intervention region. Calculated using method 2: Comparison based on hospital admissions estimated from diagnosis related group codes for 65 years in New South Wales as a whole and intervention region.

36

Recommended strategies for preventing falls included balance, gait, and strength training, a reduction in medications, management of postural hypotension, management of visual and foot problems, and hazard reduction. The 11% relative reduction in the use of fall related medical services in the intervention region compared with the control region in the three years following the intervention, translated into 1,800 fewer A & E department visits or hospital admissions. The decrease represented potential cost savings of $US21 million in healthcare calculated on the basis of an average acute care cost of $US12,000 per event. No further cost outcomes were reported.

Limitations of modelling methods

Accuracy of data The precision of economic modelling relies heavily on the assumptions made. Therefore for each intervention modelled, we have stated the assumptions clearly and where possible we have provided varying cost scenarios. The resulting total cost (or cost saving) is a good estimate of the impact of the falls prevention strategy on healthcare costs only if all the data used in the calculations are accurate, appropriate, and generalisable. Estimating benefits beyond one year A major limitation of our models is that the timeframe of one year taken is likely to underestimate the cost effectiveness of the interventions. We have used one year as this is the standard time falls are monitored in the randomised controlled trials. We consider calculating the longer term benefits in terms of the number of falls that would be prevented after one year would be purely guess work. Adherence to interventions As mentioned previously, adherence to an intervention programme will depend on many different factors, including those related to the person delivering the intervention (for example level of professional skill, rapport with older people) and the person receiving the intervention (readiness to change lifestyle and behaviour, medical conditions, access problems). It is inevitable that some older people will withdraw or die from other causes during intervention delivery. While offering an intervention only to those who will benefit and those guaranteed to adhere will optimise value for money, we are not able to factor these scenarios into our models with any certainty. Cost estimates There are inherent logistical difficulties in comparing and combining the results of economic analyses because of differences in the health system in the country of the trial, the currency, the cost items included, perspectives taken, the time frame for measurement of costs, and the methodology used to calculate incremental cost effectiveness and cost benefit ratios. The clinical trials were all powered for falls and not for costs, and hospital costs, the major cost driver, are very often skewed. As mentioned previously, in our models we have assumed that the average cost of delivering an intervention in a clinical trial is generalisable to a similar subgroup in the wider New Zealand community. The estimation of the costs of delivery for each intervention has included the cost items required for implementation, and has ignored any programme modification required or other set up costs. These will vary between interventions. For the Otago Exercise Programme an instruction manual is available and a web based training course is already established. If a programme of individual assessment and multifactorial intervention were to be initiated, substantial time and effort would be required to prepare the protocols, set up clinics, and liaise with and train geriatricians, occupational therapists, nurses, and physiotherapists.

37

Table 18. Effective falls prevention interventions reported to be cost saving: estimated impact when delivered to 1,000 older New Zealanders Intervention Target group Estimated cost for delivery to 1,000 participants $NZ 2008 $549,000
a

Estimated number of falls prevented for 1,000* participants 611

Total cost/cost saving for 1,000 participants $NZ 2008

Otago Exercise Programme 80 years Home safety assessment and modification programme 65 years, fall in previous year, recently discharged from hospital Individual assessment and home based multifactorial intervention 70 years, 1 of 8 targeted risk factors for falls

$447,400

$369,000

913

Cost saving e $241,277

$1,870,000

499

Cost saving h $1,536,453

$NZ denotes New Zealand dollars. *Extrapolated from the number of falls per person prevented during the trial. Negative values indicate cost savings. a: Extrapolated from the cost per person of delivering the intervention reported in the West Auckland trial (societal cost) (Robertson 2001b). b: From Robertson 2001b, calculated using reported incremental cost per fall prevented and number of falls per person prevented in the trial, incorporates fall related hospital admission costs averted. c: Average cost per person in the trial not reported, assumed to be the same as the home safety assessment and modification intervention in the New Zealand VIP trial (Campbell 2005). d: Reported in Salkeld 2000, incorporates total cost of healthcare services for participants during the trial (societal perspective), 15 outliers excluded (if total cost was >3 standard deviations above the group mean and/or >50 falls), monetary value of cost saving not reported. e: Estimated from cost of delivering the home safety programme (see note c), and number of falls per person prevented reported in Salkeld 2000, each fall event valued at $NZ668.43 (Robertson 2001a). f: Extrapolated from average cost per person of delivering the intervention reported in the trial (Rizzo 1996). g: Reported in Rizzo 1996 when calculated using mean costs, both when total costs of healthcare services for participants during the trial (societal perspective) were incorporated, and when total cost of medical care only was incorporated, monetary value of cost savings not reported. h: Estimated from cost of delivering the intervention (see note f) and number of falls per person prevented reported in Rizzo 1996, each fall event valued at $NZ668.43 per fall (Robertson 2001a).

38

Actual cost savings We have already pointed out that any indications of potential cost savings by our models, are not actual cost savings to the health system unless these saved resources are employed for effective and worthwhile alternatives. Does falls prevention mean injury prevention? The falls prevention literature is not able to give us a definitive answer as to whether preventing falls will prevent injuries, since no single randomised controlled falls prevention trial has been large enough to answer this question. Interestingly, the authors of a recent article in the BMJ suggest shifting the focus in fracture prevention from treatment and prevention of osteoporosis, to falls prevention (Jrvinen 2008). There is some indication that fall injuries are reduced by the same amount as falls when an effective falls prevention programme is implemented (Robertson 2002). Our own work from a meta-analysis of four trials of the Otago Exercise Programme has shown that both falls and fall injuries were reduced by 35%. It seems logical to assume that the more falls that are prevented, the more fall related injuries will also be prevented. When we pooled fracture outcomes from individual trials as part of this project, the number of trials reporting this outcome limited our findings. However, it appears from these trials that fracture reduction is likely to track falls reduction.

Modelling falls strategies for the Australian population

Dr Lesley Day from Monash University, Melbourne, Australia and members of the National Injury Prevention Working Group have recently submitted a comprehensive report to their funders titled Modelling the impact, costs and benefits of falls prevention measures to support policy-makers and program planners (Day 2008). Lesley has kindly sent us a confidential copy. She also presented some results from this project at the recent 3rd Australian and New Zealand Falls Prevention Conference Society in Melbourne, Australia in October 2008. The project includes epidemiological models for six interventions five interventions from the 2003 Cochrane review, plus expedited cataract removal which was reported more recently. The five strategies are home based strength and balance exercise, psychotropic medication withdrawal, 16 weeks of twice-weekly tai chi classes, home hazard assessment and modification, and cardiac pacing. The approach used was to model delivering each trial intervention nationwide to all those living in Australia in the same age group and with the same characteristics as trial participants, and who would be likely to take part. The intervention these models indicated was most cost effective in terms of hospital costs averted was home hazard assessment and modification, and the least cost effective was cardiac pacing. Predicted reductions in national fall related hospital admissions for five of the six intervention strategies if implemented in this manner over a one year period ranged from 0.4% to 4.6%. Issues such as competing national health resources, and the marked differences in set up costs and overall feasibility between the six interventions, were not included in the models. These and other methodological issues cast doubt on the relevance of these particular models to the New Zealand setting.

39

Table 19. Recommended falls prevention strategies for community living older people Strategy rank Intervention Target group Relative rate reduction (95% CI for IRR) 32% (0.52 to 0.90) Number of falls prevented per person receiving the intervention 0.611 Estimated cost per person to deliver the intervention $NZ 2008 $549 Rationale and comments

Strategy 1

Otago Exercise Programme 80 years

Strategy 2

Good trial evidence of healthcare cost savings. ACC currently supports this intervention. Trial evidence of moderate cost effectiveness. ACC currently supports this intervention. Trial evidence of healthcare cost savings. Target group readily identifiable. Experienced occupational therapists able to deliver intervention with minimal training. Good cost effectiveness (trial). Target group through RNZFB. Experienced occupational therapists able to deliver intervention with minimal training. Moderate cost effectiveness (estimated). Screening required. Some training needed but exercise instructors available. Moderate cost effectiveness (estimated). Experienced occupational therapist as facilitator and for home safety visits. Good trial evidence of healthcare cost savings. Delivered by geriatrician, physiotherapist and nurse. Support by DHBs required. Ranks lowest of the 8 in terms of cost effectiveness (trial). Surgery feasible at private or public hospitals if funding available.

Tai chi classes, 16 weekly classes 60 years

Strategy 3

13% (0.46 to 0.96)

0.126

$303

Home safety programme 65 years, recently discharged from hospital, fall in previous year

25% (0.58 to 0.96)

0.913

$369

Strategy 4

Home safety programme 75 years, visual acuity 6/24 or worse

25% (0.58 to 0.96)

0.913

$369

Strategy 5

Multicomponent group exercise, weekly classes all year 65 years, 1 physical performance impairment
Strategy 6

40% (0.36 to 0.99)

0.341

$547

Stepping On programme 70 years, fall in previous year

Strategy 7

31% (0.50 to 0.96)

0.527

$885

Individual multifactorial intervention 70 years, 4 of 8 targeted risk factors

Strategy 8

31% (0.52 to 0.90)

0.499

$1,870

Expedited first cataract surgery >70 years

34% (0.45 to 0.96)

0.456

$3,700

IRR denotes incidence rate ratio; RNZFB Royal New Zealand Foundation of the Blind; DHBs District Health Boards.

40

RECOMMENDED STRATEGIES
Several of the falls prevention strategies examined in this project demonstrated they would give good value for money if disseminated widely, and three strategies were shown to be cost saving in high risk groups. This is noteworthy because very few health services actually save health dollars, especially those delivered to older people. Society accepts that there is a cost associated with the benefits received from health services and treatments. A summary of the evidence indicating the potential for falls prevention strategies to be cost saving from a societal perspective is shown in Table 18. It may be that more falls prevention strategies are cost saving apart from these three, but trials so far have been powered for falls, and not for injury or cost outcomes. We recommend implementation of eight of the falls prevention strategies we modelled in the previous section. These eight are listed in Table 19 in the order that we recommend widespread implementation in New Zealand. Strategies 1 to 4 The four strategies listed above the bar include two that ACC supports currently, that is, the Otago Exercise Programme delivered at home, and tai chi classes. There is good evidence to continue implementing these two strategies, whether or not the results from the randomised controlled trial of tai chi classes currently in progress in New Zealand show a significant reduction in falls. These two programmes provide a good balance in terms of choice for older people. A home based strategy can be offered to those 80 and older, or 75 plus with a previous fall or considered at high risk of falls, who prefer to exercise at home. Tai chi classes weekly for 16 weeks were effective in a general sample of Australians 60 years and older, and provide a group exercise option for those who are not too frail to do the tai chi movements, and can travel to classes. We recommend Strategies 3 and 4 for widespread implementation in the short term. There is good evidence now to support implementation of home safety programmes delivered by experienced occupational therapists to high risk groups. The groups targeted should be those aged 65 or older being discharged from hospital who report a fall in the previous year, and people aged 75 and older with poor vision identified at low vision clinics or registered with the Royal New Zealand Foundation of the Blind. Care is needed when targeting home safety programmes because this intervention did not significantly reduce falls when delivered to a general, rather than a high risk group (Stevens 2001). Occupational therapists already deliver home safety programmes to promote rehabilitation and independent living for elderly people. A falls prevention emphasis could be built into these programmes and minimal training would be required. Strategies 5 to 8 The strategies listed below the bar are also effective options for falls prevention, but are more complex to implement and require resources for careful screening at recruitment. We recommend Strategies 5 and 6 be considered in the medium term, and suggest Strategies 7 and 8 be considered in the longer term. We chose to model multicomponent group exercise classes when delivered to a high risk subgroup (Strategy 5), but exercise classes have also been shown to be effective in a wider range of subgroups. These include people living in retirement villages (Lord 2003), and women aged 65 years with three or more falls in the previous year (Skelton 2005). A meta-analysis showed that it is the balance retraining exercises which are crucial to these classes, rather than the other components (Sherrington 2008). Many current falls prevention programmes run in the New Zealand community include various components addressing a range of fall risk factors. As these programmes have not been tested in randomised controlled trials, funders and providers should be aware that many of them may not be effective in significantly reducing falls, therefore not cost effective. For this type of approach we would recommend that Strategy 6, the Stepping On programme be used. This

41

programme has been carefully designed to provide relevant, practical information to older people, and to encourage behaviour change. It has been tested in Australia, and all the various components are available and would suit the New Zealand situation. There is now good evidence that interventions addressing the individual persons fall risk factors are effective in reducing falls, but resources will not be available to treat every older New Zealander in this way. We recommend Strategy 7, the first multifactorial programme shown to be effective in reducing falls (Tinetti 1994). This is a home based programme which requires input from geriatricians, physiotherapists and nurses. There was good evidence for potential cost savings from this approach, especially in those with four of the eight, targeted risk factors (Rizzo 1996). Other multifactorial interventions have been shown to be effective, but many would involve even more commitment and resources from District Health Boards. Strategy 8, cataract removal, which is effective in reducing falls, was the least cost effective of the eight strategies we modelled. It should be noted however that, as with most falls prevention strategies, the benefits of this intervention are much broader than preventing falls and injuries per se. Additional points for consideration The eight strategies in Table 19 recommended for widespread dissemination are all for older people in the community setting, reflecting the potential for effective falls prevention currently favouring those living independently in the community. The evidence also supports vitamin D supplementation for those in residential care homes. Multifactorial interventions have been effective in reducing falls and even hip fractures in residential care homes in Europe, and trial results are needed to show the components of these programmes that would work in New Zealand. Implementation of multifactorial programmes in hospitals also appears the most promising approach in this setting. When deciding on the proportion of the falls prevention budget for older people to allocate to those living independently in the community, ACC needs to take into account 1) the proportion and strength of the evidence available on effective falls prevention strategies (stronger evidence for effectiveness comes from the 67% of the 140 randomised controlled trials which are in those living in the community), 2) the visible costs to ACC (24% of all claims in 2006/2007 for fall injuries for people 80 and older came from those in rest homes), and 3) although falls are more common in individuals in residential care facilities, around three in four older New Zealanders live independently in the community. It would therefore seem advisable that ACC spends around 75% of their older persons falls prevention budget to support effective strategies for community living older people, and 25% on supporting or researching effective strategies in residential care or in hospitals.

42

REFERENCES
Accident Compensation Corporation. Otago exercise programme to prevent falls in older adults, August 2003. http://www.acc.co.nz/otagoexerciseprogramme [accessed 1 November 2008]. American Geriatrics Society, British Geriatrics Society, American Academy of Orthopaedic Surgeons Panel on Falls Prevention. Guideline for the prevention of falls in older persons. Journal of the American Geriatrics Society 2001; 49: 664-72. Barnett A, Smith B, Lord SR, Williams M, Baumand A. Community-based group exercise improves balance and reduces falls in at-risk older people: a randomised controlled trial. Age & Ageing 2003;32(4):407-14. Beard J, Rowell D, Scott D, van Beurden E, Barnett L, Hughes K, et al. Economic analysis of a community-based falls prevention program. Public Health 2006;120(8):742-51. Becker C, Kron M, Lindemann U, Sturm E, Eichner B, Walter-Jung B, et al. Effectiveness of a multifaceted intervention on falls in nursing home residents. Journal of the American Geriatrics Society 2003;51(3):306-13. Brown P, Njogu F, Kerse N, Willingale J. Cost effectiveness of impact absorbent flooring and hip protectors among institutionalized elderly. Paper from the School of Population Health, University of Auckland, October 2008. Butler M, Kerse N, Todd M. Circumstances and consequences of falls in residential care: the New Zealand story. New Zealand Medical Journal 2004;117(1202):U1076. URL: http://www.nzma.org.nz/journal/117-1202/1076/ Cali CM, Kiel DP. An epidemiological study of fall-related fractures among institutionalized older people. Journal of the American Geriatrics Society 1995;43:1336-40. Cameron ID, Murray GR, Gillespie LD, Cumming RG, Robertson MC, Hill K, Kerse N. Interventions for preventing falls in older people in residential care facilities and hospitals. (Protocol) Cochrane Database of Systematic Reviews 2005, Issue 3. Art. No.: CD005465. DOI: 10.1002/14651858.CD005465. Campbell AJ, Robertson MC, Gardner MM, Norton RN, Tilyard MW, Buchner DM. Randomised controlled trial of a general practice programme of home based exercise to prevent falls in elderly women. BMJ 1997;315(7115):1065-9. Campbell AJ, Robertson MC, Gardner MM, Norton RN, Buchner DM. Psychotropic medication withdrawal and a home-based exercise program to prevent falls: a randomized, controlled trial. Journal of the American Geriatrics Society 1999;47(7):850-3. Campbell AJ, Robertson MC, La Grow SJ, Kerse NM, Sanderson GF, Jacobs RJ, et al. Randomised controlled trial of prevention of falls in people aged 75 with severe visual impairment: the VIP trial. BMJ 2005;331(7520):817. Campbell AJ, Robertson MC. Rethinking individual and community fall prevention strategies: a metaregression comparing single and multifactorial interventions. Age & Ageing 2007;36:656-62. Chan DK, Hiller G, Cooke R, Monk R, Mills J, Hung WT. Effectiveness and acceptability of a newly designed hip protector. Archives of Gerontology & Geriatrics 2000;30:25-34. Clemson L, Cumming RG, Kendig H, Swann M, Heard R, Taylor K. The effectiveness of a communitybased program for reducing the incidence of falls in the elderly: a randomized trial. Journal of the American Geriatrics Society 2004;52(9):1487-94. Clemson L, Mackenzie L, Ballinger C, Close JCT, Cumming RG. Environmental interventions to prevent falls in community-dwelling older people: a meta-analysis of randomized trials. Journal of Aging and Health 2008 OnlineFirst DOI:10.1177/0898264308324672. Close J, Ellis M, Hooper R, Glucksman E, Jackson S, Swift C. Prevention of falls in the elderly trial (PROFET): a randomised controlled trial. Lancet 1999;353: 93-97. Cumming RG, Thomas M, Szonyi G, Salkeld G, O'Neill E, Westbury C, et al. Home visits by an occupational therapist for assessment and modification of environmental hazards: a randomized trial of falls prevention. Journal of the American Geriatrics Society 1999;47(12):1397-402. Cumming RG, Ivers R, Clemson L, Cullen J, Hayes MF, Tanzer M, et al. Improving vision to prevent falls in frail older people: A randomized trial. Journal of the American Geriatrics Society 2007;55(2):175-81. Dargent-Molina P, Favier F, Grandjean H, Baudoin C, Schott AM, Hausherr E, et al. Fall-related factors and risk of hip fracture: the EPIDOS prospective study. Lancet 1996;384:145-9. Davison J, Bond J, Dawson P, Steen IN, Kenny RA. Patients with recurrent falls attending Accident & Emergency benefit from multifactorial intervention a randomised controlled trial. Age & Ageing 2005;34(2):162-8.

43

Day L, Hoareau E, Finch C, Harrison J, Segal L, Bolton T, et al. Modelling the impact, costs and benefits of falls prevention measures to support policy-makers and program planners. Report from project funded by the Australian Government Department of Health and Ageing (Falls and Community Injury Grants program), September 2008. Drahota A, Gal D, Windsor J. Flooring as an intervention to reduce injuries from falls in healthcare settings: an overview. Quality Ageing 2007;8(1):3-9. Drummond MF, Sculpher MJ, Torrance GW, O'Brien BJ, Stoddart GL. Methods for the economic evaluation of health care programmes. Third edition. New York, United States of America: Oxford University Press, 2005. Dyer CA, Taylor GJ, Reed M, Dyer CA, Robertson DR, Harrington R. Falls prevention in residential care homes: a randomised controlled trial. Age & Ageing 2004;33(6):596-602. Elley CR, Robertson MC, Garrett S, Kerse NM, McKinlay E, Lawton B, et al. Effectiveness of a falls and fracture nurse coordinator to reduce falls: randomized controlled trial of at-risk older adults. Journal of the American Geriatrics Society 2008;56:1383-9. Foss AJ, Harwood R, Osborn F, Gregson RM, Zaman A, Masud T. Falls and health status in elderly women following second eye cataract surgery: a randomised controlled trial. Age & Ageing 2006;35:66-71. Garrett S, Elley CR, ODea D. The cost of falls in older adults in the community. New Zealand Family Physician 2008;35(1):22-7. Gillespie LD, Gillespie WJ, Robertson MC, Lamb SE, Cumming RG, Rowe BH. Interventions for preventing falls in elderly people. Cochrane Database of Systematic Reviews 2003, Issue 4. Art. No.: CD000340. Gillespie LD, Robertson MC, Gillespie WJ, Lamb S, Gates S, Cumming RG, et al. Interventions for preventing falls in older people living in the community. (Protocol) Cochrane Database of Systematic Reviews 2008, Issue 2. Art. No.: CD007146. DOI: 10.1002/14651858.CD007146. Haas M. Economic analysis of tai chi as a means of preventing falls and related injuries among older adults. Centre for Health Economics Research and Evaluation (CHERE), University of Technology, Sydney, Australia working paper 2006/4. Accessed 1 March 2008 at: http://datasearch.uts.edu.au/chere/research/working_papers.cfm. Haines T, Kuys S, Morrison G, Clarke J, Bew P. Is falls risk screening worthwhile? Paper presented at the 3rd Australian and New Zealand Falls Prevention Society Conference Melbourne, Australia 1214 October 2008. Haran M, Cameron I, Ivers R, Simpson J, Lee B, Porwal M, et al. The effect of providing single lens distance glasses on falls in older multifocal glasses wearers: a randomised controlled trial. Paper presented by Professor Stephen Lord at the 3rd Australian and New Zealand Falls Prevention Society Conference, Melbourne, Australia, 1214 October 2008. Harwood RH, Foss AJ, Osborn F, Gregson RM, Zaman A, Masud T. Falls and health status in elderly women following first eye cataract surgery: a randomised controlled trial. British Journal of Ophthalmology 2005;89(1):53-9. Harwood RH. Economic evaluations of complex services for older people. Age & Ageing 2008;37:491-2. Iglesias CP, Manca DJ, Torgerson DJ. The health-related quality of life and cost implications of falls in elderly women. Osteoporosis International 2008 DOI 10.1007/s00198-008-0753-5. Jrvinen TLN, Sievnen H, Khan KM, Heinonen A, Kannus P. Shifting the focus in fracture prevention from osteoporosis to falls. BMJ 2008;336:124-6. Jensen J, Lundin-Olsson L, Nyberg L, Gustafson Y. Fall and injury prevention in older people living in residential care facilities. A cluster randomized trial. Annals of Internal Medicine 2002;136(10):733-41. Kempton A, Van Beruden E, Sladden T, Garner E, Beard J. Older people can stay on their feet: final results of a community-based falls prevention programme. Health Promotion International 2000;15(1):27-33. Kerse N, Butler M, Robinson E, Todd M. Fall prevention in residential care: a cluster, randomized, controlled trial. Journal of the American Geriatrics Society 2004;52(4):524-31. King AC, Marcus B, Ahn D, Dunn AL, Rejeski WJ, Sallis JF. Identifying subgroups that succeed or fail with three levels of physical activity intervention: the Activity Counseling Trial. Health Psychology 2006;25:336-47. Liu-Ambrose T, Donaldson MG, Ahmed Y, Graf P, Cook WL, Close J, et al. Otago home-based strength and balance retraining improves executive functioning in older fallers: a randomised controlled trial. Journal of the American Geriatrics Society 2008 DOI: 10.1111/j.1532-5415.2008.01931.x (ahead of print). Lord SR, Castell S, Corcoran J, Dayhew J, Matters B, Shan A, et al. The effect of group exercise on physical functioning and falls in frail older people living in retirement villages: a randomized, controlled trial. Journal of the American Geriatrics Society 2003;51(12):1685-92.

44

McClure R, Turner C, Peel N, Spinks A, Eakin E, Hughes K. Population-based interventions for the prevention of fall-related injuries in older people. Cochrane Database of Systematic Reviews 2005, Issue 1. Art.No.:CD004441. DOI: 10.1002/14651858.CD004441.pub2. McKiernan FE. A simple gait-stabilizing device reduces outdoor falls and nonserious injurious falls in fallprone older people during the winter. Journal of the American Geriatrics Society 2005;53(6):9437. NHS National Institute for Clinical Excellence. Falls: the assessment and prevention of falls in older people, November 2004: Clinical Guideline 21. Nikolaus T, Bach M. Preventing falls in community-dwelling frail older people using a home intervention team (HIT): results from the randomized falls-HIT trial. Journal of the American Geriatrics Society 2003;51:300-5. Norton R, Campbell AJ, Lee-Joe T, Robinson E, Butler M. Circumstances of falls resulting in hip fractures among older people. Journal of the American Geriatrics Society 1997;45:1108-12. Peterson EW, Clemson L. Understanding the role of occupational therapy in fall prevention for community-dwelling older adults (Article code CEA0208). OT Practice 2008;13(3):CE-1-8. Rizzo JA, Baker DI, McAvay G, Tinetti ME. The cost-effectiveness of a multifactorial targeted prevention program for falls among community elderly persons. Medical Care 1996;34(9):954-9. Robertson MC, Devlin N, Scuffham P, Gardner MM, Buchner DM, Campbell AJ. Economic evaluation of a community based exercise programme to prevent falls. Journal of Epidemiology and Community Health 2001a;55(8):600-6. Robertson MC, Devlin N, Gardner MM, Campbell AJ. Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 1: Randomised controlled trial. BMJ 2001b;322(7288):697-701. Robertson MC, Devlin N, Gardner MM, McGee R, Campbell AJ. Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 2: Controlled trial in multiple centres. BMJ 2001c;322(7288):701-4. Robertson MC. Development of a falls prevention programme for elderly people: evaluation of efficacy, effectiveness, and efficiency [PhD thesis]. Department of Medical and Surgical Sciences, Dunedin School of Medicine, University of Otago, Dunedin, New Zealand, 2001d. Robertson MC, Campbell AJ, Gardner MM, Devlin N. Preventing injuries in older people by preventing falls: a meta-analysis of individual-level data. Journal of the American Geriatrics Society 2002;50(5):905-11. Robertson MC, Campbell AJ, Herbison P. Statistical analysis of efficacy in falls prevention trials. Journals of Gerontology Series A Biological Sciences and Medical Sciences 2005;60(4):530-4. Robertson MC, Milburn PD, Campbell AJ, Bowmar J. Introducing KradalTM a new energy absorbing floor technology to prevent fractures in elderly people. Poster presentation at the 2nd Australian Falls Prevention Society Conference, Brisbane, Australia November 2006. Robertson MC, Campbell AJ. What type of exercise reduces falls in older people? Chapter (systematic review) in: Evidence-based sports medicine, second edition. Editors Domhnall MacAuley and Thomas Best. Blackwell Publishing, Oxford, UK, 2007, pp135-66. Robertson MC, Milburn PD, Carman AB, Campbell AJ. Safety flooring: energy absorption, postural stability, and gait patterns. Paper presented at 3rd Australian and New Zealand Falls Prevention Society Conference Melbourne, Australia 1214 October 2008. Sach TH, Foss AJE, Gregson RM, Zaman A, Osborn F, Masud T, et al. Falls and health status in elderly women following first eye cataract surgery: an economic evaluation conducted alongside a randomised controlled trial. British Journal of Ophthalmology 2007;91:1675-9. Salkeld G, Cumming RG, ONeill, Thomas M, Szonyi G, Westbury C. The cost effectiveness of a home hazard reduction program to reduce falls among older persons. Australian and New Zealand Journal of Public Health 2000;24(3):265-71. Sherrington C, Whitney JC, Lord SR, Herbert RD, Cumming RG, Close JCT. Effective exercise for the prevention of falls: a systematic review and meta-analysis. Journal of the American Geriatrics Society 2008 DOI: .10.1111/j.1532-5415.2008.02014.x. Skelton D, Dinan S, Campbell M, Rutherford O. Tailored group exercise (Falls Management Exercise FaME) reduces falls in community-dwelling older frequent fallers (an RCT) [letter]. Age & Ageing 2005;34(6):636-9. Smith RD, Widiatmoko D. The cost-effectiveness of home assessment and modification to reduce falls in the elderly. Australia and New Zealand Journal of Public Health 1998;22(4):436-40. Stevens M, Holman CD, Bennett N, de Klerk N. Preventing falls in older people: outcome evaluation of a randomized controlled trial. Journal of the American Geriatrics Society 2001;49(11):1448-55. Tinetti ME, Baker DI, McAvay G, Claus EB, Garrett P, Gottschalk M, et al. A multifactorial intervention to reduce the risk of falling among elderly people living in the community. New England Journal of Medicine 1994;331(13):821-7.

