Article Text
PDF
Abstract
Objective The objective of this study is to explore whether offering an integrated self-management strategy and exercise referral scheme intervention (ERS+SMS) or a stand-alone ERS intervention is more effective in reducing healthcare service utilisation among community-dwelling older adults over time compared with a stand-alone SMS/control intervention.
Design Secondary analyses of two randomised controlled trials (RCTs) with linkage to Danish national health registries.
Setting Three Danish municipalities: Esbjerg, Slagelse and Odense.
Participants Danish community-dwelling older adults, aged 65 years or older, were included in the study. Participants originated from two completed RCTs: the Welfare Innovation in Primary Prevention (WIPP, n=115) and the SITLESS project (SITLESS, n=333).
Intervention Participants in the two trials were randomly assigned to one of three intervention groups: ERS in combination with SMS (ERS+SMS), ERS alone or an SMS/control intervention.
Main outcome measure Annual number of days with contact to general practice (GP) and days hospitalised for all causes.
Results No statistically significant between-group difference was observed in the annual number of days with contact with GP between participants randomised to ERS+SMS (p=0.759) or ERS only (p=0.156) compared with those randomised to the SMS/control group. In addition, no statistically significant between-group differences in days hospitalised were observed when comparing the ERS+SMS (p=0.222) or ERS only (p=0.060) group to the SMS/control group.
Conclusion Neither the integrated intervention (ERS+SMS) nor the ERS alone intervention was more effective in reducing healthcare service utilisation during a 2-year follow-up period compared with a stand-alone SMS/control intervention. A potential stabilising effect of ERS and/or SMS interventions on the general age-related increase in healthcare service utilisation cannot be ruled out but needs further exploration in future studies along with the exploration of the effects of long-term interventions.
- Aged
- Health Services for the Aged
- PUBLIC HEALTH
- REGISTRIES
Data availability statement
Data may be obtained from a third party and are not publicly available. The datasets underpinning the findings in this article are not publicly accessible due to privacy and confidentiality constraints associated with individual-level health information.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
The use of data from two completed randomised controlled trials in combination with Danish health registry data is a methodological strength.
Despite the inevitable risk of self-selection bias, a notable strength lies in the composition of the study population; specifically, our study encompasses a distinctive sample of community-dwelling older adults, a demographic traditionally challenging group to recruit for research purposes.
The content of the interventions was not completely identical in the two randomised controlled trials, which we addressed by including by-project stratified analyses.
It cannot be ruled out that some of the study outcomes may be influenced by the COVID-19 pandemic-induced lockdown in Denmark in 2020; however, changes in reporting to the Danish National Patient Registry are unlikely to have influenced our results, as no remarkable changes in the number of days hospitalised were identified in the matched comparison group following the shift in reporting.
Introduction
In the context of ageing, there is a well-established association between advancing age and a decline in physical functions, coupled with a rise in multimorbidity.1–3 This shift in health profile inevitably results in an increased utilisation of healthcare services, providing an incentive for policy-makers to initiate new preventive initiatives aimed at enhancing and maintaining the health and well-being of older adults.
