Physical activity has a multitude of benefits for older people, including risk reduction for type 2 diabetes, obesity, falls, cardiovascular disease, and mortality.1 In addition to the impacts on physical health, physical activity can also improve mental health and reduce cognitive decline.2 Although benefits to the individual are vitally important, one must also consider the broader economic benefits to society. Inactivity places a considerable financial burden on health care systems. Internationally, 0.3% to 4.6% of national health care expenditure is attributed to inactivity,3 and this cost is estimated at an international INT$53.8 billion (INT: International dollars) per year (2013).4 In addition to the health care costs associated with inactivity, costs associated with loss of productivity for people of working age are estimated to be an international INT$13.7 billion per year (2013).4
The outdoor built environment includes parks and other outdoor spaces that can be used for physical activity, recreation, and leisure5,6 and has the potential to impact positively on inactivity and subsequently reduce health care costs.6 Population density, diversity of land use, availability of destinations, distance to transit, design, and neighborhood walkability impact utilization of the built environment and the subsequent health care savings.6 Over the past 2 years, the COVID-19 pandemic has impacted indoor physical activities, lending further support to the importance of the outdoor built environment in engaging older people in physical activity.7 Incorporating purposefully designed and built age-friendly outdoor exercise equipment that includes exercise stations specific to the needs of older people, such as the Seniors Exercise Park,8 can further promote physical and functional benefits. The Seniors Exercise Parks are specifically designed to incorporate exercise stations targeting balance, strength, joint mobility, functional movements, and dexterity.9
The Exercise interveNtion outdoor proJect in the cOmmunitY (ENJOY) study has shown that a 12-week progressive physical activity intervention utilizing a Seniors Exercise Park provides many benefits to the end users, including improving and sustaining physical activity and well-being9 and reducing the risk of falls.10 In addition, 61% of participants continued to exercise when a supervised maintenance program was provided.11 However, it is unknown whether participation in the ENJOY study led to changes in health care utilization or productivity associated with paid employment. The current economic evaluation aims to determine the cost-utility (quality of life) and the cost-effectiveness (falls avoided) of the physical and social activity intervention by comparing the health care utilization for a 6-month period immediately prior to participation in the program and a 6-month period after the end of the 6-month program (Figure 1).
—Timeline for the project. ENJOY indicates Exercise interveNtion outdoor proJect in the cOmmunitY.
Citation: Journal of Physical Activity and Health 20, 6; 10.1123/jpah.2022-0380
Methods
Detailed methodology for the ENJOY Project has been published elsewhere.9,12 The economic evaluation has been prepared according to the Consolidated Health Economic Evaluation Reporting Standards: Consolidated; Health Economic Evaluation Reporting Standards (CHEERS) checklist13 Supplementary Table 1 (available online). Ethical approval was obtained prior to data collection from the Melbourne Health Human Research Ethics Committee, Melbourne (Application ID: HREC/18/MH/286, local number 2018.238). Participants provided written informed consent to participate in the project and separate written informed consent to allow access to participants’ linked health care utilization data.
Aims and Design
The primary aim of the economic evaluation was to determine the cost-utility of the Seniors Exercise Park physical activity intervention by comparing quality of life and the cost of health care utilization for 6 months prior to participation in the program (block 1) with the 6-month period beginning 6 months after commencing the program (block 3; Figure 1, Table 1).
Time Horizon for Health Care Utilization and Outcome Measures Before and After the ENJOY Intervention
| 6 mo Pre block 1 | Block 1: 6 mo | Block 2: 6 mo | Block 3: 6 mo | |||
|---|---|---|---|---|---|---|
| Intervention | No intervention | ENJOY primary intervention (3 mo) | ENJOY maintenance phase (optional for 6 mo) | ENJOY maintenance phase (optional for 6 mo) | No intervention | |
| Utility index | Administered at the end of block 1 | Administered at the end of block 3 | ||||
| Falls | Retrospective recall of falls for 12 mo prior to commencement of intervention participation | Prospective recording of falls for the 12 mo following commencement of intervention participation | ||||
| MBS—funded health care | ✓ | ✓ | ||||
| PBS—funded pharmaceuticals | ✓ | ✓ | ||||
| Hospitalizations (self-reported) | ✓ | ✓ | ||||
| Home nursing, allied health, and community services (self-reported) | ✓ | ✓ | ||||
| Productivity costs | Questionnaire administered at the end of block 1; extrapolated to a 6-mo time horizon | Questionnaire administered at the end of block 3; extrapolated to a 6-mo time horizon | ||||
Abbreviations: ENJOY, Exercise interveNtion outdoor proJect in the cOmmunitY; MBS, Medicare Benefits Schedule; PBS, Pharmaceutical Benefits Scheme.
The secondary aim of the economic evaluation was to determine the cost-effectiveness of the ENJOY intervention by comparing the number of falls (falls avoided in the 12 mo after ENJOY commencement compared with the 12 mo prior) and the health care utilization for a 6-month period prior to participation in the program (block 1) and a 6-month period beginning 6 months after commencing the program (block 3; Table 1).
This was a pre–post study design for a single cohort. The economic evaluation used an incremental cost-utility analysis for the primary aim and an incremental cost-effectiveness analysis for the secondary aim, wherein the outcomes used to assess effectiveness were quality-adjusted life years (QALYs) and falls, respectively. Both analyses took a societal perspective inclusive of government-funded health care and pharmaceuticals under the Medicare Benefits Schedule (MBS, a listing of the medical services subsidized by the Australian Government) and Pharmaceutical Benefits Scheme (PBS, medicines utilized at a government-subsidized price), respectively, in addition to hospitalizations, community-based nursing and allied health, and general community services. The economic analyses also included productivity costs.14 The government data (MBS/PBS) were accessed from Services Australia (https://www.servicesaustralia.gov.au/).
