A Systematic Review without Meta-Analysis (SWiM): Factors Associated with Care Transitions for Home Support Service Recipients

By Junko Fukui Innes MSc ¹

Abdul Roudsari PhD¹

Karen L. Courtney RN, PhD¹

Mary Ellen Purkis PhD²

¹Health Information Science, University of Victoria, British Columbia, Canada

² School of Nursing, University of Victoria, British Columbia, Canada

Citation: Fukui Innes, J., Courtney, K. L., & Purkis, M. E. (2024). A systematic review without meta-analysis (SWiM): Factors associated with care transitions for home support service recipients. Canadian Journal of Nursing Informatics, 19(3).  https://cjni.net/journal/?p=13485

Abstract

BACKGROUND: Home support provides services for people who require assistance with their daily living.  However, at some point, home support recipients need to transition to a higher level of care, such as acute care, and/or long-term care, due to changes in their health care status and care needs. 

METHODS: A systematic review without meta-analysis (SWiM), was conducted to investigate factors associated with home support recipients’ transition to a higher level of care.  The inclusion criteria were articles that examined factors associated with higher health care transitions for home support recipients who were sixty-five or above.  An effect direction plot was developed to synthesize the identified factors.  12,099 papers published between 2009 and 2023 were screened and eight met the inclusion criteria. 

RESULTS: Hospital, long-term care, assisted living and mortality were included as the transition destinations in the eight articles.  Across the eight articles, 116 factors were identified with the transition destinations.  These factors were categorized into eight factor groups and plotted on an effect direction plot.  Due to the heterogeneity of dependent variables, observation periods, factors and measurement of the same factors across the articles made it impossible to conduct a meta-analysis. 

RECOMMENDATIONS: Development of an international standardized definition and/or names of “home support-like” services and measures for the factors is strongly recommended.

Background

Home Support provides services to assist activities of daily living, so that people can live independently at home (BC Ministry of Health, n.d.).  The services include preparing meals, assistance with eating including tube feeding, personal care including bathing, grooming, managing incontinence, assistance with toileting, transferring from a bed to a wheel chair, laundry, medication management, and catheter care.  Also, it provides services so a family caregiver can have a break from care (Office of the Seniors Advocate BC, 2023).  Internationally, the rapid growth of aging populations has been challenging health care systems. Home support recipients can transition to a higher level of care, such as acute care, long-term care, or palliative care, depending on their health care status and care needs.  It is important to understand factors leading to a higher level of care.  Knowing the factors would help health care providers plan for preventative care and resources and support older adults to help them stay at home longer.  This systematic review was conducted to investigate factors associated with home support recipients’ transition to a higher level of care. 

Methods

Eligibility Criteria and Data Items

Any article that examined factors associated with home support for seniors requiring a transition to a higher level of care was included.   ‘Home support service’ in this study is the service that was defined by the Government of British Columbia (BC Ministry of Health, n.d.) and discussed earlier in the Introduction.  There were no restrictions for the study design or interventions to identify the significant factors.  Articles were included if:

1. the outcome measures of the article included significant factors associated with transitions to the next level of care for individuals in home support.
2. it was stated that home support was one of the services they focused on, or if they discussed what types of care and services that the articles focused on, and these types of care or services met the description of home support; and
3. the study population had to include older adults who were sixty-five years old or above and who were receiving home support, as defined earlier, before they transitioned to a higher level of care. 

Information Sources

A computerized literature search, EBSCOhost Database and IEEE were used in this systematic review. The searched databases were CINAHL Complete, MEDLINE, Ageline and IEEE.  The search was conducted based on the literature published in English between January 1, 2009, and December 31, 2023. 

Search Terms

The following terms were originally identified: home support, home care, hospital, hospitalization, acute care, facilities, long term care. After a consultation with a department librarian at the University of Victoria, additional terms were identified via CINAHL Thesaurus, MeSH and Ageline Thesaurus.

