Objectives To investigate the value for money of strategies to prevent falls in older adults living in the community.
Design Systematic review of peer reviewed journal articles reporting an economic evaluation of a falls prevention intervention as part of a randomised controlled trial or a controlled trial, or using an analytical model. MEDLINE, PUBMED, EMBASE and NHS EED databases were searched to identify cost-effectiveness, cost–utility and cost–benefit studies from 1945 through July 2008.
Main outcome measures The primary outcome measure was incremental cost-effectiveness, cost–utility and cost–benefit ratios in the reported currency and in pounds sterling at 2008 prices. The quality of the studies was assessed using two instruments: (1) an economic evaluation checklist developed by Drummond and colleagues and (2) the Quality of Health Economic Studies instrument.
Results Nine studies meeting our inclusion criteria included eight cost-effectiveness analyses, one cost–utility and one cost–benefit analysis. Three effective falls prevention strategies were cost saving in a subgroup of participants: (1) an individually customised multifactorial programme in those with four or more of the eight targeted fall risk factors, (2) the home-based Otago Exercise Programme in people ≥80 years and (3) a home safety programme in the subgroup with a previous fall. These three findings were from six studies that scored ≥75% on the Quality of Health Economic Studies instrument.
Conclusions Best value for money came from effective single factor interventions such as the Otago Exercise Programme which was cost saving in adults 80 years and older. This programme has broad applicability thus warranting warrants health policy decision-makers’ close scrutiny.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Physical activity provides a vast and well-documented range of health benefits.1 2 In a recent BJSM editorial, Church and Blair asked why, given the powerful, pleiotropic medical benefits of exercise, it is not prescribed more in clinical settings.3 4 Similarly, Sallis argued that “Exercise is medicine and physicians need to prescribe it!”5 A medical condition where there is both a substantial burden of disease6 7 and high-quality evidence that exercise improves patient outcomes is in falls prevention among seniors.8 Despite all of the foregoing elements, our clinical impression is that there is neither a universal adoption of exercise as a therapy by individual clinicians,9 10 nor any widespread concerted public health spending on exercise programmes to prevent falls in seniors. Before addressing whether this gap between what should happen and what does happen is due to economic factors (ie, the cost to deliver falls prevention interventions) we briefly update the reader on the burden of falls and current evidence to prevent them.
Cohort studies have consistently demonstrated that 30% of those aged 65 years and older experience at least one fall each year, and half of those fall recurrently.11 12 In the UK, falls by adults aged 60 years and older cost £981 million (at 2000 prices), and 59% of these costs were incurred by the National Health Service (NHS).13 The burden of disease is similar in other countries.14,–,16
Some interventions that aimed to prevent falls in seniors have proven successful. To aid analysis of the various interventions, it is useful to categorise interventions into (1) individually customised multifactorial interventions,17 (2) regimens targeting the same multiple factors to all participants and (3) single-factor interventions. The initial individually customised multifactorial intervention targeted up to eight different risk factors for falls depending on the participant’s individual risk profile; it reduced falls by 31% in 1 year.18 In multifactorial interventions, reduction in the risk of falling ranges from not being significantly different from usual care to 61% reduction.19 We distinguish this individually customised multifactorial approach from the strategy of providing multiple interventions at a community level where all individuals are exposed to a universal intervention irrespective of the person’s risk profile.20 The third category is single factor intervention—for example, strength and balance training, withdrawing use of psychotropic medication, home safety modification and cataract surgery.21 22 Overall, in four trials of community living seniors, a progressive home-based strength and balance retraining programme reduced both falls and injuries by 35%.8
Which of these approaches to falls prevention provide the best value for money? Given our ageing population, the large financial burden imposed by falls and the scarcity of healthcare system resources, economic evaluations are increasingly important to assist healthcare decision-makers in allocating resources. There are three main types of economic evaluations used to determine value for money when comparing falls prevention interventions with usual care or when comparing at least two different interventions: cost-effectiveness analysis, cost–utility analysis and cost–benefit analysis. The common feature of all these analyses is the comparison of monetary units between competing alternative interventions; however, they differ in the approach taken to measure the health benefits of the intervention. In cost-effectiveness analysis, the benefits are measured using a clinically relevant outcome such as life years gained or number of falls prevented. The primary outcome of a cost-effectiveness analysis is the incremental cost-effectiveness ratio (ICER). By definition, an ICER is the difference between the costs of providing the competing interventions divided by the difference in effectiveness (ie, the number of falls prevented) (equation 1).23
In cost–utility analyses, health benefits are measured by a quality-adjusted life year, and for cost–benefit analyses, in monetary units.