45

Tinetti ME, Baker DI, King M, Gottschalk M, Murphy TE, Acampara D, et al. Effect of dissemination of evidence in reducing injuries from falls. The New England Journal of Medicine 2008;359(3):25261. van der Velde N, Meerding WJ, Looman CW, Pols HA, van der Cammen TJ. Cost effectiveness of withdrawal of fall-risk-increasing drugs in geriatric outpatients. Drugs and Aging 2008;25(6):5219. Voukelatos A, Cumming RG, Lord SR, Rissel C. A randomized, controlled trial of tai chi for the prevention of falls: the Central Sydney tai chi Trial. Journal of the American Geriatrics Society 2007;55(8):1185-91. Williams M. Building the rationale to provide vitamin D supplementation for fall prevention in residential care. Paper presented at the 3rd Australian and New Zealand Falls Prevention Society Conference Melbourne, Australia 1214 October 2008. Yardley L, Kirby S, Ben-Shlomo Y, Gilbert R, Whitehead S, Todd C. How likely are older people to take up different falls prevention activities? Preventive Medicine 2008; ePub ahead of print. Zermansky AG, Alldred DP, Petty DR, Raynor DK, Freemantle N, Eastaugh J, et al. Clinical medication review by a pharmacist of elderly people living in care homes randomised controlled trial. Age & Ageing 2006;35(6):586-91.

46

Optimisation of ACC's fall prevention programmes for older people APPENDICES TO FINAL REPORT
November 2008

M. Clare Robertson, PhD BSc(Hons) BCom A. John Campbell, MD FRACP

Dunedin School of Medicine, University of Otago Dunedin, New Zealand

Contact details: Associate Professor Clare Robertson Department of Medical and Surgical Sciences Dunedin School of Medicine, PO Box 913 Dunedin 9054, New Zealand Phone: 03 474 7007 ext 8508 Fax: 03 474 7641 Email: clare.robertson@stonebow.otago.ac.nz

CONTENTS

APPENDICES A Project timeline B Methods used in systematic reviews C Interventions for community dwelling older people: systematic review (from progress report dated 15 September 2007) D Interventions in long term care and hospital settings: systematic review (from progress report dated 15 December 2007) E Preventing injury from falls: selected issues (from progress report dated 15 June 2008) F Relative cost effectiveness of interventions: systematic review (from progress report dated 15 March 2008) G Tables of ACC claims and costs 1 2 4 33 50 56 68

Project timeline
Month 13 Month 46 Month 79 Month 1012 Month 1315 Month 1618

Objective
1A 1 B i) 1 B ii) Specific age groups Community dwelling Rest homes

1 B iii) Long term hospital care 1 B iv) Acute hospitals 1C 1D 1E 2A 2B 3A Maori and Pacific populations Specific disabilities People who present to Emergency Departments Identify ACC costs by living situation Cost effectiveness of strategies Identify characteristics of high claimants ACC ACC ACC

3 B, C Survey high claimants (no longer planned) 3D Recommend strategies

Methods used in systematic reviews

The criteria for considering studies for the literature reviews 1) Types of studies: We included randomised controlled trials, including those in which the method of allocation to treatment or control group was inadequately concealed (e.g. trials in which patients were allocated using an open random number list or coin toss). 2) Types of participants: Trials which included elderly individuals, of either sex, a) living independently in the community, or b) the majority of whom were living in institutional care or an inpatient in a hospital. Participant characteristics of interest included falling status at entry (e.g. non-faller, single faller, multiple faller), ethnicity, age, residential status (e.g. community, institution), and where appropriate, associated co-morbidity. 3) Types of interventions: Participants randomised to receive an intervention or group of interventions versus usual care to minimise the effect of, or exposure to, any risk factor for falling. Studies comparing two types of interventions were also included. 4) Types of outcome measures: The main outcomes of interest were the rate of falls, number of fallers, and severity of falls. Severity was assessed by the number of falls resulting in injury, medical attention, or fracture. Information was also sought on adverse events as a result of the interventions employed and duration of effect of the interventions. Trials which focused on intermediate outcomes such as improved balance or strength, and did not report fall rates or number of fallers, were excluded. An improvement in a surrogate outcome does not provide direct evidence that an intervention can impact on the clinical outcome of interest, in this case falls. Therefore, only trials which reported falls or falling as an outcome were included. Search strategy for identification of studies We searched the Cochrane Bone, Joint and Muscle Trauma Group (formerly Musculoskeletal Injuries Group) specialised register, the Cochrane Central Register of Controlled Trials (The Cochrane Library), MEDLINE, EMBASE, CINAHL, AARP AgeLine, The National Research Register Current Controlled Trials (www.controlled-trials.com) and reference lists of articles. No language restrictions were applied. Further trials were identified by contact with researchers in the field. In MEDLINE (OVID ONLINE) subject specific search terms were combined with the first two sections of the optimal MEDLINE trial search strategy. This search strategy was modified for use in other databases: The Cochrane Library, EMBASE, CINAHL, and AARP AgeLine. Selecting trials for inclusion One review author screened the title, abstract and descriptors of identified studies for possible inclusion. From the full text, two authors independently assessed potentially eligible trials for inclusion and resolved any disagreement through discussion. Searches of bibliographies and texts were conducted to identify additional studies or review articles that would be useful for the optimisation project. Assessment of methodological quality Two reviewers independently assessed methodological quality using an 11 item scoring system. Reviewers were not blinded to author and source institution. Disagreement was resolved by consensus, or third party adjudication. We assessed level of concealment of allocation at randomisation using the criteria in the Cochrane Handbook. Studies were graded A if it appeared that the assigned treatment was adequately concealed prior to allocation, B if there was inadequate information to judge concealment, and C if the assigned treatment was clearly not concealed prior to allocation. Methods used to collect data from included trials Data were independently extracted by two reviewers using a data extraction form which was designed and tested prior to use. Disagreement was resolved by consensus, or third party adjudication. Classification of interventions We used the draft taxonomy which is being developed by the European Commission funded Prevention of Falls Network Europe (ProFaNE) (www.profane.eu.org) for classifying and grouping types of interventions.

Statistical analysis Statistical analyses for the meta-analyses were carried out using Excel and the statistical package Stata Release 8.0. We calculated pooled rate ratios and risk ratios with 95% confidence intervals using the fixed-effect model where appropriate. We pooled rate ratios comparing the rate of falls in the intervention group with the control group. We used the rate ratio reported by the authors (for example incidence rate ratio, in preference unadjusted for confounding variables) or, if neither an unadjusted or adjusted ratio were reported, calculated with Excel from falls per person year by group using raw data if these were provided in the article. We pooled reported risk ratios comparing the number of participants in each group with one or more falls during each trial (relative risk, odds ratio, hazard ratio, in preference unadjusted for confounding variables). If neither an unadjusted or adjusted ratio were reported, we used Stata to calculate a relative risk to pool from raw data if provided in the article. Where comparable data were reported, we calculated a pooled relative risk of fracture by pooling reported or calculated risk ratios (as above) comparing the number of participants with fractures (all fractures, hip fracture) in the intervention and the control groups. When no adjustment for clustering had been made to the rate ratio reported for a cluster randomised trial, we adjusted the 95% confidence interval before pooling using an appropriate design effect for that trial. For the trials in an institutional setting we used the intra-cluster correlation coefficient for falls per person year (0.100) reported in a trial in residential care (Dyer 2004). For the risk ratios (number of fallers, number of participants with a fracture) we adjusted the study raw data using the method for categorical data reported in the Cochrane Handbook and the intra-cluster correlation coefficients for the number of residents falling (0.071) and residents sustaining a fracture (0.026) (Dyer 2004), and then recalculated an adjusted risk ratio and confidence interval. For trials in the community we used an intra-cluster correlation of 0.01 (Smeeth 2002). Heterogeneity between pooled trials was tested using a chi-squared test and considered to be statistically significant when P < 0.10. The I2 test was also applied to estimate effects due to heterogeneity rather than sampling error. We considered a value of greater than 50% as substantial heterogeneity. If there was statistical heterogeneity, we pooled results using the random-effects model. Costs and cost effectiveness outcomes When extracting data from the randomised controlled trials we noted any cost outcomes reported. These included the cost of delivering the intervention; cost effectiveness, cost benefit and cost utility ratios; and costs of healthcare use (total or falls related). For each comprehensive economic evaluation we noted the intervention and comparator evaluated; the relative rate reduction in falls; the actual number of falls prevented during the trial; the sample of participants studied; perspectives taken; time horizon; sensitivity analyses; the country and the currency.

Interventions for community systematic review

dwelling

older

people:

Publications identified
A total of 103 randomised controlled trials testing a falls prevention programme in 41,348 community living older people have been identified and reviewed. In this project we divide interventions into single factor, multiple and multifactorial programmes. These terms are defined in the taxonomy for falls prevention interventions currently nearing publication by ProFaNE (Prevention of Falls Network Europe) funded by the European Commission. Single interventions address one type of risk factor for falls whereas multiple and multifactorial programmes have multiple components and address more than one type of risk factor. Examples of single interventions are exercise programmes and home safety programmes. Multifactorial interventions include a risk assessment with programme components to address the identified risks for each individual. In a multiple programme, the same components are delivered to each individual. We published an article in Age and Ageing that compared the effectiveness of single versus multicomponent interventions (Campbell 2007). This article provided valuable information for compiling the recommendations for this project. In the 103 trials identified and meeting the inclusion criteria, the interventions tested were: 81 single factor interventions 29 multifactorial programmes 11 multiple intervention programmes A total of 121 interventions were tested. The most common intervention was an exercise programme (48 were tested) and 12 used vitamin D supplementation. These are not the final numbers as further trials are regularly appearing in the literature. No information on falls specific to Maori and Pacific populations has been identified so far in the literature. The search for relevant publications for this project has not been restricted to articles reporting randomised controlled trials. We have also compiled a list of 47 relevant systematic reviews, metaanalyses and general reviews of falls and fall injury interventions. We used these publications to double check that we had identified all relevant randomised controlled trials, and to consider when compiling our final recommendations to ACC and the Ministry of Health. The guidelines developed by the American Geriatrics Society, the British Geriatrics Society and American Academy of Orthopaedic Surgeons Panel on Falls Prevention are currently being updated.

Interventions successful in reducing falls

Data on the rate of falls during the trial were available from one multiple and 11 multifactorial interventions tested in randomised controlled trials in community living older people and considered appropriate for pooling. We pooled data separately from the following types of exercise programme the Otago Exercise Programme, tai chi, and group exercise programmes. Data on the rate of falls were also available for pooling from three home safety programmes and two trials testing cataract removal. For the pooling of data from trials testing vitamin D supplementation (9 trials) we used number of fallers in the vitamin D and control groups. Number or rate of falls is a much superior outcome to pool but these data were available from only two of the 12 trials testing vitamin D supplementation. We include graphs for withdrawal of psychotropic medication and fitting a pacemaker, two further interventions that have been shown to reduce falls in community living older people, although evidence from only one trial of each has been published to date. The cut-off date for literature to include in these analyses was 31 August 2007. Interventions with multiple components The first intervention tested in a randomised controlled trial that was successful in reducing falls was a multifactorial programme addressing individual risk factors for falls delivered at home by a nurse and physiotherapist (Tinetti 1994). Participants were aged 70 or over and had one or more of eight specified risk factors for falling. The intervention was cost saving in those with four or more of the eight risk factors at baseline (Rizzo 1996).

We found a total of 40 interventions with multiple components tested in community living people. Pooling of 12 programmes with data on the rate of falls available, showed that the rate of falls was significantly reduced by 25% (see Figure 1 below). This finding fits with the fact that all current guidelines recommend addressing multifactorial risk factors as an effective strategy for reducing falls. We will investigate further the heterogeneity of the pooled results (see below) to see whether a particular combination of components, the particular population targeted, the setting, the professional background of the staff delivering the intervention, or methodological characteristics of the trial appear to contribute to the success or failure of these interventions. A programme that may well provide a possible multiple intervention option for ACC and the Ministry of Health to promote is Stepping On (Clemson 2004). In a 14 month randomised controlled trial in Sydney, Australia, small group learning (2-hour weekly sessions for seven weeks) led to a significant 31% reduction in falls in community living people who had had a fall in the previous 12 months. Investigation of the cost effectiveness of effective interventions to be undertaken in Phase 2 of this project will provide information on the likely value for money of multicomponent interventions compared with other effective interventions such as exercise, home safety programmes, and cataract removal.

Hornbrook 1994 Tinetti 1994 Gallagher 1996 Close 1999 Hill 2000 Hogan 2001 Nikolaus 2003 Clemson 2004 Davison 2005 Lord 2005 a) Lord 2005 b) Mahoney 2007

Combined

.1

.5 Relative rate ratio

1.5

Figure 1. Meta-analysis of programmes with multifactor components. Pooled rate ratio 0.75 (95%CI 0.65 to 0.88), P = < 0.001. Test for heterogeneity: Q = 78.65, P = 0.001.

Otago Exercise Programme The Otago Exercise Programme is currently the only falls prevention intervention that has been tested in more than one randomised controlled trial. This strength and balance retraining programme was designed specifically to reduce falls in community living older people. It has been tested in nine different cities and towns in New Zealand. Pooling of the rate ratios from the first three randomised controlled trials shows that the rate of falls was reduced by 34% (see Figure 2 below). A meta-analysis of individual data from these three trials plus a fourth controlled trial, showed that both the overall rate of falls and the rate of injuries were reduced by 35% (Robertson 2002). These four trials have shown that the programme is equally effective when delivered by an experienced physiotherapist, and by nurses trained and supervised by an expert physiotherapist. Process and impact evaluations of the programme have been published, and the cost effectiveness of the programme established in three of the four trials (Robertson 2001a, 2001b, 2001c). Over 1,000 men and women aged 65 to 97 years took part in the trials for up to 2 years and no other falls risk factors were used to select participants, however those aged 80 years and older with a previous fall benefited the most in terms of injury prevention (Robertson 2002).

Campbell 1997

Campbell 1999 b)

Robertson 2001

Combined

.1

.5 Relative rate ratio

1.5

Figure 2. Meta-analysis for Otago Exercise Programme (New Zealand trials). Pooled rate ratio 0.66 (95%CI 0.52 to 0.83), P < 0.001. Test for heterogeneity: Q = 1.003, P = 0.606.

Campbell 1997

Campbell 1999 b)

Robertson 2001

Donaldson 2007

Combined

.1

.5 Relative rate ratio

1.5

Figure 3. Meta-analysis for Otago Exercise Programme (includes a recent trial in Canada). Pooled rate ratio 0.64 (95%CI 0.51 to 0.81), P < 0.001. Test for heterogeneity: Q = 1.645, P = 0.649.

Recently a small randomised controlled trial of the Otago Exercise Programme was completed in Vancouver, Canada (Donaldson 2007). The 74 participants, aged 70 and older, were recruited after attending a dedicated referral based falls clinic service. All received a comprehensive geriatric assessment before being randomised to receive the Otago Exercise Programme at home or usual care. After 12 months falls were reduced by 54% in those receiving the Otago Exercise Programme (incidence rate ratio 0.46,

95% confidence interval 0.20 to 1.10). In this trial there were three obvious outliers (two in the control group and one in the intervention group, at least 18 falls each). With these outliers removed, and a negative binomial distribution maintained, the adjusted incidence rate ratio was 0.44 (95% confidence interval 0.23 to 0.84). Pooling the conservative results from this trial with the three New Zealand randomised controlled trials, shows that the rate of falls was reduced by 36% (see Figure 3 above). This further endorses the benefit of this programme in those older people who have fallen and are considered at risk of further falls. Tai chi Tai chi has been shown to improve balance and has been suggested as a form of exercise to reduce falls since the successful trial by Wolf and colleagues (Wolf 1996). In this trial 72 participants aged 70 and older attended tai chi classes twice a week for 15 weeks. Falls were a secondary outcome of the study and were monitored for 7 to 20 months. Compared with the 138 participants receiving computerised balance training or health education classes, the rate of falls was reduced by 47.5%. A further larger trial by Wolf, this time in a frailer group of older people and with a more intense version of tai chi, after 48 weeks showed a non-significant reduction in the rate of falls (Wolf 2003). There are now three other trials with data available for pooling. Participants in the trial with the largest reduction in the rate of falls (55%) were physically inactive 70 to 92 year olds in the US who were monitored for 6 months (Li 2005). The pooled rate ratio from the five published trials (n = 1,629 participants) shows that tai chi reduced the rate of falls by 40% (see Figure 4 below).

Wolf 1996 a)

Wolf 2003

Li 2005

Voukelatos 2007

Woo 2007 a)

Combined

.1

.5 Relative rate ratio

1.5

Figure 4. Meta-analysis for current tai chi programmes. Pooled rate ratio 0.60 (95%CI 0.50 to 0.73), P < 0.001. Test for heterogeneity: Q = 3.99, P = 0.408.

ACC has funded a randomised control trial of tai chi in people aged 65 years and older with a fall in the previous year and this pragmatic trial is in progress. We use here three scenarios to investigate whether the results of this trial are likely to change the outcome of the meta-analysis of the published tai chi trials. Scenario 1: If the New Zealand trial contributes results replicating the best trial outcome so far (Li 2005), the pooled relative rate ratio would be 0.57 (95% confidence interval 0.47 to 0.68):

Wolf 1996 a)

Wolf 2003

Li 2005

Voukelatos 2007

Woo 2007 a)

NZ (hypothetical)

Combined

.1

.5 Relative rate ratio

1.5

Figure 5. Meta-analysis for current tai chi programmes plus hypothetical results (scenario 1) from the New Zealand study. Pooled rate ratio 0.57 (95%CI 0.47 to 0.68), P < 0.001. Tests for heterogeneity: Q = 5.46, P = 0.362.

Scenario 2: If the results replicate the worst trial in the meta-analysis above (Wolf 2003), the pooled relative rate ratio would be 0.63 (95% confidence interval 0.53 to 0.75):

Wolf 1996 a)

Wolf 2003

Li 2005

Voukelatos 2007

Woo 2007 a)

NZ (hypothetical)

Combined

.1

.5 Relative rate ratio

1.5

Figure 6. Meta-analysis for current tai chi programmes plus hypothetical results (scenario 2) from the New Zealand study. Pooled rate ratio 0.63 (95%CI 0.53 to 0.75), P < 0.001. Tests for heterogeneity: Q = 5.11, P = 0.403.

Scenario 3: In a worse case scenario where the New Zealand results show tai chi in the trial significantly increased the rate of falls (hypothetical incidence rate ratio 1.50, 95% confidence interval 1.05 to 2.25), the pooled analysis would still indicate a significant reduction from intervention with tai chi (pooled relative rate ratio 0.69, 95% confidence interval 0.48 to 0.99):

Wolf 1996 a)

Wolf 2003

Li 2005

Voukelatos 2007

Woo 2007 a)

NZ (hypothetical)

Combined

.1

.5 Relative rate ratio

1.5

Figure 7. Meta-analysis for current tai chi programmes plus hypothetical results (scenario 3) from the New Zealand study. Pooled rate ratio 0.69 (95%CI 0.48 to 0.99), P = 0.043. Tests for heterogeneity: Q = 21.71, P = 0.001.

These three scenarios demonstrate that the results of the New Zealand trial, whether the outcome indicates a significant decrease or even a significant increase in falls, are unlikely to change the message from the pooled analysis from current trials shown above (Figure 4). All three scenarios gave a pooled relative rate ratio that indicated tai chi significantly reduces the rate of falls (range 31% to 43%) (see forest plots in Figures 5, 6 and 7 above). Of the tai chi trials currently published, the Central Sydney tai chi Trial best emulates the conditions expected in the New Zealand community (Voukelatos 2007). The 702 relatively healthy participants were aged 60 and older (mean 69, standard deviation 6.5, range 60 to 96 years). Those randomised to receive tai chi attended classes for one hour once a week for 16 weeks. The style of tai chi taught varied (majority Sun style, two classes Yang style) but all instructors had at least five years experience or had completed an accredited tai chi trainers course. Falls were monitored for only 24 weeks (at least 12 months is the optimal follow up time) when the rate of falls had been reduced by 33%. Funders and providers of tai chi programmes for older people in New Zealand requiring a standardised programme should continue to use a Paul Lam version of tai chi. The modified version of tai chi used in most of the classes in the Central Sydney tai chi Trial was designed by Dr Paul Lam especially for older people with arthritis (www.taichiproductions.com). Tai Chi for Arthritis has 12 movements based on the Sun style. In 2005 ACC invited Dr Lam to help design safety measures and set up training courses for class teachers to be used throughout New Zealand. The effectiveness of tai chi in reducing falls has been established, but value for money to ACC and the Ministry of Health from funding tai chi classes, and sustainability of effect longer than 6 months, are issues still to be determined. In the published trials, falls were reduced by 40% in a combination of healthy people aged 60 and older plus older, frailer people. Therefore value for money will depend on the subgroup of older people targeted, as fall injury costs in different subgroups of older people would be expected to differ. It will also depend on the cost effectiveness of providing tai chi classes compared with other successful falls prevention programmes. Economic modelling of these evaluations formed part of Objective 2 of this project.

Group exercise programmes Exercise classes, typically with strength, balance, agility and flexibility exercises included and taught by a trained instructor, were the most common single intervention in the literature (23 trials). Overall there was significant reduction in the rate of falls by 29% in the 9 trials using such a combination of exercises and with data on the rate of falls available for pooling (see Figure 8 below).

Lord 1995 Buchner 1997 McMurdo 1997 Barnett 2003 Lord 2003 Suzuki 2004 Bunout 2005 Skelton 2005 Weerdesteyn 2006

Combined

.1

.5 Relative rate ratio

1.5

Figure 8. Meta-analysis of group exercise programmes. Pooled rate ratio 0.71 (95%CI 0.62 to 0.81), P < 0.001. Test for heterogeneity: Q = 10.32, P = 0.243.

A recent systematic review of exercise programmes to prevent falls concludes with the following key messages (Robertson 2007): Many different risk factors contribute to falls but muscle weakness and poor balance underlie most falls The most common components of successful exercise programmes to reduce falls are moderate intensity strength training against resistance and dynamic balance retraining, but there are insufficient good quality negative studies to determine whether some types of exercises are ineffective Exercise programmes that are individually tailored, progress in difficulty, and target carefully selected groups at high risk have so far resulted in the greatest absolute reduction in falls and injuries Those running falls prevention programmes in particular settings or subgroups of older people should use evidence based interventions To facilitate this, researchers should make the necessary programme details available. Home safety programmes Modifications in the home aimed at reducing hazards associated with falls are a component of many of the multifactorial falls prevention programmes. However, effectiveness of this approach was not known until trials testing home assessment and modification alone were completed. We found three trials testing a home safety programme as a single intervention and with the rate of falls available for pooling. The populations tested differed considerably which may explain the significant heterogeneity of the pooled analysis. Pooling data from trials in a general community sample aged 70 or more (Stevens 2001), people aged 65 and over recently discharged from hospital (Cumming 1999), and those over 75 with severe visual impairment (Campbell 2005), gave a non-significant 17% reduction in fall rate (see Figure 9 below).

10

Cumming 1999

Stevens 2001

Campbell 2005 a)

Combined

.1

.5 Relative rate ratio

1.5

Figure 9. Meta-analysis of home safety programmes. Pooled rate ratio 0.83 (95%CI 0.64 to 1.07), P = 0.156. Test for heterogeneity: Q = 7.16, P = 0.028.

Cumming 1999*

Campbell 2005 a)

Combined

.1

.5 Relative rate ratio

1.5

Figure 10. Meta-analysis of home safety programmes in high risk older people. Pooled rate ratio 0.69 (95%CI 0.56 to 0.84), P = 0.001. Test for heterogeneity: Q = 1.23, P = 0.267.