Article Selection and Data Collection Process

Two reviewers (AR, JFI) assessed the eligibility of the articles independently in an unbiased standardized manner. A table with the article names were developed for the reviewers to use when reviewing articles. The reviewers indicated “include”, “exclude” and “undecided” next to the article titles. If a reviewer decided to exclude an article, he/she had to explain why it was excluded. Inter-rater reliability was evaluated based on Cohen’s kappa using the SAS Enterprise Guide.

After reviewing the articles, the two reviewers met to discuss their disagreements over the eligibility of some articles. The disagreements were resolved by consensus after the discussion. The reviewers also discussed the undecided articles and made a consensual decision about which of the undecided articles were to be included or excluded.

Protocol

After examining the selected articles there was sufficient variability in the selected articles in terms of the study design and the evaluated factors to permit a meta-analysis to be completed.  As a result, this study was conducted based on  the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) (Page et al., 2021) and Synthesis Without Meta-Analysis (SWiM) (Campbell et al., 2020) guidelines. 

Evaluation and Analysis

The quality of the selected articles was assessed based on Critical Appraisal and Data Extraction for Systematic Review of Prediction Modeling Studies (CHARMS) (Moons et al., 2014; Fernandez-Felix et al., 2023). The Prediction model Risk Of Bias Assessment Tool (PROBAST) was used to assess the risk of bias for the selected article (Fernandez-Felix et al., 2023 ;  Moons et al., 2019).   An effect direction plot was developed based on the high-level categories of the factors evaluated as independent variables in the selected articles. The plot was developed based on the method  discussed by Boon & Thomson (2021) .  Boon & Thompson (2021) developed an algorithm of the effect direction plot to examine the effect of variable factors in the same domain based on the proportion of effect direction without relying on statistical significance based on the 2019 Cochrane Handbook (Boon & Thomson, 2021, McKenzie & Brennan, 2019).  The direction of each arrow indicates the effect of the factor that represents the arrow. The numbers next to the arrows indicate the number of factors combined to show the effect.   If there is no number next to an arrow, the arrow direction was based only on one factor.   Where a factor group consisted of multiple factors, if 70% or more  factors favoured the same direction, the direction of the arrow will be the favoured direction (Boon & Thomson, 2021).  Otherwise, the effect of the factor group is considered as a mixed effect.

Results

Article Selection

The process of the article selection is shown in Figure 1.  As noted above, the literature search produced 12,099 articles for review.  After removing duplicates, 10,546 articles were screened based on the titles and the abstracts.  220 articles were identified after a title and abstract screening for the manuscript review.  At the point of the manuscript review, the interrater reliability based on Cohen’s Kappa was 0.7270, which indicated moderate agreement (McHugh, 2012).  After discussion among the two reviewers, eight articles were identified as meeting all the criteria for this systematic review. 

Figure 1:

PRISM Flow diagram records identified through database searching (n=12,099)

Article Characteristics

The eight articles which are included in this systematic review and the summary of the characteristics of these eight articles are shown in Table 1. 

Table 1:

The Summary of the Characteristics of the Eight Articles

Table 2 shows the service name and description that were included.  The name of home support-like services varies, depending on the article. 

Table 2:

Description of Home Support-Like Services in the Selected Articles

Three articles focused on the transition to hospital (Elkjær et al., 2021, Van Cleve et al., 2023, Dahlberg et al., 2018 )  and the other five articles focused on the transition to assisted living and/or residential/nursing home (Aspell et al., 2019, Fabius et al., 2020, Spillman & Long, 2009, Tate et al., 2022, Wu et al., 2016). Two of the eight articles also included mortality as one of the destinations (Elkjær et al., 2021, Aspell et al., 2019).  

Figure 2 shows the observation periods of the eight articles.  Timelines ranged between 7 days (Elkjær et al., 2021) and five years (Spillman & Long, 2009).   The start of the observation could be at discharge from a short emergency department stay (Elkjær et al., 2021),  interview or audit (Dahlberg et al., 2018, Aspell et al., 2019 , Spillman & Long, 2009), at the time of the data collection (Van Cleve et al., 2023) or at the day of enrollment (Tate et al., 2022 ;  Wu et al., 2016).  One article started its observations six months after individuals were discharged from a long term care institution (Fabius et al., 2020). 

