JAMDA xxx (2015) 1e7
JAMDA journal homepage: www.jamda.com
Original Study
Cost-effectiveness of a Home-Exercise Program Among Older People After Hospitalization Inez Farag MSafSc (UNSW) a, *, Kirsten Howard PhD b, Alison J. Hayes PhD b, Manuela L. Ferreira PhD a, Stephen R. Lord PhD c, Jacqueline T. Close MD, FRACP c, d, Constance Vogler PhD e, f, Catherine M. Dean PhD g, Robert G. Cumming MD, PhD b, Cathie Sherrington PhD a a
The George Institute for Global Health, University of Sydney, Sydney, NSW, Australia Sydney School of Public Health, University of Sydney, Sydney, NSW, Australia Neuroscience Research Australia, UNSW, Randwick, NSW, Australia d Prince of Wales Clinical School, UNSW, Randwick, NSW, Australia e Sydney Medical School, University of Sydney, Sydney, NSW, Australia g Faculty of Human Sciences, Macquarie University, North Ryde, NSW, Australia f Department of Aged Care, Royal North Shore Hospital Reserve Road, St Leonard’s, NSW, Australia b c
a b s t r a c t Keywords: Older people hospitalization exercise programs fall prevention
Background: Older people who have been recently discharged from hospital are at increased risk of falls and deterioration in physical functioning. Objective: To investigate the cost-effectiveness of a 12-month home-exercise program for older adults after hospitalization. Method: An economic evaluation was conducted alongside a randomized controlled trial. The analysis was conducted from the health and community service provider perspective. A total of 340 people aged 60 years and older, with a recent hospital admission, were randomized into exercise and usual care control groups. Incremental costs per extra person showing improvement in mobility performance (using the Short Physical Performance Battery), per person indicating improvement in health (self-reported using a 3-point Likert scale) and per quality-adjusted life year (QALY) gained (utility measured using the EQ-5D) were estimated. Uncertainty was represented using cost-effectiveness acceptability curves. Subgroup analyses for participants with better cognition (above the median MMSE score of 28) also were undertaken. Results: The average cost of the intervention was $A751 per participant. The incremental costeffectiveness of the program relative to usual care was $A22,958 per extra person showing an improvement in mobility, $A19,020 per extra person indicating an improvement in health, and $A77,403 per QALY. The acceptability curve demonstrates that the intervention had an 80% probability of being cost-effective relative to the control at a threshold of $A48,000 per extra person achieving mobility improvement and $A36,000 indicating an improvement in self-reported health. There was no threshold value at which the program can be considered as having an 80% probability of cost-effectiveness for the QALY outcome. Subgroup analyses for participants with better cognitive status indicated improved costeffectiveness for all outcomes. Conclusion: The exercise intervention appeared to offer reasonable value for money for mobility outcomes and self-reported health status. Value for money for all measures was greater in the higher cognitive status subgroup. Ó 2015 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
The authors declare no conflicts of interest. The main trial received project grant funding from the Australian National Health and Medical Research Council for the submitted work (NHMRC ID: 457464). CS and SRL receive salary funding from Australian National Health and Medical Research Council Research Fellowships. The Australian National Health and Medical http://dx.doi.org/10.1016/j.jamda.2015.01.075 1525-8610/Ó 2015 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
Research Council played no role in the production or approval of this manuscript. The researchers are independent of the funder. * Address correspondence to Inez Farag, MSafSc (UNSW), The George Institute for Global Health, The University of Sydney, Level 13, 321 Kent Street, Sydney 2000, Australia. E-mail address: [email protected] (I. Farag).
2
I. Farag et al. / JAMDA xxx (2015) 1e7
There is an increased risk of falls1 and persistent deterioration in physical functioning in older people who have had recent hospital admission.2 The aging of the global population and the expanding scope of medical science3 have led to a greater number of older people being admitted to hospital. The impact on quality of life and the threat to health and independence for the older person who has recently been hospitalized is considerable. Urgent attention is therefore required to reduce the economic and social burden of recent hospital admission. Exercise programs have been shown to prevent falls,4 enhance strength5 and balance,6 and prevent disability7 in older people living in the general community. Exercise interventions targeting older people with recent hospital admission have indicated improvement in mobility and balance,8,9 but are yet to show a reduction in falls.10 We undertook a trial to evaluate a balance-challenging home-exercise program on mobility-related disability and falls in people who had been recently hospitalized.11 The trial found that performancebased mobility was significantly better in the intervention group than the control group; between-group difference in continuously scored Short Physical Performance Battery (SPPB) adjusted for baseline performance (incidence rate ratio 0.13, 95% confidence interval [CI] 0.04e0.21, P ¼ .004) but found a higher fall rate in the intervention group (incidence rate ratio 1.43, 95% CI 1.07e1.93, P ¼ .017) than the control group. The main study results also indicated a greater impact on performance-based mobility for those with better Mini-Mental State Examination (MMSE) scores (P for interaction ¼ .037); between-group difference in continuously scored SPPB adjusted for baseline performance among those with MMSE scores 28 to 30 (0.19, 95% CI 0.09e0.29, P < .001, n ¼ 201). Economic evaluations conducted in association with randomized controlled trials12,13 and modeled studies14,15 indicate potential for cost savings and considerable benefits of fall prevention programs conducted in the general community. To date, however, no economic evaluations have been published examining the cost-effectiveness of a home-exercise program for older people recently discharged from hospital. Further research is also required to assess the impact of fall prevention programs on quality of life12; enabling comparison with other health care programs available to older people. The objective of the present study was to examine, from a health and community service provider perspective, the cost-effectiveness of a 12-month home-based exercise program in older people who have been recently discharged from hospital. Method Randomized Controlled Trial The trial involved 340 community-dwelling participants who were discharged from public hospital wards in Sydney, Australia, with 169 participants randomly allocated to the control group and 171 to the intervention group. Participants were consenting adults, aged 60 years and older who had sufficient English language skills to undertake study assessments and interventions.11 The exclusion criteria included a MMSE score lower than 24, a diagnosed degenerative neurological disorder, inability to walk 1 meter with assistance or a walking aid, and residence in a high-care facility (nursing home). Randomization was performed centrally by an investigator not involved in recruitment or assessment. Intervention and Control Participants randomized to the intervention group received 10 home visits and 5 phone calls from an experienced physical therapist over a 12-month period. Exercises were based on the Weight-bearing
Exercise for Better Balance (WEBB) program, targeting postural control and lower limb muscle strength. Participants were also requested to undertake a 20- to 30-minute program of exercises up to 6 times per week at home for 12 months. A full description of the exercise program is detailed in the study’s protocol paper.16 The control group received usual care from health and community service providers. All participants also received a fall prevention booklet. Outcome Measures The primary outcome measures for the trial were fall rate (selfreported using monthly calendars); performance-based mobility (the continuously scored version of the SPPB measured by an assessor masked to group allocation at baseline and 3 and 12 months) and self-reported level of difficulty in mobility task performance (assessed over 12 months using monthly questionnaires). Outcome measures for the present economic analyses were the 12-point SPPB, a 3-point self-rated measure of health status, and quality of life. The SPPB includes walking, standing balance, and sitto-stand tests.17 The self-rated measure of health status asked participants to rate their health over the past 12 months using “better, same, or worse” categories. Quality-of-life weights were determined from the EQ-5D questionnaire18 results from baseline and follow-up interviews and by participant self-completion at 3-monthly intervals, up until study end, or date of death. The proportion of participants showing an improvement from baseline to 12 months on (1) SPPB score and (2) self-rated health status was compared between intervention and control groups using the c2 test. Quality of life was compared between groups using the average quality-adjusted life year (QALY) measure. Costs All costs were valued in 2012 Australian dollars ($A). Cost of program delivery The cost of implementing the program was estimated from records of the staff time taken for service delivery and travel costs. An estimate of the average time taken on a home-visit and in travel, as well as the kilometers traveled, was derived from the information supplied by staff. The average hospital award wage for physiotherapists19 was used to estimate the cost of service delivery. The cost per kilometer of travel was estimated using information supplied by the Australian Taxation Office.20 Other intervention costs included booklets showing exercises, fall prevention booklets, and phone calls. Health service use and costs Health services included respite care, residential aged care and hospital admission, emergency department presentation, general practitioner, specialist medical services and nursing services, and physiotherapy and occupational therapy provided in the home or in consultation rooms. Social support services were counted as occasions of service of home care, transport, shopping assistance, housework, and the number of meals provided. Participants recorded use of health and social support services on monthly calendars, with follow-up via telephone if calendars were not returned. Baseline demographic data were obtained, before randomization, from medical record reviews and participant interviews. Costs were estimated by multiplying the occasions of service by the unit costs (valued in 2012 Australian dollars). Unit costs were obtained from the Medicare Benefits Schedule21 for medical and allied health services and from Australian Refined Diagnosis Related Group cost weights (AR-DRG version 5.2)22 for hospital admissions,
I. Farag et al. / JAMDA xxx (2015) 1e7
3
adjusted for length of hospital stay. An average cost per hospital day was derived ($844) and was used where the information on cause of hospital admission was not available. The cost of emergency department presentation was based on an average estimate of $451 for all triaged categories.22 The average cost per bed day for residential care was estimated at $A94.79 per bed day.23 The unit costs of social support services were obtained from a recent systematic review conducted by the authors.24 Health system contact was evaluated for all study participants. Residents of low-care facilities were excluded from analysis of social support services. Instead, the per annum cost of a low-care residential facility (hostel) was estimated at $A11,472. Unit costs of services are detailed in Table A1 in the Appendix.