Community-dwelling older adults constitute a distinctive target group of interest with considerable preventive potential. For both individual and socioeconomic benefits, the objective of many preventive interventions directed towards community-dwelling older adults is, therefore, to foster the maintenance of physical functioning and independence, thereby avoiding hospitalisations and reducing the use of healthcare services.4 5
Exercise referral schemes (ERSs) and self-management strategies (SMSs) are frequently used preventive strategies.6 7 ERSs are exercise-based interventions designed to facilitate exercise participation and promote physical activity.8 9 The efficacy of ERS may be heightened by integrating SMS, characterised by, for example, fostering the adoption of physically active behaviours.10 Numerous studies have demonstrated positive outcomes from offering ERS on various health metrics, such as body mass index,9 11 12 fat and muscle mass,13 blood pressure,9 11 12 fitness rating11 and quality of life.9 Additionally, studies among older adults in care homes indicate that ERS may significantly reduce fall risk factors14 and enhance physical function.15 Furthermore, positive findings from SMS interventions have been demonstrated on chronic pain16 and self-efficacy.17 Moreover, the combined intervention of ERS and SMS is effective in increasing functional capacity.18 19 Existing research indicates that both physical activity, which is the central element in ERS interventions, and SMS are linked to reduced healthcare utilisation, meaning fewer hospitalisations and general practice (GP) services.5 17 20–23 Such effects may in part be linked to a positive association between muscle strength and health-related quality of life.24 However, the aforementioned studies are cohort studies,20–22 cross-sectional studies23 or systematic reviews of randomised controlled trials (RCTs) measuring outcome postintervention at either 3 or 6 months.5 17
Based on the existing research, uncertainties persist regarding the long-term impact of ERS and SMS interventions, individually or in combination, on healthcare service utilisation. We hypothesise that an integrated ERS+SMS intervention or a stand-alone ERS intervention will be more effective in reducing healthcare service utilisation among community-dwelling older adults, compared with stand-alone SMS/control interventions. This will be investigated in the present study, with healthcare service utilisation measured through contacts to GP and hospitalisations.
Methods
Study population
We included participants from two RCTs, The Welfare Innovation in Primary Prevention (WIPP)18 and The SITLESS project: exercise referral schemes enhanced by self-management strategies to battle sedentary behaviour in older adults (SITLESS).25 The participants resided in the Danish municipalities of Esbjerg (WIPP), Slagelse (WIPP) and Odense (SITLESS) and were all community-dwelling older adults aged 65 years or older. The participants recruited in both WIPP18 and SITLESS25 were older adults with sedentary behaviour or other key risk factors for premature disability (eg, reduced physical function).
All participants were recruited to the WIPP and SITLESS studies through the nationally regulated preventive home visits offered by the municipalities according to Danish legislation, the Act on Social Service § 79a.26 One of the objectives of the home visit is to support older adults who do not receive personal care and/or practical help from the municipality and address individual needs that are crucial for enabling them to maintain independence.27 To be included in the present study, participants had to meet the following criteria: (1) they were included and randomised to an intervention in either the WIPP or SITLESS trials and (2) they had a valid Civil Registration Number (CPR number).
As the WIPP and SITLESS studies were based on voluntary participation, leading to a self-selected group of individuals, we included an external matched comparison group to show general trends in the population. This group was identified from the general Danish population based on information from the Danish Central Person Register and identified through Statistics Denmark. The group was matched to the study population in a 1:10 ratio and was matched on age, sex and vital status (alive) at the time of the first intervention session for the respective study participant to whom they were matched; this is referred to as the index date.
Interventions
The interventions in WIPP and SITLESS have previously been described in detail.18 25 The WIPP study (www.ClinicalTrials.gov reg. no.: NCT04531852) was a two-armed RCT, where individuals were randomised to either an ERS+SMS or SMS-only intervention.10 The index date of the first and last participants was 24 September 2018 and 9 August 2019, respectively. The ERS+SMS intervention consisted of a 12-week structured exercise programme, designed to improve physical activity and physical functioning, with two supervised exercise sessions per week each of 1-hour duration, in combination with a 24-week SMS intervention consisting of 8 group-based 1.5-hour sessions. The SMS-only intervention also lasted for 24 weeks, but with a total of 12 group-based 1.5-hour sessions. The SMS intervention focused on, for example, sedentary behaviour, physical activity, diet and additional risk factors for functional loss.6
The SITLESS study (www.ClinicalTrials.gov reg. no.: NCT02629666) was a three-armed RCT. Participants were randomised to ERS+SMS, ERS-only or a control group. The index date of the first and last participants was 3 November 2016 and 16 November 2017, respectively. The ERS+SMS intervention consisted of a 16-week ERS intervention, with two supervised 1-hour exercise sessions per week, in combination with a 30-week SMS intervention consisting of six group-based sessions, one one-to-one session and four telephone calls with a duration of 20–60 minutes per session. The ERS-only intervention consisted of two supervised 1-hour exercise sessions per week for 16 weeks. The control group received a written general booklet containing WHO's recommendations regarding physical activity and was invited to two seminars, each of 2-hour duration. The seminars focused on (1) physical activity, sedentary behaviour and exercise and (2) healthy nutrition in older adults.25 According to this, we categorise the group as having undergone an SMS/control intervention. For an overview of the interventions in WIPP and SITLESS, see table 1.