Figure 2:

The Observed Period Varied Across the Articles

A total of 116 factors were identified across the eight articles.  Figure 3 shows the summary of the included factors.  The numbers of factors included in each article ranged between four and twenty-five.  The factors were categorized under seven factor groups: Demographic, Home Support, activity of daily living (ADL)/instrumental activity of daily living (IADL), Physical & Mental Health, Socialization, Health Service availability, and Informal Caregiver.   ADL and IADL activities were categorized based on Edemekong et al. (2024). Table 3 outlines detailed information of each factor and its measurement. 

Figure 3:

Included Factors in the Selected Articles

Table 3:

Detailed Information of Each Factor and its Measurement

All articles included some demographic factors (Elkjær et al., 2021,  Van Cleve et al., 2023, Aspell et al., 2019,  Fabius et al., 2020, Spillman & Long, 2009, Tate et al., 2022, Dahlberg et al., 2018,  Wu et al., 2016).  A majority of the articles included ADL/IADL, physical/mental health related factors (Van Cleve et al., 2023,  Aspell et al., 2019,  Spillman & Long, 2009,Tate et al., 2022,  Wu et al., 2016).  Five articles included social and communication related factors (Dahlberg et al., 2018, Van Cleve et al., 2023, Fabius et al., 2020, Tate et al., 2022, Wu et al., 2016).  It can be observed that the articles which focused on home support admission tended to include health service availability and informal care giver related factors, compared to the articles which focused on hospital admissions.  In terms of the age group, seven articles included seniors 65 years old and above (Elkjær et al., 2021,  Van Cleve et al., 2023,  Aspell et al., 2019,  Fabius et al., 2020,  Fabius et al., 2020. Tate et al., 2022, Wu et al., 2016) One article included seniors who were 76 years of age and above (Dahlberg et al., 2018).

The effect direction plot based on the synthesis of the selected articles is shown in Figure 4.  This graphical synthesizing method was chosen for this SWiM because of the variability of factors, study methods, and methods chosen to present outcome variables. Some of the factor groups that are shown in Table 2 are divided further, and a total of eighteen factor groups were presented in the plot. The directions of arrows were determined based on covariates (Spillman & Long, 2009), , odds ratios (Elkjær et al., 2021,  Van Cleve et al., 2023,  Aspell et al., 2019, Fabius et al., 2020 ) or hazard ratios (Dahlberg et al., 2018, Tate et al., 2022, Wu et al., 2016) depending on the type of method used in each article.

Figure 4:

The Effect Direction Plot based on the Synthesis of the Selected Articles

All factors were adjusted so that the direction of the factors shows the effect on the dependent variable.  For example, although the social status baseline in the figure is married, if the baseline of the gender in a study was ‘unmarried’, the direction of the effect based on the dependent variable was adjusted in a way that the effect would show based on ‘married’ as a baseline.  The risk of bias based on PROBAST (Moons et al., 2019) was denoted by three colours in the figure.  

Hospital Admission as a Transition Destination

Three studies included hospital admissions as independent variables and home support service use was the only factor that was included as an independent variable across all three articles  Elkjær et al., 2021,  Dahlberg et al., 2018, , Van Cleve et al., 2023).  Age, reported gender, IADL, and physical health were included in two articles  (Dahlberg et al., 2018) Van Cleve et al., 2023) .  Race, time in home support program, ADL, mental/cognitive health, health service availability and other home care utilization were included in only one article ( Van Cleve et al., 2023). 