All cost data have been reported in AUD 2020–2021 (unless otherwise stated) to represent the most recent cost year for data collection. The Australian Bureau of Statistics consumer price index was used to convert cost data prior to 2020–2021 into a net present value of 2020–2021 using year on year overall consumer price index from the December quarter. This was to avoid the January to June 2020 impact of the COVID-19 pandemic, which was excluded from the primary analysis (https://www.abs.gov.au/statistics/economy/price-indexes-and-inflation/consumer-price-index-australia). A discount rate was not applied to the current data set as the time horizon for an individual participant was limited to 18 months.
Time Horizon for the Outcome and Health Utilization Data
The time horizon for the cost data covered 18 months, which has been divided into three 6-month blocks. Block 1 covered the baseline data in the 6 months prior to enrollment in the study, block 2 covered the ENJOY primary intervention (12-wk structured, supervised program) and part of the ENJOY maintenance phase (3 mo out of the 6-mo maintenance phase), and block 3 covered the following 6-month period (Table 1). Blocks 1 and 3 were the same months of the year to limit any effect of seasonal differences (albeit 1 y apart). Productivity data were reported at the end of block 1 and the end of block 3. The 2 time points were modeled to represent the 6-month period assuming that employment was steady over these 2 periods. Details for estimating the resources and costs are provided in Table 2.
Economic Evaluation Unit Costs and Data Source
| Unit description | Unit cost ($AUD 2020–2021) | Data source | |
|---|---|---|---|
| Cost of ENJOY; 3 mo | Participation in the ENJOY primary program over 12 wk based on attendance at 24 sessions (2 sessions per week over 12 wk) Costs: 24 sessions supervised by an exercise physiologist or physiotherapist. Sessions were an average of 1 h with an additional 30 min for morning tea and 30 min for travel time. Cost of morning tea was $2 per session (simple biscuits and instant coffee bulked purchased) | $140 × 100.9% (1 FY of CPI) = $141.26 per hour One group: 24 sessions (12 wk × 2 sessions per week) = staff $141.26 × 24 sessions × 2 h = $6780.48 One group: Morning tea = $2 × 24 sessions = $48 Total cost per group = $6828.48 Cost per participant = $6828.48/8 participants = $853.56 | Home care packages, fees, and charges 2019 https://www.myagedcare.gov.au/sites/default/files/attachments/1-X3J7G9L_2.pdf. Cost data from 2019 to 2020 inflated by CPI to 2020–2021 Ninety-five people enrolled in the 12 groups (4 at each of the 3 sites).9 Based on an intention to treat, there were 8 people per group; therefore, the cost per person was the total cost of a group, divided by 8 |
| Cost of ENJOY maintenance phase; 6 mo | Participants who used the exercise park regularly during the 6-mo post ENJOY11 Costs: 48 sessions supervised by an exercise physiologist or physiotherapist. Sessions were an average of 1 h with an additional 30 min for travel time. There was no morning tea during the maintenance phase | $140 × 100.9% (1 FY of CPI) = $141.26 per hour One group: 48 sessions (24 wk × 2 sessions per week) = staff $141.26 × 48 sessions × 1.5 h = $10,170.72 Cost per participant = $10,170.72/4 participants = $2542.68 | Home care packages, fees, and charges 2019 https://www.myagedcare.gov.au/sites/default/files/attachments/1-X3J7G9L_2.pdf. Cost data from 2019 to 2020 inflated by CPI to 2020–2021 Fifty-five people from the full study11 regularly used the exercise parks during the maintenance phase across the 12 groups (4 at each of the 3 sites).11 On average, there were 4 people per group; therefore, the cost per person was the total cost of a group, divided by 4 |
| MBS; 6-mo pre and post | Commonwealth government-funded health care. Units were measured as the number of individual health services funded | Variable depending on the MBS schedule for each service type | MBS administrative data set http://www.mbsonline.gov.au/internet/mbsonline/publishing.nsf/Content/Home. Cost data prior to 2020–2021 inflated by CPI |
| PBS; 6-mo pre and post | Commonwealth government-funded pharmaceuticals. Units were measures as the number of individual pharmaceuticals funded | Variable depending on the PBS schedule for each pharmaceutical type | PBS administrative data set https://www.pbs.gov.au/pbs/home;jsessionid=bpxlj875m89pt9fqxs7v7le. Cost data prior to 2020–2021 inflated by CPI |
| Hospitalizations; 6-mo pre and post | Self-reported. Number of days admitted to an acute or rehabilitation hospital. Acute hospitalizations, which included an intensive care unit admission, were adjusted accordingly. Acute admissions were measured in days, rehabilitation admissions were measured per admission | Acute 1 day = $5027/national average 2.4 LOS × 102.7% (2 FY of CPI) = $2151.48 Rehabilitation 1 admission = $13,974 × 102.7% (2 FY of CPI) = $14,353.56 | National Hospital Cost Data Collection Report Round 23 (Financial Year 2018–2019) https://www.ihpa.gov.au/sites/default/files/publications/round_23_2018-19_nhcdc_report_public_sector.pdf. Cost data from 2018 to 2019 inflated by CPI to 2020–2021 |
| Home nursing services; 6-mo pre and post | Self-reported. Home nursing services utilized. Units were measured as the number of individual home nursing services received | $82.00 × 100.9% (1 FY of CPI) = $82.74 per session | Home care packages, fees, and charges 2019 https://www.myagedcare.gov.au/sites/default/files/attachments/1-X3J7G9L_2.pdf. Cost data from 2019 to 2020 inflated by CPI to 2020–2021 |
| Allied health services; 6-mo pre and post | Self-reported. Allied health services utilized. This excluded allied health services funded under a chronic disease management plan approved by the treating GP as these plans were funded by MBS and have been captured under this cost category. Units were measured as the number of individual allied health services received | $140 × 100.9% (1 FY of CPI) = $141.26 per session | Home care packages, fees, and charges 2019 https://www.myagedcare.gov.au/sites/default/files/attachments/1-X3J7G9L_2.pdf. Cost data from 2019 to 2020 inflated by CPI to 2020–2021 |