Overview of interventions in WIPP and SITLESS: exercise referral scheme (ERS) interventions described using parameters (frequency, intensity, time and type)
In the present study, the intervention groups from WIPP and SITLESS were merged into the following groups: ERS+SMS, ERS and SMS/control.
Outcome variables and data sources
The outcomes of interest were: (1) annual number of days with contact with GP (all causes) and (2) annual number of days hospitalised (all causes), retrieved from the Danish National Health Service Register28 and the Danish National Patient Register (DNPR).29 By using the CPR number, a unique personal identifier, data from WIPP and SITLESS were linked with information from the Danish National Health Service Register and DNPR. Table 2 displays the different types of contacts to GP (including service codes) included in the analyses. Data were extracted from both public and private hospitals, encompassing somatic and psychiatric departments. Due to Danish national legislation about treatment guarantees at hospitals within a given time frame a free choice of hospitals is available in Denmark,30 which is the reason for the inclusion of private hospitals. As a result of policy regulations, the reporting practice to DNPR was altered in January 2019. It was mandatory for all Danish regions, private hospitals and specialised physicians to fully adopt the new reporting standard by 3 March 2019.31 The description of the reporting practices and coding details is available in online supplemental appendix 1.
Supplemental material
Included types of contacts to general practice (GP)
Baseline number of days with contact with GP and days hospitalised was calculated using data from 360 days before the index date (day −361 to −1). The index date was included in the 2-year follow-up period (days 0–720), with months defined as 30-day intervals. Individuals were not excluded in the event of death or migration during the follow-up period. Instead, we estimated the time at risk for each individual, which we accounted for in the analyses.
Data analysis
Descriptive statistics were reported as counts (n) and proportions (%). An intention-to-treat approach was applied and a linear fixed effects regression model accounting for clustering in matched comparison groups was used to estimate between-group differences in the total number of days with contact with GP and days hospitalised, from baseline to follow-up. The primary comparisons are between ERS+SMS and SMS/control, and between stand-alone ERS and SMS/control, both within the RCT setup. Since this study involves secondary analyses of RCTs, we interpret our results as exploratory; therefore, we do not perform p value adjustments for multiplicity. The group receiving SMS/control served as the reference group in all analyses. Since the fixed effects model simultaneously models outcomes at both times, the regression coefficient of interest was included as an interaction term between time (days) and group. Each participant’s time at risk was accounted for in the analyses by dividing the counts by risk time. As this study combines data from two distinct RCTs (WIPP and SITLESS), we also conducted all analyses stratified by project to determine whether there were project-specific differences. Additionally, we assessed whether the transition in reporting to DNPR had a systematic impact on the number of days hospitalised in the matched comparison group 1 year before and after the transition. All statistical analyses were conducted by using Stata/BE V.18.0 (StataCorp), with a predetermined level of statistical significance set at 5%.
Patient and public involvement
The design, conduct, reporting and dissemination plans of this study were undertaken without the involvement of patients or the public. Nonetheless, the original RCTs, which serve as the data source for this study, did include patient and public involvement.