The three trials pooled (Figure 9) give some clear messages. They differed not only in the sample recruited, but also in terms of the health professional delivering the programme, and in the follow through aspects for the home modifications. In the unsuccessful study by Stevens and coleagues, a research nurse made one home visit only, whereas in the other two trials (Cumming 1999; Campbell 2005), the intervention

11

was delivered by an experienced occupational therapist who facilitated follow through with the home modifications and behaviour changes. In the trial by Cumming, the subgroup who benefited most from the home safety programme were those reporting a fall in the previous year (hazard ratio 0.75, 95% confidence interval 0.58 to 0.96). We provide a meta-analysis of this high risk subgroup plus the trial results of people with severe visual impairment (Campbell 2005). As the meta-analysis demonstrates (Figure 10 above), there is a significant 31% reduction in the rate of falls in these two high risk groups by using home safety as an intervention. A comprehensive economic evaluation of the home safety programme in Cummings trial confirms that the programme is more likely to be cost effective in older people who have a history of falls (Salkeld 2000). The cost effectiveness of the home safety programme in people with visual impairment has been established (Campbell 2005). Psychotropic medication withdrawal Withdrawing psychotropic mediation was shown in one of our own trials to significantly reduce the risk of (all) falls by 66%. Participants were aged 65 and older and were regularly taking psychotropic medication, mainly for sleeping. All participants were provided with identical looking capsules prepared by a pharmacist this is currently the only double blind trial testing a successful falls prevention intervention. In those randomised to the withdrawal group, the active medication was gradually reduced over 14 weeks whereas the control group continued to receive their active mediation throughout the 44 week trial. For comparison purposes we provide here the same forest plot template for this one trial (Figure 11). It should be noted that recruitment for this trial was difficult and the drop out rate high (23%). Also, one month after the end of the trial 8 of the 17 (47%) who had withdrawn from their medication had resumed taking a psychotropic drug. Therefore this successful intervention is recommended for individuals, but not as a public health intervention for widespread dissemination.

Campbell 1999 a)

Combined

.1

.5 Relative rate ratio

1.5

Figure 11. Rate ratio for withdrawal of psychotropic medication compared with remaining on active medication (hazard ratio 0.34, 95%CI 0.16 to 0.74).

Vision improvement Older people with visual impairment are at double the risk of falling compared with their normally sighted peers. Three trials have tested an intervention aimed at reducing falls by improving vision. Two trials tested whether expedited cataract removal would reduce falls in women over 70 who had been referred to a hospital ophthalmology department (Harwood 2005; Foss 2006). Patients on the waiting list were

12

randomised to expedited surgery (approximately 4 weeks) or to remain on the waiting list (12 months wait). The rate of falls was significantly reduced following first eye cataract surgery compared with those on the waiting list (Harwood 2005), but not following second eye surgery (Foss 2006). However, pooling of the rate ratio from the two trials showed a significant 33% reduction in falls (Figure 12).

Harwood 2005

Foss 2006

Combined

.1

.5 Relative rate ratio

1.5

Figure 12. Meta-analysis of expedited first or second cataract removal. Pooled rate ratio 0.67 (95%CI 0.49 to 0.91), P = 0.011. Test for heterogeneity: Q = 0.008, P = 0.931.

The third trial aiming to prevent falls by improving vision, had quite the opposite effect on the fall rate (Cumming 2007). Community living people aged 70 and older recruited mainly from an aged care service, received vision and eye examinations and subsequent treatment of eye problems. After 12 months the rate of falls was significantly increased by 57% in the intervention group compared with the control group (incidence rate ratio 1.57, 95% confidence interval 1.20 to 2.05). Among the reasons for this increase is the fact that bifocal and multifocal glasses appear to contribute to falls by blurring the lower visual field at critical distances for identifying hazards when walking. The most common treatment in the intervention group was new glasses, therefore the main message here appears to be that older people should be very cautious after major prescription changes to their multifocal or bifocal glasses. A trial is in progress in Sydney testing whether supplying people with monofocal glasses for walking will reduce falls. No publications were found on the cost effectiveness of cataract removal in reducing falls. This investigation will form part of the economic modeling in Objective 2 of this project. Cardiac pacing Cardioinhibitory carotid sinus syndrome causes syncope, and symptoms can be relieved by cardiac pacing. One trial showed that falls were significantly reduced by 58% when eligible patients were fitted with a pacemaker. For comparison purposes we provide the same forest plot template for this one trial (Figure 13 below), so that further studies can be added if available for the final report. From 71,299 people aged 50 years or over attending a UK accident and emergency department, carotid sinus massage was performed on 1,624 patients and of these, 257 were diagnosed with cardioinhibitory or mixed carotid sinus hypersensitivity and therefore eligible for the trial (0.4% of the 71,299). Of these 175 agreed to take part. The average age of those eligible was 74. Falls are the most common reason for older people to attend an accident and emergency department and this intervention provides an effective treatment for the very small percentage of older people who will benefit from a pacemaker.

13

Kenny 2001

Combined .1 .5 Relative rate ratio 1 1.5 2

Figure 13. Rate ratio for pacemaker implant in older people with carotid sinus hypersensitivity 0.42 (95%CI 0.23 to 0.75).

Vitamin D supplementation

Pfeifer 2000

Latham 2003 b)

Trivedi 2003

Dukas 2004

Harwood 2004

Grant 2005 a)

Porthouse 2005

Bischoff-Ferrari 2006

Gallagher 2007 a)

Combined

.1

.5 Relative risk

1.5

Figure 14. Meta-analysis of trials testing the relative risk of falling for vitamin D supplementation. Pooled relative risk from random effects model 0.86 (95%CI 0.75 to 1.00), P = 0.044. Test for heterogeneity: Q = 14.31, P = 0.074.

14

Vitamin D has been suggested as a falls prevention strategy because low levels have been associated with increased body sway and loss of muscle strength. Levels of vitamin D are known to be low in many older people and there is some previous evidence that increasing the level will reduce falls (Bischoff-Ferrari 2004). An incidence rate ratio comparing a group taking vitamin D supplementation (with or without calcium) compared with a placebo or calcium group was available for only two of the 12 trials identified. Therefore we provide here the risk ratio obtained from pooling the risk of one or more falls (number of fallers) in the intervention versus the control group from the nine trials with these data available. Overall vitamin D supplementation significantly reduced the risk of a fall by 14% (see Figure 14 above). In the two trials most successful in reducing the number of fallers, participants were 489 healthy women aged 65 to 77 years (Gallagher 2007), and 150 previously independent elderly women recruited following surgery for hip fracture (Harwood 2004). Vitamin D supplementation forms part of accepted management of people with osteoporosis, and together with calcium nutrition is important in the maintenance of musculoskeletal health. High dose preparations of vitamin D that enable less frequent dosing are likely to be effective, cheaper and have higher adherence.

Interventions not successful in reducing falls

Currently there are several interventions trialed that were not effective in reducing falls. These include exercise programmes with only one type of exercise, for example resistance training only, walking groups, impact exercises, computerised balance exercises, agility training only, weight bearing exercises only, step up step down exercise only, and stretching and weight shifting exercises. Other interventions that have been tested and were not effective include hormone replacement treatment, nutrient supplementation, falls education alone, and cognitive behavioural programmes.

Implications of findings from the literature review

With current evidence, ACC and the Ministry of Health are justified in continuing to fund the Otago Exercise Programme for those 80 and older, and those 75 and older with a previous fall. A small study carried out in Vancouver, Canada has added to the evidence from the New Zealand trials and endorsed the effectiveness of the Otago Exercise Programme in over 70 year olds considered at high risk of falling. With current evidence, ACC is justified in continuing to fund tai chi classes although long term sustainable benefit for more than 6 months has not been confirmed. There is some evidence that frailer older people do not benefit as much in terms of falls prevention as relatively healthy or physically inactive people aged 60 years and older. Comparative cost effectiveness also needs to be established. The most common components of successful group exercise programmes to reduce falls are moderate intensity strength training against resistance and dynamic balance retraining, but there are insufficient good quality negative studies to determine whether some types of exercises are ineffective. Exercise programmes that are individually tailored, progress in difficulty, and target carefully selected groups at high risk have so far resulted in the greatest absolute reduction in falls and injuries. Elderly people with vision loss are at double the risk of falls compared with those who have normal sight. Vision should be optimal and cataract removal does prevent falls. Older people should be advised to be extremely careful while adjusting to major changes in lens prescriptions. Monofocal glasses are recommended while walking. Certain multifactorial programmes reduce falls but any advantage over single interventions in terms of effectiveness and cost effectiveness has not been demonstrated. Those running falls prevention programmes in particular settings or subgroups of older people should use the successful programmes that are available.

15

In funding research on injury prevention, ACC should look to answer questions where: a) there is evidence in the international literature but New Zealand conditions are likely to be different, or b) the answer cannot be provided from the literature by meta-analysis of existing studies, and primary or confirmatory research is needed. Preventing falls prevents injuries resulting from falls, but the lack of information on injury outcomes in most of the trials makes estimation of the potential benefit in terms of injury prevention difficult.

16

References
Bischoff-Ferrari HA, Dawson-Hughes B, Willett WC, Staehelin HB, Bazemore MG, Zee RY, et al. Effect of Vitamin D on falls: a meta-analysis. JAMA 2004;291(16):1999-2006. Campbell AJ, Robertson MC, La Grow SJ, Kerse NM, Sanderson GF, Jacobs RJ, et al. Randomised controlled trial of prevention of falls in people aged 75 with severe visual impairment: the VIP trial. BMJ 2005;331(7520):817. Campbell AJ, Robertson MC. Rethinking individual and community fall prevention strategies: a metaregression comparing single and multifactorial interventions. Age & Ageing 2007;36:656-62. Clemson L, Cumming RG, Kendig H, Swann M, Heard R, Taylor K. The effectiveness of a communitybased program for reducing the incidence of falls in the elderly: a randomized trial. Journal of the American Geriatrics Society 2004;52(9):1487-94. Cumming RG, Thomas M, Szonyi G, Salkeld G, O'Neill E, Westbury C, et al. Home visits by an occupational therapist for assessment and modification of environmental hazards: a randomized trial of falls prevention. Journal of the American Geriatrics Society 1999;47(12):1397-402. Cumming RG, Ivers R, Clemson L, Cullen J, Hayes MF, Tanzer M, et al. Improving vision to prevent falls in frail older people: a randomized trial. Journal of the American Geriatrics Society 2007;55(2):17581. Rizzo JA, Baker DI, McAvay G, Tinetti ME. The cost-effectiveness of a multifactorial targeted prevention program for falls among community elderly persons. Medical Care 1996;34(9):954-9. Robertson MC, Devlin N, Gardner MM, Campbell AJ. Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 1: Randomised controlled trial. BMJ 2001a;322(7288):697-701. Robertson MC, Devlin N, Gardner MM, McGee R, Campbell AJ. Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 2: Controlled trial in multiple centres. BMJ 2001b;322(7288):701-4. Robertson MC, Devlin N, Scuffham P, Gardner MM, Buchner DM, Campbell AJ. Economic evaluation of a community based exercise programme to prevent falls. Journal of Epidemiology and Community Health 2001c;55(8):600-6. Robertson MC, Campbell AJ, Gardner MM, Devlin N. Preventing injuries in older people by preventing falls: a meta-analysis of individual-level data. Journal of the American Geriatrics Society 2002;50(5):905-11. Robertson MC, Campbell AJ, Herbison P. Statistical analysis of efficacy in falls prevention trials. Journals of Gerontology Series A Biological Sciences and Medical Sciences 2005;60(4):530-4. Robertson MC, Campbell AJ. What type of exercise reduces falls in older people? Chapter (systematic review) in: Evidence-based sports medicine, second edition. Editors Domhnall MacAuley and Thomas Best. Blackwell Publishing, Oxford, UK, 2007, pp135-166. Salkeld G, Cumming RG, ONeill, Thomas M, Szonyi G, Westbury C. The cost effectiveness of a home hazard reduction program to reduce falls among older persons. Australian and New Zealand Journal of Public Health 2000;24(3):265-71. Stevens M, Holman CD, Bennett N, de Klerk N. Preventing falls in older people: outcome evaluation of a randomized controlled trial. Journal of the American Geriatrics Society 2001;49(11):1448-55. Tinetti ME, Baker DI, McAvay G, Claus EB, Garrett P, Gottschalk M, et al. A multifactorial intervention to reduce the risk of falling among elderly people living in the community. New England Journal of Medicine 1994;331(13):821-7. Voukelatos A, Cumming RG, Lord SR, Rissel C. A randomized, controlled trial of tai chi for the prevention of falls: the Central Sydney tai chi Trial. Journal of the American Geriatrics Society 2007;55(8):1185-91. Wolf SL, Barnhart HX, Kutner NG, McNeely E, Coogler C, Xu T, et al. Reducing frailty and falls in older persons: an investigation of Tai Chi and computerized balance training. Journal of the American Geriatrics Society 1996;44(5):489-97. Wolf SL, Sattin RW, Kutner M, O'Grady M, Greenspan AI, Gregor RJ. Intense tai chi exercise training and fall occurrences in older, transitionally frail adults: a randomized, controlled trial. Journal of the American Geriatrics Society 2003;51(12):1693-701.

17

References for systematic reviews, meta-analyses, reviews

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Allain H, Bentue-Ferrer D, Polard E, Akwa Y, Patat A. Postural instability and consequent falls and hip fractures associated with use of hypnotics in the elderly: a comparative review. Drugs & Aging 2005;22(9):749-65. American Geriatrics Society, British Geriatrics Society, American Academy of Orthopaedic Surgeons Panel on Falls Prevention. Guideline for the prevention of falls in older persons. Journal of the American Geriatrics Society 2001;49(5):664-72. Avenell A, Gillespie WJ, Gillespie LD, O'Connell DL. Vitamin D and vitamin D analogues for preventing fractures associated with involutional and post-menopausal osteoporosis. Cochrane Database of Systematic Reviews 2005(3):CD000227. Bischoff-Ferrari HA, Dawson-Hughes B, Willett WC, Staehelin HB, Bazemore MG, Zee RY, et al. Effect of Vitamin D on falls: a meta-analysis. JAMA 2004;291(16):1999-2006. Bischoff-Ferrari HA, Willett WC, Wong JB, Giovannucci E, Dietrich T, Dawson-Hughes B. Fracture prevention with vitamin D supplementation: a meta-analysis of randomized controlled trials. JAMA 2005;293(18):2257-64. Campbell AJ, Robertson MC. Rethinking individual and community fall prevention strategies: a meta-regression comparing single and multifactorial interventions. Age & Ageing 2007;36:656-62. Carter ND, Kannus P, Khan KM. Exercise in the prevention of falls in older people: a systematic literature review examining the rationale and the evidence. Sports Medicine 2001;31(6):427-38. Chang JT, Morton SC, Rubenstein LZ, Mojica WA, Maglione M, Suttorp MJ, et al. Interventions for the prevention of falls in older adults: systematic review and meta-analysis of randomised clinical trials. BMJ 2004;328(7441):680. Cumming RG. Intervention strategies and risk-factor modification for falls prevention. A review of recent intervention studies. Clinics in Geriatric Medicine 2002;18(2):175-89. Cumming RG, Le Couteur DG. Benzodiazepines and risk of hip fractures in older people: a review of the evidence. CNS Drugs 2003;17(11):825-37. Feder G, Carter Y, Donovan S, Cryer C. Preventing osteoporosis, falls, and fractures among elderly people. Few exercise programmes studied have prevented falls. BMJ 1999;318(7199):1695; author reply -6. Feder G, Cryer C, Donovan S, Carter Y. Guidelines for the prevention of falls in people over 65. The Guidelines' Development Group. BMJ 2000;321(7267):1007-11. Gardner MM, Robertson MC, Campbell AJ. Exercise in preventing falls and fall related injuries in older people: a review of randomised controlled trials. British Journal of Sports Medicine 2000;34(1):7-17. Gillespie LD, Gillespie WJ, Robertson MC, Lamb SE, Cumming RG, Rowe BH. Interventions for preventing falls in elderly people. Cochrane Database of Systematic Reviews 2003(4):CD000340. Gillespie WJ, Avenell A, Henry DA, O'Connell DL, Robertson J. Vitamin D and vitamin D analogues for preventing fractures associated with involutional and post-menopausal osteoporosis. Cochrane Database of Systematic Reviews 2001(1):CD000227. Gregg EW, Pereira MA, Caspersen CJ. Physical activity, falls, and fractures among older adults: a review of the epidemiologic evidence. Journal of the American Geriatrics Society 2000;48(8):88393. Hauer K, Becker C, Lindemann U, Beyer N. Effectiveness of physical training on motor performance and fall prevention in cognitively impaired older persons: a systematic review. American Journal of Physical Medicine & Rehabilitation 2006;85(10):847-57. Hill-Westmoreland EE, Soeken K, Spellbring AM. A meta-analysis of fall prevention programs for the elderly: how effective are they? Nursing Research 2002;51(1):1-8. Jackson C, Gaugris S, Sen SS, Hosking D. The effect of cholecalciferol (vitamin D3) on the risk of fall and fracture: a meta-analysis. QJM 2007;100(4):185-92. Jorstad EC, Hauer K, Becker C, Lamb SE, ProFaNE Group. Measuring the psychological outcomes of falling: a systematic review. Journal of the American Geriatrics Society 2005;53(3):501-10. Judge JO. Balance training to maintain mobility and prevent disability. American Journal of Preventive Medicine 2003;25(3 Suppl 2):150-6. Kannus P, Sievanen H, Palvanen M, Jarvinen T, Parkkari J. Prevention of falls and consequent injuries in elderly people. Lancet 2005;366(9500):1885-93.

18

23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

Kannus P, Uusi-Rasi K, Palvanen M, Parkkari J. Non-pharmacological means to prevent fractures among older adults. Annals of Medicine 2005;37(4):303-10. Kellogg International Work Group on the Prevention of Falls by the Elderly. The prevention of falls in later life. A report of the Kellogg International Work Group on the Prevention of Falls by the Elderly. Danish Medical Bulletin 1987;34 (suppl 4):1-24. Latham NK, Anderson CS, Reid IR. Effects of vitamin D supplementation on strength, physical performance, and falls in older persons: a systematic review. Journal of the American Geriatrics Society 2003;51(9):1219-26. Latham NK, Bennett DA, Stretton CM, Anderson CS. Systematic review of progressive resistance strength training in older adults. Journals of Gerontology Series A-Biological Sciences & Medical Sciences 2004;59(1):48-61. Leipzig RM, Cumming RG, Tinetti ME. Drugs and falls in older people: a systematic review and meta-analysis: II. Cardiac and analgesic drugs. Journal of the American Geriatrics Society 1999;47(1):40-50. Leipzig RM, Cumming RG, Tinetti ME. Drugs and falls in older people: a systematic review and meta-analysis: I. Psychotropic drugs. Journal of the American Geriatrics Society 1999;47(1):30-9. McClure R, Turner C, Peel N, Spinks A, Eakin E, Hughes K. Population-based interventions for the prevention of fall-related injuries in older people. Cochrane Database of Systematic Reviews 2005(1):CD004441. Moreland J, Richardson J, Chan DH, O'Neill J, Bellissimo A, Grum RM, et al. Evidence-based guidelines for the secondary prevention of falls in older adults. Gerontology 2003;49(2):93-116. Moreland JD, Richardson JA, Goldsmith CH, Clase CM. Muscle weakness and falls in older adults: a systematic review and meta-analysis. Journal of the American Geriatrics Society 2004;52(7):11219. Myers AH, Young Y, Langlois JA. Prevention of falls in the elderly. Bone 1996;18(1 Suppl):87S101S. NHS National Institute for Clinical Excellence. Falls: the assessment and prevention of falls in older people. Clinical Guideline 21, November 2004. Available at http://www.nice.org.uk/CG021NICEguideline Accessed May 2005. Oakley A, Dawson MF, Holland J, Arnold S, Cryer C, Doyle Y, et al. Preventing falls and subsequent injury in older people. Quality in Health Care 1996;5(4):243-9. Oliver D, Connelly JB, Victor CR, Shaw FE, Whitehead A, Genc Y, et al. Strategies to prevent falls and fractures in hospitals and care homes and effect of cognitive impairment: systematic review and meta-analyses. BMJ 2007;334(7584):82. Oliver S, Peersman G, Harden A, Oakley A. Discrepancies in findings from effectiveness reviews: the case of health promotion for older people in accident and injury prevention. Health Education Journal 1999;58:66-77. Parker MJ, Gillespie WJ, Gillespie LD. Hip protectors for preventing hip fractures in older people. Cochrane Database of Systematic Reviews 2005(3):CD001255. Parker MJ, Gillespie WJ, Gillespie LD. Effectiveness of hip protectors for preventing hip fractures in elderly people: systematic review. BMJ 2006;332(7541):571-4. Province MA, Hadley EC, Hornbrook MC, Lipsitz LA, Miller JP, Mulrow CD, et al. The effects of exercise on falls in elderly patients. A preplanned meta-analysis of the FICSIT trials. JAMA 1995;273:1341-7. Robertson MC, Campbell AJ. What type of exercise reduces falls in older people? In: Domhnall MacAuley, Thomas Best, editors. Evidence-based sports medicine, second edition. Oxford, UK: Blackwell Publishing, 2007. Robertson MC, Campbell AJ, Gardner MM, Devlin N. Preventing injuries in older people by preventing falls: a meta-analysis of individual-level data. Journal of the American Geriatrics Society 2002;50(5):905-11. Rubenstein LZ, Powers CM, MacLean CH. Quality indicators for the management and prevention of falls and mobility problems in vulnerable elders. Annals of Internal Medicine 2001;135(8 Pt 2):68693. Sherrington C, Lord SR, Finch CF. Physical activity interventions to prevent falls among older people: update of the evidence. Journal of Science & Medicine in Sport 2004;7(1 Suppl):43-51. Vu MQ, Weintraub N, Rubenstein LZ. Falls in the nursing home: Are they preventable? Journal of the American Medical Directors Association 2005;6(3 Suppl):S82-7.

19

45 46 47

Weatherall M. Prevention of falls and fall-related fractures in community-dwelling older adults: a meta-analysis of estimates of effectiveness based on recent guidelines. Internal Medicine Journal 2004;34(3):102-8. Weigand JV, Gerson LW. Preventive care in the emergency department: should emergency departments institute a falls prevention program for elder patients? A systematic review. Academic Emergency Medicine 2001;8(8):823-6. Wu G. Evaluation of the effectiveness of Tai Chi for improving balance and preventing falls in the older populationa review. Journal of the American Geriatrics Society 2002;50(4):746-54.

20

References for falls prevention randomised controlled trials in the community included in systematic review
1 2 3 4 5 6 7 8 Alexander NB, Bentur N, Strasburg D, Nyquist LV. Fall risk reduction in Israeli day care center attendees using exercise and behavior strategies [abstract]. Journal of the American Geriatrics Society 2003;51(Suppl 4):S117. Armstrong AL, Oborne J, Coupland CAC, Macpherson MB, Bassey EJ, Wallace WA. Effects of hormone replacement therapy on muscle performance and balance in post-menopausal women. Clinical Science 1996;91(6):685-90. Assantachai P, Chatthanawaree W, Thamlikitkul V, Praditsuwan R, Pisalsarakij D. Strategy to prevent falls in the Thai elderly: a controlled study integrated health research program for the Thai elderly. Journal of the Medical Association of Thailand 2002;85(2):215-22. Ballard JE, McFarland C, Wallace LS, Holiday DB, Roberson G. The effect of 15 weeks of exercise on balance, leg strength, and reduction in falls in 40 women aged 65 to 89 years. Journal of the American Medical Womens Association 2004;59(4):255-61. Barnett A, Smith B, Lord SR, Williams M, Baumand A. Community-based group exercise improves balance and reduces falls in at-risk older people: a randomised controlled trial. Age & Ageing 2003;32(4):407-14. Bischoff-Ferrari HA, Orav EJ, Dawson-Hughes B. Effect of cholecalciferol plus calcium on falling in ambulatory older men and women: a 3-year randomized controlled trial. Archives of Internal Medicine 2006;166(4):424-30. Brown AI. Functional adaptation to exercise in elderly subjects [thesis]. Curtin University of Technology, 2002. Buchner DM, Cress ME, de Lateur BJ, Esselman PC, Margherita AJ, Price R, et al. The effect of strength and endurance training on gait, balance, fall risk, and health services use in communityliving older adults. Journals of Gerontology Series A-Biological Sciences & Medical Sciences 1997;52A(4):M218-24. Bunout D, Barrera G, Avendano M, de la Maza P, Gattas V, Leiva L, et al. Results of a communitybased weight-bearing resistance training programme for healthy Chilean subjects [letter]. Age & Ageing 2005;34(1):80-3. Campbell AJ, Robertson MC, Gardner MM, Norton RN, Tilyard MW, Buchner DM. Randomised controlled trial of a general practice programme of home based exercise to prevent falls in elderly women. BMJ 1997;315(7115):1065-9. Campbell AJ, Robertson MC, Gardner MM, Norton RN, Buchner DM. Psychotropic medication withdrawal and a home-based exercise program to prevent falls: a randomized, controlled trial. Journal of the American Geriatrics Society 1999;47(7):850-3. Campbell AJ, Robertson MC, La Grow SJ, Kerse NM, Sanderson GF, Jacobs RJ, et al. Randomised controlled trial of prevention of falls in people aged 75 with severe visual impairment: the VIP trial. BMJ 2005;331(7520):817. Carpenter GI, Demopoulos GR. Screening the elderly in the community: controlled trial of dependency surveillance using a questionnaire administered by volunteers. BMJ 1990;300(6734):1253-6. Carter S, Campbell E, Sanson-Fisher R, Tiller K, Gillespie WJ. A randomised controlled trial of two strategies aimed at reducing falls and other unintentional events through home modification and medication review. Unpublished work 1997. Carter ND, Khan KM, McKay HA, Petit MA, Waterman C, Heinonen A, et al. Community-based exercise program reduces risk factors for falls in 65- to 75-year-old women with osteoporosis: randomized controlled trial. Canadian Medical Association Journal 2002;167(9):997-1004. Cerny K, Blanks R, Mohamed O, Schwab D, Robinson B, Russo A, et al. The effect of a multidimensional exercise program on strength, range of motion, balance and gait in the well elderly [abstract]. Gait & Posture 1998;7(2):185-6. Clemson L, Cumming RG, Kendig H, Swann M, Heard R, Taylor K. The effectiveness of a community-based program for reducing the incidence of falls in the elderly: a randomized trial. Journal of the American Geriatrics Society 2004;52(9):1487-94. Close J, Ellis M, Hooper R, Glucksman E, Jackson S, Swift C. Prevention of falls in the elderly trial (PROFET): a randomised controlled trial. Lancet 1999;353(9147):93-7.