| Community services; 6-mo pre and post | Self-reported. Community services included a range of services, such as cleaning, gardening, and home help. Units were measured as the number of individual community services received | $46.00 × 100.9% (1 FY of CPI) = $46.41 per session | Home care packages, fees, and charges 2019 https://www.myagedcare.gov.au/sites/default/files/attachments/1-X3J7G9L_2.pdf. Cost data from 2019 to 2020 inflated by CPI to 2020–2021 |
| Productivity costs; 6-mo pre and post | Self-reported. Reported as hours per week employed | $1711.60/40 h = $42.79 per hour | The Australian Bureau of Statistics reported an average weekly earnings of $1711.60 per full-time week in 2020/2021 https://www.abs.gov.au/statistics/labour/earnings-and-work-hours/average-weekly-earnings-australia/latest-release |
Abbreviations: CPI, consumer price index; ENJOY, Exercise interveNtion outdoor proJect in the cOmmunitY; FY, financial year; GP, general practitioner; LOS, length of stay; MBS, Medicare Benefits Schedule; PBS, Pharmaceutical Benefits Scheme.
Outcomes
The primary outcome was the 5-level EQ-5D (EQ-5D-5L) utility score. The EQ-5D-5L is a multiattribute quality of life measure with scores that can be converted into a utility index to calculate QALYs.15 The EQ-5D-5L was reported at the end of block 1 and the end of block 3.
The secondary outcome was based on the number of falls reported for 12 months prior to participation (retrospective recall) and 12 months post commencement of participation (prospectively recorded using monthly calendars10). Falls were defined as an event when the participant “inadvertently comes to rest on the ground, floor or other lower level.”16
Setting
The study took place in 3 outdoor spaces in the community in Melbourne, Australia, which were freely accessible to participants for the duration of the study. Two sites were at public areas, and a third site was at a residential aged care facility.12
Intervention
The Seniors Exercise Park physical and social activity intervention has been detailed elsewhere.9,12 In summary, the ENJOY project was a multisite study, and the intervention included 12 weeks of a structured, supervised exercise program (referred to here as the ENJOY primary intervention) at 3 designated outdoor exercise parks (each with Seniors Exercise Park equipment installed) followed by 6 months of an optional maintenance exercise program, which offered unstructured supervised exercises at the same exercise park (ENJOY maintenance phase).
Target Population
Inclusion and exclusion criteria are described in the main study protocol.12 Participants were Australians aged 60 years or older and were either concerned about falling or had experienced one or more falls in the 12 months before participating. In addition, participants were independent with their mobility, did not use a gait aid more than a single point stick, and did not have cognitive impairment.
Participants from the main study were included in the economic evaluation if they had complete cost, falls, and quality of life data and provided additional consent to authorize access to their health care and medications utilization (MBS and PBS data). During data collection, participants enrolled in the latter parts of the ENJOY study were impacted by the COVID-19 pandemic lockdown restrictions, reducing the ability to participate in the optional maintenance exercise program between February 2020 and November 2020. Due to the potential impact of the restrictions on accessing the parks and on utilization of health care services by older people,17,18 only the data of the participants who were not impacted by the COVID-19 pandemic lockdown restrictions were included in the analysis.
Statistical Analysis
Quality of life data were presented as a utility index to calculate QALYs. To calculate the utility index, the EQ-5D-5L scores were converted via an algorithm into a utility index using normative UK data,15 as there was no algorithm available at the time of the analysis to base this analysis on normative Australian data. A utility index ranges from 0 to 1, where 0 = a state equal to death and 1 = a state of perfect health; however, negative values are possible when a health state is considered worse than death.19 Quality of life and cost data were analyzed using a paired t test for parametric data as this format supported the subsequent incremental cost-effectiveness ratio analysis. Due to the nonnormal distribution of the falls data, a related-samples Wilcoxon signed-rank test for nonparametric data was used. However, as the falls data were also used in the incremental cost-effectiveness ratio analysis, it was also reported as a mean value via a paired t test.
The incremental cost-effectiveness ratio was calculated via bootstrapping (5000 repetitions), utilizing the central limit theorem.20 Results were presented graphically with a point estimate and 50%, 75%, and 95% confidence ellipses. The graph is divided into 4 quadrants, with the 2 lower quadrants indicating that block 3 (postintervention) was associated with lower costs and the 2 right-sided quadrants indicating that block 3 (postintervention) was associated with greater effect compared with block 1 (preintervention) and vice versa. This meant that the bottom right-hand quadrant was associated with lower costs and greater effect for block 3 (postintervention), making it the dominant quadrant. The economic evaluation did not have a power calculation as it was a sample of convenience based on the main study.12 Significance was assumed at P < .05. The economic evaluation used SPSS (version 27)21 and a customized version of Microsoft Excel20 for the analyses.