Results
The combined participant count from the WIPP and SITLESS studies totalled 448 individuals. Among these, 175 were randomly assigned to ERS+SMS, 110 to ERS-only and 163 to SMS/control (figure 1). Most of the participants were women (n=273 (61%)). Across all intervention groups, the predominant age range was between 75 and 84 years. Additionally, the proportion of participants with one or more chronic diseases was comparable between the three intervention groups, ranging from 17% to 19% (table 3). The median number of days with contact with GP and days hospitalised is close to similar among the three intervention groups and the matched comparison group at baseline. However, we find that the baseline maximum number of days hospitalised varies to a larger degree between the groups than the maximum number of days with contact with GP (online supplemental appendix 2). From the project-stratified baseline characteristics, we found statistically significant differences between the participants included in WIPP and SITLESS in terms of sex, age, marital status and number of chronic diseases (online supplemental appendix 3).
Baseline characteristics of the study population given in counts and percentages (N (%))
Flow chart of the study population. WIPP, Welfare Innovation in Primary Prevention. SITLESS, SITLESS project: exercise referral schemes enhanced by self-management strategies to battle sedentary behaviour in older adults.
Contacts to GP
The contact pattern to GP was similar across the three intervention groups at baseline as well as follow-up. At baseline, the estimated annual mean number of days with contact with GP ranged from 12.4 to 13.7 and at follow-up, it ranged from 12.5 to 13.8. No statistically significant differences in days with contact with GP were observed between or within the intervention groups. However, a small but statistically significant increase of 0.9 days with contact with GP was identified in the matched comparison group (95% CI 0.61; 1.24) (table 4).
Within-group changes and between-group differences in contacts to general practice and days hospitalised, including by project stratified analyses
Days hospitalised
The estimated annual mean number of days hospitalised ranged from 0.8 to 1.7 at baseline and increased to a range of 1.7 to 4.6 at follow-up among the three intervention groups and the matched comparison group. No statistically significant between-group differences in days hospitalised were identified. The estimated within-group changes revealed a statistically significant increase from baseline to follow-up of 3.7 days hospitalised in the SMS/control group (95% CI 1.21; 6.14), 1.8 days hospitalised in the ERS+SMS group (95% CI 0.12; 3.50) and 2.2 days hospitalised in the matched comparison group (95% CI 1.72; 2.77) (table 4). Online supplemental appendix 4 illustrates days with contact with GP and days hospitalised over time for the three intervention groups.
We found no systematic changes in the number of days hospitalised within the matched comparison group following the shift in reporting to DNPR (online supplemental appendix 5).
By-project stratified analyses of days with contact with GP and days hospitalised
Welfare Innovation in Primary Prevention
In the WIPP study, a statistically significant between-group difference of 4.8 days, indicating more GP contacts, was observed in the matched comparison group (95% CI 1.39; 8.29) compared with the SMS/control group. A statistically significant within-group increase of 1 day in the annual mean number of days with contact with GP was observed in the matched comparison group (95% CI 0.30; 1.71).
No between-group differences were identified in days hospitalised among the WIPP participants. The within-group changes showed statistically significant increases of 7.2 days in the SMS/control group (95% CI 0.86; 13.63) and 2.7 days in the matched comparison group (95% CI 1.48; 4.01) (table 4).
SITLESS
No statistically significant between-group differences in the number of days with contact with GP were identified. A statistically significant within-group increase of 0.9 days with contact with GP from baseline to follow-up was observed in the matched comparison group (95% CI 0.54; 1.25).
We found no statistically significant between-group differences in the number of days hospitalised. The within-group changes revealed a statistically significant increase of 1.6 days hospitalised from baseline to follow-up in the ERS+SMS group (95% CI 0.42; 2.86) and 2.08 days in the matched comparison group (95% CI 1.52; 2.63) (table 4). Online supplemental appendix 6 illustrates days with contact with GP and days hospitalised over time for WIPP and SITLESS participants separately, and their matched comparison group.