To our knowledge, no systematic review of economic evaluations has been published for falls prevention interventions. Therefore, to help physicians, other health professionals and health managers make better decisions, we systematically reviewed economic evaluations of falls prevention studies to answer the question “Which falls prevention strategy provides the best value for money?”
Literature search strategy
In accordance with Quality of Reporting of Meta-analyses (QUOROM)24 and Cochrane Collaboration guidelines,25 we searched MEDLINE, PUBMED, EMBASE and NHS EED databases to identify cost-effectiveness, cost–utility or cost–benefit studies based on falls prevention interventions published in the English language from 1945 through July 2008. We limited our search to the English language and to studies of people aged 60 years and older. Included in our search terms were keywords: fall prevention, economic evaluation, cost effectiveness, cost–utility and cost–benefit analysis.
Selection of studies
We selected peer-reviewed, published studies that included a comprehensive (full) cost-effectiveness, cost–benefit or cost–utility analysis, that is, they reported the costs and consequences of two alternatives and an incremental cost-effectiveness, cost–benefit or cost–utility ratio. We included only studies in community dwelling older adults aged 60 years and older. We excluded studies if they were set in a hospital or long-term care facility, were still in progress, or the intervention was aimed primarily at fracture prevention rather than falls prevention.
Abstraction of data
Two raters (JCD, MCA) independently extracted data from each study, and any discrepancies were discussed and reviewed by a third party (one of MCR, CAM). We noted the country, type of economic analysis (analytical model or evaluation within a clinical trial), journal, study sample, intervention and comparator evaluated, length of intervention phase, effectiveness in terms of falls reduction, perspectives, type of currency, year of currency, analytical time horizon, results of sensitivity analyses, discount rate, cost items measured, intervention costs and incremental cost-effectiveness, cost–benefit and cost–utility ratios.
Data synthesis and analysis
We emailed authors of the papers that are included in this systematic review for additional information that was not available from the publications themselves. Due to the heterogeneity of reported cost items, we were not able to complete a meta-analysis; therefore, we provide a qualitative description of outcomes.
Standardised cost outcomes
The studies reported costs in different currencies and from different years. To attenuate this variation, we express monetary values in two ways: (1) by year and currency as reported in the manuscript, and (2) converted to 2008 pounds sterling, by inflating the values to 2008 prices using the Consumer Price Index and then using Purchasing Power Parity values in 2008 to convert the currency to pounds sterling.26 27 The Consumer Price Index (CPI), specific to each country, is an indicator of consumer prices in that country. It is obtained by calculating, on a monthly basis, the cost of a fixed list of goods and services purchased by a “typical” consumer during a given month. The CPI is a widely used indicator of inflation and deflation. Purchasing power parity assumes there is equilibrium between exchange rates when the purchasing power is equal in the comparator countries. Therefore, the ratio of the comparator countries’ price levels for the goods and services of interest should equal the exchange rate between the comparator countries. We used this ratio in order to convert all monetary values into pounds sterling.
We assessed the quality of each study using (1) a checklist developed by Drummond and colleagues23 and (2) the Quality of Health Economic Studies instrument,28 29 which were designed to evaluate all common types of health economic analyses (eg, cost-effectiveness analysis, cost–utility analysis). The Quality of Health Economics Studies instrument was originally developed from a comprehensive literature search of all existing checklists and guidelines for economic evaluations. Weightings for each question were determined by conjoint analysis based on survey results from an international panel of health economists. The Quality of Health Economic Studies instrument comprises 16 criteria that cover presentation of study objectives, description of methods and comprehensive reporting of results. We excluded item 4, which was not applicable for all the included studies, so the weighted total for the 15 items was a score of 99. We considered studies with a score of 75% or greater as “good” quality.29
Two raters (JCD, MCA) independently evaluated each study, and any discrepancies were discussed and reviewed by a third party (one of MCR, CAM). One author (MCR) did not review or rate her own studies.30,–,33
Overview of studies identified
After a critical review of the 29 full text manuscripts, nine studies met our inclusion criteria (fig 1). One study reported a cost–benefit analysis, seven a cost-effectiveness analysis and one both a cost-effectiveness and a cost–utility analysis (table 1). Five of the cost-effectiveness analyses reported the incremental cost of delivering the intervention per fall prevented, and five reported a cost-effectiveness ratio which incorporated incremental healthcare costs related to falls during the study, or total healthcare costs (table 2). The studies differed in the particular cost items included (table 3) and the methods used for valuing these items. The cost–utility analysis used quality-adjusted life years (QALYs) gained as the measure of effectiveness. The cost–benefit analysis used dollars to quantify both effectiveness and costs.