MMSE, Mini-Mental Status Examination score.
Cost-Effectiveness Analysis
Costs
The economic evaluation was conducted from the health and community care funder perspective and estimated the incremental costs and incremental health outcomes of the intervention compared with usual care. The total costs of health and social support services and the cost of program provision in the intervention and control groups was used to calculate incremental cost-effectiveness ratios (ICERs), given by the formula¼ (CieCc)/(EieEc), where Ci is the program and other resource use costs for the intervention group; Cc is the resource use costs for the control group; Ei and Ec are the health effects of the intervention and control groups, respectively. ICERs were calculated for the incremental cost per person showing an improvement on the SPPB (identified as an improvement on the baseline SPPB score), the incremental cost per person showing an improvement in health status (identified as an improvement on the baseline self-rated measure of health status) and the incremental cost per QALY gained in the exercise group compared with the control group. The probability that the intervention was cost-effective given a decision-maker’s willingness to pay for each additional unit of health outcome achieved was estimated by bootstrapping costs and health outcomes. Bootstrapping allows estimation of joint uncertainty in costs and effects, and presents this as the plot of incremental costs and health outcomes on the cost-effectiveness plane. As the underlying willingness to pay is unknown, the probability of the intervention being cost-effective is presented as a function of a varying willingness to pay in cost-effectiveness acceptability curves.25 A subgroup analysis for participants with better cognition (above the median MMSE score of 28) was conducted. Sensitivity analyses were also performed to examine the effects of varying costs on the base case cost-effectiveness results. Sensitivity analyses included removal of hostel costs ($A11,472) and exclusion of participants who were hostel residents at baseline.
Cost of program delivery The total cost to deliver the 12-month home-based exercise intervention was $A125,417 corresponding to an average cost of $A751 per participant. This estimate was based on an average of 8.8 sessions per participant. In total, 1493 home visits were conducted. If all participants had received the planned number of home visits (n ¼ 10), then the average cost of the program increases to $A861 per participant. The mean therapist travel time per visit was 25 minutes and the estimated average cost of a home-visit session, excluding travel, was $A82.
Results Participant characteristics are described in Table 1. At baseline, exercise and control group participants were similar in terms of demographic descriptors and comorbidities. Of the 340 participants in the trial, 9 were excluded from the economic analysis because of missing cost and outcome data (4 from the intervention and 5 from the control group). In addition, 19 participants had missing SPPB and self-rated health status data (10 from the intervention group and 9 from the control group). These participants were allocated the baseline SPPB and self-rated health status score (conservatively assuming no improvement in mobility or in quality of life over the 12-month period). When participants died during the study follow-up period, they were allocated a utility value of zero for the remainder of follow-up.
Table 1 Participant Characteristics (n ¼ 340) Characteristic
Intervention Mean (SD) or n (%) n ¼ 171
Control Mean (SD) or n (%) n ¼ 169
Gender, % female Age, y Comorbidities Medications Hospital length of stay, d No. of falls since leaving hospital Living in low-care residential facility MMSE 27
123 82 7.2 7.5 17.0 0.25 8 9
128 81 6.9 7.5 15.6 0.18 12 7
(72%) (8) (2.9) (3.3) (19.3) (0.55) (5%) (5%)
(76%) (8) (2.7) (3.3) (15.7) (0.46) (7%) (4%)
Cost of health and social support services Total costs of health and social support services for the intervention and control groups were $A3,400,790. Fifty-six percent of total costs were associated with hospital admission and 22% with social support services. More detail on types of services used and costs is presented elsewhere (Farag I, Howard K, O’Rourke S, et al. Health and social support services in older adults recently discharged from hospital; utilization, costs and impact of a home-exercise intervention. J Aging Health; Under review). A summary of the costs of health and social support services for the intervention and control groups is presented in Table 2. Significance testing did not indicate a betweengroup difference in costs or occasions of service for any of the service categories.
Table 2 Program and Health Service Costs: Base Case, Subgroup and Sensitivity Analysis (Base Year 2012, $A)
Program Costs (as delivered) Health service cost Residential aged care Hospital costs Outpatient services Social support services Total costs base case (program þ service use) Total costs (sensitivity analysis 1)y Total costs (sensitivity analysis 2)z Total costs (subgroup analysis)x
Intervention n ¼ 167 Average per Participant Mean (SD)
Control n ¼ 164 Average per Participant Mean (SD)
$751*
$0
$950 $6700 $1430 $2947 $12,780
($3447) ($14,909) ($1240) ($5138) ($18,449)
$11,956 ($18,004) $11,612($16,221) $11,108 ($15,327)
$850 $4979 $1587 $2907 $10,327
($2816) ($15,689) ($2717) ($5622) ($17,432)
$9,697 ($17,384) $9,935 ($17,743) $10,151 ($17,774)
*Range of program costs vary from $751 to $861, depending on the number of home visits (8.8 to 10.0). y Removal of annual hostel costs ($A11,472) for participants in a hostel at baseline. z Participants not in a hostel at baseline n ¼ 310 (intervention n ¼ 155, control n ¼ 155). x Higher cognitive status-program þ health service use, intervention n ¼ 108, control n ¼ 103.
$4650 per person indicating improvement in self-rated health status 16 (0.16)
The outcome measures used in the cost-effectiveness analysis are summarized in Table 3. Significantly more people in the intervention group had an improvement in mobility measures than the controls (49% versus 38%, difference in proportions 0.11, 95% CI 0.01e0.22, P ¼ .038). The proportion of participants indicating an improvement in health status was also greater in the intervention group (28% versus 15%, difference in proportions 0.13, 95% CI 0.04e0.22, P ¼ .005). There was a QALY difference between the 2 groups, favoring the intervention group that did not reach statistical significance (0.03, 95% CI e0.014e0.078, P ¼ .17).
$20,346 QALY gained
$6754 per person showing mobility improvement 35 (0.34)
Outcomes
0.66 (0.21)
ICER ($A) Cost per Health Gain Control Mean (SD) n ¼ 103
0.71 (0.15) $44,496 QALY gained *Removal of annual hostel costs ($A11,472) for participants in a hostel at baseline. y Participants not in a hostel at baseline n ¼ 310 (intervention n ¼ 155, control n ¼ 155).