The study done by Dahlberg et al. (2018) investigated planned and unplanned hospital admissions as dependent variables.  Eight factor groups were included in the article and age, home support service use and IADL showed opposite directions depending on the type of hospitalization.  For example, home support use was positively associated with unplanned hospitalization, but it was negatively associated with planned hospitalization (Dahlberg et al., 2018).   The positive association of home support service with hospital admission was identified by Elkjær et al. (2021)).   Elkjær et al. (2021) focused on readmissions to hospital between seven and thirty days after a short ED admission and some of the factors showed the same trend as some of the factors for unplanned admission that were identified by Dahlberg et al. (2018).  For example, there was a positive association between home support service use and the chance of readmission to hospital within seven and thirty days (Elkjær et al., 2021).

Van Cleve et al. (2023), on the other hand, showed similar results for hospitalization as the planned hospitalization of  Dahlberg et al. (2018)2018) .   Although they did not agree on the gender group factors, the four factor groups that Dahlberg et al. (2018)) and Van Cleve et al. (2023)() both investigated, showed the same trend.  All the factor groups that Dahlberg et al. (2018) and Van Cleve et al.  (2023) agreed on, except Physical health, did not show the same direction of trend between unplanned and planned hospital admissions in the article of Dahlberg et al. (2018). 

The difference and similarity in the trend of the group factors that were discovered in (Elkjær et al., (2021), Dahlberg et al. (2018), and Van Cleve et al. (2023) may have revealed a trend of factors that were associated with planned and unplanned hospitalization.  However, there is not enough evidence to conclude that this discovered trend is true because the study of Van Cleve et al. (2023) focused on general hospitalization and did not differentiate between planned and unplanned hospitalization for its dependent variables.    

Long Term Care Admission as a Transition Destination

Five of the selected articles focused on long term care admissions as the dependent variables.  Four articles included age as a factor group and all showed an upward trend of a change of the transition to long term care as individuals got older (Fabius et al., 2020, Spillman & Long, 2009, Tate et al., 2022, Wu et al., 2016).  In terms of gender, females were shown to be less likely to be admitted to long term care, according to three articles (Spillman & Long, 2009,  Tate et al., 2022, Wu et al., 2016).  In terms of home support service use or time in service use, three articles showed a positive association with transition to long-term care (Aspell et al., 2019 ,  Spillman & Long, 2009,  Wu et al., 2016)).  Three articles showed a positive association between transition to long term care and an increase in ADL needs (Aspell et al., 2019, Fabius et al., 2020,  Spillman & Long, 2009, Tate et al., 2022,2).  Three articles included IADL related factors and two studies showed a positive association between IADL and transition to long term care (Fabius et al., 2020, Tate et al., 2022,) and one article had a mixed outcome (Wu et al., 2016). 

Deterioration in both physical and mental health status showed an increased chance in transition to long term care in most of the articles (Aspell et al., 2019 ,  Fabius et al., 2020, Spillman & Long, 2009, Tate et al., 2022, Wu et al., 2016).  Falls-related factors showed a higher chance of transition to long term care consistently across the three articles which included falls as independent variables (Aspell et al., 2019, Fabius et al., 2020, Tate et al., 2022). In terms of factors associated with informal caregivers, two articles included informal caregiver stress as one of the independent variables and both articles showed a positive association between the burden on informal caregivers and transition to long term care (Spillman & Long, 2009, Tate et al., 2022,).  At the same time, two articles showed a negative association between informal caregiver support hours and transition to long term care (Fabius et al., 2020, Tate et al., 2022).

Most of the factors showed similar trends for long term care admission and re-institutionalization. However, unlike long term care transition, both home support service use and mental health deterioration were negatively associated with re-institutionalization (Fabius et al., 2020) ).  Assisted living transition also showed a negative association with home support service time (Tate et al., 2022).  Also, females as reported gender have a higher chance of re-institutionalization or transitioning to assisted living (Fabius et al., 2020 , Tate et al., 2022). 

Mortality as a Transition Destination

Two articles included mortality as dependent variables.  Home support service group factor was the only common factor between the two articles and both articles indicated a positive association between an increase in home support services and a higher rate of mortality (Elkjær et al. , 2021, Aspell et al., 2019).