Discussion
To the best of our knowledge, this study is the first to investigate whether an integrated SMS and ERS intervention (ERS+SMS) or a stand-alone ERS intervention is more effective in reducing healthcare service utilisation among community-dwelling older adults, compared with a stand-alone SMS/control intervention. We observed no significant differences between the intervention groups, suggesting that an integrated ERS+SMS or a stand-alone ERS intervention is not more effective at reducing healthcare service utilisation, measured as days with contact with GP and days hospitalised.
Some previous studies on the effects of exercise programmes, alone or in combination with telephone follow-up, contradict our findings by showing fewer readmissions, GP visits32 and a shorter overall length of stay in hospital.33 Another study aligns with our results by finding no differences in hospital utilisation over time. However, it also found that the control group was significantly more prone to hospitalisations lasting more than 3 days.34 It is worth noting that the mentioned studies vary considerably in terms of intervention content and target group, and only one of the studies included a combined intervention.
WIPP and SITLESS provided complex lifestyle interventions, featuring an ERS+SMS component to enhance physical functioning and activity in community-dwelling older adults. While not directly addressing healthcare service utilisation, results from WIPP and SITLESS, along with numerous other studies on complex lifestyle interventions, support their positive health effects.9 11 12 18 19 Therefore, complex lifestyle interventions may potentially reduce healthcare service utilisation by affecting various influential factors. However, common causes for admission and readmissions in older adults, such as polypharmacy, infectious diseases, dehydration, new acute diseases1 and adverse events associated with the use of analgesics,35 were not specifically targeted by the WIPP and SITLESS interventions.
Participants from WIPP and SITLESS were recruited through the preventive home visits offered by the municipalities. The participants were, therefore, living independently without municipal assistance in terms of personal and/or practical help. Moreover, all participants voluntarily joined WIPP and SITLESS, indicating that they had the energy in their everyday lives to participate in a research project. This may indicate that we have recruited a selected group of individuals, and their general use of healthcare services might be expected to be lower compared with older adults receiving assistance from the municipality or residing in an assisted living facility.
A significant rise in the number of days with contact to a GP was evident only in the matched comparison group. This rise aligns with the overall trend in the general Danish population of older adults.36 The absence of a corresponding rise in GP contacts within the intervention groups might be explained by the fact that community-dwelling older adults attending research projects may be selected and healthier group of individuals compared with multimorbid older adults in nursing homes. Nevertheless, the observation that the overall intervention group exhibited a higher percentage of individuals with multimorbidity compared with the externally matched comparison group suggests that the intervention offered by WIPP and SITLESS may in fact have a stabilising impact on healthcare service utilisation. Such a potential effect must be explored further in future studies specifically designed for that purpose. It is also worth investigating whether a potential stabilising effect might occur with the use of SMS interventions alone.
Strengths and limitations
Although we contend that our sample reasonably reflects the characteristics of typical community-dwelling older adults in Denmark, it is important to acknowledge the potential self-selection bias in voluntary research projects. Older adults residing in the community, without municipal assistance in terms of personal or practical help, who willingly participate in research projects, might be more resourceful, potentially leading to distinct patterns of healthcare system engagement compared with the general population. To address this concern, we incorporated a matched comparison group. Despite this risk of self-selection bias, a notable strength lies in the composition of the study population. Specifically, our study encompasses a distinctive sample of community-dwelling older adults, a demographic traditionally challenging to recruit for research purposes. The successful recruitment in both WIPP and SITLESS may be attributed to the integration of recruitment within the framework of preventive home visits.
Both WIPP and SITLESS are completed RCTs with similar intervention components, including an ERS+SMS intervention and an SMS/control intervention. However, the SITLESS study also included a stand-alone ERS intervention. The rationale behind combining the SMS programme from the WIPP study with the SMS/control group in the SITLESS study was to form an overall intervention group comprising a stand-alone less comprehensive intervention, contrasting with the integrated ERS+SMS intervention. However, as the content of the interventions was not completely identical in WIPP and SITLESS, we also conducted by-project stratified analyses.