Characterizing Uncertainty
Three sensitivity analyses were undertaken to characterize uncertainty for both the cost and effect data.
- (1)Effect: Compared with prospective falls data, the retrospective recall of falls data for the 12 months prior to the intervention was subject to recall bias associated with underreporting, with a previous study reporting that only 85% of community falls were recalled over a 12-month period22 (indicating that falls numbers should be multiplied by 17.6% for a number comparable with that identified through prospective reporting). As such, the retrospective falls data were inflated by 17.6% in a sensitivity analysis.
- (2)Cost: It was recognized that the self-recalled health service utilization might have been subject to recall bias. A previous study has shown over a 6-month period that the recall of health service utilization can be associated with both underreporting (36%) and overreporting (35%).23 Within the current study, general practitioner (GP) visits were both self-reported and reported via the MBS administration system. GP visits were compared between the 2 data sets to determine whether recall bias was present in the current cohort. The degree of recall bias then informed adjustments to the self-reported health service utilization in a sensitivity analysis.
- (3)Cost: Costs for the maintenance phase were based on the total costs divided by the actual attendees during the 6-month ENJOY maintenance phase. In the real world, these sessions would host 6 to 10 participants, with an average of 8. As such, the cost per person to participate in the ENJOY maintenance phase was adjusted to reflect between 2 and 12 participants per session in a sensitivity analysis.
- (4)Cost and effect: Eleven participants were excluded due to missing data. This included all missing MBS/PBS data (as participants did not consent to this additional data collection) as well as 4 missing quality of life scores at the end of block 3. To investigate the robustness of the findings, a sensitivity analysis was conducted that included these additional 11 participants. Missing MBS/PBS data were imputed using multiple imputations.21 Missing quality of life scores at the end of block 3 were imputed using the quality of life at baseline to assume no change to quality of life. Key cost and effect variables were analyzed as independent t tests to report differences between the primary cohort (n = 50) and the additional participants (n = 11) as well as by repeating the paired t tests to report baseline to follow-up differences between the primary cohort (n = 50) and the newly combined cohort (n = 61).
Results
From the 80 participants who completed the intervention, complete economic data were available for 50 participants, including the cost, falls, and quality of life data. These 50 participants had an average age of 72.8 years (SD 7.4 y), and 78.0% (n = 39/50) were women. Although 100% (n = 50/50) participated in the Seniors Exercise Park intervention, only 60% (n = 30/50) utilized the Seniors Exercise Park during the maintenance phase.
There was a small, nonsignificant, observed increase in utility index of 0.011 (95% CI, −0.034 to 0.056; P = .631) between blocks 1 and 3, indicating that 100 participants were needed to treat for a QALY gained (Table 3). There was a nonsignificant reduction in falls of −0.50 (95% CI, 0.00 to −0.50; P = .160), favoring the 12 months following commencement of the intervention, indicating that 2 participants were needed to treat for a fall avoided (Table 3).
Falls and Utility Index Scores
| Block 1 | Block 3 | Mean difference; post block 3 minus block 1 (95% CI) | P | |
|---|---|---|---|---|
| Utility index | 0.755 (SD 0.203) | 0.766 (SD 0.180) | 0.011 (−0.034 to 0.056) | .631 |
| Falls (over 12 mo) | ||||
| Median (IQR) | 0.5 (0.5–1) | 0.0 (0.0 to 1.0) | −0.5 (0.0 to −0.5) | .160 (nonpara) |
| Mean (SD) | 0.76 (SD 0.96) | 0.54 (SD 1.13) | −0.22 (−0.61 to 0.17) | .263 (para) |
Abbreviations: CI, confidence interval; IQR, interquartile range.
The cost of the ENJOY primary intervention (12-wk structured supervised physical activity) was $853.56 (SD $0.00; n = 50) per participant as the total cost was equally attributed to all participants. The cost of the ENJOY maintenance program was $2542.68 (SD $0.00; n = 30) per participant when only attributing costs to the 30 participants who completed the maintenance period. However, the ENJOY maintenance program cost was $1525.61 (SD $1258.30; n = 50) per participant when attributing costs to all 50 participants who had the option of completing the maintenance period with a combined intervention/maintenance cost of $2379.17.
The Overall Cost of ENJOY (12-wk Primary Intervention and 24-wk Maintenance Phase)
Across all health and service utilization categories, block 1 average cost was $9764.49 (SD $26,033.35), and block 3 was $5179.30 (SD $3826.64), indicating a nonsignificant reduction of −$4585.20 (95% CI, −$12,113.99 to $2943.59; P = .227). The main cost difference was attributable to higher acute and rehabilitation admissions in block 1; reduced MBS benefits, allied health, and community services in block 3; and improved productivity in block 3 (Table 4).