Risks of Bias

Risk of bias was assessed based on CHARMS and PROBST (Fernandez-Felix et al., 2023) (Moons et al., 2019) and the assessment results were colour coded in Figure 2.  All articles indicated potential risks that could be introduced by the analysis.  Two articles mentioned how they evaluated the model performance and the performance measures (Van Cleve et al. , 2023, Spillman & Long, 2009).  Two of the three articles identified as high risk were due to participants with missing data being excluded from the final analyses (Dahlberg et al., 2018, Aspell et al., 2019) One article had a small sample size considering the number of covariates that were included in the final model (Fabius et al., 2020).  One study was identified as a moderate risk because the results of an adjusted model, which focused on independent variables and their odd ratios, were reported in the article. Other covariates and the performance of the original model before being adjusted, were not shared (Elkjær et al., 2021).

Discussion

Although only eight articles were included in this systematic review, the eight articles examined a total of 116 factors that could affect transitions between levels of care for home support recipients.  Although some of the factors were common across all articles, like age, how the articles used those factors in their analyses varied. For example, age was used as a continuous variable in some articles while others used it as a categorical variable.  The 116 factors were categorized under twelve factor groups and the direction of effect on transition to the next level of care was studied.  There were a variety of care transition destinations studied as dependent variables, including hospitalization, in general, planned hospitalization, unplanned hospitalization, hospital readmission, long term care admission, long term care re-institutionalization, assisted living, and mortality.  In addition, the timeline of focus also varied depending on the study.  Although there were some variability among factors within the factor groups, many factor groups like demographic factors, such as age and gender, home support related factors, ADL, physical health, mental health, caregiver burden and falls showed that most articles trended in the same direction in the association on the transition of level of care. 

Although this study shed some light on trends in the research for transition of level of care for home support recipients, this study has several limitations. 

The major limitation is due to the heterogeneity of the articles.  As mentioned earlier, there were a variety of dependent variables, observation periods and factors discovered and synthetized for this study.  Some trends on the association between the included factors as independent variables and dependent variables were identified even though there was a large variability in terms of measures developed for each factor and study method.  It is not possible to conduct meta-analysis on these articles due to the variability. 

First, a variability in the observational periods may affect the outcomes.  There may be differences in the health conditions, and the support they need, for home support recipients at various points in time while receiving home support services.  The difference of their needs and health conditions may affect the association between dependent and independent variables.  For example, Elkjær et al. (2021). included individuals who received home support after a short term ED admission ( Fabius et al. (2020) studied individuals transitioning to home and community based care for the Money Follows the Person (MFP) Rebalancing Demonstration program at the time of the study. Tate et al. (2022) included individuals who were long-stay home care clients requiring home care support for more than six months.  The observation period of the study done by Wu et al. started at the time that an individual was enrolled in home and community care (Wu et al., 2016).  Other articles used the interview or audit as the start of the study  (Dahlberg et al., 2018, Aspell et al., 2019 ,   Spillman & Long, 2009) .  Among those articles, two articles included the time in home support service as dependent variables which may affect the starting point of the observation period (Van Cleve et al., 2023, Spillman & Long, 2009). 

Another limitation is that no standard measures were included in the analysis.  For example, support for ADL is one of the main services for home support.  ADL factors included in the articles were ADL related factors from the administrative evaluation data (Van Cleve et al., 2023),   self-reported ADL impairment measure (Fabius et al., 2020, Spillman & Long, 2009), Barthel index (BI) (Aspell et al., 2019)  and a Resident Assessment Instrument–Home Care (RAI-HC) (Tate et al., 2022,  Wu et al., 2016).  Two articles did not include ADL related factors in their studies (Elkjær et al., 2021, Dahlberg et al., 2018).  Standardized ADL measures, for example, inter-RAI which are mandated for some countries including Canada (interRAI, n.d.), were not included as common measures in the selected articles which makes it difficult to evaluate ADL needs and association of care transition across the studies.  The inconsistency of measures was also observed for IADL, physical and mental or cognitive health and home support service utilization across the selected studies.