9 10 11 12 13 14 15 16 17 18

21

19 20

21 22 23 24 25 26

27 28 29 30 31 32 33

34 35 36

Coleman EA, Grothaus LC, Sandhu N, Wagner EH. Chronic care clinics: a randomized controlled trial of a new model of primary care for frail older adults. Journal of the American Geriatrics Society 1999;47(7):775-83. Cornillon E, Blanchon MA, Ramboatsisetraina P, Braize C, Beauchet O, Dubost V, et al. Impact d'un programme de prevention multidisciplinaire de la chute chez le sujet age autonome vivant a domicile, avec analyse avant-apres des performances physiques (Effectiveness of a multidisciplinary falls prevention programme for elderly subjects who live in the community, with before-after assessment of physical activities). Annales de Readaptation et de Medecine Physique 2002;45(9):493-504. Cumming RG, Thomas M, Szonyi G, Salkeld G, O'Neill E, Westbury C, et al. Home visits by an occupational therapist for assessment and modification of environmental hazards: a randomized trial of falls prevention. Journal of the American Geriatrics Society 1999;47(12):1397-402. Cumming RG, Ivers R, Clemson L, Cullen J, Hayes MF, Tanzer M, et al. Improving vision to prevent falls in frail older people: a randomized trial. Journal of the American Geriatrics Society 2007;55(2):175-81. Davison J, Bond J, Dawson P, Steen IN, Kenny RA. Patients with recurrent falls attending Accident & Emergency benefit from multifactorial intervention a randomised controlled trial. Age & Ageing 2005;34(2):162-8. Day L, Fildes B, Gordon I, Fitzharris M, Flamer H, Lord S. Randomised factorial trial of falls prevention among older people living in their own homes. BMJ 2002;325(7356):128. Dhesi JK, Jackson SH, Bearne LM, Moniz C, Hurley MV, Swift CG, et al. Vitamin D supplementation improves neuromuscular function in older people who fall. Age & Ageing 2004;33(6):589-95. Dukas L, Bischoff HA, Lindpaintner LS, Schacht E, Birkner-Binder D, Damm TN, et al. Alfacalcidol reduces the number of fallers in a community-dwelling elderly population with a minimum calcium intake of more than 500 mg daily. Journal of the American Geriatrics Society 2004;52(2):230-6. Ebrahim S, Thompson PW, Baskaran V, Evans K. Randomized placebo-controlled trial of brisk walking in the prevention of postmenopausal osteoporosis. Age & Ageing 1997;26(4):253-60. Fabacher D, Josephson K, Pietruszka F, Linderborn K, Morley JE, Rubenstein LZ. An in-home preventive assessment program for independent older adults: a randomized controlled trial. Journal of the American Geriatrics Society 1994;42(6):630-8. Fiatarone MA, O'Neill EF, Doyle RN, Clements K. Efficacy of home-based resistance training in frail elders [abstract]. Abstracts of the 16th Congress of the International Association of Gerontology 1997;323:Abstract 985. Foss AJ, Harwood RH, Osborn F, Gregson RM, Zaman A, Masud T. Falls and health status in elderly women following second eye cataract surgery: a randomised controlled trial. Age & Ageing 2006;35(1):66-71. Gallagher EM, Brunt H. Head over heels: impact of a health promotion program to reduce falls in the elderly. Canadian Journal on Aging 1996;15(1):84-96. Gallagher JC, Rapuri PB, Smith LM. An age-related decrease in creatinine clearance is associated with an increase in number of falls in untreated women but not in women receiving calcitriol treatment. Journal of Clinical Endocrinology & Metabolism 2007;92(1):51-8. Grant AM, Avenell A, Campbell MK, McDonald AM, MacLennan GS, McPherson GC, et al. Oral vitamin D3 and calcium for secondary prevention of low-trauma fractures in elderly people (Randomised Evaluation of Calcium Or vitamin D, RECORD): a randomised placebo-controlled trial. Lancet 2005;365(9471):1621-8. Gray-Donald K, Payette H, Boutier V. Randomized clinical trial of nutritional supplementation shows little effect on functional status among free-living frail elderly. Journal of Nutrition 1995;125(12):2965-71. Greenspan SL, Resnick NM, Parker RA. The effect of hormone replacement on physical performance in community-dwelling elderly women. American Journal of Medicine 2005;118(11):1232-9. Harwood RH, Sahota O, Gaynor K, Masud T, Hosking DJ, The Nottingham Neck of Femur S. A randomised, controlled comparison of different calcium and vitamin D supplementation regimens in elderly women after hip fracture: The Nottingham Neck of Femur (NONOF) Study. Age & Ageing 2004;33(1):45-51.

22

37 38 39 40 41 42 43 44 45 46 47 48 49

50 51 52 53 54 55 56

Harwood RH, Foss AJ, Osborn F, Gregson RM, Zaman A, Masud T. Falls and health status in elderly women following first eye cataract surgery: a randomised controlled trial. British Journal of Ophthalmology 2005;89(1):53-9. Hauer K, Rost B, Rtschle K, Opitz H, Specht N, Brtsch P, et al. Exercise training for rehabilitation and secondary prevention of falls in geriatric patients with a history of injurious falls. Journal of the American Geriatrics Society 2001;49(1):10-20. Helbostad JL, Sletvold O, Moe-Nilssen R. Effects of home exercises and group training on functional abilities in home-dwelling older persons with mobility and balance problems. A randomized study. Aging-Clinical & Experimental Research 2004;16(2):113-21. Hill S, Mossman J, Stockdale P, Crome P. A randomised controlled trial of a nurse-led falls prevention clinic [abstract]. Age & Ageing 2000;29(Suppl 2):20. Hogan DB, MacDonald FA, Betts J, Bricker S, Ebly EM, Delarue B, et al. A randomized controlled trial of a community-based consultation service to prevent falls. CMAJ Canadian Medical Association Journal 2001;165(5):537-43. Hornbrook MC, Stevens VJ, Wingfield DJ, Hollis JF, Greenlick MR, Ory MG. Preventing falls among community-dwelling older persons: results from a randomized trial. Gerontologist 1994;34(1):16-23. Huang TT, Acton GJ. Effectiveness of home visit falls prevention strategy for Taiwanese community-dwelling elders: randomized trial. Public Health Nursing 2004;21(3):247-56. Huang TT, Liang SH. A randomized clinical trial of the effectiveness of a discharge planning intervention in hospitalized elders with hip fracture due to falling. Journal of Clinical Nursing 2005;14(10):1193-201. Jitapunkul S. A randomised controlled trial of regular surveillance in Thai elderly using a simple questionnaire administered by non-professional personnel. Journal of the Medical Association of Thailand 1998;81(5):352-6. Kenny RA, Richardson DA, Steen N, Bexton RS, Shaw FE, Bond J. Carotid sinus syndrome: a modifiable risk factor for nonaccidental falls in older adults (SAFE PACE). Journal of the American College of Cardiology 2001;38(5):1491-6. Kingston P, Jones M, Lally F, Crome P. Older people and falls: a randomized controlled trial of a health visitor (HV) intervention. Reviews in Clinical Gerontology 2001;11(3):209-14. Korpelainen R, Keinanen-Kiukaanniemi S, Heikkinen J, Vaananen K, Korpelainen J. Effect of impact exercise on bone mineral density in elderly women with low BMD: a population-based randomized controlled 30-month intervention. Osteoporosis International 2006;17(1):109-18. Latham NK, Anderson CS, Lee A, Bennett DA, Moseley A, Cameron ID, et al. A randomized, controlled trial of quadriceps resistance exercise and vitamin D in frail older people: the Frailty Interventions Trial in Elderly Subjects (FITNESS). Journal of the American Geriatrics Society 2003;51(3):291-9. Lehtola S, Hnninen L, Ptolo M. The incidence of fall during six-months exercise trial and fourmonths follow-up among home-dwelling persons aged 70-75 years. Liikunta & Tiede 2000;6:41-7. Li F, Harmer P, Fisher KJ, McAuley E, Chaumeton N, Eckstrom E, et al. Tai Chi and fall reductions in older adults: a randomized controlled trial. Journals of Gerontology Series A-Biological Sciences & Medical Sciences 2005;60(2):187-94. Lightbody E, Watkins C, Leathley M, Sharma A, Lye M. Evaluation of a nurse-led falls prevention programme versus usual care: a randomized controlled trial. Age & Ageing 2002;31(3):203-10. Lin MR, Wolf SL, Hwang HF, Gong SY, Chen CY. A randomized, controlled trial of fall prevention programs and quality of life in older fallers. Journal of the American Geriatrics Society 2007;55(4):499-506. Liu-Ambrose T, Khan KM, Eng JJ, Janssen PA, Lord SR, McKay HA. Resistance and agility training reduce fall risk in women aged 75 to 85 with low bone mass: a 6-month randomized, controlled trial. Journal of the American Geriatrics Society 2004;52(5):657-65. Lord SR, Ward JA, Williams P, Strudwick M. The effect of a 12-month exercise trial on balance, strength, and falls in older women: a randomized controlled trial. Journal of the American Geriatrics Society 1995;43(11):1198-206. Lord SR, Castell S, Corcoran J, Dayhew J, Matters B, Shan A, et al. The effect of group exercise on physical functioning and falls in frail older people living in retirement villages: a randomized, controlled trial. Journal of the American Geriatrics Society 2003;51(12):1685-92.

23

57 58 59

60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76

Lord SR, Tiedemann A, Chapman K, Munro B, Murray SM, Gerontology M, et al. The effect of an individualized fall prevention program on fall risk and falls in older people: a randomized, controlled trial. Journal of the American Geriatrics Society 2005;53(8):1296-304. Luukinen H, Lehtola S, Jokelainen J, Vaananen-Sainio R, Lotvonen S, Koistinen P. Pragmatic exercise-oriented prevention of falls among the elderly: a population-based, randomized, controlled trial. Preventive Medicine 2007;44(3):265-71. Madureira MM, Takayama L, Gallinaro AL, Caparbo VF, Costa RA, Pereira RM. Balance training program is highly effective in improving functional status and reducing the risk of falls in elderly women with osteoporosis: a randomized controlled trial. Osteoporosis International 2007;18(4):41925. Mahoney JE, Shea TA, Przybelski R, Jaros L, Gangnon R, Cech S, et al. Kenosha County falls prevention study: a randomized, controlled trial of an intermediate-intensity, community-based multifactorial falls intervention. Journal of the American Geriatrics Society 2007;55(4):489-98. Marigold DS, Eng JJ, Dawson AS, Inglis JT, Harris JE, Gylfadottir S. Exercise leads to faster postural reflexes, improved balance and mobility, and fewer falls in older persons with chronic stroke. Journal of the American Geriatrics Society 2005;53(3):416-23. McKiernan FE. A simple gait-stabilizing device reduces outdoor falls and nonserious injurious falls in fall-prone older people during the winter. Journal of the American Geriatrics Society 2005;53(6):943-7. McMurdo MET, Mole PA, Paterson CR. Controlled trial of weight bearing exercise in older women in relation to bone density and falls. BMJ 1997;314(2):569. Means KM, Rodell DE, O'Sullivan PS. Balance, mobility, and falls among community-dwelling elderly persons: effects of a rehabilitation exercise program. American Journal of Physical Medicine & Rehabilitation 2005;84(4):238-50. Morgan RO, Virnig BA, Duque M, Abdel-Moty E, Devito CA. Low-intensity exercise and reduction of the risk for falls among at-risk elders. Journals of Gerontology Series A-Biological Sciences & Medical Sciences 2004;59(10):1062-7. Newbury JW, Marley JE, Beilby JJ. A randomised controlled trial of the outcome of health assessment of people aged 75 years and over. Medical Journal of Australia 2001;175(2):104-7. Nikolaus T, Bach M. Preventing falls in community-dwelling frail older people using a home intervention team (HIT): results from the randomized Falls-HIT trial. Journal of the American Geriatrics Society 2003;51(3):300-5. Nitz JC, Choy NL. The efficacy of a specific balance-strategy training programme for preventing falls among older people: a pilot randomised controlled trial. Age & Ageing 2004;33(1):52-8. Pardessus V, Puisieux F, Di Pompeo C, Gaudefroy C, Thevenon A, Dewailly P. Benefits of home visits for falls and autonomy in the elderly: a randomized trial study. American Journal of Physical Medicine & Rehabilitation 2002;81(4):247-52. Pereira MA, Kriska AM, Day RD, Cauley JA, LaPorte RE, Kuller LH. A randomized walking trial in postmenopausal women: effects on physical activity and health 10 years later. Archives of Internal Medicine 1998;158(15):1695-701. Pfeifer M, Begerow B, Minne HW, Abrams C, Nachtigall D, Hansen C. Effects of a short-term vitamin D and calcium supplementation on body sway and secondary hyperparathyroidism in elderly women. Journal of Bone & Mineral Research 2000;15(6):1113-8. Porthouse J, Cockayne S, King C, Saxon L, Steele E, Aspray T, et al. Randomised controlled trial of calcium and supplementation with cholecalciferol (vitamin D3) for prevention of fractures in primary care. BMJ 2005;330(7498):1003. Reinsch S, MacRae P, Lachenbruch PA, Tobis JS. Attempts to prevent falls and injury: a prospective community study. Gerontologist 1992;32(4):450-6. Resnick B. Testing the effect of the WALC intervention on exercise adherence in older adults. Journal of Gerontological Nursing 2002;28(6):40-9. Robertson MC, Devlin N, Gardner MM, Campbell AJ. Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 1: Randomised controlled trial. BMJ 2001;322(7288):697-701. Robson E, Edwards J, Gallagher E, Baker D. Steady as you go (SAYGO): A falls-prevention program for seniors living in the community. Canadian Journal on Aging 2003;22(2):207-16.

24

77 78 79 80 81 82

83 84 85 86 87 88 89 90 91 92 93 94 95 96

Rubenstein LZ, Josephson KR, Trueblood PR, Loy S, Harker JO, Pietruszka FM, et al. Effects of a group exercise program on strength, mobility, and falls among fall-prone elderly men. Journals of Gerontology Series A-Biological Sciences & Medical Sciences 2000;55(6):M317-21. Ryan JW, Spellbring AM. Implementing strategies to decrease risk of falls in older women. Journal of Gerontological Nursing 1996;22(12):25-31. Sato Y, Manabe S, Kuno H, Oizumi K. Amelioration of osteopenia and hypovitaminosis D by 1ahydroxyvitamin D3 in elderly patients with Parkinson's disease. Journal of Neurology, Neurosurgery & Psychiatry 1999;66(1):64-8. Schrijnemaekers VJ, Haveman MJ. Effects of preventive outpatient geriatric assessment: short-term results of a randomized controlled study. Home Health Care Services Quarterly 1995;15(2):81-97. Sherrington C, Lord SR, Herbert RD. A randomized controlled trial of weight-bearing versus nonweight-bearing exercise for improving physical ability after usual care for hip fracture. Archives of Physical Medicine & Rehabilitation 2004;85(5):710-6. Singh NA, Stavrinos TM, Scarbek Y, Galambos G, Liber C, Fiatarone Singh MA. A randomized controlled trial of high versus low intensity weight training versus general practitioner care for clinical depression in older adults. Journals of Gerontology Series A-Biological Sciences & Medical Sciences 2005;60(6):768-76. Skelton D, Dinan S, Campbell M, Rutherford O. Tailored group exercise (Falls Management Exercise FaME) reduces falls in communty-dwelling older frequent fallers (an RCT) [letter]. Age & Ageing 2005;34(6):636-9. Steadman J, Donaldson N, Kalra L. A randomized controlled trial of an enhanced balance training program to improve mobility and reduce falls in elderly patients. Journal of the American Geriatrics Society 2003;51:847-52. Steinberg M, Cartwright C, Peel N, Williams G. A sustainable programme to prevent falls and near falls in community dwelling older people: results of a randomised trial. Journal of Epidemiology & Community Health 2000;54(3):227-32. Stevens M, Holman CD, Bennett N, de Klerk N. Preventing falls in older people: outcome evaluation of a randomized controlled trial. Journal of the American Geriatrics Society 2001;49(11):1448-55. Suzuki T, Kim H, Yoshida H, Ishizaki T. Randomized controlled trial of exercise intervention for the prevention of falls in community-dwelling elderly Japanese women. Journal of Bone & Mineral Metabolism 2004;22(6):602-11. Swanenburg J, Douwe de Bruin E, Stauffacher M, Mulder T, Uebelhart D. Effects of exercise and nutrition on postural balance and risk of falling in elderly people with decreased bone mineral density: randomized controlled trial pilot study. Clinical Rehabilitation 2007;21:523-34. Tinetti ME, Baker DI, McAvay G, Claus EB, Garrett P, Gottschalk M, et al. A multifactorial intervention to reduce the risk of falling among elderly people living in the community. New England Journal of Medicine 1994;331(13):821-7. Trivedi DP, Doll R, Khaw KT. Effect of four monthly oral vitamin D3 (cholecalciferol) supplementation on fractures and mortality in men and women living in the community: randomised double blind controlled trial. BMJ 2003;326(7387):469. van Haastregt JC, Diederiks JP, van Rossum E, de Witte LP, Voorhoeve PM, Crebolder HF. Effects of a programme of multifactorial home visits on falls and mobility impairments in elderly people at risk: randomised controlled trial. BMJ 2000;321(7267):994-8. van Rossum E, Frederiks CMA, Philipsen H, Portengen K, Wiskerke J, Knipschild P. Effects of preventive home visits to elderly people. BMJ 1993;307(6895):27-32. Vellas B, Albarde JL. A randomized clinical trial on the value of raubasine-dihydroergocristine (Iskdyl) in the prevention of post fall syndrome. Psycholgie Medicale 1991;23(7):831-9. Vetter NJ, Lewis PA, Ford D. Can health visitors prevent fractures in elderly people? BMJ 1992;304(4):888-90. von Koch L, de Pedro-Cuesta J, Kostulas V, Almazan J, Widen Holmqvist L. Randomized controlled trial of rehabilitation at home after stroke: one-year follow-up of patient outcome, resource use and cost. Cerebrovascular Diseases 2001;12(2):131-8. Voukelatos A, Cumming RG, Lord SR, Rissel C. A randomized, controlled trial of tai chi for the prevention of falls: The Central Sydney tai chi Trial. Journal of the American Geriatrics Society 2007;55(8):1185-91.

25

97 98 99 100 101 102 103

Wagner EH, LaCroix AZ, Grothaus L, Leveille SG, Hecht JA, Artz K, et al. Preventing disability and falls in older adults: a population-based randomized trial. American Journal of Public Health 1994;84(11):1800-6. Weerdesteyn V, Rijken H, Geurts AC, Smits-Engelsman BC, Mulder T, Duysens J. A five-week exercise program can reduce falls and improve obstacle avoidance in the elderly. Gerontology 2006. Whitehead C, Wundke R, Crotty M, Finucane P. Evidence-based clinical practice in falls prevention: a randomised controlled trial of a falls prevention service. Australian Health Review 2003;26(3):8897. Wilder P. Seniors to seniors exercise program: a cost effective way to prevent falls in the frail elderly living at home. Journal of Geriatric Physical Therapy 2001;24(3):13. Wolf SL, Barnhart HX, Kutner NG, McNeely E, Coogler C, Xu T, et al. Reducing frailty and falls in older persons: an investigation of Tai Chi and computerized balance training. Journal of the American Geriatrics Society 1996;44(5):489-97. Wolf SL, Sattin RW, Kutner M, O'Grady M, Greenspan AI, Gregor RJ. Intense tai chi exercise training and fall occurrences in older, transitionally frail adults: a randomized, controlled trial. Journal of the American Geriatrics Society 2003;51(12):1693-701. Woo J, Hong A, Lau E, Lynn H. A randomised controlled trial of Tai Chi and resistance exercise on bone health, muscle strength and balance in community-living elderly people. Age & Ageing 2007;36(3):262-8.

26

FALL PREVENTION INTERVENTIONS TESTED IN RANDOMISED CONTROLLED TRIALS IN THE COMMUNITY 31 August 2007
Study Alexander 2003 Armstrong 1996 Assantachai 2002 Ballard 2005 Barnett 2003 Bischoff-Ferrari 2006 Brown a) 2002 Brown b) 2002 Buchner 1997 Bunout 2005 Campbell 1997 Campbell 1999 a) Campbell 1999 b) Campbell 2005 a) Campbell 2005 b) Carpenter 1990 Carter 2002 Carter (unpublished) a) Carter (unpublished) b) Cerny 1998 Clemson 2004 Close 1999 Coleman 1999 Cornillon 2002 Trial no. 1 2 3 4 5 6 7 7 8 9 10 11 11 12 12 13 14 15 15 16 17 18 19 20 Year 2003 1996 2002 2005 2003 2006 2002 2002 1997 2005 1997 1999 1999 2005 2005 1990 2002 1998 2004 1999 1999 2002 Intervention Group balance training & educational seminars HRT + calcium vs calcium Leaflet+free access to geriatric clinic Group exercise Group exercise Vitamin D + calcium vs placebo Group exercise Social intervention Group exercise Group exercise Otago Exercise Programme (OEP) OEP (taking psychotropic drugs) Withdraw psychotropic drugs OEP + vitamin D (visual impairment) Home safety (visual impairment) Home visits by volunteers Group exercise Experimental group 1 vs control Experimental group 2 vs control Group exercise Small group learning Comprehensive geriatric assessment and home safety Clinic focusing on chronic disease & medications Group exercise Intervention code MPLE ADRUG MPLE EX EX VITD EX SOC EX EX EX EX WDRUG MPLE HS MFACT EX MPLE MPLE EX MPLE MFACT MFACT EX 28 310 397 169 303 539 93 657 391 105 298 233 93 No. randomised 44 116 1043 40 163 445 149

27

Cumming 1999 Cumming 2007 Davison 2005 Day 2002 a) Day 2002 b) Day 2002 c) Dhesi 2004 Dukas 2004 Ebrahim 1997 Fabacher 1994 Fiatarone 1997 Foss 2006 Gallagher 1996 Gallagher 2007 a) Gallagher 2007 b) Gallagher 2007 c) Grant 2005 a) Grant 2005 b) Gray-Donald 2005 Greenspan 2005

21 22 23 24 24 24 25 26 27 28 29 30 31 32 32 32 33 33 34 35 36 37 38 39 40

1999 2007 2005 2002 2002 2002 2004 2004 1997 1994 1997 2006 1996 2001 2001 2001 2005 2005 2005 2005 2004 2005 2001 2004 2000

Home safety Vision improvement Multifactorial post-fall assessment & intervention Group exercise Home hazard management Vision improvement Vitamin D Alfacalcidol Walking group (year 1) Assessment at home + advice Home resistance training Cataract removal Risk assessment HRT vs placebo Calcitriol vs placebo HRT +calcitriol vs placebo Vitamin D3 or vitamin D3 + calcium vs calcium or placebo Calcium or vitamin D3 + calcium vs vitamin D3 or placebo Nutrient supplements HRT vs placebo Vitamin D injection, Calcium + vitamin D injection, Calcium + vitamin D tablets vs no treatment Cataract removal Group exercise NOTE: Home exercise + group vs home exercise Group exercise + advice

HS VI MFACT EX HS VI VITD VITD EX MFACT EX VI MFACT ADRUG VITD ADRUG VITD VITD NUTR ADRUG VITD VI EX EX MPLE

530 616 313 1090

139 378 165 254 34 239 100 489

5292

50 373 150 306 57 77

Harwood 2004 Harwood 2005 Hauer 2001 Helbostad 2004 Hill 2000

28

Hogan 2001 Hornbrook 1994 Huang 2004 Huang 2005 Jitapunkul 1998 Kenny 2001 Kingston 2001 Korpelainen 2006 Latham 2003 a) Latham 2003 b) Lehtola 2000 Li 2005 Lightbody 2002 Lin 2007 a) Lin 2007 b) Lin 2007 c) Liu-Ambrose 2004 a) Liu-Ambrose 2004 b) Lord 1995 Lord 2003 Lord 2005 a) Lord 2005 b) Luukinen 2007

41 42 43 44 45 46 47 48 49 49 50 51 52 53 53 53 54 54 55 56 57 57 58

2001 1994 2004 2005 1998 2001 2001 2006 2003 2003 2000 2005 2002 2007 2007 2007 2004 2004 1995 2003 2005 2005 2007

Geriatric assessment and referral Home visit & advice Individualised advice Hospital discharge planning vs usual care Home visits (non-professionals) & referrals Pacemaker Health visitor Impact exercise Resistance training (individual at home) Vitamin D vs placebo Group exercise group + home exercise Tai chi group Falls nurse assessment at home 3 intervention groups: education, home safety, home exercise 3 intervention groups: education, home safety, home exercise 3 intervention groups: education, home safety, home exercise Group resistance training Group agility training Group exercise Group exercise "Extensive intervention group" vs control "Minimum intervention group" vs control Received OT & "suggestions for a programme of exercise" from PT

MFACT MFACT MFACT MFACT MFACT PACE MFACT EX EX VITD EX EX MFACT ED HS EX EX EX EX EX MFACT MFACT MFACT

163 3182 120 126 160 175 109 160 243

131 256 348 150

98 197 551 620

486

29

Madureira 2007 Mahoney 2007

59 60 63 61 62 64 65 66 67 68 69 70 71 72 73 73 74 75 76 77 78 79 80 81 81 82

2007 2007 2005 2005 1997 2005 2004 2001 2003 2004 2002 1998 2000 1992 1992 2002 2001 2003 2000 1996 1999 1995 2004 2004 2005

Group balance training Assessment at home + referrals NOTE: Group agility vs stretching/weight shifting (like Tai chi) - chronic stroke Yaktrax walker Group exercise + calcium vs calcium Group exercise Group exercise Assessment at home by nurse, referral Assessment + home intervention NOTE: Balance training vs group exercise Home visit by physician, OT, ergotherapist Walking group (trial completed 10 years previously) Calcium + vitamin D vs calcium Calcium + vitamin D vs control Group exercise Cognitive-behavioural Walking group Otago Exercise Programme Group sessions, exercise video Group exercise Educational intervention group, oneon-one, control Vitamin D3 vs placebo Assessment and referral Weight bearing (after hip #) Non weight bearing (after hip #) NOTE: High vs low intensity weight training vs GP care

EX MFACT EX FOOTW EX EX EX MFACT MFACT EX MFACT EX VITD VITD EX COG EX EX MPLE EX MFACT VITD MFACT EX EX EX

66 349 61 109 118 338 294 100 360 73 60 229 148 3314 230 20 240 660 59 45 86 223 120

Marigold 2005 McKiernan 2005 McMurdo 1997 Means 2005 Morgan 2004 Newbury 2001 Nikolaus 2003 Nitz 2004 Pardessus 2002 Pereira 1998 Pfeifer 2000 Porthouse 2005 Reinsch 1992 a) Reinsch 1992 b) Resnick 2002 Robertson 2001 Robson 2003 Rubenstein 2000 Ryan 1996 Sato 1999 Schrijnemaekers 1995 Sherrington 2004 a) Sherrington 2004 b) Singh 2005

60

30

Skelton 2005 Steadman 2003 Steinberg 2000 Stevens 2001 Suzuki 2004 Swanenburg 2007 Tinetti 1994 Trivedi 2003 van Haastregt 2000 van Rossum 1993 Vellas 1991 Vetter 1992 von Koch 2001 Voukelatos 2006, 2007 Wagner 1994 Weerdestyn 2006 Whitehead 2003 Wilder 2001 Wolf 1996 a) Wolf 1996 b) Wolf 2003 Woo 2007 a) Woo 2007 b)

83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 101 102 103 103

2005 2003 2000 2001 2004 2007 1994 2003 2000 1993 1991 1992 2001 2006 1994 2006 2003 2001 1996 1996 2003 2007 2007

Group exercise Enhanced balance exercise vs usual physiotherapy ED; ED+EX; ED+EX+HS; ED+EX+HS+clinical assessment Home safety Group exercise Exercise + nitrition OT and physiotherapist at home Vitamin D3 vs placebo Home visits Multifactorial home visits Iskedyl (medication) Multifactorial home visits Rehabilitation at home after stroke Tai chi group 3 groups: multifactorial, nurse home visit, control Group exercise Falls prevention service 3 groups: control, simple home modifications, Homes safety +ex Tai chi group Balance training Tai chi group Tai chi group Group resistance training

EX EX MPLE HS EX MPLE MFACT VITD MFACT MFACT ADRUG MFACT MFACT EX MFACT EX MFACT HS, MPLE EX EX EX EX EX

100 199 252 1737 52 24 301 2686 316 580 96 674 83 702 1559 113 140 60 200 311 180

31

No. of intervention group comparisons: Single interventions Add drug (other than vitamin D) Cognitive/behavioural Education Exercise Footwear Home safety Nutrient supplements Pacemaker Social Vision improvement Vitamin D Withdraw drug(s) Total Multiple interventions Multifactorial interventions Total no. of group comparisons Total no. of participants: ADRUG COG ED EX FOOTW HS NUTR PACE SOC VI VITD WDRUG 5 1 1 48 1 5 1 1 1 4 12 1 81 11 29 121 No. of trials to date: 41,348 103

MPLE MFACT

Other abbreviations: HRT Hormone replacement treatment OEP Otago Exercise Programme OT Occupational therapist PT Physiotherapist

32

D Interventions in long term care and hospital settings: systematic review

Falls in long term care facilities
Falls are even more common in residential care facilities than for older people living in the community (Rubenstein 1996). A study in 14 New Zealand residential care facilities recorded falls over 18 months and reported that there were 954 falls by 271 of the 680 residents (almost half of the residents), 68% resulting in injury (Butler 2004). The number of falls increased in the afternoon. Environmental hazards such as cords, steps, mats and dim light were involved with 15% of the falls. The number of injuries was higher during the day when the residents were active, and the more active residents fell more but were less likely to injure themselves. Prevention is a challenge as many residents are physically frail and have cognitive impairment, and facility staff are not routinely trained in prevention strategies. ACC has recognised the costs and resources implicated and the lack of successful programmes in a New Zealand residential care setting, and is currently funding a pilot study of an exercise and a multifactorial programme led by Associate Professor Ngaire Kerse, School of Population Health, University of Auckland.