Utilization and Cost Data (MBS, PBS, Hospitalizations, Home Nursing, Allied Health, Community Support, and Productivity)
| Utilization | Costs | |||||||
|---|---|---|---|---|---|---|---|---|
| Block 1 Mean (SD) | Block 3 Mean (SD) | Mean difference; block 3 minus block 1 (95% CI) | P | Block 1 Mean (SD) | Block 3 Mean (SD) | Mean difference; block 3 minus block 1 (95% CI) | P | |
| MBS, number of services; benefit paid | 20.98 (13.99) | 21.34 (13.50) | 0.36 (−2.89 to 3.61) | .825 | $1422.54 ($1715.76) | $1185.46 (824.42) | −$237.07 (−662.47 to $188.33) | .268 |
| MBS; patient out of pocket | $351.56 ($1103.11) | $182.89 ($374.49) | −$168.67 (−$498.09 to $160.74) | .309 | ||||
| PBS, number of prescriptions; benefit paid | 25.20 (20.19) | 26.00 (20.32) | 0.80 (−2.15 to 3.75) | .588 | $838.27 ($1365.31) | $839.52 ($1171.77) | $1.25 (−$143.01 to $145.51) | .986 |
| PBS; patient out of pocket | $151.91 ($132.34) | $153.53 ($150.99) | $1.63 (−$22.38 to 25.65) | .892 | ||||
| Acute hospitalizations, number of days | 2.64 (10.98) | 0.22 (0.82) | −2.42 (−5.52 to 0.678) | .123 | $5649.91 ($23,434.67) | $473.33 ($1754.34) | −$5206.58 (−$11,872.74 to $1459.58) | .123 |
| Rehabilitation hospitalizations, number of admissions | n = 2 admissions | n = 0 admissions | $574.14 ($2841) | $0.00 ($0.00) | −$574.14 (−1381.62 to $233.34) | .159 | ||
| Home nursing services, number of services | 0.02 (0.14) | 0.02 (0.14) | 0.00 (−0.06 to 0.06) | 1.000 | $1.65 (11.70) | $1.65 (11.70) | $0.00 (−$4.75 to $4.75) | 1.000 |
| Allied health, number of services | 4.62 (8.53) | 2.90 (5.72) | −1.72 (−3.58 to 0.24) | .083 | $652.62 ($1205.03) | $409.65 ($807.83) | −$242.97 (−$519.21 to 22.38) | .083 |
| Community services, number of services | 1.98 (5.19) | 0.46 (2.04) | −1.52 (−2.90 to −0.15) | .031 | $91.89 ($240.79) | $21.35 ($94.79) | −$70.54 (−$134.34 to −$6.74) | .031 |
| Productivity (represented as a negative number as productivity reduces costs–society) | n = 3 in paid employment | n = 2 in paid employment | −$1430.41 ($6690.67) | −$1907.21 ($9352.40) | −$676.80 (−$1.435.48 to $481.87) | .322 | ||
| Combined costs (excluding cost of ENJOY intervention and maintenance) | $9764.49 (SD $26,033.35) | $2800.13 (SD $3565.90) | −$6964.36 (95% CI −$14,450.39 to $521.66) | .068 | ||||
| Combined costs (including cost of ENJOY intervention and maintenance) | $9764.49 (SD $26,033.35) | $5179.30 (SD $3826.64) | −$4585.20 (95% CI −$12,113.99 to $2943.59) | .227 | ||||
Abbreviations: CI, confidence interval; ENJOY, Exercise interveNtion outdoor proJect in the cOmmunitY; MBS, Medicare Benefits Schedule; PBS, Pharmaceutical Benefits Scheme. Note: Bold values are significant.
The cost-utility analysis (quality of life) and the cost-effectiveness analysis (falls) reported that ENJOY was a dominant intervention (Figures 2 and 3, respectively). Although the point estimate was observed in the dominant quadrant (ie, the bottom right-hand quadrant, which represented lesser cost and greater effect) for both the cost-utility (quality of life) and the cost-effectiveness (falls) analyses, the confidence ellipses extended into all quadrants, indicating that the results were not statistically significant (Figures 2 and 3, respectively).
—QALYs gained—confidence ellipses for cost-utility pre (block 1) to post (block 3) participation. ENJOY indicates Exercise interveNtion outdoor proJect in the cOmmunitY; QALY, quality-adjusted life year. Note: 95% confidence ellipse is the outer ellipse, 75% is the middle ellipse, and 50% is the inner ellipse.
Citation: Journal of Physical Activity and Health 20, 6; 10.1123/jpah.2022-0380
—Falls avoided—confidence ellipses for cost-effectiveness pre (12 mo) to post (12 mo) commencement of Exercise interveNtion outdoor proJect in the cOmmunitY (ENJOY). Note: 95% confidence ellipse is the outer ellipse, 75% is the middle ellipse, and 50% is the inner ellipse.
Citation: Journal of Physical Activity and Health 20, 6; 10.1123/jpah.2022-0380
Characterizing Uncertainty
All 3 sensitivity analyses had a negligible impact on the magnitude of the findings. Only the falls analysis altered the significance, changing the primary analysis postintervention trend in falls reduction (P = .160) to a significant falls reduction (P = .018).
Falls Data: Inflating the Retrospective Falls Data by 17.6%
There was a significant reduction in falls of −0.59 (95% CI, −0.59 to 0.00; P = .018), favoring the 12 months following commencement of the intervention (Table 5).
Falls Data Sensitivity Analysis
| Block 1 | Block 3 | Mean difference; post block 3 minus block 1 (95% CI) | P | |
|---|---|---|---|---|
| Falls (over 12 mo) | ||||
| Median (IQR) | 0.59 (0.00–1.18) | 0.00 (0.00–1.00) | −0.59 (−0.59 to 0.00) | .018 (nonpara) |
| Mean (SD) | 0.89 falls (SD 1.13) | 0.54 falls (SD 1.13) | −0.35 (−0.77 to 0.07) | .097 (para) |
Abbreviations: CI, confidence interval; IQR, interquartile range. Note: Bold value is significant.