$71,259 QALY gained $77,403 QALY gained 0.66 (0.23) 0.69 (0.19)
39 (0.36) $10,475 per person indicating improvement in self-rated health status $17,510 per person indicating improvement in self-rated health status $19,020 per person indicating improvement in self-rated health status 47 (0.28)
25 (0.15)
52 (0.48) $18,622 per person showing mobility improvement $22,958 per person showing mobility improvement 82 (0.49)
62 (0.38)
$21,136 per person showing mobility improvement
Cost-Effectiveness Analysis
SPPB: Number and proportion of participants showing improvement Self-rated measure of health status: Number and proportion of participants indicating an improvement QALYs over 12 months Mean (SD)
Intervention Mean (SD) n ¼ 108 ICER ($A) Cost per Health Gain ICER ($A) Cost per Health Gain Intervention Mean (SD) n ¼ 167
Control Mean (SD) n ¼ 164
ICER ($A) Cost per Health Gain
Sensitivity Analysis 1* Base Case Analysis
Total Group Outcome Measure
Table 3 Outcome Measures and the Incremental Cost-Effectiveness Ratio at 12 Months: Base Case and Subgroup Analysis
Sensitivity Analysis 2y
Higher Cognitive Status
I. Farag et al. / JAMDA xxx (2015) 1e7
Base Case Analysis
4
The incremental cost-effectiveness of the program relative to usual care was $A22,958 per person showing an improvement in mobility performance. Figure 1 illustrates the results of the bootstrapping in a scatterplot and cost-effectiveness acceptability curve. The acceptability curve demonstrates that the intervention has an 80% probability of being cost-effective relative to the control at a threshold of $A48,000 per extra person achieving mobility improvement. The cost-effectiveness analysis also indicated an ICER of $A19,020 per extra person reporting an improvement in overall health status with an 80% probability of cost-effectiveness at a threshold value of $A36,000. The ICER for the QALY outcome was $A77,403 per QALY gained in the intervention compared with the control group (Table 3). The acceptability curve demonstrates that there is no threshold value at which the program can be considered as having an 80% probability of cost-effectiveness for the QALY outcome (Appendix). The subgroup analysis of participants with better cognitive status (MMSE >28) demonstrated improved cost-effectiveness for all outcome measures. The ICER was $A6754 per extra person experiencing a mobility improvement, with an 80% probability of costeffectiveness at a threshold value of $A22,000. The ICER was $A4650 per extra person experiencing an improvement in selfreported health with an 80% probability of cost-effectiveness at a threshold value of $A12,000. For the QALY measure, the obtained ICER was $A20,346 per QALY gained (Table 3) with an 80% probability of cost-effectiveness at a threshold of $A80,000 per QALY gained. Removal of the low-care residential costs of $A11,472 from the cost data for participants residing in a hostel at baseline did not markedly affect the ICER value for all outcome measures (Table 3). Full exclusion of those participants who were hostel residents at baseline results in lower cost-effectiveness ratios with an ICER of $A44,496 per QALY gained. Discussion This study estimated a cost of $A22,958 per person showing an improvement in mobility and an 80% probability of cost-effectiveness at a willingness to pay value of $A48,000. For other outcome measures, there was a cost of $A19,020 per person indicating an improvement in self-reported health and $A77,403 per QALY gained. With a decision-maker’s willingness to pay $A36,000 for an additional health outcome, there was an 80% probability of being costeffective relative to the control for the self-reported health status measure. There was no threshold value at which the intervention could be considered as having an 80% probability of cost-effectiveness for the QALY outcome. Cost-effectiveness for all outcomes was greater in the subgroup with better cognitive status, with the results indicating an incremental cost of $A6754 per person showing mobility improvement, $A4650 per person indicating an improvement in selfreported health and $A20,346 per QALY gained.
I. Farag et al. / JAMDA xxx (2015) 1e7
B
12000
Incremental cost $A
8000 4000
-0.1
0.1
0.2
0.3
0.4
-4000 -8000 -12000
Incremental patients with improvement in mobility
Probability cost-effective
A
5
1.0 0.8 0.6 0.4 0.2 0.0
0
20000
40000
60000
80000
100000
Willingness to pay per person with improved mobility $A
Fig. 1. Cost-effectiveness of the program for mobility improvement. (A) Incremental cost-effectiveness plane for point estimate (black circle) and 1000 bootstrapped pairs (gray circles); (B) cost-effectiveness acceptability curve for mobility improvement. Solid line, all participants; dashed line, subgroup with higher cognition MMSE >28.
There are a limited number of economic evaluations conducted to date that have considered cost-effectiveness in terms of functional improvements.26 Several studies have used quality of life as an outcome measure,27e29 and many more studies have considered costeffectiveness in terms of the cost per fall prevented.12,13,28,30,31 There is wide variation in the results of the incremental cost per fall prevented, with the range spanning from $A294 (2008)32 to $NZ1803 (1998).33 The variation in the values presented is reflective of heterogeneity between studies, with differences in the health care systems between countries, in the study perspective, which in turn influences the cost consideration and in the type of program intervention. Nevertheless, for many of the studies there is an indication of significant potential for cost-effectiveness or cost savings of fall prevention programs.4 In our study, we have not been able to provide a cost per fall averted, as falls increased in the intervention group. It is interesting to note, however, that the higher fall rates in the intervention group did not necessarily translate to a decrement in quality of life or in the self-reported measure of health status. Our exercise program resulted in a greater number of participants showing improvement on the mobility measure in the intervention group. These results suggest a substantial improvement in the objective assessment of functional ability using the SPPB measure, which has been shown to be an accurate indicator of risk of disability, mortality, and nursing home admission in the older population.17 Our results also indicate an improvement in self-rated health and in quality of life for the intervention group; whether this is deemed cost-effective is dependent on the “willingness to pay” of the funder for the health outcomes gained. Across all health outcomes, the ICERs for the subgroup with better cognitive status were more favorable. In addition, cost per QALY gained from our trial was also comparable to values obtained in modeling studies.34 The improved cost-effectiveness for the subgroup with better cognitive status is due to larger incremental differences in the outcome measures. This suggests that better cognitive capacity may be associated with a greater potential to benefit from the intervention. As the study excluded people with cognitive impairment, no conclusions can be drawn about the potential benefits of the intervention in this group. The sensitivity analysis with exclusion of participants residing in a hostel at baseline also produced improved cost-effectiveness. This was largely driven by the impact on the cost data, with higher nonehostel-related health care costs accrued by the small number of hostel residents. Removal of the annual hostel costs only from the analysis did not affect cost-effectiveness. There appears to be substantial use of services in older people after hospital discharge. This is consistent with the previously noted high growth in expenditures for nonhospital35 and hospital-related36
services by older people in the latter decades of life, but it is not yet clear whether any of these costs are avoidable. The cost of fall prevention programs vary from (2009, $A) $A500 for Tai Chi, to home-based exercise programs, which are estimated to cost an average of $A1233 per participant.37 Multifactorial interventions are estimated to cost $A1886 per participant.37 At a program cost of $A751 (2012, $A) per participant, our program cost is at the lower end of this range. Previous modeling studies undertaken by the authors indicate that cost-effectiveness is markedly susceptible to the cost of the program.38 The factors that influence program costs include frequency and mode of delivery (eg, face-to-face or by telephone or telemedicine), whether group-based or individual, wage rate of the professional delivering the program. Cost-effectiveness would be optimized by reducing the cost of the program delivery, providing that the effectiveness of the program is not compromised with alternative delivery modes. In this study, costs were considered only from the service provider perspective and therefore no estimate was made of the cost of informal care services. Further research with inclusion of costs from the societal perspective would provide a better indicator of the impact of falls on the wider community. Conclusion The exercise intervention appeared to offer reasonable value for money for mobility outcomes and self-reported health status but whether this is deemed to be cost-effective is dependent on the “willingness to pay” of the funder for the health outcomes gained. Value for money improved for all measures in the subgroup with higher cognitive status. References 1. Mahoney J, Sager M, Dunham NC, Johnson J. Risk of falls after hospital discharge. J Am Geriatr Soc 1994;42:269e274. 2. Gill TM, Allore HG, Holford TR, Guo Z. Hospitalization, restricted activity, and the development of disability among older persons. JAMA 2004;292: 2115e2124. 3. Australian hospital statistics 2011e12. Canberra, Australia: Australian Institute of Health and Welfare; 2013. 4. Gillespie LD, Robertson CM, Gillespie WJ, et al. Interventions for preventing falls in older people living in the community. Cochrane Database Syst Rev; 2012:CD007146. 5. Liu CJ, Latham NK. Progressive resistance strength training for improving physical function in older adults. Cochrane Database Syst Rev; 2009:CD002759. 6. Howe TE, Shea B, Dawson LJ, et al. Exercise for preventing and treating osteoporosis in postmenopausal women. Cochrane Database Syst Rev; 2011: CD000333. 7. Investigators LS, Pahor M, Blair SN, et al. Effects of a physical activity intervention on measures of physical performance: Results of the lifestyle
6
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18. 19.