Falls in hospital
Falls account for over 80% of all injury related admissions to hospital for people over 65 years (Lord 2001). Paradoxically, a hospital setting is not a safe place for elderly people but is actually associated with an increased risk of falling (Kannus 2006). On admission, the older patient accumulates additional falls risk factors including a new, strange environment with poorly recognised external dangers for falling. This is often combined with confusion, acute illness, and balanceaffecting medication, in addition to chronic risk factors such as comorbidities, muscle weakness and impaired balance and gait. Most hospitals have a falls prevention policy in place in an attempt to minimise these risks.

33

Publications identified
A total of 37 randomised controlled trials testing a falls prevention programme in a hospital or long term care setting in 20,153 residents have been identified and reviewed. Eight of the trials were carried out in acute- or sub-acute wards in hospitals, or a combination of these wards (2,862 inpatients; 1,146 men; 1,716 women), and the remaining 29 trials were in long term care facilities (17,291 residents; 3,765 men; 13,481 women; 45 gender not specified). In this project we divide interventions into single factor, multiple and multifactorial programmes. These terms are defined in the taxonomy for falls prevention interventions currently nearing publication by ProFaNE (Prevention of Falls Network Europe) funded by the European Commission. Single interventions address one type of risk factor for falls whereas multiple and multifactorial programmes have multiple components and address more than one type of risk factor. Examples of single interventions are exercise programmes and vitamin D supplementation. Multifactorial interventions include a risk assessment with programme components to address the identified risks for each individual. In a multiple programme, the same components are delivered to each individual. In the 37 trials identified and meeting the inclusion criteria, the interventions tested were: 26 single factor interventions (6 were tested in hospital wards) 11 multifactorial programmes (3 were tested in hospital wards) 4 multiple intervention programmes (none was tested in a hospital) A total of 41 interventions were tested. The most common single factor intervention was an exercise programme (11 were tested) and six trials used vitamin D supplementation. These are not the final numbers as further trials are regularly appearing in the literature. No information on falls specific to Maori and Pacific populations has been identified so far in the literature. The cut-off date for literature to include in these analyses was 15 December 2007. A new systematic review on interventions for preventing falls in hospitals has since been published but it did not include any randomised controlled trials we had not already identified (Coussement 2008).

Interventions to reduce falls in long term care

Data on the rate of falls during the trial were available from 25 interventions tested in 23 of the 29 randomised controlled trials in residential care. Although we also extracted the risk ratio for one or more falls (number of fallers) during the trial for each intervention, the rate of falls is a much superior outcome to pool. This is especially the case in the trials in institutions as it is important to take into account the number of bed days and not just the number of people who fall in each group during the monitoring period. We pooled data on the rate ratio for falls separately from the following types of interventions multifactorial, exercise programmes, and vitamin D supplementation. There were also two interventions using medication review and modification with a risk ratio (but not a rate ratio) available for pooling. Interventions using a multifactorial approach We found a total of eight interventions with multifactorial components tested in long term care residents. Pooling of seven of these programmes with data on the rate of falls available, showed that overall the rate of falls was reduced by 21% (see Figure 1 below). This reduction did not quite reach statistical significance (P = 0.058). The meta-analysis demonstrated significant heterogeneity (P < 0.001). One potential source of heterogeneity is the multifactorial programme tested in New Zealand rest homes which showed that the rate of falls was significantly increased by 34% (Kerse 2004). This was the only trial in our systematic search that significantly increased the rate of falls in institutions. The programme tested by Shaw was targeted at those with marked cognitive impairment only (Shaw 2003) and therefore differed from the remainder of the trials pooled. A meta-analysis with these two trial excluded, although now demonstrating a significant reduction in falls, still showed substantial heterogeneity (pooled rate ratio 0.66 (95%CI 0.52 to 0.85), P = 0.001. Test for heterogeneity: Q = 18.728, P = 0.001).

34

Becker 2003

Dyer 2004

Jensen 2002

Kerse 2004

McMurdo 2000

Rubenstein 1990

Shaw 2003

Combined

.1

.5 Relative rate ratio

1.5

Figure 1. Meta-analysis of programmes with multifactor components. Participants received interventions based on a falls risk assessment. Pooled rate ratio 0.79 (95%CI 0.61 to 1.01), P = 0.058. Test for heterogeneity: Q = 57.455, P < 0.001. | Weights Study 95% CI Study | (Random) Est Lower Upper ---------------------------------------------------------------------------------Becker 2003 | 8.52 0.55 0.41 0.73 Dyer 2004 | 8.95 0.54 0.42 0.69 Jensen 2002 | 9.52 0.60 0.50 0.72 Kerse 2004 | 9.01 1.34 1.05 1.71 McMurdo 2000 | 5.99 0.78 0.46 1.32 Rubenstein 1990 | 9.47 0.95 0.78 1.15 Shaw 2003 | 10.14 0.99 0.89 1.10

All current guidelines recommend addressing multifactorial risk factors as an effective strategy for reducing falls in all elderly people. We will investigate further the heterogeneity of the pooled results to see whether a particular combination of components, the particular subgroup targeted, the professional background of the staff delivering the intervention, or methodological characteristics of the trial appear to contribute to the success or failure of these interventions. The interventions used in Becker (2003), Dyer (2004), and Jensen (2002) were effective in reducing falls. Components included educating staff on fall prevention, implementing exercise programmes, modifying the environment, supplying and repairing aids, reviewing drug regimens, providing free hip protectors, and having post-fall problem solving conferences. However, it may not be possible to deliver these interventions in a similar way in New Zealand given the different staffing levels available.

35

Exercise interventions Exercise programmes have been shown to be successful in reducing falls in community living older people but this success is not mirrored in trials in long term care. Maintaining muscle strength and balance is very important but more challenges are involved in exercise instruction to long term care residents, for example due to the disparate levels of frailty and cognitive function. Pooling of the results from testing seven exercise programmes with data on the rate of falls available, showed that overall the rate of falls was unchanged (Figure 2). Again this meta-analysis showed substantial heterogeneity (P < 0.001).

Faber 2006 a)

Faber 2006 b)

Mulrow 1994

Sakamoto 2006

Schoenfelder 2000

Shimada 2004

Sihvonen 2004

Combined

.1

.5 1 Relative rate ratio

1.5

Figure 2. Meta-analysis for trials testing exercise programmes. Pooled rate ratio 1.05 (95%CI 0.80 to 1.39), P = 0.726. Test for heterogeneity: Q = 27.366, P < 0.001. | Weights Study 95% CI Study | (Random) Est Lower Upper ----------------------------------------------------------------------------Faber 2006 a) | 9.92 1.32 1.09 1.60 Faber 2006 b) | 9.77 0.96 0.78 1.18 Mulrow 1994 | 8.29 1.32 0.94 1.85 Sakamoto 2006 | 9.38 0.82 0.64 1.05 Schoenfelder 2000 | 5.03 2.72 1.43 5.17 Shimada 2004 | 3.20 0.53 0.21 1.33 Sihvonen 2004 | 3.64 0.40 0.17 0.93

The types of exercise programme tested were diverse and included walking programmes, treadmill training, balance programmes, ankle strengthening, and extra sit-to-stands. Few programmes included a combination of flexibility, muscle strengthening, and balance training, the type of exercise programme most likely to be successful in reducing falls in the community setting. No effect on the rate of falls was seen when we pooled the results from the three trials with walking programmes (Faber 2006 functional walking, Schoenfelder 2000, Shimada 2004) Shimada (2004) tested a treadmill versus a usual exercise programme): pooled rate ratio 1.33 (95%CI 0.68 to 2.61), P = 0.405. There was significant heterogeneity in this meta-analysis (Q = 8.561, P = 0.014). Similarly there was no effect on the rate of falls shown with pooling the results from three trials testing balance exercises alone (Faber 2006 balance programme, Sakamoto 2006 standing on

36

one leg, Sihvonen 2004 visual feedback training) (pooled rate ratio 0.83 (95%CI 0.63 to 1.09), P = 0.174. Test for heterogeneity: Q = 4.390, P = 0.111). We know that all elderly people, even into the nineties, can improve their strength and balance with appropriate exercises, therefore there is potential for preventing falls by delivering exercise programmes in residential care facilities. Only one of the exercise programmes in the systematic review reported a significant reduction in falls, but this was a small trial (27 women in 2 residential care homes) testing individualised visual feedback training on a computerized balance platform and the results may not be generalisable (Sihvonen 2004). Although the three successful multifactorial trials all had an exercise programme component (Becker 2003, Dyer 2004, Jensen 2002), it is not known how much the exercise programme contributed to the overall effectiveness in reducing falls. Vitamin D supplementation Vitamin D has been suggested as a falls prevention strategy because low levels have been associated with increased body sway and loss of muscle strength. Levels of vitamin D are known to be low in many older people especially those living in long term care facilities and there is some previous evidence that increasing the level will reduce falls (Bischoff-Ferrari 2004). An incidence rate ratio comparing a group taking vitamin D supplementation (with or without calcium) compared with a placebo or calcium group was available for all five trials identified meeting the inclusion criteria. Overall vitamin D supplementation significantly reduced the rate of falls by 35% (Figure 3).

Bischoff 2003

Broe 2007

Flicker 2005

Law 2006

Sato 2005

Combined

.1

.5 Relative rate ratio

1.5

Figure 3. Meta-analysis for trials testing vitamin D supplements. Pooled rate ratio 0.65 (95%CI 0.48 to 0.87), P = 0.004. Test for heterogeneity: Q = 13.089, P = 0.011. | Weights Study 95% CI Study | (Random) Est Lower Upper ---------------------------------------------------------------------------------Bischoff 2003 | 4.37 0.51 0.23 1.15 Broe 2007 | 3.11 0.28 0.11 0.73 Flicker 2005 | 13.48 0.73 0.57 0.94 Law 2006 | 17.05 0.87 0.81 0.94 Sato 2005 | 7.66 0.47 0.28 0.80

37

We have emailed Dr Yoshihiro Sato about some inconsistencies in the results he reports in the article included in this review and he has admitted to there being a mistake in the results reported. We are also aware of there being question marks over the results from several other trials where Dr Sato is the first author. For this reason we are hesitant in including his trial in our pooling and provide a metaanalysis below which excludes his trial results (Figure 4). This shows that falls were significantly reduced by 28% (rather than the 34% with Sato 2005 trial included).

Bischoff 2003

Broe 2007

Flicker 2005

Law 2006

Combined

.1

.5 Relative rate ratio

1.5

Figure 4. Meta-analysis for trials testing vitamin D supplements (Sato 2005 excluded, see text above for justification). Pooled rate ratio 0.72 (95%CI 0.55 to 0.95), P = 0.021. Test for heterogeneity: Q = 7.927, P = 0.048. | Weights Study 95% CI Study | (Random) Est Lower Upper --------------------------------------------------------------------------------Bischoff 2003 | 4.75 0.51 0.23 1.15 Broe 2007 | 3.30 0.28 0.10 0.77 Flicker 2005 | 17.85 0.73 0.57 0.94 Law 2006 | 24.71 0.87 0.81 0.94

Vitamin D supplementation forms part of accepted management of people with osteoporosis, and together with calcium nutrition is important in the maintenance of musculoskeletal health. High dose preparations of vitamin D that enable less frequent dosing are likely to be effective, cheaper and have higher adherence. Medication review and modification It is accepted that certain medications, particularly psychotropic drugs, contribute to the risk of falling and that these medications are commonly prescribed to long term care residents. In two studies a pharmacist reviewed medication related to falling in long term care residents (Crotty 1 2004, Zermansky 2006). In one study, the intervention resulted in a significant reduction in the mean number of drug changes per patient for the 661 residents, and a 38% reduction in falls (incidence rate ratio 0.62, 95% confidence interval 0.53 to 0.72) (Zermansky 2006). No rate ratio was available for the other study (Crotty 2004 1).

38

We pooled the risk ratios (based on number of fallers) from these two studies (Figure 5). The number of residents falling was reduced by 15% but this was not a significant reduction (P = 0.153). One of the programmes used a pharmacist coordinator when patients were transferred from hospital to long term care (Crotty 2004 1). We classified another trial by Crotty and colleagues (Crotty 2004 2) as an education intervention. Here the aim was to assess whether outreach visits by a pharmacist to doctors and residential care staff would improve falls reduction and stroke prevention. There was no reduction in the risk of falling.

Crotty 2004 1

Zermansky 2006

Combined

.1

.5 Relative risk ratio

1.5

Figure 5. Meta-analysis for trials testing medication review and modification. Pooled risk ratio 0.85 (95%CI 0.68 to 1.06), P = 0.153. Test for heterogeneity: Q = 1.983, P = 0.159. | Weights Study 95% CI Study | (Fixed) Est Lower Upper -------------------------------------------------------------------------------Crotty 2004 1 | 14.47 1.19 0.71 1.99 Zermansky 2006 | 64.53 0.79 0.62 1.01

Other intervention strategies in residential care For the remaining seven studies, there were no available data for pooling, or pooling was not appropriate as the interventions were not similar. These included three trials testing education programmes, two for restraint use, and one was aimed at fracture prevention. An additional education programme for medications (an outreach programme) is mentioned above. There were four studies with multiple interventions and each had an exercise component. One trial included the use of a sensory mat, two added in a falls prevention education programme, and one addressed incontinence in addition to the exercise programme. None of these seven studies demonstrated a significant reduction in the number of falls or the number of fallers.

Interventions to reduce falls in hospitals

There were eight trials testing nine interventions carried out in a hospital setting that met our systematic review criteria. Three tested a multifactorial programme, two an exercise programme, one compared falls for patients in rooms with carpet versus vinyl flooring, one tested a bed alarm, one an

39

identification bracelet, and one vitamin D supplementation. Only one of these eight studies produced a significant reduction in the number of falls (Haines 2004), and one in the number of fallers (Stenvall 2007). Both these two trials tested a multifactorial approach. Carpeted rooms appeared to increase the rate of falls compared with vinyl flooring (incidence rate ratio 14.77, 95% confidence interval 1.89 to 115.36). For the trials in a hospital setting, data were not available or interventions were too dissimilar for pooling, except for two trials testing a multifactorial intervention. Multifactorial interventions We pooled the two studies testing multifactorial programmes that were targeted at all patients on admission (Haines 2004) or those with a previous fall (Healey 2004), The rate of falls was significantly reduced by 31 % (see Figure 6 below). Of note in Haines study is the fact that the difference in fall rate between the intervention and control groups in the trial in the three subacute wards was the most obvious after 45 days of observation (Haines 2004). The successful combination of components in this trial was the use of fall risk alert cards with an information brochure, an exercise programme (description available on the web), and hip protectors. The third trial tested a successful programme comprising an assessment and treatment of fall risk factors aimed at reducing postoperative complications in 199 patients after hip surgery (Stenvall 2007). In this study the rate of falling in hospital by all patients in the study was significantly reduced by 62%, and was also reduced in the subgroup with dementia (incidence rate ratio for all participants 0.38, 95% confidence interval 0.20 to 0.76).

Haines 2004

Healey 2004

Combined

.1

.5 Relative rate ratio

1.5

Figure 6. Meta-analysis of trials testing multifactorial interventions in a hospital setting. Pooled relative risk from fixed effects model 0.69 (95%CI 0.54 to 0.88), P = 0.002. Test for heterogeneity: Q = 0.152, P = 0.697. | Weights Study 95% CI Study | (Fixed) Est Lower Upper ----------------------------------------------------------------------------------Haines 2004 | 61.55 0.70 0.55 0.90 Healey 2004 | 5.66 0.59 0.26 1.34

40

Implications of findings from the institutional literature review

Although falls are three times more common in long term care facilities than in community living older people, fewer randomised controlled trials testing effectiveness were identified in this setting (29 versus 103 in the community). Vitamin D supplementation was the only intervention, when trials testing this approach were pooled, to show a significant benefit in terms of falls reduced in long term care. In three individual trials, falls were reduced by using a multifactorial approach in a long term care setting. Delivery of these programmes may differ in a New Zealand healthcare setting. Overall the current trials demonstrate no benefit on falls in using an exercise progamme alone. However, the types of exercise programmes tested have varied and do not appear to emulate successful community based exercise programmes. In a hospital setting there is an increased risk of falling for older people. Two multifactorial approaches have been successful in reducing falls in a mixture of acute and sub-acute wards. Those running falls prevention programmes in long term care and hospital settings should use the successful programmes and guidelines that are available but be aware that none of the evidence for successfully reducing falls has been carried out in New Zealand. No information on falls prevention specific to Maori and Pacific populations has been identified so far in the literature. Preventing falls prevents injuries resulting from falls, but the lack of information on injury outcomes in most of the trials makes estimation of the potential benefit in terms of injury prevention difficult. A recent article in the BMJ suggests shifting the focus in fracture prevention from treatment and prevention of osteoporosis, to falls prevention. ACC is currently funding a pilot research trial on multifactorial and exercise programmes in New Zealand long term care facilities.

41

References
Bischoff-Ferrari HA, Dawson-Hughes B, Willett WC, Staehelin HB, Bazemore MG, Zee RY, et al. Effect of Vitamin D on falls: a meta-analysis. JAMA 2004;291(16):1999-2006. Butler M, Kerse N, Todd M. Circumstances and consequences of falls in residential care: the New Zealand story. New Zealand Medical Journal 2004;117(1202):U1076. URL: http://www.nzma.org.nz/journal/117-1202/1076/ Cali CM, Kiel DP. An epidemiological study of fall-related fractures among institutionalized older people. Journal of the American Geriatrics Society 1995;43:1336-1340. Chan DK, Hiller G, Cooke R, Monk R, Mills J, Hung WT. Effectiveness and acceptability of a newly designed hip protector. Archives of Gerontology and Geriatrics 2000;30:25-34. Coussement J, De Paepe L, Schwendimann R, Denhaerynck K, Dejaeger E, Milisen K. Interventions for preventing falls in acute- and chronic-care hospitals: a systematic review and metaanalysis. Journal of the American Geriatrics Society 2008:46:29-36. Cumming RG, Sherrington C, Lord SR, Simpson JM, Vogler C, Cameron ID, Naganathan V. Cluster randomised trial of a targeted multifactorial intervention to prevent falls among older people in hospital. BMJ 2008;336:758-60. Dargent-Molina P, Favie F, Grandjean H, et al. Fall-related factors and risk of hip fracture: the EPIDOS prospective study. Lancet 1996;384:145-49. Drahota A, Gal D, Windsor J. Flooring as an intervention to reduce injuries from falls in healthcare settings: an overview. Quality in Ageing 2007;8(1):3-9. Jrvinen TLN, Sievnen H, Khan KM, Heinonen A, Kannus P. Shifting the focus in fracture prevention from osteoporosis to falls. BMJ 2008;336:124-126. Kannus P, Khan KM, Lord SR. Preventing falls among elderly people in the hospital environment. MJA 2006;184:372-373. Lord S, Sherrington C, Menz H, Close J. Falls in Older People: Risk Factors and Strategies for Prevention. 2nd edition. Cambridge University Press, Australia, 2007. Luukinen H, Koski K, Honkanen R, Kivela SL. Incidence of injury-causing falls among older adults by place of residence: a population based study. Journal of the American Geriatrics Society 1995;43:871-876. Norton R, Campbell AJ, Lee-Joe T, Robinson E, Butler M. Circumstances of falls resulting in hip fractures among older people. Journal of the American Geriatrics Society 1997;45:1108-12. Robertson MC, Milburn PD, Campbell AJ, Bowmar J. Introducing KradalTM a new energy absorbing floor technology to prevent fractures in elderly people. Poster presentation at the Australian Falls Prevention Conference, Brisbane, November 2006. Rubenstein LZ, Josephson KR, Osterweil D. Falls and fall prevention in the nursing home. Clinics in Geriatric Medicine 1996;12:881-902.

42

References for falls prevention randomised controlled trials in institutions included in systematic review
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Barreca S, Sigouin CS, Lambert C, Ansley B. Effects of extra training on the ability of stroke survivors to perform an independent sit-to-stand: a randomized controlled trial. Journal of Geriatric Physical Therapy 2004;27(2):59-68. Becker C, Kron M, Lindemann U, et al. Effectiveness of a multifaceted intervention on falls in nursing home residents. Journal of the American Geriatrics Society 2003;51(3):306-313. Bischoff HA, Stahelin HB, Dick W, et al. Effects of vitamin D and calcium supplementation on falls: a randomized controlled trial. Journal of Bone & Mineral Research 2003;18(2):343351. Broe KE, Chen TC, Weinberg J, Bischoff-Ferrari HA, Holick MF, Kiel DP. A higher dose of vitamin D reduces the risk of falls in nursing home residents: a randomized, multiple-dose study. Journal of the American Geriatrics Society 2007;55(2):234-239. Buettner LL. Falls prevention in dementia populations. Provider 2002;28(2):41-43. Burleigh E, McColl J, Potter J. Does vitamin D stop inpatients falling? A randomised controlled trial. Age & Ageing 2007;36:507-513. Capezuti E, Strumpf NE, Evans LK, Grisso JA, Maislin G. The relationship between physical restraint removal and falls and injuries among nursing home residents. Journals of Gerontology Series A-Biological Sciences & Medical Sciences 1998;53(1):M47-52. Choi JH, Moon JS, Song R. Effects of Sun-style Tai Chi exercise on physical fitness and fall prevention in fall-prone older adults. Journal of Advanced Nursing 2005;51(2):150-157. Cox H, Puffer S, Morton V, et al. Educating nursing home staff on fracture prevention: a cluster randomised trial. Age & Ageing 2007; Advance Access doi:10.1093/ageing/afm168. Crotty M, Rowett D, Spurling L, Giles LC, Phillips PA. Does the addition of a pharmacist transition coordinator improve evidence-based medication management and health outcomes in older adults moving from the hospital to a long-term care facility? Results of a randomized, controlled trial. American Journal Geriatric Pharmacotherapy 2004;2(4):257-264. Crotty M, Halbert J, Rowett D, et al. An outreach geriatric medication advisory service in residential aged care: a randomised controlled trial of case conferencing. Age & Ageing 2004;33(6):612-617. Donald IP, Pitt K, Armstrong E, Shuttleworth H. Preventing falls on an elderly care rehabilitation ward. Clinical Rehabilitation 2000;14(2):178-185. Dyer CA, Taylor GJ, Reed M, Dyer CA, Robertson DR, Harrington R. Falls prevention in residential care homes: a randomised controlled trial. Age & Ageing 2004;33(6):596-602. Faber MJ, Bosscher RJ, Chin APMJ, van Wieringen PC. Effects of exercise programs on falls and mobility in frail and pre-frail older adults: A multicenter randomized controlled trial. Archives of Physical Medicine & Rehabilitation 2006;87(7):885-896. Flicker L, MacInnis RJ, Stein MS, et al. Should older people in residential care receive vitamin D to prevent falls? Results of a randomized trial. Journal of the American Geriatrics Society 2005;53(11):1881-1888. Haines TP, Bennell KL, Osborne RH, Hill KD. Effectiveness of targeted falls prevention programme in subacute hospital setting: randomised controlled trial. BMJ 2004;328(7441):676. Healey F, Monro A, Cockram A, Adams V, Heseltine D. Using targeted risk factor reduction to prevent falls in older in-patients: a randomised controlled trial. Age & Ageing 2004;33(4):390-395. Jensen J, Lundin-Olsson L, Nyberg L, Gustafson Y. Fall and injury prevention in older people living in residential care facilities. A cluster randomized trial. Annals of Internal Medicine 2002;136(10):733-741. Kerse N, Butler M, Robinson E, Todd M. Fall prevention in residential care: a cluster, randomized, controlled trial. Journal of the American Geriatrics Society 2004;52(4):524-531.