Cost Data: Reporting Recall Bias and Completing Adjustments to the Self-Reported Health Service Utilization
Within the current study, GP visits were both self-reported and reported via the MBS administration system. In comparison, it was found that recall bias was present in the current cohort (n = 50). The multiplication factor was determined for the self-reported data to represent MBS data. We found that retrospective recall of GP visits over 6 months was overreported, and prospective recall of GP visits was underreported (Table 6).
Recall Bias Associated With GP Visits
| Block 1 (n = 50) | Block 3 (n = 50) | |
|---|---|---|
| Self-reported occasions of GP services | 393 | 242 |
| MBS data for occasions of GP services | 358 | 316 |
| Multiplication for the MBS data to represent self-reported data | 1.098 | 0.766 |
| Multiplication for the self-reported data to represent MBS data | 0.911 | 1.306 |
| Summary | Self-reported GP visits are overreported by 9.8% | Self-reported GP visits are underreported by 23.4% |
Abbreviations: GP, general practitioner; MBS, Medicare Benefits Schedule.
The multiplication factor to self-reported health care utilization was applied to block 1 (0.911) and block 3 (1.306). Results were consistent with the primary analysis wherein there was a trend toward lower costs in block 3 following participation in the intervention. Excluding the cost of overall ENJOY (12-wk primary intervention and 24-wk maintenance phase), across all health and service utilization categories, block 1 average cost was $9698.08 (SD $26,037.58), and block 3 was $2932.52 (SD $3541.87), indicating that block 3 had a nonsignificant reduction in cost of −$6765.56 (95% CI, −$14,255.07 to $723.95; P = .076). Including the cost of ENJOY, block 1 average cost remained the same; however, block 3 cost was $5311.69 (SD $3781.75), indicating that block 3 had a nonsignificant reduction of −$4386.39 (95% CI, −$11,917.59 to $3144.80; P = .247).
Cost Data: The Cost per Person to Participate in the ENJOY Maintenance Program was Adjusted to Reflect Between 1 and 12 Participants per Session
Costs for the ENJOY maintenance program were attributed to between 1 and 12 participants in a sensitivity analysis, then added to the cost of health service utilization, productivity, and ENJOY intervention for the combined cost (Figure 4). With 1 participant per maintenance group, the combined costs were greater for block 3 ($3206), however, with 2 to 12 participants per maintenance group, the costs were less for block 3 (range $1978–$6117).
—Cost per participant for block 1 and block 3 (including cost of health service utilization, productivity, and ENJOY intervention and ENJOY maintenance) when the cost for the ENJOY maintenance program was attributed to between 1 and 12 participants. ENJOY indicates Exercise interveNtion outdoor proJect in the cOmmunitY. Note: Block 1 is consistent at $9,764; Block 2 is the higher curved line, and Block 3 is the lower curved line.
Citation: Journal of Physical Activity and Health 20, 6; 10.1123/jpah.2022-0380
Cost and Effect Data: To Investigate the Robustness of the Findings, a Sensitivity Analysis was Conducted That Included the 11 Participants Who Were Excluded Due to Missing MBS/PBS Data
Key cost and effect variables were analyzed as independent t tests to report differences between the primary cohort (n = 50) and the additional participants (n = 11). There was no difference between the primary cohort and the additional participants for quality of life (utility index), falls, employment income, home nursing costs, allied health costs, community service costs, nursing costs, MBS benefits paid, or PBS benefits paid at baseline or at follow-up (P > .05). However, acute and rehabilitation costs at follow-up were higher in the additional participants (MD $6763; 95% CI, $3025 to $10,502; P < .001 and MD $2051; 95% CI, 185 to $3916; P = .03, respectively).
Paired t tests were repeated for key cost and effect variables to report differences in significance between the primary cohort (n = 50) and the newly combined cohort (n = 61). The newly combined cohort did not change the significance reported in quality of life (utility index), falls, or any of the cost elements except for allied health service costs. The primary cohort reported a nonsignificant lower cost in block 3 ($409.65 [SD $807.83]) compared with block 1 ($652.62 [SD $1205.03]), with a mean difference of −$242.97 (95% CI, −$519.21 to 22.38), P = .083. However, the newly combined cohort reported significantly lower costs in block 3 ($384.41 [SD $764.25]) compared with block 1 ($620.62 [SD $1128.69]), with a mean difference of −$236.21 (95% CI, −469.24 to −$3.17), P = .047. Acute and rehabilitation costs were not significantly different between baseline and follow-up for the primary cohort (n = 50) and the newly combined cohort (n = 61).
Although small variations were reported between the primary cohort (n = 50) and the additional participants (n = 11), they did not change the direction of the initial findings.
Discussion
The ENJOY project utilized the Seniors Exercise Parks to successfully engage older people in physical activity, improving their function and well-being.9 Using a subsample of available data (n = 50), the current economic evaluation demonstrated that following participation in the ENJOY program, there was a likely reduction in health care utilization and productivity losses with a possible reduction in falls. Although early economic evaluations alongside clinical trials provide vital information for future intervention scaling and cost projections,24 they are often not powered with adequate sample size, as appears to be the case in the current study. Should a larger sample size confirm significance for the observed reduction in health service costs, there is potential for an almost AUD$7000 reduction in health care costs per person over the 6 months postintervention. Globally, this has the potential to offset some of the 0.3% to 4.6% national health care expenditure attributed to inactivity.3
In 2013, physical inactivity cost the health care systems $53.8 billion (INT$) worldwide.4 Physical activity interventions could substantially improve population health at less than AUD$50,000 per QALY saved.25 Physical activity in well-designed outdoor spaces can be an inexpensive method for an individual to maintain their health. The potential health gains and health care cost savings associated with a well-designed outdoor built environment has been suggested to range between AUD$4698 and AUD$36,777 per 100,000 adults per year.6 The ENJOY project involved installing the Seniors Exercise Park and delivering a physical and social activity program using this specialized equipment. With the potential economic benefits associated with well-designed outdoor built environments, the utilization of age-suitable equipment can provide added benefits to older people.