20.
21.
22.
I. Farag et al. / JAMDA xxx (2015) 1e7 interventions and independence for Elders Pilot (LIFE-P) study. J Gerontol A Biol Sci Med Sci 2006;61:1157e1165. Timonen L, Rantanen T, Ryynanen OP, et al. A randomized controlled trial of rehabilitation after hospitalization in frail older women: Effects on strength, balance and mobility. Scand J Med Sci Sports 2002;12:186e192. Vogler CM, Sherrington C, Ogle SJ, Lord SR. Reducing risk of falling in older people discharged from hospital: A randomized controlled trial comparing seated exercises, weight-bearing exercises, and social visits. Arch Phys Med Rehabil 2009;90:1317e1324. Latham N, Anderson C, Bennett D, Stretton C. Progressive resistance strength training for physical disability in older people. Cochrane Database Syst Rev; 2003:CD002759. Sherrington CLS, Vogler CM, Close JCT, et al. A post-hospital home exercise program improved mobility but increased falls in older people: A randomised controlled trial. PLOS One 2014;9:e104412. Irvine L, Conroy SP, Sach T, et al. Cost-effectiveness of a day hospital falls prevention programme for screened community-dwelling older people at high risk of falls. Age Ageing 2010;39:710e716. Davis JC, Robertson MC, Ashe MC, et al. Does a home-based strength and balance programme in people aged > or ¼80 years provide the best value for money to prevent falls? A systematic review of economic evaluations of falls prevention interventions. Br J Sports Med 2010;44:80e89. Smith RD, Widiatmoko D. The cost-effectiveness of home assessment and modification to reduce falls in the elderly. Aust N Z J Public Health 1998;22: 436e440. Church J, Goodall S, Norman R, Haas M. The cost-effectiveness of falls prevention interventions for older community-dwelling Australians. Aust N Z J Public Health 2012;36:241e248. Sherrington C, Lord SR, Vogler CM, et al. Minimising disability and falls in older people through a post-hospital exercise program: A protocol for a randomised controlled trial and economic evaluation. BMC Geriatr 2009;9:8. Guralnik JM, Simonsick EM, Ferrucci L, et al. A short physical performance battery assessing lower extremity function:association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol 1994;49:M85eM94. Dolan P. Modeling valuations for EuroQol health states. Med Care 1997;35: 1095e1108. NSW Health Service Health Professionals (State) Award. NSW Department of Health. Available at: www.health.nsw.gov.au/careers/conditions/Pages/P.aspx; 2013. Accessed December 6, 2014. Individuals’ income deduction 2013. Available at: https://www.ato.gov.au/ Individuals/Income-and-deductions/In-detail/Deductions-for-work-related-ex penses/Claiming-a-deduction-for-car-expenses-using-the-cents-per-kilometremethod/. Accessed July 15, 2013. Medicare Benefits Schedule, Commonwealth of Australia. Department of Health. 2011. Available at: www.medicareaustralia.gov.au/provider/medicare/ mbs.jsp. Accessed July 2, 2013. National Hospital Cost Data Collection Cost Report Round 14 (2009e10). Canberra, Australia: Commonwealth Department of Health and Ageing; 2011.
23. Cost of residential aged care in Australia. Available at: https://www.dss.gov.au/ our-responsibilities/ageing-and-aged-care/aged-care-funding/aged-care-feesand-charges; 2013. Accessed December 12, 2013. 24. Farag I, Sherrington C, Ferreira M, Howard K. A systematic review of the unit costs of allied health and community services used by older people in Australia. BMC Health Serv Res 2013;13:69. 25. World Health Organization. Macroeconomics and health: Investing in health economic developmentdreport of the commission on macroeconomics and health. Geneva, Switzerland: World Health Organization; 2001. 26. Melis RJ, Adang E, Teerenstra S, et al. Cost-effectiveness of a multidisciplinary intervention model for community-dwelling frail older people. J Gerontol A Biol Sci Med Sci 2008;63:275e282. 27. Hendriks MR, Evers SM, Bleijlevens MH, et al. Cost-effectiveness of a multidisciplinary fall prevention program in community-dwelling elderly people: A randomized controlled trial (ISRCTN 64716113). Int J Technol Assess Health Care 2008;24:193e202. 28. Sach TH, Foss AJ, Gregson RM, et al. Falls and health status in elderly women following first eye cataract surgery: An economic evaluation conducted alongside a randomised controlled trial. Br J Ophthalmol 2007;91:1675e1679. 29. Sach TH, Logan PA, Coupland CA, et al. Community falls prevention for people who call an emergency ambulance after a fall: An economic evaluation alongside a randomised controlled trial. Age Ageing 2012;41:635e641. 30. Robertson MC, Gardner MM, Devlin N, et al. Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 2: Controlled trial in multiple centres. BMJ 2001;322:701e704. 31. Davis JC, Robertson MC, Ashe MC, et al. International comparison of cost of falls in older adults living in the community: A systematic review. Osteoporos Int 2010;21:1295e1306. 32. Haines TP, Hill AM, Hill KD, et al. Cost effectiveness of patient education for the prevention of falls in hospital: Economic evaluation from a randomized controlled trial. BMC Med 2013;11:135. 33. Campbell AJ, Robertson MC, Gardner MM, et al. Falls prevention over 2 years: A randomized controlled trial in women 80 years and older. Age Ageing 1999;28: 513e518. 34. Church J, Goodall S, Norman R, Haas M. An economic evaluation of community and residential aged care falls prevention strategies in NSW. N S W Public Health Bull 2011;22:60e68. 35. Payne G, Laperte A, Deber R, Coyte PC. Counting backwards to health care’s future: Using time to death modelling to identify changes in end-of-life morbidity and the impact of aging on health care expenditures. Milbank Q 2007;85:213e257. 36. Kardamanidis K, Lim K, Da Cunha C, et al. Hospital costs of older people in New South Wales in the last year of life. Med J Aust 2007;187:383e386. 37. Day L, Hoareau E, Finch C, et al. Modelling the impact, costs and benefits of falls prevention measures to support policy-makers and program planners. Monash University Accident Research Centre. Available at: www.monash.edu.au/miri/ research/reports/muarc286.html; Accessed December 15, 2014. 38. Farag I, Howard K, Ferreira M, Sherrington C. Economic modelling of a public health program of fall prevention [published online ahead of print December 18, 2014]. Age Ageing
I. Farag et al. / JAMDA xxx (2015) 1e7
7
Appendix
Table A1 Unit Costs of Services (Base Year 2012, $A) Service
Unit Cost (Average)
Residential aged care (high) Residential aged care (low) Hospital Emergency Department presentation General practitioner consultation room General practitioner consultation home-visit Specialist Nursing Allied health service consultation room Allied health service home-visit Showering/dressing Meals Domestic services Transport Shopping
$94.79 per bed day23 $11,472 per annum23 AR-DRG version 5.222 $45122 $70.3021 $81.8521 $105.4821 $56.1821 $55.4421 $55.4421 $36.4036 $11.1036 $39.0736 $12.3936 $36.4036
A
B Probability cost-effective
Probability cost-effective
1.0 0.8 0.6 0.4 0.2 0.0
0
20000
40000
60000
80000
100000
Willingness to pay per person with improved health status $A
1.0 0.8 0.6 0.4 0.2 0.0
0
20000
40000
60000
80000
100000
Willingness to pay per QALY gained $A
Fig. A1. Cost-effectiveness acceptability curves (A) per person showing improvement in self-reported health status and (B) per QALY gained. Solid line, all participants; dashed line, subgroup with higher cognition MMSE >28.