11. 12. 13. 14. 15. 16. 17. 18. 19.

43

20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.

33. 34. 35. 36. 37.

Law M, Withers H, Morris J, Anderson F. Vitamin D supplementation and the prevention of fractures and falls: results of a randomised trial in elderly people in residential accommodation. Age & Ageing 2006;35(5):482-486. Mador JE, Giles L, Whitehead C, Crotty M. A randomized controlled trial of a behavior advisory service for hospitalized older patients with confusion. International Journal of Geriatric Psychiatry 2004;19(9):858-863. Mayo NE, Gloutney L, Levy AR. A randomized trial of identification bracelets to prevent falls among patients in a rehabilitation hospital. Archives of Physical Medicine & Rehabilitation 1994;75(12):1302-1308. McMurdo MET, Millar AM, Daly F. A randomized controlled trial of fall prevention strategies in old peoples' homes. Gerontology 2000;46(2):83-87. Mulrow CD, Gerety MB, Kanten D, et al. A randomized trial of physical rehabilitation for very frail nursing home residents. JAMA 1994;271(3):519-524. Nowalk MP, Prendergast JM, Bayles CM, D'Amico FJ, Colvin GC. A randomized trial of exercise programs among older individuals living in two long-term care facilities: the FallsFREE Program. Journal of the American Geriatrics Society 2001;49(7):859-865. Rubenstein LZ, Robbins AS, Josephson KR, Schulman BL, Osterweil D. The value of assessing falls in an elderly population. A randomized clinical trial. Annals of Internal Medicine 1990;113(4):308-316. Sakamoto K, Nakamura T, Hagino H, et al. Effects of unipedal standing balance exercise on the prevention of falls and hip fracture among clinically defined high-risk elderly individuals: a randomized controlled trial. Journal of Orthopaedic Science 2006;11(5):467-472. Sato Y, Iwamoto J, Kanoko T, Satoh K. Low-dose vitamin D prevents muscular atrophy and reduces falls and hip fractures in women after stroke: a randomized controlled trial. Cerebrovascular Diseases 2005;20(3):187-192. Schnelle JF, Kapur K, Alessi C, et al. Does an exercise and incontinence intervention save healthcare costs in a nursing home population? Journal of the American Geriatrics Society 2003;51(2):161-168. Schoenfelder DP. A fall prevention program for elderly individuals. Exercise in long-term care settings. Journal of Gerontological Nursing 2000;26(3):43-51. Shaw FE, Bond J, Richardson DA, et al. Multifactorial intervention after a fall in older people with cognitive impairment and dementia presenting to the accident and emergency department: randomised controlled trial. BMJ 2003;326(7380):73. Shimada H, Obuchi S, Furuna T, Suzuki T. New intervention program for preventing falls among frail elderly people: the effects of perturbed walking exercise using a bilateral separated treadmill. American Journal of Physical Medicine & Rehabilitation 2004;83(7):493499. Sihvonen S, Sipila S, Taskinen S, Era P. Fall incidence in frail older women after individualized visual feedback-based balance training. Gerontology 2004;50(6):411-416. Stenvall M, Olofsson B, Lundstrom M, et al. A multidisciplinary, multifactorial intervention program reduces postoperative falls and injuries after femoral neck fracture. Osteoporosis International 2007;18(2):167-175. Tideiksaar R, Feiner CF, Maby J. Falls prevention: the efficacy of a bed alarm system in an acute-care setting. The Mount Sinai Journal of Medicine 1993;60(6):522-527. Toulotte C, Fabre C, Dangremont B, Lensel G, Thevenon A. Effects of physical training on the physical capacity of frail, demented patients with a history of falling: a randomised controlled trial. Age & Ageing 2003;32(1):67-73. Zermansky AG, Alldred DP, Petty DR, et al. Clinical medication review by a pharmacist of elderly people living in care homes--randomised controlled trial. Age & Ageing 2006;35(6):586-591.

44

Study Barreca 2004 (hospital)

Intervention

Setting

Target group

No. of men

No. of Total women randomised

Injury data Y/N

Cost data Y/N

EX (extra practice sit-to-stands) MFACT (staff training, education for residents, environmental adaptations, exercise, hip protectors)

Stroke rehabilitation unit

Stroke survivors

31

17

48

Becker 2003

6 nursing homes

All long stay residents Very few exclusion criteria Very few exclusion criteria

209

772

981

3 4 5

Bischoff 2003 Broe 2007 Buettner 2002 Burleigh 2007 (hospital)

VITD VITD MPLE (EX + sensory mat)

Long-stay geriatric care Long term care facility

0 34 ?

122 90 ?

122 124 25

N N N

N N Y

Nursing home residents with 2 falls in past 2 dementia months Acute admissions 65 years Restrained residents

VITD ED (restraint education) vs control ED (restraint education + consultation) vs control ED (2 intervention homes combined) vs control home EX (tai chi)

Hospital

84

121

205

Capezuti 1998 a)

Nursing home

95

538

633

Capezuti 1998 b)

7 8

Capezuti 1998 c) Choi 2005

Residential care

1 of 5 risk factors

15

44

59

Cox 2008

ED (staff training on fracture prevention)

Advised staff to assess all Residential care and nursing residents for falls homes, and EMI risk

1402

4642

6044

45

10

Crotty 2004 1

MED (transition pharmacist) ED (outreach intervention: 2 pharmacist visits to physicians, link nurse) EX (additional exercises vs conventional) EN (carpet vs vinyl) MFACT (exercise, medication review, podiatry, optometry)

Long term care facilities

Transfer from hospital to long term care

43

67

110

11

Crotty 2004 2 Donald 2000 (hospital) a) Donald 2000 (hospital) b)

Residential care facilities

All residents

115

600

715

12 12

Hospital

All patients

44

10

54

13

Dyer 2004

20 residential care homes

All residents All residents (excluding severe dementia)

43

153

196

14 14 14 15

Faber 2006 a) Faber 2006 b) Faber 2006 c)

EX ("functional walking" vs control) EX ("in balance" vs control) EX (both vs control) VITD MFACT (falls alert card, exercise, education, hip protectors) MFACT (brief screen using checklist and suggested interventions)

Long term care (self-care, nursing homes)

50

188

238

Flicker 2005 Haines 2004 (hospital)

Residential care (assisted living, nursing homes)

Vitamin D lower half of lab reference range All patients on admission Patients with fall history (previous or in hospital)

32

593

625

16

Hospital (subacute)

206

420

626

17

Healey 2004 (hospital)

Hospital (6 wards)

662

992

1654

46

18

Jensen 2002

MFACT (staff education, screening, modify environment, exercise programmes, supplying & repairing aids, drug reviews, free hip protectors, post-fall conferences, guiding staff) MFACT (risk assessment, information on strategies)

Residential care facilities

All residents and all permanent staff

121

288

409

19

Kerse 2004

New Zealand rest homes Residential care, nursing care, care for elderly mentally infirm; 223 units (118 homes)

All residents

152

476

628

20

Law 2006

VITD MFACT (assessment and advice from nurse)

All residents Patients with confusion Patients with stroke, urinary incontinence, history of falls

892

2825

3717

21

Mador 2004 (hospital)

Hospital

37

34

71

22

Mayo 1994 (hospital)

ID bracelet MFACT (assessment, referrals, seated exercise)

Rehabilitation hospital

72

62

134

23

McMurdo 2000 EX (individual ROM, strength, balance, transfer, mobility) MPLE (3 ED + EX vs the 3 ED programmes) MPLE (3 ED + tai chi vs the 3 ED programmes) MFACT (assessment and recommendations to GP)

9 residential homes

Residents 70 years Residents 3 months with 2 ADLs All ambulating residents

25

108

133

24

Mulrow 1994

Nursing homes

57

137

194

25

Nowalk 2001 a)

2 long term care facilities

15

95

110

25

Nowalk 2001 b)

26

Rubenstein 1990

A long term care facility

Within 7 days after a fall

24

136

160

47

27

Sakamoto 2006

EX (stand on one leg for 1 minute 3 times a day)

Nursing care facilities; users of outpatient rehabilitation centres (21%)

All residents Stroke 2 years previously Incontinence Able to ambulate, MMSE 20 Cognitive impairment or dementia Receiving rehabilitation All residents invited Patients with femoral neck fracture "At increased risk of bed falls" Dementia + fall(s) in the previous 3 months All residents

142

411

553

28 29

Sato 2005 Schnelle 2003

VITD MPLE (EX + incontinence care) EX (ankle strengthening and walking programme) MFACT (assessment, medication modification, exercise, home safety by OT) EX (treadmill vs usual programme) EX (balance training)

Long term nursing care 4 nursing homes

0 28

96 162

96 190

Y Y

N Y

30

Schoenfelder 2000

2 nursing homes Recruited from 3 A&E departments, presenting after a fall Residents long term care or outpatients 2 residential care homes

12

16

31

Shaw 2003

55

219

274

32 33

Shimada 2004 Sihvonen 2004

7 0

25 27

32 27

N N

N N

34

Stenvall 2007 Tideiksaar 1993 (hospital)

MFACT (assessment, treatment of fall risk factors, address 1 geriatric and 1 orthopaedic postoperative complications) ward at a university hospital Geriatric evaluation and treatment unit

51

148

199

35

BED ALARM

10

60

70

36 37

Toulotte 2003 Zermansky 2006

EX (balance training) MED (medication review by pharmacist)

Institutional care Nursing, residential and mixed care homes

? 154 Men 4,911 45

? 507 Women 15,197 (2 trials)

20 661 Total 20,153

N N

N Y

Total: Gender not specified:

48

Note: Studies entered in italics denote a hospital setting, the remainder were in long term care facilities. ADL denotes activities of daily living. Codes for types of interventions tested: ED Education programme EN Environmental change EX Exercise programme MED Medication review and/or modification MFACT Multifactorial programme (participants received interventions based on individual choice or an assessment of fall risk factors) MPLE The same combination of two or more interventions received by all participants VITD Vitamin D supplements

49

Preventing injury from falls: selected issues

Extracting and pooling the information on injury as a result of a fall from the falls intervention trials identified proved problematic as the definitions of injuries differed markedly in the trials, or there was no definition of injury provided. A limited number of studies included in the systematic reviews reported the number of participants in each group who had a fracture as a result of a fall during the trial. This outcome is more robust as fractures can be verified by X-ray. Here we present the risk ratio data pooled when appropriate, and when there were results from two or more studies available to pool. Interventions with multiple components in the community and residential care We found a total of 40 interventions with multiple components tested in community living people. Pooling of 12 programmes with data on the rate of falls available, showed that the rate of falls was significantly reduced by 25% (see Appendix C). All current guidelines recommend addressing multifactorial risk factors as an effective strategy for reducing falls, however our own work has established that single factor approaches are equally effective (Campbell 2007). Pooling the six risk ratios for fracture (any site) reported in these trials showed that the number sustaining a fracture was not significantly reduced (see Figure 1).

Davison 2005

Hogan 2001

Lightbody 2002

Nikolaus 2003

Tinetti 1994

Vetter 1992

Combined

.1

.5 Risk ratio

1.5

Figure 1. Meta-analysis for proportion of participants sustaining a fracture (any site) resulting from a fall in trials testing multifactorial programmes in community living older people. Pooled risk ratio 0.74 (95%CI 0.47 to 1.17), P = 0.195. Test for heterogeneity: Q = 3.067, P = 0.690. | Weights Study 95% CI Study | Fixed Random Est Lower Upper ------------------------------------------------------------------------------Davison 2005 | 4.09 4.09 0.53 0.20 1.40 Hogan 2001 | 1.99 1.99 0.64 0.16 2.57 Lightbody 2002 | 0.82 0.82 0.26 0.03 2.26 Nikolaus 2003 | 1.75 1.75 1.32 0.30 5.81 Tinetti 1994 | 2.12 2.12 0.49 0.13 1.88 Vetter 1992 | 8.00 8.00 1.00 0.50 2.00

50

We found a total of eight interventions with multifactorial components tested in long term care residents. Pooling of seven of these programmes with data on the rate of falls available, showed that overall the rate of falls was reduced by 21%, but this reduction did not quite reach statistical significance (P = 0.058) (see Appendix D). The pooling of data from three of these trials showed that the number sustaining a hip fracture was significantly reduced by 51% from multifactorial approaches in a residential care setting (P = 0.045, Figure 2). With so few of the studies pooled, and so few hip fracture events recorded, this result should be viewed with caution. Although there is no statistical heterogeneity in this model, it must also be noted that the programme tested by Shaw et al was targeted at those with marked cognitive impairment only (Shaw 2003) and therefore differed from the other two trials pooled.

Becker 2003

Jensen 2002

Shaw 2003

Combined .1 .5 Risk ratio 1 1.5 2

Figure 2. Meta-analysis for proportion of participants sustaining a hip fracture resulting from a fall in trials testing multifactorial programmes in residential care. Pooled risk ratio 0.49 (95%CI 0.24 to 0.98), P = 0.045. Test for heterogeneity: Q = 1.848, P = 0.397. Weights Study 95% CI Study | Fixed Random Est Lower Upper -------------------------------------------------------------------------Becker 2003 | 1.54 1.54 0.93 0.19 4.52 Jensen 2002 | 2.03 2.03 0.23 0.06 0.91 Shaw 2003 | 4.18 4.18 0.55 0.21 1.44

Exercise programmes in the community Exercise classes, typically with strength, balance, agility and flexibility exercises included and taught by a trained instructor, were the most common single intervention in the literature (23 trials). Overall there was a significant reduction in the rate of falls by 29% in the 9 trials using such a combination of exercises and with data on the rate of falls available for pooling (see Appendix C). Two of these trials reported the number of people sustaining a fracture, as did a trial of a home based exercise programme for people with Parkinsons Disease (Ashburn 2007). Pooling the three risk ratios showed that fractures were significantly reduced by 66% (P = 0.005, see Figure 3). This result needs to be viewed with caution as it represents information from only 6% (3 of 48) of the trials that

51

tested an exercise programme in community living older people, and the fact that the participants in one of the trials differed markedly (clinical heterogeneity) in that all had Parkinsons Disease.

Ashburn 2007

Korpleinen 2006

McMurdo 1997

Combined

.1

Risk ratio

.5

1.5 2

Figure 3. Meta-analysis for proportion of participants sustaining a fracture (any site) resulting from a fall in trials testing a group exercise programme in community living older people. Pooled risk ratio 0.34 (95%CI 0.16 to 0.72), P = 0.005. Test for heterogeneity: Q = 0.097, P = 0.953. | Weights Study 95% CI Study | Fixed Random Est Lower Upper --------------------------------------------------------------------------------------Ashburn 2007 | 1.58 1.58 0.33 0.07 1.57 Korpleinen 2006 | 4.85 4.85 0.36 0.15 0.88 McMurdo 1997 | 0.42 0.42 0.22 0.01 4.48 An incidence rate ratio comparing a community living group taking vitamin D supplementation (with or without calcium) compared with a placebo or calcium group was available for only two of the 12 trials identified. Therefore we provided the risk ratio obtained from pooling the risk of one or more falls (number of fallers) in the intervention versus the control group from the nine trials with these data available (see Appendix C). Overall vitamin D supplementation significantly reduced the risk of a fall by 14%. We pool here fracture risk ratios from nine trials testing supplementation with vitamin D or an analogue (Figure 4). The pooled risk ratio showed no significant reduction in the numbers sustaining a fracture in the vitamin D groups versus the control groups.

52

Bischoff-Ferarri 2006

Gallagher 2001

Grant 2005

Harwood 2004

Pfeiffer 2000

Porthouse 2005

Sato 1999

Smith 2007

Trivedi 2003

Combined

.1

Risk ratio

.5

1.5

Figure 4. Meta-analysis for proportion of participants sustaining a fracture (any site) resulting from a fall in trials testing vitamin D supplements (with or without calcium supplementation) in community living older people. Pooled risk ratio 0.86 (95%CI 0.71 to 1.04), P = 0.129. Test for heterogeneity: Q = 18.286, P = 0.019. | Weights Study 95% CI Study | Fixed Random Est Lower Upper ------------------------------------------------------------------------------------------Bischoff-Ferarri 2006 | 8.26 6.47 0.46 0.23 0.91 Gallagher 2001 | 6.72 5.49 0.60 0.28 1.28 Grant 2005 |168.72 25.43 1.02 0.88 1.19 Harwood 2004 | 3.01 2.74 0.50 0.16 1.55 Pfeiffer 2000 | 2.06 1.93 0.48 0.12 1.88 Porthouse 2005 | 31.34 15.31 1.01 0.71 1.43 Sato 1999 | 1.03 1.00 0.13 0.02 0.90 Smith 2007 |150.59 24.97 1.09 0.93 1.28 Trivedi 2003 | 65.53 20.55 0.78 0.61 0.99 Vision improvement Three trials tested an intervention aimed at reducing falls by improving vision. Two trials tested whether expedited cataract removal would reduce falls in women over 70 who had been referred to a hospital ophthalmology department (Harwood 2005; Foss 2006). Pooling the rate ratios from the two trials showed a significant 33% reduction in falls (see Appendix C). The third trial aiming to prevent falls by improving vision, had quite the opposite effect on the fall rate (Cumming 2007). Community living people aged 70 and older recruited mainly from an aged care service, received vision and eye examinations and subsequent treatment of eye problems. After 12 months the rate of falls was significantly increased by 57% in the intervention group compared with the control group. Pooling risk ratios for fracture for the treatment versus the control groups during these three trials showed no difference in the number with fractures sustained as a result of a fall (Figure 5).

53

Cumming 2007

Foss 2006

Harwood 2005

Combined

.1

.5

1 1.5 Risk ratio

Figure 5. Meta-analysis for proportion of participants with a fracture (any site) resulting from a fall in trials testing interventions to improve vision in community living older people. Pooled risk ratio 1.11 (95%CI 0.35 to 3.54), P = 0.861. Test for heterogeneity: Q = 7.038, P = 0.030. | Weights Study 95% CI Study | Fixed Random Est Lower Upper ---------------------------------------------------------------------------------------Cumming 2007 | 11.32 1.22 1.74 0.97 3.12 Foss 2006 | 1.48 0.71 2.50 0.50 12.50 Harwood 2005 | 2.90 0.93 0.33 0.10 1.04

Implications of findings from fall related injuries reported in the randomised controlled falls prevention trials

Although there has been a rapid growth in the number of randomised controlled trials published that tested falls prevention interventions in older people, only 38 (27%) of the trials reported injuries events or falls resulting in medical care being sought. Marked variations in the definition of a fall related injury event in the trials meant that these data, even if reported, could not be pooled. Information gained from the falls prevention trials about the number of participants who sustained a fracture as a result of a fall is more robust since fractures can be verified by X-ray, but a limited number of trials reported this outcome. Three trials of multifactorial programmes in residential care reported the number of participants with a hip fracture as a result of a fall during the trial. The pooled risk ratio showed a significant reduction in fractures, but this finding needs to be viewed with caution due to the very small number of hip fracture events recorded. Pooling the risk ratio for sustaining a fracture from the three trials of exercise programmes in the community showed the number with fractures were significantly reduced. However, this result also must be viewed with caution as it represents only 6% (3 of 48) of the trials testing

54

exercise programmes in the community. In addition, the exercise programmes differed in that two were group exercise programmes and one tested a home based programme for people with Parkinsons Disease.

55

Relative cost effectiveness systematic review

of

interventions:

Cost effectiveness of falls prevention strategies

ACC and the Ministry of health require not only evidence of effectiveness but also value for money when funding and supporting falls prevention strategies. Economic evaluations provide a measure of efficiency for the purposes of decision making by comparing the costs and benefits of alternative programmes. There are four main types of economic evaluations 1) cost analysis, 2) cost effectiveness analysis, 3) cost utility analysis, and 4) cost benefit analysis. When extracting data from the 103 randomised controlled trials in the community and 37 in an institutional setting identified in our systematic reviews, we noted the following economic evaluations had been carried out within these trials: For 8 trials testing an intervention in the community setting, the costs of delivering the intervention were reported For 1 intervention tested in a long term care setting, the costs of delivering the intervention were reported For 1 intervention in the community and 2 in a long term care setting, the costs of some health care use during the trial was reported, but no comprehensive economic evaluations were undertaken No information was reported on the costs or cost effectiveness of delivering a falls prevention programme in a hospital setting In a total of 8 trials in a community setting, a comprehensive economic evaluation was reported for the 7 different interventions tested (8 cost effectiveness analyses, one cost utility analysis). In this section we describe the results of the cost analyses, and cost effectiveness and cost utility evaluations carried out within the 140 trials. Investigation of the cost effectiveness of effective interventions was undertaken to provide information on the likely value for money of each type of approach in the community setting, and in long term care and hospitals.

Costs, cost effectiveness and cost utility of delivering effective falls prevention programmes to community dwelling older people
The majority of randomised controlled trials have taken place in the community setting (103 of 140) and several different approaches have been shown to significantly reduce the rate of falls. Comprehensive economic evaluations have been carried out within these trials for evaluating the cost effectiveness of six of the successful interventions. A cost effectiveness analysis compares the costs as well as the outcomes for alternative treatments or approaches. This provides information for decision making to enable comparisons between programmes with the same clinically relevant outcome measure in this case number of falls prevented. The intervention with the lowest cost per fall prevented is preferred as this will give the best value for money. A cost analysis compares only the costs of alternative treatments or therapies, but does not capture the effectiveness of the interventions. Cost effectiveness analyses have been reported for the Otago Exercise Programme (Robertson 2001a, Robertson 2001b), tai chi classes (Haas 2006), a multifactorial programme addressing individual risk factors delivered mainly at home (Rizzo 1996), a home safety assessment and modification programme for people with severe visual impairment (Campbell 2005) and those recently discharged from hospital (Salkeld 2000), for psychotropic medication withdrawal (Robertson 2001c), and for elderly women following first eye cataract surgery (Sach 2007). The time period for all these economic evaluations was within the trial duration. The perspectives of the economic studies varied from a health system, health care provider or a societal perspective. The results from three trials showed the potential for cost savings as a result of delivering a falls prevention intervention (Rizzo 1996, Robertson 2001a, Salkeld 2000). In the West Auckland trial of the Otago Exercise Programme there were five hospital admissions as a result of a fall in the control

56

group and none in the intervention group. This resulted in a cost saving during one year of $47,818 (actual inpatient costs at 1998 prices). The incremental cost effective ratios for particular high risk subgroups of older people was less than zero (indicating cost savings) in two further trials. The multifactorial intervention tested by Tinetti and colleagues (Tinetti 1994) was cost saving when the incremental cost effective ratio was calculated using mean costs, both in terms of number of falls prevented and falls resulting in medical treatment prevented (Rizzo 1996). The incremental cost effectiveness ratio indicated cost savings when the subgroup with a previous fall was considered in Cumming and colleagues trial (Cumming 1999) testing a home safety programme (Salkeld 2000). This indicates that delivering a home safety programme to this subgroup was cost effective, whereas to a broader group of older people it was not. A further trial described the use of health care resources during the trial and found that hospitalised control group participants were more likely to have hospital costs of US$5,000 during the trial than those taking part in group exercise (Buchner 1997). The results from this study and the cost effectiveness analyses are summarised in Table 1. One cost utility analysis has been reported (in addition to a cost effectiveness analysis) as part of the trial testing expedited cataract surgery (Sach 2007). Cost utility analysis compares outcomes in terms of quality adjusted life years (QALYs). This type of analysis enables comparison between treatments or healthcare services where the outcomes are quite different, for example a way to compare value for money of a falls prevention programme with say a programme to improve quality of life in people with osteoarthritis. However measuring QALYs may be imprecise as individuals value different health states differently, and people with and without a condition (such as osteoarthritis for example) give quite different responses. The study by Sach et al reported that the incremental cost utility ratio was 35,704 which is above a currently accepted UK threshold of willingness to pay per QALY gained of 30,000. If however the time period of the analysis was extended from the trial period only to the persons expected lifetime, the incremental cost per QALY gained was much lower at 13,172. In this study the increase in QALYs for participants following surgery can be attributed to quality of life gains resulting from improved vision, including falls prevented. The information provided by these analyses indicates there is some, although limited, evidence that falls prevention strategies can be cost saving during the trial period, and may also be cost effective over the participants remaining lifetime. The results endorse the fact that, to obtain maximum value for money, effective strategies need to be targeted at particular subgroups of older people.

Costs of delivering falls prevention programmes in long term care facilities

Falls are even more common in residential care facilities than for older people living in the community (Rubenstein 1996). A study in 14 New Zealand residential care facilities recorded falls over 18 months and reported that there were 954 falls by 271 of the 680 residents (almost half of the residents), 68% resulting in injury (Butler 2004). The number of falls increased in the afternoon. Environmental hazards such as cords, steps, mats and dim light were involved with 15% of the falls. The number of injuries was higher during the day when the residents were active, and the more active residents fell more but were less likely to injure themselves. Prevention in this setting is a challenge as many residents are physically frail and have cognitive impairment, and facility staff are not routinely trained in prevention strategies. A total of 29 randomised controlled trials in a residential care setting were identified, and multifactorial interventions and vitamin D trials when pooled showed a significant reduction in falls. The cost of delivering the intervention and health service costs for participants during the trial were available for only one of these trials (Mulrow 1994). This trial tested a physical rehabilitation programme in very frail nursing home residents and showed an improvement in the physical disability index score but no reduction in falls. There was no difference in the cost per participant for health care service use in the two groups, In two other trials, the cost of some health care use was reported. Schnelle and colleagues reported the costs of treating the acute episodes which the intervention aimed to reduce (Schnelle 2003). After recommendations by a pharmacist regarding medicines, there was no difference between the intervention and control groups in the cost of 28 days of repeat medicines at

57

the end of the trial (Zermansky 2006). The results from these three studies are summarised in Table 2. No cost effectiveness or cost utility analyses were included in any of the study designs.