By combining data from the two RCTs with information from the Danish health registries, we gain valuable insights into additional relevant outcomes. Using register-based outcome measures in the present study is a strength, as the validity and accuracy of the DNPR are high37 and the coverage of the Danish National Health Service Register is satisfactory.28 38
Originally, our plan included home nursing assistance as an additional outcome. However, due to insufficient data completeness, we decided not to include this outcome in the study. Yet, incorporating home nursing assistance would have been interesting and relevant, given the study’s focus on community-dwelling older adults.
To elucidate the absence of statistically significant findings, a post hoc power analysis could have been conducted. However, this study represents a secondary and exploratory analysis of data from two previously completed RCTs, where power calculations were based on the primary outcomes of those trials. We acknowledge that a post hoc power analysis could highlight the potential lack of statistical power in this study to detect statistically significant changes. Nevertheless, given that the observed effect size in this study is relatively small and, in our judgement, not clinically relevant, conducting a post hoc power analysis would not be meaningful.
It is unlikely that the results observed in our study have been influenced by the policy changes in reporting to the DNPR, as we did not identify systematic changes in the number of days hospitalised following the shift in reporting to DNPR. With respect to the potential influence of the COVID-19 lockdown, statistical documentation from Statistics Denmark found that there was an increase in the total number of contacts to GP during COVID-19.39 This is opposite to our findings in the by-project stratified analysis for the WIPP population, which was the study population potentially affected by COVID-19. Yet, we do find a statistically significant within-group increase in contacts to GP in the matched comparison group corresponding to the overall increase found by Statistics Denmark.39 Therefore, we cannot rule out that COVID-19 may have affected the WIPP population leading to fewer GP contacts.
Conclusion
An integrated ERS+SMS or stand-alone ERS intervention was not more effective in reducing healthcare service utilisation among community-dwelling older adults, as compared with a stand-alone SMS/control intervention. However, our results indicated a potential stabilising effect on the use of healthcare services compared with the external matched comparison group. Such effects should be explored further in future studies.
Data availability statement
Data may be obtained from a third party and are not publicly available. The datasets underpinning the findings in this article are not publicly accessible due to privacy and confidentiality constraints associated with individual-level health information.
Ethics statements
Patient consent for publication
Ethics approval
In compliance with European data protection regulations, the legal services at the University of Southern Denmark (Research and Innovation Organisation) approved the data processing activities for the project (journal number 11.352). Additionally, the Research Ethics Committee at the University of Southern Denmark (case no. 21/30425) granted approval. According to section 14.2 of the Act on Research Ethics Review of Health Research Projects and section 10 of the Data Protection Act, approval from the ethics committee and informed consent are not required, as the study is based solely on register data.41 42
Acknowledgments
This study is based on data from the WIPP and SITLESS studies and we would, therefore, like to acknowledge and express our gratitude to the consortiums behind the two studies. In addition, we would like to thank the committed staff from the municipalities of Esbjerg and Slagelse for their invaluable assistance in contributing data. Finally, special thanks go to Mathias Skjødt from the University of Southern Denmark for his support in providing data and sharing his expertise on the SITLESS study.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
X @jbthorlund
Contributors NHS, JBT, JS, SW, PC and TT designed the study. NHS and HSA performed data management. NHS, HSA and SW performed the statistical analyses. NHS and HSA prepared all figures and tables. NHS, JBT, PØØO, JS, SW, HSA, PC and TT interpreted the data. NHS, JBT, PØØO, PC, JS and TT prepared the manuscript. All authors read, provided critical reviews and approved the final manuscript. NHS is the guarantor.
Funding The work was supported by The University of Southern Denmark; Møllerens Fund; Aase og Ejner Danielsens Fund (grant number 21-20-0083) and Østifterne (grant number 2022-0023).
Disclaimer The sponsors were not involved in the review or approval of the manuscript.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.