JAMDA journal homepage: www.jamda.com
Original Study
Cost-effectiveness of a Home-Exercise Program Among Older People After Hospitalization Inez Farag MSafSc (UNSW) a, *, Kirsten Howard PhD b, Alison J. Hayes PhD b, Manuela L. Ferreira PhD a, Stephen R. Lord PhD c, Jacqueline T. Close MD, FRACP c, d, Constance Vogler PhD e, f, Catherine M. Dean PhD g, Robert G. Cumming MD, PhD b, Cathie Sherrington PhD a a
The George Institute for Global Health, University of Sydney, Sydney, NSW, Australia Sydney School of Public Health, University of Sydney, Sydney, NSW, Australia Neuroscience Research Australia, UNSW, Randwick, NSW, Australia d Prince of Wales Clinical School, UNSW, Randwick, NSW, Australia e Sydney Medical School, University of Sydney, Sydney, NSW, Australia g Faculty of Human Sciences, Macquarie University, North Ryde, NSW, Australia f Department of Aged Care, Royal North Shore Hospital Reserve Road, St Leonard’s, NSW, Australia b c
a b s t r a c t Keywords: Older people hospitalization exercise programs fall prevention
Background: Older people who have been recently discharged from hospital are at increased risk of falls and deterioration in physical functioning. Objective: To investigate the cost-effectiveness of a 12-month home-exercise program for older adults after hospitalization. Method: An economic evaluation was conducted alongside a randomized controlled trial. The analysis was conducted from the health and community service provider perspective. A total of 340 people aged 60 years and older, with a recent hospital admission, were randomized into exercise and usual care control groups. Incremental costs per extra person showing improvement in mobility performance (using the Short Physical Performance Battery), per person indicating improvement in health (self-reported using a 3-point Likert scale) and per quality-adjusted life year (QALY) gained (utility measured using the EQ-5D) were estimated. Uncertainty was represented using cost-effectiveness acceptability curves. Subgroup analyses for participants with better cognition (above the median MMSE score of 28) also were undertaken. Results: The average cost of the intervention was $A751 per participant. The incremental costeffectiveness of the program relative to usual care was $A22,958 per extra person showing an improvement in mobility, $A19,020 per extra person indicating an improvement in health, and $A77,403 per QALY. The acceptability curve demonstrates that the intervention had an 80% probability of being cost-effective relative to the control at a threshold of $A48,000 per extra person achieving mobility improvement and $A36,000 indicating an improvement in self-reported health. There was no threshold value at which the program can be considered as having an 80% probability of cost-effectiveness for the QALY outcome. Subgroup analyses for participants with better cognitive status indicated improved costeffectiveness for all outcomes. Conclusion: The exercise intervention appeared to offer reasonable value for money for mobility outcomes and self-reported health status. Value for money for all measures was greater in the higher cognitive status subgroup. Ó 2015 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
The authors declare no conflicts of interest. The main trial received project grant funding from the Australian National Health and Medical Research Council for the submitted work (NHMRC ID: 457464). CS and SRL receive salary funding from Australian National Health and Medical Research Council Research Fellowships. The Australian National Health and Medical http://dx.doi.org/10.1016/j.jamda.2015.01.075 1525-8610/Ó 2015 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
Research Council played no role in the production or approval of this manuscript. The researchers are independent of the funder. * Address correspondence to Inez Farag, MSafSc (UNSW), The George Institute for Global Health, The University of Sydney, Level 13, 321 Kent Street, Sydney 2000, Australia. E-mail address: [email protected] (I. Farag).
2
I. Farag et al. / JAMDA xxx (2015) 1e7
There is an increased risk of falls1 and persistent deterioration in physical functioning in older people who have had recent hospital admission.2 The aging of the global population and the expanding scope of medical science3 have led to a greater number of older people being admitted to hospital. The impact on quality of life and the threat to health and independence for the older person who has recently been hospitalized is considerable. Urgent attention is therefore required to reduce the economic and social burden of recent hospital admission. Exercise programs have been shown to prevent falls,4 enhance strength5 and balance,6 and prevent disability7 in older people living in the general community. Exercise interventions targeting older people with recent hospital admission have indicated improvement in mobility and balance,8,9 but are yet to show a reduction in falls.10 We undertook a trial to evaluate a balance-challenging home-exercise program on mobility-related disability and falls in people who had been recently hospitalized.11 The trial found that performancebased mobility was significantly better in the intervention group than the control group; between-group difference in continuously scored Short Physical Performance Battery (SPPB) adjusted for baseline performance (incidence rate ratio 0.13, 95% confidence interval [CI] 0.04e0.21, P ¼ .004) but found a higher fall rate in the intervention group (incidence rate ratio 1.43, 95% CI 1.07e1.93, P ¼ .017) than the control group. The main study results also indicated a greater impact on performance-based mobility for those with better Mini-Mental State Examination (MMSE) scores (P for interaction ¼ .037); between-group difference in continuously scored SPPB adjusted for baseline performance among those with MMSE scores 28 to 30 (0.19, 95% CI 0.09e0.29, P < .001, n ¼ 201). Economic evaluations conducted in association with randomized controlled trials12,13 and modeled studies14,15 indicate potential for cost savings and considerable benefits of fall prevention programs conducted in the general community. To date, however, no economic evaluations have been published examining the cost-effectiveness of a home-exercise program for older people recently discharged from hospital. Further research is also required to assess the impact of fall prevention programs on quality of life12; enabling comparison with other health care programs available to older people. The objective of the present study was to examine, from a health and community service provider perspective, the cost-effectiveness of a 12-month home-based exercise program in older people who have been recently discharged from hospital. Method Randomized Controlled Trial The trial involved 340 community-dwelling participants who were discharged from public hospital wards in Sydney, Australia, with 169 participants randomly allocated to the control group and 171 to the intervention group. Participants were consenting adults, aged 60 years and older who had sufficient English language skills to undertake study assessments and interventions.11 The exclusion criteria included a MMSE score lower than 24, a diagnosed degenerative neurological disorder, inability to walk 1 meter with assistance or a walking aid, and residence in a high-care facility (nursing home). Randomization was performed centrally by an investigator not involved in recruitment or assessment. Intervention and Control Participants randomized to the intervention group received 10 home visits and 5 phone calls from an experienced physical therapist over a 12-month period. Exercises were based on the Weight-bearing
Exercise for Better Balance (WEBB) program, targeting postural control and lower limb muscle strength. Participants were also requested to undertake a 20- to 30-minute program of exercises up to 6 times per week at home for 12 months. A full description of the exercise program is detailed in the study’s protocol paper.16 The control group received usual care from health and community service providers. All participants also received a fall prevention booklet. Outcome Measures The primary outcome measures for the trial were fall rate (selfreported using monthly calendars); performance-based mobility (the continuously scored version of the SPPB measured by an assessor masked to group allocation at baseline and 3 and 12 months) and self-reported level of difficulty in mobility task performance (assessed over 12 months using monthly questionnaires). Outcome measures for the present economic analyses were the 12-point SPPB, a 3-point self-rated measure of health status, and quality of life. The SPPB includes walking, standing balance, and sitto-stand tests.17 The self-rated measure of health status asked participants to rate their health over the past 12 months using “better, same, or worse” categories. Quality-of-life weights were determined from the EQ-5D questionnaire18 results from baseline and follow-up interviews and by participant self-completion at 3-monthly intervals, up until study end, or date of death. The proportion of participants showing an improvement from baseline to 12 months on (1) SPPB score and (2) self-rated health status was compared between intervention and control groups using the c2 test. Quality of life was compared between groups using the average quality-adjusted life year (QALY) measure. Costs All costs were valued in 2012 Australian dollars ($A). Cost of program delivery The cost of implementing the program was estimated from records of the staff time taken for service delivery and travel costs. An estimate of the average time taken on a home-visit and in travel, as well as the kilometers traveled, was derived from the information supplied by staff. The average hospital award wage for physiotherapists19 was used to estimate the cost of service delivery. The cost per kilometer of travel was estimated using information supplied by the Australian Taxation Office.20 Other intervention costs included booklets showing exercises, fall prevention booklets, and phone calls. Health service use and costs Health services included respite care, residential aged care and hospital admission, emergency department presentation, general practitioner, specialist medical services and nursing services, and physiotherapy and occupational therapy provided in the home or in consultation rooms. Social support services were counted as occasions of service of home care, transport, shopping assistance, housework, and the number of meals provided. Participants recorded use of health and social support services on monthly calendars, with follow-up via telephone if calendars were not returned. Baseline demographic data were obtained, before randomization, from medical record reviews and participant interviews. Costs were estimated by multiplying the occasions of service by the unit costs (valued in 2012 Australian dollars). Unit costs were obtained from the Medicare Benefits Schedule21 for medical and allied health services and from Australian Refined Diagnosis Related Group cost weights (AR-DRG version 5.2)22 for hospital admissions,
I. Farag et al. / JAMDA xxx (2015) 1e7
3
adjusted for length of hospital stay. An average cost per hospital day was derived ($844) and was used where the information on cause of hospital admission was not available. The cost of emergency department presentation was based on an average estimate of $451 for all triaged categories.22 The average cost per bed day for residential care was estimated at $A94.79 per bed day.23 The unit costs of social support services were obtained from a recent systematic review conducted by the authors.24 Health system contact was evaluated for all study participants. Residents of low-care facilities were excluded from analysis of social support services. Instead, the per annum cost of a low-care residential facility (hostel) was estimated at $A11,472. Unit costs of services are detailed in Table A1 in the Appendix.