Costs of delivering falls prevention programmes in hospital

Falls account for over 80% of all injury related admissions to hospital for people over 65 years (Lord 2001). Paradoxically, a hospital setting is not a safe place for elderly people but is actually associated with an increased risk of falling (Kannus 2006). On admission, the older patient accumulates additional falls risk factors including a new, strange environment with poorly recognised external dangers for falling. This is often combined with confusion, acute illness, and balanceaffecting medication, in addition to chronic risk factors such as comorbidities, muscle weakness and impaired balance and gait. Most hospitals have a falls prevention policy in place in an attempt to minimise these risks. There were eight trials testing nine interventions carried out in a hospital setting that met our systematic review criteria. Three tested a multifactorial programme, two an exercise programme, one compared falls for patients in rooms with carpet versus vinyl flooring, one tested a bed alarm, one an identification bracelet, and one vitamin D supplementation. Only one of these eight studies produced a significant reduction in the number of falls (Haines 2004), and one in the number of fallers (Stenvall 2007). Both these two trials tested a multifactorial approach. Carpeted rooms appeared to increase the rate of falls compared with vinyl flooring (incidence rate ratio 14.77, 95% confidence interval 1.89 to 115.36). For the trials in a hospital setting, no cost or cost effectiveness data were reported.

Interventions to reduce fall injuries in hospitals and long term care facilities
Currently there are several biomechanical means of reducing the trauma resulting from falls being developed and tested. Two that aim to prevent fractures from falls are hip protectors and safety flooring. Both these approaches have cost effectiveness evaluations available that were carried out in New Zealand. Hip protectors Hip fractures, the most costly and traumatic injury from a fall, result almost exclusively from a simple fall and the impact of the greater trochanter with the floor (Dargent-Molina 1996, Norton 1997). Studies have addressed the problem by using hip protectors, but compliance has been estimated at only 25-30% (Chan 2000). Further problems with hip protectors result from displacement of the pads from the greater trochanter region prior to the fall. In the New Zealand study of the circumstances and consequences of falls in residential care, 35 falls occurred with hip protectors being worn at the time (Butler 2004). There were no hip fractures when residents fell wearing hip protectors, but over the 18 month period there were 12 hip fractures in the 917 falls without a hip protector in place. ACC commissioned a review on the potential effectiveness of hip protectors in New Zealand residential care facilities, plus a cost effectiveness evaluation, to investigate the value for money in providing hip protectors to all residents of long term care facilities in New Zealand. This work will be used in Phase 2 of this project to rank the cost effectiveness of this approach against falls prevention strategies in institutions and in community living older people. Safety flooring The researchers have been collaborating with the New Zealand inventor in the development and testing of an innovative new flooring material designed specifically to absorb energy on impact (Robertson 2006). This has the potential to reduce the number of fractures and other injuries as a result of falls in elderly people, especially those living in long term care (Cali 1995). Compared with known injury prevention programmes, this strategy has the advantage of being a long term approach plus eliminating the problems of individual compliance (Drahota 2007). In progress is our own study comparing gait patterns and balance of elderly women walking and standing on three different types of flooring a standard non-compliant flooring (vinyl), a

58

compliant flooring (carpet with good quality underlay) and Kradal, the new safety flooring manufactured by Acma Industries in Upper Hutt, Wellington. The biomechanics laboratory of Professor Steve Robinovitch, Simon Fraser University, Vancouver, Canada is comparing the impact attenuation properties of Kradal using impact on an artificial hip to simulate a fall onto the hip. Kradal has undergone all the standard flooring material tests and passed with flying colours. A cost effectiveness evaluation of the new flooring has been carried out at the School of Population Health, University of Auckland by health economist Dr Paul Brown which show that this flooring material has the potential to be cost saving from ACCs perspective.

Implications of findings from the costs, cost effectiveness and cost utility analyses in the randomised controlled trials

Although there has been a rapid growth in the number of randomised controlled trials published that tested falls prevention interventions in older people, analyses of the costs and cost effectiveness of the interventions within the trials to date is limited. Comprehensive cost effectiveness analyses were carried out in only eight of the 140 randomised controlled trials included in the systematic reviews of falls prevention interventions. All eight trials tested strategies that were effective in reducing falls in community living older people. For three interventions, the Otago Exercise Programme psychotropic medication withdrawal and a home safety programme effective in reducing falls in people with severe visual impairment, the costs of delivering the programmes and the cost effectiveness in New Zealand have been established. In a trial in West Auckland in 1998, the cost of delivering the Otago Exercise Programme was $432 per person for one year and the incremental cost effectiveness ratio was $1,803 per fall prevented. When hospital costs averted were considered, the incremental cost effectiveness ratio dropped to $155 per fall prevented. The home safety programme cost $325 per person to deliver and the incremental cost effectiveness ratio was $650 per fall prevented (at 2004 prices). The psychotropic medication withdrawal intervention cost $258 per person to deliver in a research setting at 1996 prices and the incremental cost per fall prevented was $538. One trial in community living older people reported a cost utility analysis and extended the time period of the analysis from the trial duration to the participants remaining lifetime. The information provided by these comprehensive economic evaluations indicates there is some, although limited, evidence that falls prevention strategies can be cost saving during the trial period, and may also be cost effective over the participants remaining lifetime. The results of these analyses also endorse the fact that, to obtain maximum value for money, effective strategies need to be targeted at particular subgroups of older people. The cost of delivering the intervention and health service costs for participants during the trial were available for only one of the trials in residential care. No cost effectiveness analyses in this setting were reported. For the randomised controlled trials in a hospital setting, no cost data or cost effectiveness analyses were reported.

59

References
Buchner DM, Cress ME, de Lateur BJ, Esselman PC, Margherita AJ, Price R, et al. The effect of strength and endurance training on gait, balance, fall risk, and health services use in community-living older adults. Journals of Gerontology Series A-Biological Sciences & Medical Sciences 1997;52A(4):M218-24. Butler M, Kerse N, Todd M. Circumstances and consequences of falls in residential care: the New Zealand story. New Zealand Medical Journal 2004;117(1202):U1076. URL: http://www.nzma.org.nz/journal/117-1202/1076/. Cali CM, Kiel DP. An epidemiological study of fall-related fractures among institutionalized older people. Journal of the American Geriatrics Society 1995;43:1336-1340. Chan DK, Hiller G, Cooke R, Monk R, Mills J, Hung WT. Effectiveness and acceptability of a newly designed hip protector. Archives of Gerontology and Geriatrics 2000;30:25-34. Cameron I, Murray GR, Gillespie LD, Cumming RG, Robertson MC, Hill K, Kerse N. Interventions for preventing falls in older people in residential care facilities and hospitals. (Protocol) Cochrane Database of Systematic Reviews 2005, Issue 3. Art. No.: CD005465. DOI: 10.1002/14651858.CD005465. Campbell AJ, Robertson MC, Gardner MM, Norton RN, Tilyard MW, Buchner DM. Randomised controlled trial of a general practice programme of home based exercise to prevent falls in elderly women. BMJ 1997;315(7115):1065-9. Campbell AJ, Robertson MC, Gardner MM, Norton RN, Buchner DM. Falls prevention over 2 years: a randomized controlled trial in women 80 years and older. Age & Ageing 1999a;28:513-8. Campbell AJ, Robertson MC, Gardner MM, Norton RN, Buchner DM. Psychotropic medication withdrawal and a home-based exercise program to prevent falls: a randomized, controlled trial. Journal of the American Geriatrics Society 1999b;47(7):850-3. Campbell AJ, Robertson MC, La Grow SJ, Kerse NM, Sanderson GF, Jacobs RJ, et al. Randomised controlled trial of prevention of falls in people aged 75 with severe visual impairment: the VIP trial. BMJ 2005;331(7520):817. Campbell AJ, Robertson MC. Rethinking individual and community fall prevention strategies: a metaregression comparing single and multifactorial interventions. Age & Aeing 2007;36:656-62. Cumming RG, Thomas M, Szonyi G, Salkeld G, O'Neill E, Westbury C, et al. Home visits by an occupational therapist for assessment and modification of environmental hazards: a randomized trial of falls prevention. Journal of the American Geriatrics Society 1999;47(12):1397-402. Dargent-Molina P, Favie F, Grandjean H, et al. Fall-related factors and risk of hip fracture: the EPIDOS prospective study. Lancet 1996;384:145-49. Drahota A, Gal D, Windsor J. Flooring as an intervention to reduce injuries from falls in healthcare settings: an overview. Quality Ageing 2007;8(1):3-9. Dyer CA, Taylor GJ, Reed M, Dyer CA, Robertson DR, Harrington R. Falls prevention in residential care homes: a randomised controlled trial. Age & Ageing 2004;33(6):596-602. Gillespie LD, Gillespie WJ, Robertson MC et al. Interventions for preventing falls in elderly people. Cochrane Database of Systematic Reviews 2003;(4):CD000340 Haas M. Economic analysis of tai chi as a means of preventing falls and related injuries among older adults. Centre for Health Economics Research and Evaluation (CHERE), University of Technology, Sydney, Australia working paper 2006/4. Accessed 1 March 2008 at: http://datasearch.uts.edu.au/chere/research/working_papers.cfm. Harwood RH, Foss AJ, Osborn F, Gregson RM, Zaman A, Masud T. Falls and health status in elderly women following first eye cataract surgery: a randomised controlled trial. British Journal of Ophthalmology 2005;89(1):53-9. Jrvinen TLN, Sievnen H, Khan KM, Heinonen A, Kannus P. Shifting the focus in fracture prevention from osteoporosis to falls. BMJ 2008;336:124-126. Mulrow CD, Gerety MB, Kanten D, et al. A randomized trial of physical rehabilitation for very frail nursing home residents. JAMA 1994;271(3):519-24. Norton R, Campbell AJ, Lee-Joe T, Robinson E, Butler M. Circumstances of falls resulting in hip fractures among older people. Journal of the American Geriatrics Society 1997;45:1108-12.

60

Rizzo JA, Baker DI, McAvay G, Tinetti ME. The cost-effectiveness of a multifactorial targeted prevention program for falls among community elderly persons. Medical Care 1996;34(9):954-9. Robertson MC, Devlin N, Scuffham P, Gardner MM, Buchner DM, Campbell AJ. Economic evaluation of a community based exercise programme to prevent falls. Journal of Epidemiology and Community Health 2001a;55(8):600-6. Robertson MC, Devlin N, Gardner MM, Campbell AJ. Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 1: Randomised controlled trial. BMJ 2001b;322(7288):697-701. Robertson MC. Development of a falls prevention programme for elderly people: evaluation of efficacy, effectiveness, and efficiency [PhD thesis]. Department of Medical and Surgical Sciences, Dunedin School of Medicine, University of Otago, Dunedin, New Zealand, 2001c. Robertson MC, Devlin N, Gardner MM, McGee R, Campbell AJ. Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 2: Controlled trial in multiple centres. BMJ 2001d;322(7288):701-4. Robertson MC, Milburn PD, Campbell AJ, Bowmar J. Introducing KradalTM a new energy absorbing floor technology to prevent fractures in elderly people. Poster presentation at the Australian Falls Prevention Conference, Brisbane, November 2006. Robertson MC, Campbell AJ. What type of exercise reduces falls in older people? Chapter (systematic review) in: Evidence-based sports medicine, second edition. Editors Domhnall MacAuley and Thomas Best. Blackwell Publishing, Oxford, UK, 2007, pp135-166. Rubenstein LZ, Josephson KR, Osterweil D. Falls and fall prevention in the nursing home. Clinics in Geriatric Medicine 1996;12:881-902. Sach TH, Foss AJE, Gregson RM, Zaman A, Osborn F, Masud T, Harwood RH. Falls and health status in elderly women following first eye cataract surgery: an economic evaluation conducted alongside a randomised controlled trial. British Journal of Ophthalmology 2007;91:1675-9. Salkeld G, Cumming RG, ONeill, Thomas M, Szonyi G, Westbury C. The cost effectiveness of a home hazard reduction program to reduce falls among older persons. Australian and New Zealand Journal of Public Health 2000;24(3):265-71. Schnelle JF, Kapur K, Alessi C, et al. Does an exercise and incontinence intervention save healthcare costs in a nursing home population? Journal of the American Geriatrics Society 2003;51(2):161-168. Tinetti ME, Baker DI, McAvay G, Claus EB, Garrett P, Gottschalk M, et al. A multifactorial intervention to reduce the risk of falling among elderly people living in the community. New England Journal of Medicine 1994;331(13):821-7. Voukelatos A, Cumming RG, Lord SR, Rissel C. A randomized, controlled trial of tai chi for the prevention of falls: the Central Sydney tai chi Trial. Journal of the American Geriatrics Society 2007;55(8):1185-91. Zermansky AG, Alldred DP, Petty DR, et al. Clinical medication review by a pharmacist of elderly people living in care homes--randomised controlled trial. Age & Ageing 2006;35(6):586-591.

61

Table 1 Results from studies reporting costs of intervention and healthcare resource use community living older people
Article, study sample, length of time falls monitored Interventions and number being compared, length of intervention phase Perspective, type of currency, year of costs, time period costs measured HMO US dollars Randomisation 19921993 Period 7 to 18 months after randomisation Health system New Zealand dollars 2004 During trial period Cost items measured Intervention costs Healthcare service costs Measures of cost effectiveness

Buchner et al 1997 Patients from a HMO, mild deficits in strength and balance, mean age 75 years Up to 25 months Campbell et al 2005 People aged 75 years with severe visual impairment, mean age 83.6 (SD 4.8) years 1 year

Centre based endurance training and/or strength training (n = 75) vs no intervention (n = 30) Supervised for 2426 weeks then self supervised Home safety assessment and modification programme (n = 198) vs no home safety programme (n = 193) 1 to 2 home visits by experienced occupational therapist Tai chi classes weekly for 1 hour (n = 347) vs no intervention (n = 337) 16 weeks

Hospital costs, ancillary outpatient costs (from HMO computerised records)

Hospitalised control participants more likely to have hospital costs >$5,000 (p < 0.05) Total cost $64,337 Mean (SD) cost per home safety group participant $325 ($292) Total cost $81,232 Mean $245 per intervention group participant plus charge $44 per participant Not calculated (preplanned, no significant difference in number of fall injuries in the two groups) Total $24,795 (tai chi group $18,915, control group $5,880) Cost of delivering the home safety programme per fall prevented $650

Intervention costs (training costs; recruitment; occupational therapists time, transport, administration; services and equipment installed in homes; overhead costs) Intervention costs (cost of venues, advertising, instructor) Health service use related to falls from health care use diary and hospital records, valued at standard costs (GP, specialist, tests, hospitalisations, medications)

Haas 2006 (effectiveness of the intervention is reported in Voukelatos et al 2006): Healthy community living people aged 60 years, mean age 69 (SD 6.5) years 24 weeks

Health system (NSW Dept of Health) Australian dollars Recruitment June 2001 to March 2003 During trial period

Cost per fall prevented $1,683 (includes cost offset by charging $44 per course)

62

Article, study sample, length of time falls monitored

Interventions and number being compared, length of intervention phase

Perspective, type of currency, year of costs, time period costs measured Health system US dollars 1993 prices The year after study enrolment

Cost items measured

Intervention costs

Healthcare service costs

Measures of cost effectiveness

Rizzo et al 1996 (effectiveness of the intervention is reported in Tinetti et al 1994): Aged 70 years with 1 risk factor for falling (postural hypotension; use of sedatives; 4 medications; impairment in arm or leg strength or range of motion, balance, gait, transfer skills; environmental hazards) 1 year Robertson et al 2001a (effectiveness of the intervention is reported in Campbell et al 1997 and Campbell et al 1999a): Women from 17 general practices, mean age 84.1 (SD 3.3) years Up to 2 years

Multifactorial targeted intervention (behavioural instructions, exercise programmes, adjustment to medications, home safety) delivered by physician and at home by nurse and physiotherapist (n = 148) vs home visits by social work student (n = 140) 3 months after the baseline assessment, extended if health problems had interfered with ability to exercise Specific set of muscle strengthening and balance retraining exercises individually prescribed at home* by physiotherapist during 4 visits plus monthly phone calls (n = 116) vs social visits and usual care (n = 117) Up to 2 years

Intervention costs (developmental and training costs, recruitment costs, overheads, equipment, and staff related costs) Health care use (hospitalisation and emergency department, outpatient, home care, skilled nursing facilities)

Mean $905 (range $588 to $1,346)

Mean total costs $8,310 intervention group and $10,439 control group

$2,668 per fall prevented (using mean costs) for intervention costs only <$0 per fall prevented (using mean costs) for total health care costs <$0 per medical fall prevented (using mean costs) for total health care costs

Societal New Zealand dollars 1995 prices During participation in trial

Intervention costs (recruitment, programme delivery, overheads) Healthcare costs resulting from falls during trial (actual costs of hospital admissions and outpatient services, estimates of general practice and other costs) Total healthcare resource use during trial (actual costs of hospital admissions and outpatient services)

In research setting: $173 per person in year 1 $22 per person in year 2

No difference between the 2 groups for healthcare costs resulting from falls or for total healthcare costs 27% of hospital admission costs resulted from falls during trial

For 1 year: $314 per fall prevented (programme implementation costs only) For 2 years: $265 per fall prevented (programme implementation costs only)

63

Article, study sample, length of time falls monitored

Interventions and number being compared, length of intervention phase

Perspective, type of currency, year of costs, time period costs measured Health system New Zealand dollars 1998 prices During participation in trial

Cost items measured

Intervention costs

Healthcare service costs

Measures of cost effectiveness

Robertson et al 2001b From 17 general practices, community living, mean (SD) age 80.9 (4.2) years 1 year

Specific set of muscle strengthening and balance retraining exercises individually prescribed at home* by trained district nurse during 5 visits plus monthly phone calls, supervised by physiotherapist (n = 121) vs usual care (n = 119) 1 year Gradual withdrawal of psychotropic medication over 14 weeks (n=48) vs continuing to take psychotropic medication (n=45) (double blind)

Intervention costs (training course, recruitment, programme delivery, supervision of exercise instructor, overheads) Hospital admission costs resulting from fall injuries during trial (actual costs of hospital admissions)

In community health service setting: $432 per person for 1 year

5 hospital admissions due to fall injuries in control group, none in exercise group (cost savings of $47,818)

$1,803 per fall prevented (programme implementation costs only) $155 per fall prevented (programme implementation costs and hospital admission cost savings)

Robertson 2001c (effectiveness of the intervention is reported in Campbell et al 1999b): Men and women 65 years currently taking psychotropic medication, mean age 74.7 (SD 7.2) years 44 weeks

Health system New Zealand dollars 1996 prices During participation in trial

Intervention costs (recruitment, medication preparation and delivery, overheads)

In research setting: $258 per person for 44 weeks

$538 per fall prevented (programme implementation costs only)

64

Article, study sample, length of time falls monitored

Interventions and number being compared, length of intervention phase

Perspective, type of currency, year of costs, time period costs measured NHS, personal social services Pounds sterling 2004 prices During participation in trial

Cost items measured

Intervention costs

Healthcare service costs

Measures of cost effectiveness

Sach et al 2007 (effectiveness of the intervention is reported in Harwood et al 2005): Women >70 years with bilateral cataracts, mean age 84.1 (SD 3.3) years 1 year

Expedited (approximately 4 weeks) first cataract surgery (n = 148) vs control (routine, 12 months wait) (n = 140) Surgery and routine post surgery care

Secondary health care (cataract operation, bed days, outpatient, A&E, lower and upper limb fractures), primary health care (GP visits, practice/district nurse visits), personal social services (home care, day care centre, residential and nursing home care, meals on wheels, special equipment), patient and carers costs (home care, time costs) Hospitalisation, other health care costs provided in an institutional setting (eg outpatients), other health care costs provided in the home (eg home nursing), informal care costs (eg personal care provided by a relative or friend and help around the home), home modification costs, occupational therapist (intervention costs) in subsample of 103 in intervention group and 109 in the control group (last 212 recruited into trial)

Mean (SD) cataract operation 672 (0)

Total costs intervention group mean 2,004 95%CI 1,363 to 2,833 less than control group

4,390 per fall prevented (excluding carer costs) 3,983 per fall prevented (with carer costs included)

Salkeld et al 2000 (effectiveness of the intervention is reported in Cumming et al 1999): Recruited primarily before discharge from selected hospital wards, mean age 77 years 1 year

Home visit by experienced occupational therapist, environmental hazard assessment, facilitation for necessary modifications (n = 264) vs routine care (n = 266) 1 home visit, follow up telephone call 2 weeks later

Societal Australian dollars 1998 prices 1 year from trial entry

Mean $223 in intervention group, $15 in control group (home modification and occupational therapist intervention costs)

Mean total cost $10,084 in intervention group, $8,279 in control group (NS difference in median costs)

Average cost per fall prevented $4,986 (all n = 527 participants) For participants reporting a fall in the previous year average cost per fall prevented $3,980 (n = 203 participants) For participants reporting a fall in the previous year with outliers removed average cost per fall prevented <$0

HMO, health maintenance organisation. *Same home exercise programme (Otago Exercise Programme) as in Campbell et al 1997.

65

Table 2 Results from studies reporting costs of intervention and healthcare resource use long term care
Article, study sample, length of time falls monitored Interventions and number being compared, length of intervention phase Perspective, type of currency, year of costs, time period costs measured Health system US dollars Participants recruited 1992 4 months from study entry Cost items measured Intervention costs Healthcare service costs Measures of cost effectiveness

Mulrow et al 1994 Residents ( 3 months) from 9 nursing homes; dependent in 2 activities of daily living; mean age intervention group 79.7 (SD 8.5) years, control group 81.4 (SD 7.9) years 4 months

One on one sessions with physical therapist (n = 97) vs friendly visits (n = 97) 4 months

Intervention charges (wages and fringe benefits for personnel time, travel expenses, equipment based on annual depreciation, overhead costs) Nursing home, hospitalisation, physician and other health professional visits, emergency department visits, procedures, and medication charges (estimated from reimbursement fees, reference prices, prevailing allowable charges) Costs of treatment and tests related to episode (dermatological, respiratory, gastrointestinal, musculoskeletal, psychiatric, genitourinary, nutrition, cardiovascular, endocrine, neurological) using Current Procedural Terminology Center and Medicare allowable cost reimbursement at a rate of 80%

Mean charge per intervention participant $1,220 (95% CI $412 to $1,832) Mean charge per control participant $189 (95% CI $80 to $298)

Mean per participant (excluding intervention costs) $11,398; 95% CI $10 ,929 to $11,849, no difference between groups

Schnelle et al 2003 Long stay residents with incontinence in 4 nursing homes 8 months

Low-intensity exercise and incontinence care 5 days a week, every 2 hours, between 8 am and 4 pm (n = 92) vs usual care (n = 98) 8 months

Health system US dollars Not stated During 6 month baseline and 8 month intervention period

Average cost per resident per week to evaluate and treat the selected acute conditions $30.98 for intervention and $36.81 for control group

66

Article, study sample, length of time falls monitored

Interventions and number being compared, length of intervention phase

Perspective, type of currency, year of costs, time period costs measured NHS Pounds sterling Recruitment April 2002 to June 2003 28 days at end of study

Cost items measured

Intervention costs

Healthcare service costs

Measures of cost effectiveness

Zermansky et al 2006 Residents in 65 care homes aged 65 years on 1 repeat medicines 6 months

Clinical medicine review of GP record, consultation with patient and carer, recommendations by pharmacist (n =315 of 331 randomised to intervention group) vs usual care (n = 62 of 330 had a review by the GP in 6 months) Clinical medicine review by pharmacist was within 28 days of randomisation

Cost of 28 days of repeat medicines per participant at end date (using net ingredient cost NHS drug price not including dispensing cost)

Mean (SD) drug cost per patient for 28 days 42.24 (38.33) intervention group and 42.95 (41.01) control group

NHS, National Health Service

67

Tables of ACC claims and costs

68

All claims for falls for claimants aged 55 or over registered between July 2005 and June 2007 In residential care 2006-07/2007-06 Total Number of Cost to date Number of claims Cost to date claims 250 $641,673 491 $894,242 230 $429,502 459 $724,644 229 $129,208 541 $586,113 485 $354,675 942 $673,892 1,094 $861,087 2,252 $2,312,976 1,966 $1,803,583 4,090 $4,374,576 2,522 $2,560,891 4,994 $5,549,380 1,776 $1,687,101 3,658 $3,741,428 670 $569,556 1,346 $1,154,667 9,222 $9,037,277 18,773 $20,011,918

2005-07/2006-06 Number of claims Cost to date Aged 55 - 59 Aged 60 - 64 Aged 65 - 79 Aged 70 - 74 Aged 75 - 79 Aged 80 - 84 Aged 85 - 89 Aged 90 - 94 Aged 95+ Total 241 229 312 457 1,158 2,124 2,472 1,882 676 9,551 $252,569 $295,142 $456,905 $319,217 $1,451,889 $2,570,992 $2,988,489 $2,054,327 $585,111 $10,974,641

Aged 55 - 59 Aged 60 - 64 Aged 65 - 79 Aged 70 - 74 Aged 75 - 79 Aged 80 - 84 Aged 85 - 89 Aged 90 - 94 Aged 95+ Total

Community dwelling 2005-07/2006-06 2006-07/2007-06 Total Number of claims Cost to date Number of claims Cost to date Number of claims Cost to date 30,843 $37,073,392 29,238 $23,483,511 60,081 $60,556,904 23,107 $27,416,196 23,107 $18,747,358 46,214 $46,163,554 18,429 $14,112,332 19,065 $11,849,294 37,494 $25,961,626 15,839 $11,128,497 15,933 $8,861,728 31,772 $19,990,225 16,690 $11,651,415 17,054 $11,234,117 33,744 $22,885,531 14,302 $11,867,759 15,446 $11,762,500 29,748 $23,630,259 8,703 $8,843,892 9,568 $9,380,640 18,271 $18,224,532 3,197 $3,783,235 3,754 $4,807,449 6,951 $8,590,684 687 $939,776 753 $923,266 1,440 $1,863,042 131,797 $126,816,494 133,918 $101,049,862 265,715 $227,866,357