There were also distinct variations across studies with respect to time horizon, interventions evaluated, methods used for sensitivity analyses and discounting rates. The comparators evaluated for all nine studies were similar and included social visits or usual care, which in five studies was assigned a monetary value of zero. Six studies scored 75% or more on the Quality of Health Economic Studies instrument (table 4). Two studies that fell just below 75% did not provide a comprehensive presentation of the economic model or a full explanation of all assumptions made.
Individually customised multifactorial interventions: one cost-effectiveness study
Assessment, exercise, behaviour modification, medication
In the only US cost-effectiveness analysis, the mean cost of delivering an individually targeted intervention was £880 (at 2008 prices) per participant.34 For all participants, the targeted intervention was cost saving compared with usual care and social visits when the incremental cost-effectiveness ratio was calculated using mean total healthcare costs or mean total cost for falls requiring medical care. In a subgroup analysis, the targeted intervention was cost saving for the 54% (156 of 288) of participants at high risk (defined as four or more of the eight specified risk factors for falls), whereas for participants at lower risk of falls (three or fewer risk factors), the incremental cost-effectiveness ratio (total healthcare costs per fall prevented) was £2696. The high-risk group represented 8% (156 of 1950) of those originally screened for eligibility for this trial; the cost of screening was incorporated in the reported cost-effectiveness ratios.
Multiple intervention at a community level
The “Stay on your Feet” intervention targeted all older adults within the North Coast of New South Wales, Australia with a falls-related knowledge, attitudes, behaviours, and risk factor awareness campaign.35 The authors performed a cost–benefit analysis using a different region and the whole state as the comparators. Although the programmes required a large budget, both comparisons yielded a net monetary benefit to cost ratio for the intervention of 20.6 to 1.
Single factor interventions: five cost-effectiveness studies and one cost–utility study
Strength and balance training intervention
Three cost-effectiveness analyses of a nurse- or physiotherapist-delivered home exercise programme compared the programme with usual care for a 1-year period.31,–,33 These two randomised controlled trials and one controlled trial in New Zealand recruited participants on the basis of age only. The mean cost of delivering the exercise programme for 1 year ranged from £173 in a research setting to £942 per fall prevented in a community healthcare setting. Importantly, the incremental cost-effectiveness ratio for the intervention in the study incorporating the costs of hospital admissions averted as a result of the intervention demonstrated that the Otago Exercise Programme was not cost saving for all participants but was cost saving for those aged 80 years and older.31
Three home safety interventions
Community dwelling women and men in New Zealand aged ≥75 who were at risk of falling because of severely impaired vision (acuity of 6/24 or worse) were randomised to receive a home safety assessment and modification intervention.30 The delivery of the home safety programme by an experienced occupational therapist cost £304 per fall prevented in 1 year.
An Australian study compared falls prevented and costs in participants recently discharged from hospital who had a home safety intervention delivered by an occupational therapist, or no intervention.36 The incremental cost-effectiveness ratio, which incorporated health, home and community care costs was £3040 per fall prevented. A sensitivity analysis with outliers removed demonstrated that the intervention was cost saving in the subgroup of participants (39%, 203 of 527) who had a fall in the previous year.
A decision-analytical model with assumptions for fall, injury and hospitalisation rates obtained from the published literature was used to assess the cost-effectiveness of a hypothetical home assessment and modification programme.37 The authors estimated that the intervention compared with usual care would cost £1052 per fall prevented over a 1-year period.37 Selective one-way sensitivity analyses indicated that the assumed probability of a fall and assumed effectiveness of the intervention were key variables that affected the magnitude of the incremental cost-effectiveness ratio.
Expedited cataract surgery
For the cost–utility analysis of expedited first eye cataract surgery, effectiveness was estimated from quality-adjusted life years (QALYs) determined from the EuroQol-5D (EQ-5D) scores at baseline and after 6 months.38 The cost–utility ratios of £38 482 per QALY from the NHS perspective, and £34 911 from the personal social services perspective for 1 year, were above the currently accepted willingness-to-pay value of £20 000. However, when the costs and QALYs were modelled over the participants’ expected lifetime, the incremental cost per QALY was within this limit at £14 197. The base-case 1-year incremental cost-effectiveness ratio incorporating primary and secondary healthcare, and social services, personal and carer costs was £4732 per fall prevented. The authors also provided cost-effectiveness acceptability curves.