MMSE, Mini-Mental Status Examination score.
Cost-Effectiveness Analysis
Costs
The economic evaluation was conducted from the health and community care funder perspective and estimated the incremental costs and incremental health outcomes of the intervention compared with usual care. The total costs of health and social support services and the cost of program provision in the intervention and control groups was used to calculate incremental cost-effectiveness ratios (ICERs), given by the formula¼ (CieCc)/(EieEc), where Ci is the program and other resource use costs for the intervention group; Cc is the resource use costs for the control group; Ei and Ec are the health effects of the intervention and control groups, respectively. ICERs were calculated for the incremental cost per person showing an improvement on the SPPB (identified as an improvement on the baseline SPPB score), the incremental cost per person showing an improvement in health status (identified as an improvement on the baseline self-rated measure of health status) and the incremental cost per QALY gained in the exercise group compared with the control group. The probability that the intervention was cost-effective given a decision-maker’s willingness to pay for each additional unit of health outcome achieved was estimated by bootstrapping costs and health outcomes. Bootstrapping allows estimation of joint uncertainty in costs and effects, and presents this as the plot of incremental costs and health outcomes on the cost-effectiveness plane. As the underlying willingness to pay is unknown, the probability of the intervention being cost-effective is presented as a function of a varying willingness to pay in cost-effectiveness acceptability curves.25 A subgroup analysis for participants with better cognition (above the median MMSE score of 28) was conducted. Sensitivity analyses were also performed to examine the effects of varying costs on the base case cost-effectiveness results. Sensitivity analyses included removal of hostel costs ($A11,472) and exclusion of participants who were hostel residents at baseline.
Cost of program delivery The total cost to deliver the 12-month home-based exercise intervention was $A125,417 corresponding to an average cost of $A751 per participant. This estimate was based on an average of 8.8 sessions per participant. In total, 1493 home visits were conducted. If all participants had received the planned number of home visits (n ¼ 10), then the average cost of the program increases to $A861 per participant. The mean therapist travel time per visit was 25 minutes and the estimated average cost of a home-visit session, excluding travel, was $A82.
Results Participant characteristics are described in Table 1. At baseline, exercise and control group participants were similar in terms of demographic descriptors and comorbidities. Of the 340 participants in the trial, 9 were excluded from the economic analysis because of missing cost and outcome data (4 from the intervention and 5 from the control group). In addition, 19 participants had missing SPPB and self-rated health status data (10 from the intervention group and 9 from the control group). These participants were allocated the baseline SPPB and self-rated health status score (conservatively assuming no improvement in mobility or in quality of life over the 12-month period). When participants died during the study follow-up period, they were allocated a utility value of zero for the remainder of follow-up.
Table 1 Participant Characteristics (n ¼ 340) Characteristic
Intervention Mean (SD) or n (%) n ¼ 171
Control Mean (SD) or n (%) n ¼ 169
Gender, % female Age, y Comorbidities Medications Hospital length of stay, d No. of falls since leaving hospital Living in low-care residential facility MMSE 27
123 82 7.2 7.5 17.0 0.25 8 9
128 81 6.9 7.5 15.6 0.18 12 7
(72%) (8) (2.9) (3.3) (19.3) (0.55) (5%) (5%)
(76%) (8) (2.7) (3.3) (15.7) (0.46) (7%) (4%)
Cost of health and social support services Total costs of health and social support services for the intervention and control groups were $A3,400,790. Fifty-six percent of total costs were associated with hospital admission and 22% with social support services. More detail on types of services used and costs is presented elsewhere (Farag I, Howard K, O’Rourke S, et al. Health and social support services in older adults recently discharged from hospital; utilization, costs and impact of a home-exercise intervention. J Aging Health; Under review). A summary of the costs of health and social support services for the intervention and control groups is presented in Table 2. Significance testing did not indicate a betweengroup difference in costs or occasions of service for any of the service categories.
Table 2 Program and Health Service Costs: Base Case, Subgroup and Sensitivity Analysis (Base Year 2012, $A)
Program Costs (as delivered) Health service cost Residential aged care Hospital costs Outpatient services Social support services Total costs base case (program þ service use) Total costs (sensitivity analysis 1)y Total costs (sensitivity analysis 2)z Total costs (subgroup analysis)x
Intervention n ¼ 167 Average per Participant Mean (SD)
Control n ¼ 164 Average per Participant Mean (SD)
$751*
$0
$950 $6700 $1430 $2947 $12,780
($3447) ($14,909) ($1240) ($5138) ($18,449)
$11,956 ($18,004) $11,612($16,221) $11,108 ($15,327)
$850 $4979 $1587 $2907 $10,327
($2816) ($15,689) ($2717) ($5622) ($17,432)
$9,697 ($17,384) $9,935 ($17,743) $10,151 ($17,774)
*Range of program costs vary from $751 to $861, depending on the number of home visits (8.8 to 10.0). y Removal of annual hostel costs ($A11,472) for participants in a hostel at baseline. z Participants not in a hostel at baseline n ¼ 310 (intervention n ¼ 155, control n ¼ 155). x Higher cognitive status-program þ health service use, intervention n ¼ 108, control n ¼ 103.
$4650 per person indicating improvement in self-rated health status 16 (0.16)
The outcome measures used in the cost-effectiveness analysis are summarized in Table 3. Significantly more people in the intervention group had an improvement in mobility measures than the controls (49% versus 38%, difference in proportions 0.11, 95% CI 0.01e0.22, P ¼ .038). The proportion of participants indicating an improvement in health status was also greater in the intervention group (28% versus 15%, difference in proportions 0.13, 95% CI 0.04e0.22, P ¼ .005). There was a QALY difference between the 2 groups, favoring the intervention group that did not reach statistical significance (0.03, 95% CI e0.014e0.078, P ¼ .17).
$20,346 QALY gained
$6754 per person showing mobility improvement 35 (0.34)
Outcomes
0.66 (0.21)
ICER ($A) Cost per Health Gain Control Mean (SD) n ¼ 103
0.71 (0.15) $44,496 QALY gained *Removal of annual hostel costs ($A11,472) for participants in a hostel at baseline. y Participants not in a hostel at baseline n ¼ 310 (intervention n ¼ 155, control n ¼ 155).