The model of intervention used for the ENJOY program for this study utilized a relatively high level of supervision for the 12-week program (health professional supervision twice weekly for the 12 wk) as well as regular optional health professional supervision for the following 6 months. For relatively well older people, safe use of the Seniors Exercise Park may require minimal or no direct health professional supervision for safe use of the equipment, and for those with mild balance or strength impairments, a reduced level of supervision than that used in this study may be sufficient. Reduced supervision levels would impact the costs associated with program implementation and improve cost-effectiveness estimates for the intervention. Frailer older people and those with higher levels of falls risk would likely require the level of supervision provided in this study, although the health benefits would potentially be higher.
When reporting on older people, productivity through employment has implications for inclusion in an economic evaluation. In Australia, more than half the people aged over 55 have retired, with women tending to retire sooner than men.26 In the current study cohort, the average age was 73 years, and 78% were women, suggesting a low probability of paid employment at the time of data collection. Of the 50 participants, only 3 were engaged in paid employment prior to enrollment in the study, with only 2 engaged in paid employment following completion of participation. Of interest, although 1 participant retired from work (aged 75; from 4 to 0 h/wk), 2 participants actually increased their work (aged 65, from 35 to full-time hours per week; and aged 78, from 24 to full-time hours per week). The implication for the current study is that over a 12-month period, it would be expected that productivity may decrease due to time alone, and therefore, the productivity analysis should be treated with caution.
Elsewhere, we have reported the study outcomes for the falls data,10 which demonstrated that this intervention significantly reduced the number of fallers and falls incidence in the 12-month follow-up (P < .05, n = 54). This is aligned to the current study, with a slightly smaller cohort (because a small number of participants did not have all the required data for the analyses in this paper), wherein falls trended toward decreasing in the 12-month follow-up in the primary analysis (P = .160, n = 50) and significantly reduced falls in the sensitivity analysis (P = .018, n = 50). Balance training is a critical element of falls prevention programs in older people at higher risk of falling,27 and the Seniors Exercise Park kit has several exercise stations that focus on balance training. This is in contrast to the majority of fixed exercise equipment used in community parks in Australia and elsewhere, which focus primarily on strength or cardiorespiratory exercise.
Key strengths of this study included comprehensive cost data sets across 2 periods of 6 months as well as self-reported health service utilization, which was cross-checked with administrative data from Services Australia (MBS). Sensitivity analyses were applied to characterize cost and effect uncertainty, and these had a small impact on the magnitude of the findings, adding to the robustness of the results. Key limitations included the small sample size (n = 50) and the relatively short 6-month follow-up period; however, these early economic evaluations alongside intervention studies are important and inform data collection in future fully powered economic evaluations.24 Another limitation was that unpaid productivity, such as caring for other people and/or volunteer work, was not included. Future studies should be fully powered for an economic evaluation and consider 1- to 2-year follow-up periods.
It is noted that there were minor variations from the economic evaluation methodology detailed in the published protocol.12 This included the methods in costing self-reported hospitalization with the protocol stated using the National Weighted Activity Unit funding approach and the final analysis using the National Hospital Cost Data as well as a simplified calculation for productivity costs due to participants leaving data fields blank.
Conclusions
It is likely that participation in the ENJOY program reduces health care costs in the 6 months following the intervention without compromising quality of life or increasing the likelihood of a fall; however, fully powered studies are required to confirm these findings. Planning for shared community spaces should consider the benefits of a Seniors Exercise Park as a part of the built environment.
Acknowledgments
The authors would like to acknowledge Whittlesea City Council, Wyndham City Council, and Abound Communities for their collaboration and partnership in this project. Moreover, the authors would like to thank Lark Industries for the Seniors Exercise Park equipment installation and associated advice and support. Availability of Data and Materials: The data sets generated and/or analyzed during the current study are not publicly available due to ethical restrictions but are available from the corresponding author on reasonable request. Funding: This study was funded by Gandel Philanthropy. This funding source had no role in the design of the study, its execution, analyses, interpretation of the data, or writing the manuscript for publication. Trial Registration: This trial and economic evaluation was prospectively registered with the Australian New Zealand Clinical Trials Registry. Trial registration number ACTRN12618001727235 registered 19/10/2018 https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=375979.
References
- 1.↑
Koster A, Stenholm S, Schrack JA. The benefits of physical activity for older people. In: Samuel RN, Anna B, Terry H, Khim H, Charles M, Geeske P, Christina RV, Julia KW, eds. The Palgrave Handbook of Ageing and Physical Activity Promotion. Springer; 2018:43–60.
- 2.↑
Musich S, Wang SS, Hawkins K, Greame C. The frequency and health benefits of physical activity for older adults. Popul Health Manag. 2017;20(3):199–207. doi:10.1089/pop.2016.0071
- 3.↑
Ding D, Kolbe-Alexander T, Nguyen B, Katzmarzyk PT, Pratt M, Lawson KD. The economic burden of physical inactivity: a systematic review and critical appraisal. Br J Sports Med. 2017;51(19):1392–1409. doi:10.1136/bjsports-2016-097385
- 4.↑
Ding D, Lawson KD, Kolbe-Alexander TL, et al. The economic burden of physical inactivity: a global analysis of major non-communicable diseases. Lancet. 2016;388(10051):1311–1324. doi:10.1016/S0140-6736(16)30383-X
- 5.↑
World Health Organization. Interventions on Diet and Physical Activity: What Works: Summary Report. 2009.