There have been comprehensive economic evaluations published from only nine randomised controlled trials of falls prevention strategies in community-dwelling older adults. Some of the strategies included exercise, while others did not. There were considerable differences in economic evaluation methodologies used across studies (table 3). Before discussing the implications of our findings, we review the methodological issues that must be taken into account to interpret this systematic review.
In eight studies that reported incremental cost per fall prevented, effective interventions included strength and balance retraining, cataract surgery and home safety interventions. Of these, the studies testing strength and balance retraining, which had the highest quality assessment scores, prevented the greatest number of falls at the least cost. A multifactorial programme was cost saving in a narrow range of individuals with four or more of the eight specified risk factors for falls. The most favourable incremental and widely applicable cost-effectiveness ratio was £173 (at 2008 prices) per fall prevented for the Otago Exercise Programme, although this trial was in a research setting.32
The incremental cost-effectiveness ratios for home safety interventions varied from £304 per fall prevented for delivering the programme to people with severe visual impairment,30 to £3040 (incorporating all healthcare costs) for those recently discharged from hospital.36 Cataract surgery was the least favourable with a cost (incorporating all healthcare, personal and carer costs) per fall averted of £4732.38
There are often inherent logistical difficulties in comparing and combining the results of economic analyses because of differences in: the health system in the country of the study, the currency, 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 powered for falls and not for costs, and hospital costs particularly are very often skewed. We facilitate some comparability between studies and report cost per fall prevented for delivering the intervention where this was reported (converted to 2008 pounds sterling, table 2).
Quality of included studies
The Quality of Health Economic Studies instrument used for the quality assessment underscored five strong papers with scores of 92% and greater (table 4). Additionally, it provided a reference list of components that economic evaluations should include.28 29 This instrument has three main limitations. It does not specifically examine details on study design and study population, or include an item on price adjustments for inflation or currency conversion. A further key limitation is that the scoring of all items is dichotomous. For example, a selective one-way sensitivity analysis of a small number of factors over implausible ranges is scored the same as a comprehensive one-way sensitivity analysis or a probabilistic sensitivity analysis. Given that six of the nine studies included in our systematic review were published prior to the routine use of probabilistic sensitivity analyses, we anticipated that the dichotomous nature of the instrument would have minimal impact on our quality assessment. However, to mitigate this problem, we also used a checklist developed by Drummond and colleagues to support our findings from the Quality of Health Economic Studies instrument.23
How do economic evaluations aid in decision-making?
Economic analyses are important for policy makers, funders and providers making decisions among falls prevention strategies.23 Cost-effectiveness analyses allow decision-makers and clinicians to compare value for money from a multifactorial intervention with a multiple- or single-factor approach. For example, the multifactorial intervention cost of £2696 per fall prevented for delivering the one-off intervention.34 In contrast, the single factor strength and balance retraining programme compared with a control group cost £942 per fall prevented for delivering the programme in a community healthcare setting for 1 year.32 We encourage all authors of randomised controlled trials reporting effective falls prevention strategies to report details of the cost items required to deliver the intervention (table 3) and the associated incremental cost-effectiveness ratio to allow comparison of value for money. Detailed reporting of all falls and cost items for (1) delivering the intervention, (2) fall-related healthcare resource utilisation, and (3) total healthcare resource utilisation will enable comparison of the incremental cost per fall prevented for an intervention compared with the control activity. Detailed reporting of these three cost categories will provide information on value for money among different falls prevention strategies. These details are essential for decision-making.
Policy makers are faced with difficult resource allocation choices among different medical conditions. Ideally, these choices require cost–benefit or cost–utility analyses such that outcomes are on the same metric, but both these approaches are problematic for evaluating falls prevention interventions. One study using a population health approach valued the benefits in dollar terms based on the number of hospital admissions averted.35 Caution is needed when interpreting hospital costs in randomised controlled trials powered for falls and not hospital admissions. The EQ-5D was used to estimate incremental cost per QALYs gained after expedited cataract surgery, which was £14 197 over the participants’ lifetime.38 However, there are particular problems with using QALYs as an outcome for a complex intervention potentially resulting in multiple benefits for older people.39 We have not found quality-of-life measures to be sensitive to change in our falls prevention studies despite the beneficial outcomes of the trials.8
Implications for future research
We recommend that future economic evaluations be guided in part by the checklists available for assessing economic evaluations.23 Key components are: the establishment of effectiveness, comprehensive inclusion and reporting of all relevant costs and consequences for each comparator, accurate and reliable valuation of costs and consequences, discounting when appropriate, reporting an incremental analysis of costs and consequences, reporting a comprehensive one way and probabilistic sensitivity analysis and a discussion of the results.23 One major problem with comparing the economic evaluations in our systematic review was that some incremental cost-effectiveness ratios incorporated intervention costs only, some included fall-related costs, and some included total healthcare costs. Therefore, we recommend that all studies report (1) total healthcare resource utilisation costs, (2) fall-related healthcare resource utilisation costs, and (3) intervention costs only to enable a base case comparison of value for money. If future economic evaluations of falls prevention strategies were to follow these recommendations, it would help reduce heterogeneity among economic evaluations and facilitate comparison between studies.