Finally, in this study, “home support” is the term used to describe specific services that individuals can get at home to live independently.  The definition and description of home support service used in this study is based on the BC Government’s definition (n.d.).  However, services provided or even the names of home support-like service vary across countries, as shown in Table 1.  This made it difficult to search articles for this review and it is possible that some articles may have not been included due to lack of clear definitions and consistent service and/or names for home support.

Conclusion

It is important to understand the reasons why home support service recipients need to move to a higher level of care.  Trends of association between some factors of home support service recipients and care transition destinations were revealed through this SWiM study.  There were some common factors that were included as independent variables across the studies.  However, measurements of common factors varied and that made it impossible to conduct meta-analysis.  Also, a variety of names and definitions and/or descriptions of home support like services will not allow researchers to compare home support service and recipients across regions or countries and to conduct rigorous analysis, such as a meta-analysis.

To understand home support service recipients’ needs to help them stay independently at home as long as they would like, some standardized measures are needed to conduct more rigorous research. Also, a standardized name and description of home support-like service internationally is encouraged to advance research in this area.

References

Aspell, N., O’Sullivan, M., O’Shea, E., Irving, K., Duffy, C., Gorman, R., & Warters, A. (2019). Predicting admission to long-term care and mortality among community-based, dependent older people in Ireland. International Journal of Geriatric Psychiatry, 34(7), 999–1007. https://doi.org/10.1002/gps.5101

BC Ministry of Health. (n.d.). Home Support—Province of British Columbia. Province of British Columbia. https://www2.gov.bc.ca/gov/content/health/accessing-health-care/home-community-care/care-options-and-cost/home-support

Boon, M. H., & Thomson, H. (2021). The effect direction plot revisited: Application of the 2019 Cochrane Handbook guidance on alternative synthesis methods. Research Synthesis Methods, 12(1), 29–33. https://doi.org/10.1002/jrsm.1458

Campbell, M., McKenzie, J. E., Sowden, A., Katikireddi, S. V., Brennan, S. E., Ellis, S., Hartmann-Boyce, J., Ryan, R., Shepperd, S., Thomas, J., Welch, V., & Thomson, H. (2020). Synthesis without meta-analysis (SWiM) in systematic reviews: Reporting guideline. BMJ, 368, l6890. https://doi.org/10.1136/bmj.l6890

Dahlberg, L., Agahi, N., Schön, P., & Lennartsson, C. (2018). Planned and unplanned hospital admissions and their relationship with social factors: Findings from a national, prospective Study of people aged 76 years or older. Health Services Research, 53(6), 4248–4267.

Edemekong, P. F., Bomgaars, D. L., Sukumaran, S., & Schoo, C. (2024). Activities of Daily Living. In StatPearls. StatPearls Publishing. http://www.ncbi.nlm.nih.gov/books/NBK470404/

Elkjær, M., Wolff, D. L., Primdahl, J., Mogensen, C. B., Brabrand, M., & Gram, B. (2021). Older adults who receive homecare are at increased risk of readmission and mortality following a short ED admission: A nationally register-based cohort study. BMC Geriatrics, 21(1), 696. https://doi.org/10.1186/s12877-021-02644-6

Fabius, C., Shugrue, N., & Robison, J. T. (2020). Outcomes Associated with Home and Community-based Service Use Among Older Adults following a Nursing Home Transition. Journal of Gerontological Social Work, 63(8), 807–821. https://doi.org/10.1080/01634372.2020.1830328

Fernandez-Felix, B. M., López-Alcalde, J., Roqué, M., Muriel, A., & Zamora, J. (2023). CHARMS and PROBAST at your fingertips: A template for data extraction and risk of bias assessment in systematic reviews of predictive models. BMC Medical Research Methodology, 23(1), 44. https://doi.org/10.1186/s12874-023-01849-0

interRAI. (n.d.). interRAI. https://interrai.org/

McHugh, M. L. (2012). Interrater reliability: The Kappa statistic. Biochemia Medica, 22(3), 276–282.