$71,259 QALY gained $77,403 QALY gained 0.66 (0.23) 0.69 (0.19)
39 (0.36) $10,475 per person indicating improvement in self-rated health status $17,510 per person indicating improvement in self-rated health status $19,020 per person indicating improvement in self-rated health status 47 (0.28)
25 (0.15)
52 (0.48) $18,622 per person showing mobility improvement $22,958 per person showing mobility improvement 82 (0.49)
62 (0.38)
$21,136 per person showing mobility improvement
Cost-Effectiveness Analysis
SPPB: Number and proportion of participants showing improvement Self-rated measure of health status: Number and proportion of participants indicating an improvement QALYs over 12 months Mean (SD)
Intervention Mean (SD) n ¼ 108 ICER ($A) Cost per Health Gain ICER ($A) Cost per Health Gain Intervention Mean (SD) n ¼ 167
Control Mean (SD) n ¼ 164
ICER ($A) Cost per Health Gain
Sensitivity Analysis 1* Base Case Analysis
Total Group Outcome Measure
Table 3 Outcome Measures and the Incremental Cost-Effectiveness Ratio at 12 Months: Base Case and Subgroup Analysis
Sensitivity Analysis 2y
Higher Cognitive Status
I. Farag et al. / JAMDA xxx (2015) 1e7
Base Case Analysis
4
The incremental cost-effectiveness of the program relative to usual care was $A22,958 per person showing an improvement in mobility performance. Figure 1 illustrates the results of the bootstrapping in a scatterplot and cost-effectiveness acceptability curve. The acceptability curve demonstrates that the intervention has an 80% probability of being cost-effective relative to the control at a threshold of $A48,000 per extra person achieving mobility improvement. The cost-effectiveness analysis also indicated an ICER of $A19,020 per extra person reporting an improvement in overall health status with an 80% probability of cost-effectiveness at a threshold value of $A36,000. The ICER for the QALY outcome was $A77,403 per QALY gained in the intervention compared with the control group (Table 3). The acceptability curve demonstrates that there is no threshold value at which the program can be considered as having an 80% probability of cost-effectiveness for the QALY outcome (Appendix). The subgroup analysis of participants with better cognitive status (MMSE >28) demonstrated improved cost-effectiveness for all outcome measures. The ICER was $A6754 per extra person experiencing a mobility improvement, with an 80% probability of costeffectiveness at a threshold value of $A22,000. The ICER was $A4650 per extra person experiencing an improvement in selfreported health with an 80% probability of cost-effectiveness at a threshold value of $A12,000. For the QALY measure, the obtained ICER was $A20,346 per QALY gained (Table 3) with an 80% probability of cost-effectiveness at a threshold of $A80,000 per QALY gained. Removal of the low-care residential costs of $A11,472 from the cost data for participants residing in a hostel at baseline did not markedly affect the ICER value for all outcome measures (Table 3). Full exclusion of those participants who were hostel residents at baseline results in lower cost-effectiveness ratios with an ICER of $A44,496 per QALY gained. Discussion This study estimated a cost of $A22,958 per person showing an improvement in mobility and an 80% probability of cost-effectiveness at a willingness to pay value of $A48,000. For other outcome measures, there was a cost of $A19,020 per person indicating an improvement in self-reported health and $A77,403 per QALY gained. With a decision-maker’s willingness to pay $A36,000 for an additional health outcome, there was an 80% probability of being costeffective relative to the control for the self-reported health status measure. There was no threshold value at which the intervention could be considered as having an 80% probability of cost-effectiveness for the QALY outcome. Cost-effectiveness for all outcomes was greater in the subgroup with better cognitive status, with the results indicating an incremental cost of $A6754 per person showing mobility improvement, $A4650 per person indicating an improvement in selfreported health and $A20,346 per QALY gained.
I. Farag et al. / JAMDA xxx (2015) 1e7
B
12000
Incremental cost $A
8000 4000
-0.1
0.1
0.2
0.3
0.4
-4000 -8000 -12000
Incremental patients with improvement in mobility
Probability cost-effective
A
5
1.0 0.8 0.6 0.4 0.2 0.0
0
20000
40000
60000
80000
100000
Willingness to pay per person with improved mobility $A
Fig. 1. Cost-effectiveness of the program for mobility improvement. (A) Incremental cost-effectiveness plane for point estimate (black circle) and 1000 bootstrapped pairs (gray circles); (B) cost-effectiveness acceptability curve for mobility improvement. Solid line, all participants; dashed line, subgroup with higher cognition MMSE >28.
There are a limited number of economic evaluations conducted to date that have considered cost-effectiveness in terms of functional improvements.26 Several studies have used quality of life as an outcome measure,27e29 and many more studies have considered costeffectiveness in terms of the cost per fall prevented.12,13,28,30,31 There is wide variation in the results of the incremental cost per fall prevented, with the range spanning from $A294 (2008)32 to $NZ1803 (1998).33 The variation in the values presented is reflective of heterogeneity between studies, with differences in the health care systems between countries, in the study perspective, which in turn influences the cost consideration and in the type of program intervention. Nevertheless, for many of the studies there is an indication of significant potential for cost-effectiveness or cost savings of fall prevention programs.4 In our study, we have not been able to provide a cost per fall averted, as falls increased in the intervention group. It is interesting to note, however, that the higher fall rates in the intervention group did not necessarily translate to a decrement in quality of life or in the self-reported measure of health status. Our exercise program resulted in a greater number of participants showing improvement on the mobility measure in the intervention group. These results suggest a substantial improvement in the objective assessment of functional ability using the SPPB measure, which has been shown to be an accurate indicator of risk of disability, mortality, and nursing home admission in the older population.17 Our results also indicate an improvement in self-rated health and in quality of life for the intervention group; whether this is deemed cost-effective is dependent on the “willingness to pay” of the funder for the health outcomes gained. Across all health outcomes, the ICERs for the subgroup with better cognitive status were more favorable. In addition, cost per QALY gained from our trial was also comparable to values obtained in modeling studies.34 The improved cost-effectiveness for the subgroup with better cognitive status is due to larger incremental differences in the outcome measures. This suggests that better cognitive capacity may be associated with a greater potential to benefit from the intervention. As the study excluded people with cognitive impairment, no conclusions can be drawn about the potential benefits of the intervention in this group. The sensitivity analysis with exclusion of participants residing in a hostel at baseline also produced improved cost-effectiveness. This was largely driven by the impact on the cost data, with higher nonehostel-related health care costs accrued by the small number of hostel residents. Removal of the annual hostel costs only from the analysis did not affect cost-effectiveness. There appears to be substantial use of services in older people after hospital discharge. This is consistent with the previously noted high growth in expenditures for nonhospital35 and hospital-related36
services by older people in the latter decades of life, but it is not yet clear whether any of these costs are avoidable. The cost of fall prevention programs vary from (2009, $A) $A500 for Tai Chi, to home-based exercise programs, which are estimated to cost an average of $A1233 per participant.37 Multifactorial interventions are estimated to cost $A1886 per participant.37 At a program cost of $A751 (2012, $A) per participant, our program cost is at the lower end of this range. Previous modeling studies undertaken by the authors indicate that cost-effectiveness is markedly susceptible to the cost of the program.38 The factors that influence program costs include frequency and mode of delivery (eg, face-to-face or by telephone or telemedicine), whether group-based or individual, wage rate of the professional delivering the program. Cost-effectiveness would be optimized by reducing the cost of the program delivery, providing that the effectiveness of the program is not compromised with alternative delivery modes. In this study, costs were considered only from the service provider perspective and therefore no estimate was made of the cost of informal care services. Further research with inclusion of costs from the societal perspective would provide a better indicator of the impact of falls on the wider community. Conclusion The exercise intervention appeared to offer reasonable value for money for mobility outcomes and self-reported health status but whether this is deemed to be cost-effective is dependent on the “willingness to pay” of the funder for the health outcomes gained. Value for money improved for all measures in the subgroup with higher cognitive status. References 1. Mahoney J, Sager M, Dunham NC, Johnson J. Risk of falls after hospital discharge. J Am Geriatr Soc 1994;42:269e274. 2. Gill TM, Allore HG, Holford TR, Guo Z. Hospitalization, restricted activity, and the development of disability among older persons. JAMA 2004;292: 2115e2124. 3. Australian hospital statistics 2011e12. Canberra, Australia: Australian Institute of Health and Welfare; 2013. 4. Gillespie LD, Robertson CM, Gillespie WJ, et al. Interventions for preventing falls in older people living in the community. Cochrane Database Syst Rev; 2012:CD007146. 5. Liu CJ, Latham NK. Progressive resistance strength training for improving physical function in older adults. Cochrane Database Syst Rev; 2009:CD002759. 6. Howe TE, Shea B, Dawson LJ, et al. Exercise for preventing and treating osteoporosis in postmenopausal women. Cochrane Database Syst Rev; 2011: CD000333. 7. Investigators LS, Pahor M, Blair SN, et al. Effects of a physical activity intervention on measures of physical performance: Results of the lifestyle
6
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18. 19.