All claims for falls for claimants aged 55 or over registered between July 2005 and June 2007 2005-07/2006-06 Number of Cost to date claims 118,525 $113,326,864 5,069 $5,411,726 2,317 $1,800,357 2,831 $2,318,550 190 $109,966 11,187 1,221 8 141,348 $13,713,637 $1,108,959 $1,077 $137,791,136 2006-07/2007-06 Total Number of Cost to date Number of claims Cost to date claims 119,782 $88,181,056 238,307 $201,507,919 5,451 $4,720,985 10,520 $10,132,711 2,322 $1,385,249 4,639 $3,185,606 3,137 $1,599,722 5,968 $3,918,272 219 $179,715 409 $289,681 10,002 2,223 4 143,140 $11,601,939 $2,417,104 $1,369 $110,087,139 21,189 3,444 12 284,488 $25,315,576 $3,526,063 $2,447 $247,878,275

European Maori Pacific Peoples Asian Middle Eastern/Latin American/African Other Ethnicity Residual Categories Total

69

2005-07/2006-06 Number of claims Cost to date Maori or Pacific Peoples Aged 55 - 59 Aged 60 - 64 Aged 65 - 79 Aged 70 - 74 Aged 75 - 79 Aged 80 - 84 Aged 85 - 89 Aged 90 - 94 Aged 95+ Total Aged 55 - 59 Aged 60 - 64 Aged 65 - 79 Aged 70 - 74 Aged 75 - 79 Aged 80 - 84 Aged 85 - 89 Aged 90 - 94 Aged 95+ Total 2,446 1,644 1,274 854 615 334 152 48 19 7,386 28,638 21,692 17,467 15,442 17,233 16,092 11,023 5,031 1,344 133,962 $3,050,400 $1,773,526 $842,403 $428,483 $361,182 $535,395 $184,878 $24,469 $11,348 $7,212,082 $34,275,562 $25,937,812 $13,726,834 $11,019,231 $12,742,122 $13,903,357 $11,647,504 $5,813,094 $1,513,539 $130,579,054

2006-07/2007-06 Number of claims Cost to date 2,475 1,733 1,246 995 672 398 160 72 22 7,773 27,013 21,604 18,048 15,423 17,476 17,014 11,930 5,458 1,401 135,367 $2,131,293 $1,255,702 $666,557 $443,346 $1,241,350 $201,791 $111,073 $41,705 $13,416 $6,106,234 $21,993,891 $17,921,158 $11,311,945 $8,773,057 $10,853,854 $13,364,292 $11,830,458 $6,452,845 $1,479,406 $103,980,905

Total Number of claims Cost to date 4,921 3,377 2,520 1,849 1,287 732 312 120 41 15,159 55,651 43,296 35,515 30,865 34,709 33,106 22,953 10,489 2,745 269,329 $5,181,693 $3,029,228 $1,508,960 $871,829 $1,602,531 $737,186 $295,951 $66,174 $24,764 $13,318,317 $56,269,452 $43,858,970 $25,038,779 $19,792,288 $23,595,976 $27,267,649 $23,477,961 $12,265,938 $2,992,945 $234,559,959

Other ethnicities

Data as at 06 Aug 2007

70

All claims for falls for claimants aged 55 or over between July 2005 and June 2007 In residential care 2006-07/2007-06 Total Number of Cost to date Number of claims claims 550 $333,317 1,068 584 $173,880 1,175 249 $72,071 493 26 $7,815 61 82 $14,042 165 84 $802,132 187 325 $230,453 655 1,012 $585,500 2,031 193 $86,014 408 484 $291,252 983 132 $38,815 317 531 $173,940 1,039 609 $258,867 1,328 137 $266,218 261 2,117 $4,647,355 4,153 649 $369,208 1,283 711 $325,085 1,640 204 $135,593 422 97 $22,952 207 51 $8,179 119 18 $11,172 43 50 $8,167 90 327 $175,248 645 9,222 $9,037,277 18,773

Head (Except Face) Face Eye Ear Nose Neck/ Back of Head Vertebrae Shoulder (Including Clavicle/ Blade) Upper and Lower Arm Elbow Hand/ Wrist Finger/ Thumb Chest Back/Spine Abdomen/ Pelvis Hip/ Upper Leg/ Thigh Knee Lower Leg Ankle Foot Toe Internal Organ Multiple Locations Unknown Total

2005-07/2006-06 Number of Cost to date claims 518 $348,068 591 $195,159 244 $246,128 35 $11,845 83 $16,008 103 $636,255 330 $262,233 1,019 $796,731 215 $127,105 499 $320,881 185 $119,867 508 $225,094 719 $303,635 124 $228,349 2,036 $5,612,468 634 $289,767 929 $454,537 218 $221,308 110 $51,730 68 $9,432 25 $19,312 40 $66,871 318 $411,858 9,551 $10,974,641

Cost to date $681,384 $369,039 $318,200 $19,659 $30,050 $1,438,386 $492,686 $1,382,231 $213,119 $612,133 $158,682 $399,035 $562,503 $494,567 $10,259,823 $658,975 $779,622 $356,901 $74,682 $17,611 $30,484 $75,038 $587,106 $20,011,918

71

Head (Except Face) Face Eye Ear Nose Neck/ Back of Head Vertebrae Shoulder (Including Clavicle/ Blade) Upper and Lower Arm Elbow Hand/ Wrist Finger/ Thumb Chest Back/Spine Abdomen/ Pelvis Hip/ Upper Leg/ Thigh Knee Lower Leg Ankle Foot Toe Internal Organ Multiple Locations Unknown Total

Community dwelling 2005-07/2006-06 2006-07/2007-06 Total Number of claims Cost to date Number of claims Cost to date Number of claims Cost to date 2,989 $2,840,866 3,328 $2,963,882 6,317 $5,804,747 4,646 $2,458,760 4,899 $2,117,289 9,545 $4,576,049 2,980 $844,582 2,705 $846,017 5,685 $1,690,598 297 $384,038 310 $187,212 607 $571,250 677 $351,405 861 $302,800 1,538 $654,205 4,082 $3,277,793 4,280 $2,934,598 8,362 $6,212,391 8,527 $18,052,066 9,152 $11,384,459 17,679 $29,436,525 9,501 $9,830,243 10,275 $8,837,230 19,776 $18,667,473 2,612 $1,940,103 2,677 $1,809,272 5,289 $3,749,375 7,834 $5,950,083 8,144 $5,119,320 15,978 $11,069,403 6,263 $2,839,218 5,831 $2,310,989 12,094 $5,150,207 6,709 $2,739,573 7,262 $2,411,986 13,971 $5,151,558 14,212 $11,820,502 14,516 $8,246,770 28,728 $20,067,272 1,302 $1,716,373 1,389 $1,722,570 2,691 $3,438,943 8,859 $18,632,507 9,578 $18,659,008 18,437 $37,291,515 15,207 $12,465,694 15,025 $9,983,420 30,232 $22,449,115 13,292 $8,524,821 12,608 $7,267,908 25,900 $15,792,729 8,121 $8,686,772 8,022 $6,665,995 16,143 $15,352,768 4,322 $2,991,466 4,013 $2,161,904 8,335 $5,153,370 2,281 $486,228 2,226 $485,447 4,507 $971,675 399 $753,824 345 $177,454 744 $931,278 542 $345,209 622 $295,470 1,164 $640,679 6,143 $8,884,370 5,850 $4,158,862 11,993 $13,043,232 131,797 $126,816,494 133,918 $101,049,862 265,715 $227,866,357

72

All claims for over 65s

Table of rhflag by ageyr Rest Home Flag 0 1 Total

ageyr Aged 65 - 79 464533 15646 480179

Total Aged 80+ 202847 49059 251906 667380 64705 732085

Data as at 3 March 2008 All claims for falls between July 2004 and June 2007 In rest home Number % 7,276 24,702 31,978 Data as at 3 March 2008 Entitlement claims for falls between July 2004 and June 2007 In rest home Number % 770 3,168 3,938 Data as at 3 March 2008 Claimants who made a claim between July 2004 and June 2007 In rest home Number % 3,598 11,750 15,348 Data as at 3 March 2008 73 Not in rest home Number % 98,104 96.46 42,178 78.21 140,282 90.14 Not in rest home Number % 14,204 94.86 10,276 76.44 24,480 86.14 Not in rest home Number % 154,106 95.49 80,023 76.41 234,129 87.98

Aged 65 - 79 Aged 80+ Total

4.51 23.59 12.02

Aged 65 - 79 Aged 80+ Total

5.14 23.56 13.86

Aged 65 - 79 Aged 80+ Total

3.54 21.79 9.86

Proportion of claims from people in resthomes between July 2004 and June 2007 In rest home Number % 885 833 811 856 896 936 840 842 787 842 873 842 778 908 938 844 822 988 880 904 875 879 1,000 987 954 1,033 922 857 957 874 792 872 902 952 866 951 31,978 Data as at 3 March 2008 74 Not in rest home Number % 6,562 6,264 6,817 4,935 6,229 6,925 5,709 6,566 6,382 6,462 6,822 7,302 6,056 7,240 7,778 4,785 5,613 6,467 5,659 6,542 6,679 6,441 7,896 8,424 6,046 7,334 7,872 6,136 7,289 6,829 5,123 5,957 7,085 5,169 6,242 6,492 234,129

Apr-05 Apr-06 Apr-07 Aug-04 Aug-05 Aug-06 Dec-04 Dec-05 Dec-06 Feb-05 Feb-06 Feb-07 Jan-05 Jan-06 Jan-07 Jul-04 Jul-05 Jul-06 Jun-05 Jun-06 Jun-07 Mar-05 Mar-06 Mar-07 May-05 May-06 May-07 Nov-04 Nov-05 Nov-06 Oct-04 Oct-05 Oct-06 Sep-04 Sep-05 Sep-06 Total

11.88 11.74 10.63 14.78 12.58 11.91 12.83 11.37 10.98 11.53 11.35 10.34 11.38 11.14 10.76 14.99 12.77 13.25 13.46 12.14 11.58 12.01 11.24 10.49 13.63 12.35 10.48 12.26 11.61 11.35 13.39 12.77 11.29 15.55 12.18 12.78 12.02

88.12 88.26 89.37 85.22 87.42 88.09 87.17 88.63 89.02 88.47 88.65 89.66 88.62 88.86 89.24 85.01 87.23 86.75 86.54 87.86 88.42 87.99 88.76 89.51 86.37 87.65 89.52 87.74 88.39 88.65 86.61 87.23 88.71 84.45 87.82 87.22 87.98

All Claims for those aged 65 years or over in 2005/06 to 2006/07 In rest home Not in rest home Number of claims Cost to date Number of claims Cost to date Number of claims Total Median Total Median 150 $78,558 $125 8,876 $8,099,311 $165 9,026 141 $222,175 $189 8,405 $7,130,819 $169 8,546 140 $89,008 $149 7,566 $5,966,819 $163 7,706 147 $195,330 $145 7,368 $4,891,640 $165 7,515 175 $296,212 $129 7,071 $5,107,573 $163 7,246 176 $102,164 $123 6,883 $4,405,357 $163 7,059 222 $133,596 $116 6,550 $4,332,744 $169 6,772 253 $113,870 $139 6,431 $4,865,077 $168 6,684 269 $234,099 $141 6,583 $4,524,618 $156 6,852 323 $301,993 $160 6,653 $4,860,595 $168 6,976 369 $371,174 $152 6,926 $4,437,016 $164 7,295 456 $390,106 $173 7,164 $5,884,484 $170 7,620 602 $609,929 $137 7,086 $5,041,274 $171 7,688 653 $659,141 $126 7,055 $5,680,053 $174 7,708 716 $693,784 $132 6,725 $4,704,037 $174 7,441 832 $804,495 $148 6,759 $5,154,286 $178 7,591 934 $917,437 $146 6,459 $5,199,823 $180 7,393 1,020 $979,730 $178 6,213 $5,232,551 $186 7,233 1,104 $1,439,365 $179 5,725 $4,861,841 $186 6,829 1,125 $1,239,721 $166 5,433 $4,895,512 $189 6,558 1,255 $1,448,648 $189 5,086 $4,795,461 $197 6,341 1,292 $1,528,507 $188 4,280 $4,039,710 $203 5,572 1,194 $1,191,551 $143 3,527 $3,785,557 $201 4,721 1,165 $1,207,689 $185 3,039 $3,261,320 $217 4,204 1,098 $1,314,318 $202 2,674 $3,452,603 $218 3,772 1,072 $1,060,255 $177 2,124 $2,414,699 $235 3,196 979 $1,096,812 $152 1,821 $2,213,792 $223 2,800 821 $884,638 $168 1,345 $1,643,751 $257 2,166 785 $815,630 $178 1,033 $1,486,972 $259 1,818 618 $523,120 $137 729 $1,091,121 $274 1,347 447 $428,869 $152 496 $605,255 $248 943 349 $337,212 $171 364 $615,916 $334 713 243 $180,766 $186 200 $265,842 $271 443 180 $115,784 $149 126 $168,084 $276 306 127 $72,631 $145 92 $58,654 $258 219 79 $63,383 $190 71 $58,865 $203 150 41 $71,019 $174 21 $7,537 $127 62 30 $33,194 $185 23 $22,446 $305 53 16 $8,663 $215 20 $14,448 $259 36 8 $1,441 $77 3 $1,117 $371 11 Total Cost to date Total Median $8,177,868 $164 $7,352,994 $170 $6,055,827 $163 $5,086,970 $165 $5,403,784 $162 $4,507,521 $161 $4,466,340 $168 $4,978,947 $167 $4,758,718 $155 $5,162,589 $167 $4,808,190 $162 $6,274,590 $170 $5,651,202 $169 $6,339,194 $172 $5,397,821 $171 $5,958,781 $176 $6,117,260 $177 $6,212,281 $186 $6,301,206 $184 $6,135,233 $187 $6,244,109 $195 $5,568,217 $199 $4,977,109 $186 $4,469,009 $209 $4,766,921 $214 $3,474,953 $214 $3,310,604 $201 $2,528,389 $230 $2,302,602 $223 $1,614,242 $194 $1,034,124 $200 $953,128 $237 $446,609 $220 $283,868 $198 $131,285 $234 $122,248 $203 $78,555 $172 $55,639 $200 $23,111 $238 $2,557 $193 75

65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104

105 106 107 108 112 134 Total

6 5 1 1 21,619

$3,319 $402 $40 $29 $22,259,806

$89 $88 $40 $29 $163

14 15 8 3 1 165,046

$2,868 $2,177 $1,028 $950 $1,830 $135,287,430

$99 $127 $142 $363 $1,830 $176

20 20 9 3 1 1 186,665

$6,187 $2,580 $1,068 $950 $29 $1,830 $157,547,236

$91 $110 $109 $363 $29 $1,830 $175

Data as at 3 March 2008 All Claims for those aged 65 years or over 2004-07/2005-06 2005-07/2006-06 2006-07/2007-06 Number of claims Cost to date Number of claims Cost to date Number of claims Cost to date Total Median Total Median Total Median 10,359 $12,713,029 $155 10,806 $11,608,404 $164 10,813 $10,651,403 $162 Claims for falls 9,292 $5,062,672 $122 7,575 $4,514,693 $156 5,909 $4,321,334 $151 Other claims 19,651 $17,775,701 $138 18,381 $16,123,096 $160 16,722 $14,972,737 $158 Total 69,083 $55,103,277 $144 79,994 $63,472,723 $169 85,052 $71,814,707 $183 Claims for falls 95,189 $64,312,752 $131 96,979 $69,808,210 $152 105,102 $76,474,244 $166 Other claims 164,272$119,416,029 $136 176,973$133,280,933 $157 190,154$148,288,951 $174 Total 79,442 $67,816,306 $145 90,800 $75,081,126 $169 95,865 $82,466,110 $181 Claims for falls 104,481 $69,375,424 $131 104,554 $74,322,902 $152 111,011 $80,795,578 $165 Other claims 183,923$137,191,730 $136 195,354$149,404,029 $158 206,876$163,261,688 $173 Total Data as at 3 March 2008 Entitlement Claims for those aged 65 years or over 2004-07/2005-06 2005-07/2006-06 2006-07/2007-06 Number of claims Cost to date Number of claims Cost to date Number of claims Cost to date Total Median Total Median Total Median 1,445 $10,665,622 $5,442 1,298 $9,365,199 $5,361 1,195 $8,371,463 $5,607 Claims for falls 516 $3,050,679 $3,601 382 $2,707,851 $4,064 318 $2,868,709 $5,086 Other claims 1,961 $13,716,301 $5,031 1,680 $12,073,050 $5,080 1,513 $11,240,173 $5,324 Total 7,698 $40,856,898 $2,608 8,155 $44,483,523 $2,788 8,627 $50,069,501 $2,965 Claims for falls 6,817 $44,165,295 $4,258 6,473 $46,564,586 $5,254 7,202 $48,212,025 $4,707 Other claims 14,515 $85,022,193 $3,239 14,628 $91,048,109 $3,696 15,829 $98,281,526 $3,751 Total 9,143 $51,522,519 $2,945 9,453 $53,848,723 $3,068 9,822 $58,440,964 $3,269 Claims for falls 7,333 $47,215,975 $4,209 6,855 $49,272,437 $5,183 7,520 $51,080,735 $4,729 Other claims 16,476 $98,738,494 $3,493 16,308$103,121,160 $3,844 17,342$109,521,699 $3,896 Total Data as at 3 March 2008 76

In rest home

Not in rest home

Total

In rest home

Not in rest home

Total

Entitlement Claims for those aged 65 years or over 2004-07/2005-06 2005-07/2006-06 2006-07/2007-06 Number of claims Cost to date Number of claims Cost to date Number of claims Cost to date Total Median Total Median Total Median 1,445 $10,665,622 $5,442 1,298 $9,365,199 $5,361 1,195 $8,371,463 $5,607 Claims for falls 516 $3,050,679 $3,601 382 $2,707,851 $4,064 318 $2,868,709 $5,086 Other claims 1,961 $13,716,301 $5,031 1,680 $12,073,050 $5,080 1,513 $11,240,173 $5,324 Total 7,698 $40,856,898 $2,608 8,155 $44,483,523 $2,788 8,627 $50,069,501 $2,965 Claims for falls 6,817 $44,165,295 $4,258 6,473 $46,564,586 $5,254 7,202 $48,212,025 $4,707 Other claims 14,515 $85,022,193 $3,239 14,628 $91,048,109 $3,696 15,829 $98,281,526 $3,751 Total 9,143 $51,522,519 $2,945 9,453 $53,848,723 $3,068 9,822 $58,440,964 $3,269 Claims for falls 7,333 $47,215,975 $4,209 6,855 $49,272,437 $5,183 7,520 $51,080,735 $4,729 Other claims 16,476 $98,738,494 $3,493 16,308$103,121,160 $3,844 17,342$109,521,699 $3,896 Total Data as at 3 March 2008 All Claims for those aged 65-79 years 2004-07/2005-06 2005-07/2006-06 2006-07/2007-06 Number of claims Cost to date Number of claims Cost to date Number of claims Cost to date Total Median Total Median Total Median 2,484 $3,242,863 $146 2,409 $2,583,523 $141 2,383 $1,907,616 $143 Claims for falls 2,444 $1,502,493 $122 1,993 $1,324,763 $164 1,594 $1,616,693 $170 Other claims 4,928 $4,745,356 $136 4,402 $3,908,286 $150 3,977 $3,524,310 $155 Total 46,764 $34,154,430 $136 52,769 $38,200,044 $159 54,573 $41,731,371 $174 Claims for falls 69,222 $50,774,113 $127 71,111 $54,442,305 $145 77,277 $59,246,552 $159 Other claims 115,986 $84,928,543 $131 123,880 $92,642,349 $151 131,850$100,977,923 $166 Total 49,248 $37,397,293 $137 55,178 $40,783,567 $158 56,956 $43,638,988 $173 Claims for falls 71,666 $52,276,606 $127 73,104 $55,767,068 $146 78,871 $60,863,246 $160 Other claims 120,914 $89,673,899 $131 128,282 $96,550,635 $151 135,827$104,502,233 $165 Total Data as at 3 March 2008 Entitlement Claims for those aged 65 79 years 2004-07/2005-06 2005-07/2006-06 2006-07/2007-06 Number of claims Cost to date Number of claims Cost to date Number of claims Cost to date Total Median Total Median Total Median 316 $2,757,784 $5,522 238 $2,099,404 $5,247 216 $1,412,607 $4,433 Claims for falls 144 $961,064 $3,123 110 $869,951 $5,238 94 $1,196,728 $4,826 Other claims 460 $3,718,848 $4,800 348 $2,969,355 $5,238 310 $2,609,335 $4,446 Total 77

In rest home

Not in rest home

Total

In rest home

Not in rest home

Total

In rest home

Not in rest home

Total

Claims for falls Other claims Total Claims for falls Other claims Total

4,606 5,125 9,731 4,922 5,269 10,191

$24,615,716 $36,617,765 $61,233,481 $27,373,500 $37,578,829 $64,952,329

$2,518 $5,099 $3,528 $2,623 $5,000 $3,563 Data as at 3 March 2008

4,744 4,980 9,724 4,982 5,090 10,072

$25,768,385 $37,981,406 $63,749,791 $27,867,789 $38,851,357 $66,719,146

$2,739 $5,690 $3,934 $2,804 $5,671 $3,972

4,854 5,509 10,363 5,070 5,603 10,673

$27,844,929 $38,882,960 $66,727,889 $29,257,535 $40,079,689 $69,337,224

$2,825 $5,350 $3,891 $2,881 $5,321 $3,909

All Claims for those aged 80 years or over 2004-07/2005-06 2005-07/2006-06 2006-07/2007-06 Number of claims Cost to date Number of claims Cost to date Number of claims Cost to date Total Median Total Median Total Median 7,875 $9,470,166 $156 8,397 $9,024,881 $171 8,430 $8,743,786 $169 Claims for falls 6,848 $3,560,179 $122 5,582 $3,189,929 $154 4,315 $2,704,641 $143 Other claims 14,723 $13,030,345 $139 13,979 $12,214,810 $164 12,745 $11,448,427 $158 Total 22,319 $20,948,847 $165 27,225 $25,272,679 $188 30,479 $30,083,336 $205 Claims for falls 25,967 $13,538,640 $143 25,868 $15,365,905 $172 27,825 $17,227,692 $183 Other claims 48,286 $34,487,487 $152 53,093 $40,638,584 $180 58,304 $47,311,028 $191 Total 30,194 $30,419,013 $163 35,622 $34,297,560 $185 38,909 $38,827,122 $196 Claims for falls 32,815 $17,098,818 $139 31,450 $18,555,834 $170 32,140 $19,932,332 $177 Other claims 63,009 $47,517,831 $150 67,072 $52,853,394 $178 71,049 $58,759,455 $187 Total Data as at 3 March 2008 Entitlement Claims for those aged 80 years or over 2004-07/2005-06 2005-07/2006-06 2006-07/2007-06 Number of claims Cost to date Number of claims Cost to date Number of claims Cost to date Total Median Total Median Total Median 1,129 $7,907,838 $5,437 1,060 $7,265,795 $5,406 979 $6,958,856 $5,794 Claims for falls 372 $2,089,616 $3,843 272 $1,837,900 $3,685 224 $1,671,981 $5,147 Other claims 1,501 $9,997,454 $5,061 1,332 $9,103,695 $5,024 1,203 $8,630,838 $5,612 Total 3,092 $16,241,181 $2,895 3,411 $18,715,139 $2,896 3,773 $22,224,572 $3,274 Claims for falls 1,692 $7,547,531 $2,321 1,493 $8,583,180 $3,360 1,693 $9,329,065 $3,506 Other claims 4,784 $23,788,712 $2,651 4,904 $27,298,318 $3,068 5,466 $31,553,637 $3,383 Total 4,221 $24,149,019 $3,695 4,471 $25,980,934 $3,591 4,752 $29,183,429 $3,908 Claims for falls 2,064 $9,637,146 $2,503 1,765 $10,421,080 $3,392 1,917 $11,001,046 $3,685 Other claims 6,285 $33,786,165 $3,250 6,236 $36,402,014 $3,533 6,669 $40,184,475 $3,882 Total Data as at 3 March 2008 78

In rest home

Not in rest home

Total

In rest home

Not in rest home

Total

All Claims registered in 2006/07 for those aged 80 years or over Number of claims In rest home Claims for falls Other claims Total Claims for falls Other claims Total Claims for falls Other claims Total 8,430 4,315 12,745 30,479 27,825 58,304 38,909 32,140 71,049 Total $8,743,786 $2,704,641 $11,448,427 $30,083,336 $17,227,692 $47,311,028 $38,827,122 $19,932,332 $58,759,455 P25 $58 $58 $58 $65 $64 $64 $62 $64 $62 Cost to date Median P75 $169 $534 $143 $516 $158 $521 $205 $579 $183 $495 $191 $529 $196 $559 $177 $501 $187 $522

P90 $2,227 $803 $1,325 $1,661 $1,001 $1,291 $1,721 $976 $1,294

P95 $7,256 $1,752 $5,741 $5,423 $2,097 $3,540 $6,029 $2,071 $3,989

Not in rest home

Total

Data as at 3 March 2008 Entitlement Claims registered in 2006/07 for those aged 80 years or over Number of claims In rest home Claims for falls Other claims Total Claims for falls Other claims Total Claims for falls Other claims Total 979 224 1,203 3,773 1,693 5,466 4,752 1,917 6,669 Total $6,958,856 $1,671,981 $8,630,838 $22,224,572 $9,329,065 $31,553,637 $29,183,429 $11,001,046 $40,184,475 P25 $2,641 $1,786 $2,419 $1,193 $1,151 $1,180 $1,312 $1,250 $1,294 Cost to date Median P75 $5,794 $9,578 $5,147 $8,894 $5,612 $9,453 $3,274 $8,611 $3,506 $7,282 $3,383 $7,711 $3,908 $8,900 $3,685 $7,350 $3,882 $8,124

Not in rest home

Total

P90 $14,033 $14,162 $14,033 $13,954 $10,338 $13,116 $13,956 $10,761 $13,370

P95 $17,482 $19,072 $17,482 $18,637 $15,440 $17,786 $18,472 $15,706 $17,786

Data as at 3 March 2008

79