Implications for policy-makers and clinical practice
In practice, clinically appropriate risk-reduction strategies will continue to determine the most appropriate intervention for a particular patient. However, from a public health perspective, this systematic review suggests that single interventions (such as the Otago Exercise Programme) targeted at high-risk groups can prevent the greatest number of falls at the lowest incremental costs. Our study also suggests that individual clinicians can be assured that prescribing strength and balance training programmes is evidence-based—not only for effective fall prevention but also as an economically valid choice. Clinicians in the physical activity and medicine setting (ie, various relevant physician groups, physiotherapists, clinical physiologists, athletic trainers) should advocate for this treatment to be included judiciously in patient prescriptions. We acknowledge that just as pharmaceutical prescription can have adverse effects40 and is contraindicated in certain patients, clinical judgement needs to apply when prescribing exercise. Nevertheless, the Otago Exercise Programme has proven to be both safe and effective in over 500 patient years of data in studies as well as in subsequent implementation in the community setting.8 31,–,33
What is already known on this topic
Systematic reviews of randomised controlled trials testing the efficacy of falls prevention strategies indicate that both multifactorial and single factor interventions are effective in reducing falls in older people living independently in the community.
No systematic review of economic evaluations of falls prevention strategies has been reported.
What this study adds
The best value for money comes from targeting effective falls prevention strategies at high-risk groups.
Three programmes were cost saving in subgroups of seniors: (1) a multifactorial programme targeted at eight fall risk factors, (2) the home-based Otago Exercise Programme delivered to people aged ≥80 years, and (3) a home safety programme for those recently discharged from hospital, if delivered to the subgroup of participants with a previous fall.
Of these three programmes, the home-based exercise programme in people aged ≥80 years may have the broadest applicability and thus provide the best value for National Health Services money.
A factor limiting the strength of our conclusion is that incremental cost-effectiveness ratios were estimated from different economic perspectives using different cost items.
In summary, as with every field of research, additional data will provide a more solid foundation for recommendations. Recommendations may need to be changed in time. However, at this point in time, we conclude that single interventions (such as the Otago Exercise Programme) targeted at high-risk groups can prevent the greatest number of falls at the lowest incremental costs.
We thank LD Gillespie for sharing the list of studies included in the current update of the Cochrane community falls review. We classified interventions using the taxonomy developed by S Lamb and Work Package One of the Prevention of Falls Network Europe (ProFaNE).
Funding Centre for Hip Health and Mobility, Michael Smith Foundation for Health Research and the Canadian Institute for Health Research. CAM is funded by a Canada Research Chair in Pharmaceutical Outcomes and a Michael Smith Foundation for Health Research Scholar Award. MCA and TLA are funded by a Michael Smith Foundation for Health Research Scholar Award. JCD is funded by a Michael Smith Foundation for Health Research Senior Graduate Studentship and a Canadian Institute for Health Research Canada Graduate Scholarship.
Competing interests AJ Campbell, MCR and colleagues developed the Otago Exercise Programme and authored the studies cited in this systematic review. KMK and TL-A have published a paper where they have tested the Otago Exercise Programme in Canada. JCD is currently testing the cost-effectiveness of the Otago Exercise Programme in Canada. KMK has received funding from various drug companies to be on a speaker panel and to undertake research.
JCD was principal investigator, was responsible for design, quality assessment, literature search, collation and summary of papers, retrieval of articles, review of studies and writing the manuscript, and is guarantor. MCR, MCA, TLA, KMK and CAM were also responsible for design, quality assessment, review of studies and writing the manuscript. MCA, JCD and MCR were responsible for quality assessment.
Provenance and Peer review not commissioned; externally peer reviewed