20.
21.
22.
I. Farag et al. / JAMDA xxx (2015) 1e7 interventions and independence for Elders Pilot (LIFE-P) study. J Gerontol A Biol Sci Med Sci 2006;61:1157e1165. Timonen L, Rantanen T, Ryynanen OP, et al. A randomized controlled trial of rehabilitation after hospitalization in frail older women: Effects on strength, balance and mobility. Scand J Med Sci Sports 2002;12:186e192. Vogler CM, Sherrington C, Ogle SJ, Lord SR. Reducing risk of falling in older people discharged from hospital: A randomized controlled trial comparing seated exercises, weight-bearing exercises, and social visits. Arch Phys Med Rehabil 2009;90:1317e1324. Latham N, Anderson C, Bennett D, Stretton C. Progressive resistance strength training for physical disability in older people. Cochrane Database Syst Rev; 2003:CD002759. Sherrington CLS, Vogler CM, Close JCT, et al. A post-hospital home exercise program improved mobility but increased falls in older people: A randomised controlled trial. PLOS One 2014;9:e104412. Irvine L, Conroy SP, Sach T, et al. Cost-effectiveness of a day hospital falls prevention programme for screened community-dwelling older people at high risk of falls. Age Ageing 2010;39:710e716. Davis JC, Robertson MC, Ashe MC, et al. Does a home-based strength and balance programme in people aged > or ¼80 years provide the best value for money to prevent falls? A systematic review of economic evaluations of falls prevention interventions. Br J Sports Med 2010;44:80e89. Smith RD, Widiatmoko D. The cost-effectiveness of home assessment and modification to reduce falls in the elderly. Aust N Z J Public Health 1998;22: 436e440. Church J, Goodall S, Norman R, Haas M. The cost-effectiveness of falls prevention interventions for older community-dwelling Australians. Aust N Z J Public Health 2012;36:241e248. Sherrington C, Lord SR, Vogler CM, et al. Minimising disability and falls in older people through a post-hospital exercise program: A protocol for a randomised controlled trial and economic evaluation. BMC Geriatr 2009;9:8. Guralnik JM, Simonsick EM, Ferrucci L, et al. A short physical performance battery assessing lower extremity function:association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol 1994;49:M85eM94. Dolan P. Modeling valuations for EuroQol health states. Med Care 1997;35: 1095e1108. NSW Health Service Health Professionals (State) Award. NSW Department of Health. Available at: www.health.nsw.gov.au/careers/conditions/Pages/P.aspx; 2013. Accessed December 6, 2014. Individuals’ income deduction 2013. Available at: https://www.ato.gov.au/ Individuals/Income-and-deductions/In-detail/Deductions-for-work-related-ex penses/Claiming-a-deduction-for-car-expenses-using-the-cents-per-kilometremethod/. Accessed July 15, 2013. Medicare Benefits Schedule, Commonwealth of Australia. Department of Health. 2011. Available at: www.medicareaustralia.gov.au/provider/medicare/ mbs.jsp. Accessed July 2, 2013. National Hospital Cost Data Collection Cost Report Round 14 (2009e10). Canberra, Australia: Commonwealth Department of Health and Ageing; 2011.
23. Cost of residential aged care in Australia. Available at: https://www.dss.gov.au/ our-responsibilities/ageing-and-aged-care/aged-care-funding/aged-care-feesand-charges; 2013. Accessed December 12, 2013. 24. Farag I, Sherrington C, Ferreira M, Howard K. A systematic review of the unit costs of allied health and community services used by older people in Australia. BMC Health Serv Res 2013;13:69. 25. World Health Organization. Macroeconomics and health: Investing in health economic developmentdreport of the commission on macroeconomics and health. Geneva, Switzerland: World Health Organization; 2001. 26. Melis RJ, Adang E, Teerenstra S, et al. Cost-effectiveness of a multidisciplinary intervention model for community-dwelling frail older people. J Gerontol A Biol Sci Med Sci 2008;63:275e282. 27. Hendriks MR, Evers SM, Bleijlevens MH, et al. Cost-effectiveness of a multidisciplinary fall prevention program in community-dwelling elderly people: A randomized controlled trial (ISRCTN 64716113). Int J Technol Assess Health Care 2008;24:193e202. 28. Sach TH, Foss AJ, Gregson RM, et al. Falls and health status in elderly women following first eye cataract surgery: An economic evaluation conducted alongside a randomised controlled trial. Br J Ophthalmol 2007;91:1675e1679. 29. Sach TH, Logan PA, Coupland CA, et al. Community falls prevention for people who call an emergency ambulance after a fall: An economic evaluation alongside a randomised controlled trial. Age Ageing 2012;41:635e641. 30. Robertson MC, Gardner MM, Devlin N, et al. Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 2: Controlled trial in multiple centres. BMJ 2001;322:701e704. 31. Davis JC, Robertson MC, Ashe MC, et al. International comparison of cost of falls in older adults living in the community: A systematic review. Osteoporos Int 2010;21:1295e1306. 32. Haines TP, Hill AM, Hill KD, et al. Cost effectiveness of patient education for the prevention of falls in hospital: Economic evaluation from a randomized controlled trial. BMC Med 2013;11:135. 33. Campbell AJ, Robertson MC, Gardner MM, et al. Falls prevention over 2 years: A randomized controlled trial in women 80 years and older. Age Ageing 1999;28: 513e518. 34. Church J, Goodall S, Norman R, Haas M. An economic evaluation of community and residential aged care falls prevention strategies in NSW. N S W Public Health Bull 2011;22:60e68. 35. Payne G, Laperte A, Deber R, Coyte PC. Counting backwards to health care’s future: Using time to death modelling to identify changes in end-of-life morbidity and the impact of aging on health care expenditures. Milbank Q 2007;85:213e257. 36. Kardamanidis K, Lim K, Da Cunha C, et al. Hospital costs of older people in New South Wales in the last year of life. Med J Aust 2007;187:383e386. 37. Day L, Hoareau E, Finch C, et al. Modelling the impact, costs and benefits of falls prevention measures to support policy-makers and program planners. Monash University Accident Research Centre. Available at: www.monash.edu.au/miri/ research/reports/muarc286.html; Accessed December 15, 2014. 38. Farag I, Howard K, Ferreira M, Sherrington C. Economic modelling of a public health program of fall prevention [published online ahead of print December 18, 2014]. Age Ageing
I. Farag et al. / JAMDA xxx (2015) 1e7
7
Appendix
Table A1 Unit Costs of Services (Base Year 2012, $A) Service
Unit Cost (Average)
Residential aged care (high) Residential aged care (low) Hospital Emergency Department presentation General practitioner consultation room General practitioner consultation home-visit Specialist Nursing Allied health service consultation room Allied health service home-visit Showering/dressing Meals Domestic services Transport Shopping
$94.79 per bed day23 $11,472 per annum23 AR-DRG version 5.222 $45122 $70.3021 $81.8521 $105.4821 $56.1821 $55.4421 $55.4421 $36.4036 $11.1036 $39.0736 $12.3936 $36.4036
A
B Probability cost-effective
Probability cost-effective
1.0 0.8 0.6 0.4 0.2 0.0
0
20000
40000
60000
80000
100000
Willingness to pay per person with improved health status $A
1.0 0.8 0.6 0.4 0.2 0.0
0
20000
40000
60000
80000
100000
Willingness to pay per QALY gained $A
Fig. A1. Cost-effectiveness acceptability curves (A) per person showing improvement in self-reported health status and (B) per QALY gained. Solid line, all participants; dashed line, subgroup with higher cognition MMSE >28.