- 6.↑
Zapata-Diomedi B, Herrera AMM, Veerman JL. The effects of built environment attributes on physical activity-related health and health care costs outcomes in Australia. Health Place. 2016;42:19–29. doi:10.1016/j.healthplace.2016.08.010
- 7.↑
Levinger P, Cerin E, Milner C, Hill KD. Older people and nature: the benefits of outdoors, parks and nature in light of COVID-19 and beyond–where to from here? Int J Environ Health Res. 2022;32(6):1329–1336. doi:10.1080/09603123.2021.1879739
- 8.↑
Levinger P, Panisset M, Parker H, Batchelor F, Tye M, Hill KD. Guidance about age-friendly outdoor exercise equipment and associated strategies to maximise usability for older people. Health Promot J Austr. 2021;32:475–482. doi:10.1002/hpja.367
- 9.↑
Levinger P, Panisset M, Dunn J, et al. Exercise interveNtion outdoor proJect in the cOmmunitY for older people–results from the ENJOY Seniors Exercise Park project translation research in the community. BMC Geriatr. 2020;20(1):446. doi:10.1186/s12877-020-01824-0
- 10.↑
Levinger P, Dunn J, Panisset M, et al. The effect of the ENJOY Seniors Exercise Park physical activity program on falls in older people in the Community: A prospective pre-post study design. J Nutr Health Aging. 2022;26:217–221. doi:10.1007/s12603-021-1724-1
- 11.↑
Levinger P, Dunn J, Panisset M, et al. The ENJOY project: usage and factors to support adherence and physical activity participation. Trans J Am Col Sports Med. 2021;6:163. doi:10.1249/TJX.0000000000000163
- 12.↑
Levinger P, Panisset M, Dunn J, et al. Exercise interveNtion outdoor proJect in the cOmmunitY for older people—the ENJOY Senior Exercise Park project translation research protocol. BMC Public Health. 2019;19(1):933. doi:10.1186/s12889-019-7125-2
- 13.↑
Husereau D, Drummond M, Petrou S, et al. Consolidated health economic evaluation reporting standards (CHEERS)—explanation and elaboration: a report of the ISPOR health economic evaluation publication guidelines good reporting practices task force. Value Health. 2013;16(2):231–250. doi:10.1016/j.jval.2013.02.002
- 14.↑
Bouwmans C, Krol M, Severens H, Koopmanschap M, Brouwer W, Hakkaart-van Roijen L. The iMTA productivity cost questionnaire: a standardized instrument for measuring and valuing health-related productivity losses. Value Health. 2015;18(6):753–758. doi:10.1016/j.jval.2015.05.009
- 17.↑
Czeisler MÉ, Marynak K, Clarke KE, et al. Delay or avoidance of medical care because of COVID-19–related concerns—United States, June 2020. Morb Mortal Wkly Rep. 2020;69(36):1250–1257. doi:10.15585/mmwr.mm6936a4
- 18.↑
Schuster NA, de Breij S, Schaap LA, et al. Older adults report cancellation or avoidance of medical care during the COVID-19 pandemic: results from the Longitudinal Aging Study Amsterdam. Eur Geriatr Med. 2021;12(5):1075–1083. doi:10.1007/s41999-021-00514-3
- 19.↑
Grobet C, Marks M, Tecklenburg L, Audigé L. Application and measurement properties of EQ-5D to measure quality of life in patients with upper extremity orthopaedic disorders: a systematic literature review. Arch Orthop Trauma Surg. 2018;138(7):953–961. doi:10.1007/s00402-018-2933-x
- 20.↑
Nixon RM, Wonderling D, Grieve RD. Non‐parametric methods for cost‐effectiveness analysis: the central limit theorem and the bootstrap compared. Health Econ. 2010;19(3):316–333. doi:10.1002/hec.1477
- 22.↑
Hill K, Schwarz J, Flicker L, Carroll S. Falls among healthy, community‐dwelling, older women: a prospective study of frequency, circumstances, consequences and prediction accuracy. Aust N Z J Public Health. 1999;23(1):41–48. doi:10.1111/j.1467-842X.1999.tb01203.x
- 23.↑
Brusco NK, Watts JJ. Empirical evidence of recall bias for primary health care visits. BMC Health Serv Res. 2015;15:381. doi:10.1186/s12913-015-1039-1
- 24.↑
Edmunds K, Scuffham P, Reeves P, et al. Demonstrating the value of early economic evaluation alongside clinical trials: exercise medicine for men with metastatic prostate cancer. Eur J Cancer Care. 2021;30:e13479. doi:10.1111/ecc.13479
- 25.↑
Cobiac LJ, Vos T, Barendregt JJ. Cost-effectiveness of interventions to promote physical activity: a modelling study. PLoS Med. 2009;6(7):e1000110. doi:10.1371/journal.pmed.100011
- 26.↑
Australian Bureau of Statistics. Retirement and Retirement Intentions, Australia. [Accessed June 2021]. 2020. https://www.abs.gov.au/statistics/labour/employment-and-unemployment/retirement-and-retirement-intentions-australia/latest-release
- 27.↑
Sherrington C, Fairhall NJ, Wallbank GK, et al. Exercise for preventing falls in older people living in the community. Cochrane Database Syst Rev. 2019;1:CD012424. doi:10.1002/14651858.CD012424.pub2