McKenzie, J. E., & Brennan, S. E. (2019). Synthesizing and presenting findings using other methods. In Cochrane Handbook for Systematic Reviews of Interventions (pp. 321–347). John Wiley & Sons, Ltd. https://doi.org/10.1002/9781119536604.ch12

Moons, K. G. M., Groot, J. A. H. de, Bouwmeester, W., Vergouwe, Y., Mallett, S., Altman, D. G., Reitsma, J. B., & Collins, G. S. (2014). Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies: The CHARMS Checklist. PLOS Medicine, 11(10), e1001744. https://doi.org/10.1371/journal.pmed.1001744

Moons, K. G. M., Wolff, R. F., Riley, R. D., Whiting, P. F., Westwood, M., Collins, G. S., Reitsma, J. B., Kleijnen, J., & Mallett, S. (2019). PROBAST: A tool to assess rsk of bias and applicability of prediction model studies: Explanation and elaboration. Annals of Internal Medicine, 170(1), W1–W33. https://www.acpjournals.org/doi/10.7326/m18-1377

Office of the Seniors Advocate BC. (2023, February). We must do better: Home support services for BC seniors. https://www.seniorsadvocatebc.ca/app/uploads/sites/4/2023/02/OSA-HOME-SUPPORT-REPORT-2023-ONLINE.pdf

Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., … Moher, D. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 372, n71. https://doi.org/10.1136/bmj.n71

Spillman, B. C., & Long, S. K. (2009). Does high caregiver stress predict nursing home entry? Inquiry: A Journal Of Medical Care Organization, Provision And Financing, 46(2), 140–161.

Tate, K., Hoben, M., Grabusic, C., Bailey, S., & Cummings, G. G. (2022). The association of service use and other client factors with the time to transition from home care to facility-based care. Journal of the American Medical Directors Association, 23(1), 133-140.e3. https://doi.org/10.1016/j.jamda.2021.06.027

Van Cleve, R., Cole, E., & Degenholtz, H. B. (2023). Risk of hospitalization associated with different constellations of home & community-based services. BMC Geriatrics, 23(1), 36. https://doi.org/10.1186/s12877-022-03676-2

World Health Organization. (2015). The growing need for  home health care for  the elderly: Home health care for the elderly as an integral part of primary health care services. https://applications.emro.who.int/dsaf/EMROPUB_2015_EN_1901.pdf?ua=1

Wu, X., Li, C., Oberst, K., & Given, C. (2016). Predicting long-term nursing home transfer from MI choice waiver program. Geriatric Nursing (New York, N.Y.), 37(6), 446–452. https://doi.org/10.1016/j.gerinurse.2016.06.013

Acknowledgement

The authors wish to acknowledge the Denis and Pat Protti Endowment Fund, the University of Victoria Faculty of Human and Social Development, the Eldercare Foundation, and Vancouver Island Health Authority, BC Canada, for their support of this research.

Author Bios

Junko Fukui Innes MSc

Junko Fukui Innes is a PhD student in the School of Health Information Science at the University of Victoria. Her research focus area is home and community care, particularly around home support, and machine learning.  She has extensive experience in advanced analytics in health care settings.

Abdul Roudsari PhD

Abdul Roudsari is a Professor of Health Informatics at the University of Victoria. Abdul’s specific areas of interest and expertise: modelling in healthcare; modelling methodology for health resource management; clinical decision support, machine learning and artificial intelligence – development and evaluation of decision support systems; evaluation methodologies with application in telemedicine.

Karen L. Courtney RN PhD

Dr. Karen Courtney is a Professor in the School of Health Information Science at the University of Victoria. She conducts research in virtual care and community-based health informatics projects. Her current funded work focuses on modernizing gender, sex and sexual orientation terminology and associated clinical practices with digital health information systems.

Mary Ellen Purkis PhD

Mary Ellen Purkis is Professor Emerita in the School of Nursing at University of Victoria. She has published in the social science and nursing literature on the topics of home care practice and aging. Her work has focused on pursuing ethnographic methods in theorizing health care practices.