The effectiveness of alternatives to residential care for older people with on-going health and social care needs: a systematic review

Abstract

Introduction

Older adults are more likely than younger people to have multiple chronic health conditions and increased health and/or social needs. As older people generally prefer living at home in the community as they age and residential care can be expensive, there is a need for effective alternatives to residential care in the community. The objective of this review was to synthesize evidence about programs aimed at enabling older people with ongoing health and social care needs to remain in the community.

Methods

This review followed the JBI methodology for systematic reviews of effectiveness. Included studies reported on complex, multifactorial interventions that were based in the community and included more than one type of service. Six databases and gray literature were searched for published and unpublished research. Titles and abstracts, and full-text selections were screened by two or more reviewers and assessed for methodological quality using JBI critical appraisal tools. Results related to quality of life and healthcare outcomes were extracted.

Results

Fifty-five full text articles, reporting on 51 unique complex interventions, were included in the review. Studies were predominantly randomized controlled trials (n=24) and quasi-experimental studies (n=23), with five cohort and three case series studies included. The overall quality of the included studies was moderate. Key characteristics of the interventions included case management, care planning, a comprehensive assessment, and in-home visits. Comparative meta-analyses were completed for five of the outcomes (hospital admission, emergency department visits, long-term care use, primary care use and quality of life). The results showed effects in the direction of interventions for the number of hospital admissions and LTC use, however, none of the meta-analyses were statistically significant.

Conclusions

There is little agreement about the effectiveness of complex interventions on quality of life and health system outcomes. Jurisdictional differences may make the integration of literature reporting on such interventions particularly difficult. There is an ongoing need to understand what helps older people with complex needs live well in the community and what level of health system engagement is optimal.

Systematic review registration

PROSPERO reference number CRD42022324061.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12877-026-07011-x.

Introduction

Globally people are living longer, and while public health advances have contributed to the decline in infant mortality and childhood illnesses, there remains a high prevalence of chronic illness in later life; this is particularly evident in high-income countries [1]. Older adults are more likely to have chronic health conditions such as heart disease, chronic obstructive pulmonary disease, diabetes, hypertension, and dementia [2, 3]. They are also more likely to experience functional changes, such as decreased mobility, communication challenges, and cognitive decline [4]. The combination of these factors often leads to the need for additional support such as help preparing meals, maintaining their home, and assisting with activities of daily living to live well in the community [5]. This type of support is often provided in nursing homes or similar residential facilities. Such residential care is costly and older adults generally prefer to remain at home in their communities as they age [6]. Care in residential facilities remains an important part of the healthcare continuum, however, there is a need to identify programs that can effectively support older people experiencing on-going health and functional needs at home in their community to support the quality of life of older people, and the sustainability of the health care system.

The older adult population, those over the age of 60, is increasingly diverse with notable variations in health status, social connections, life experiences and exposure to environmental risks [7]. However, a growing portion of the older adult population is at risk of functional decline, increased social needs and other negative health outcomes. Older adults identified as frail may experience multimorbidity, the occurrence of two or more chronic conditions [8]; still, many people who experience frailty, or functional decline, have no formal medical diagnoses [9]. For this review, we take an inclusive approach and operationalized the concept of frailty by including people aged 60 and older, with either multiple chronic conditions or functional impairments that required ongoing support for activities of daily living (ADL). ADLs are those activities that are required to care for oneself, and include tasks such as dressing, bathing, using the toilet, and eating.

Older adults experiencing frailty often need support with ADLs and ongoing health management [10]. While many older adults experiencing frailty already have informal supports such as family or friends that enable them to live in the community, there is often a need for additional support from sources such as health professionals, community groups, and private service organizations [11]. Support needs can change over time, sometimes very quickly, contributing to concerns about caregiver burden. Lack of support can contribute to people moving to facilities that offer full time caregiving, such as nursing homes and long-term care facilities [12]. Thus, this review focused on interventions or programs that offered personalized services and supports that extended across sectors and organizations, and that responded to both social and physical needs, to enable older people with complex needs to remain at home in the community. Such programs, often called complex interventions, were defined as those that included a personalized combination of support and services to address individual needs to remain at home in the community. Complex interventions often include care coordination and the development of a care plan that addresses the physical, functional, and instrumental needs of the individual, and may include goal setting and self-management strategies [13]. Complex interventions may encourage older adults to continue doing as much as they can and provide support with ADLs, preventing physical and mental health decline and reducing the impact on family and friend caregivers and the health care system [14].

There has been a growing interest in evaluating the effectiveness of complex interventions on individual and health system outcomes, yet to date there has not been a systematic review and metanalysis of this body of literature. More specifically, there is a gap in the literature on whether such programs specifically focused on frail older people are effective. Accordingly, the objective of this review was to assess the effectiveness of programs that offer individualized, multifactorial support to community-dwelling older people with ongoing health and social care needs, and report how these programs affect quality of life and health system outcomes.

A preliminary search of MEDLINE, the Cochrane Database of Systematic Reviews and JBI Evidence Synthesis was conducted and no current or underway systematic reviews on this specific topic were identified. Nevertheless, reviews on similar topics have been completed and provide further justification for the need for this review. Ho and colleagues (2023) [15] conducted a review examining the effectiveness of complex interventions for improving independent living and quality of life for people over the age of 65 in the community, however, the population of focus was not limited to those with multimorbidity or functional decline. Two other reviews focused on interventions to prevent/delay placement in long-term care (LTC), however, did not place age or health status limits on the included populations [13, 16]. In their systematic review of reviews, Duan-Porter and colleagues (2020) [13] focused on prevention of admission to LTC facilities for a broad population, without health or functional limitations. Similarly, a review by Luker and colleagues in 2019 [16] synthesized the evidence on the effectiveness of interventions to prevent nursing home admission for older people. They reported on several types of interventions, including complex multifactorial interventions [16], but likewise did not place any limitations on the health status of participants.

A review by Young and colleagues (2017) [17] synthesized the literature on the effectiveness of long-term home care interventions for functionally dependent older people. However, this review limited its focus to home care interventions providing formal care such as assistance with ADLs but did not include social care type assistance, thus excluding multifactorial interventions that may include additional services such as respite care, meal delivery or yard maintenance help. Our review expands on this work by considering multifactorial interventions for a similar population. Another systematic review [18] proposed to synthesize the literature regarding the effectiveness of community-based, complex interventions to promote independence in older adults. Again, the population was not limited by health or functional status, and the review only included randomized controlled trials (RCT). Our review furthers this body of knowledge by examining a specific sub-population of older adults—specifically people who have multiple ongoing or chronic conditions and functional impairment that require assistance to remain in the community.

Review question

To what extent do programs that offer individualized, multifactorial support to community-dwelling older people with ongoing health and social care needs affect quality of life and health system outcomes?

Methods

This systematic review was conducted in accordance with JBI methodology for systematic reviews of effectiveness [19] and an a priori protocol [20] registered with PROSPERO, reference number CRD42022324061.

Search strategy

The search strategy aimed to find both published and unpublished studies. A three-step search strategy was used in this review. First, an initial limited search of MEDLINE (PubMed) and CINAHL (EBSCO) was undertaken, followed by analysis of the text words contained in the title and abstract and the index terms used to describe the articles. The search strategy, including all identified keywords and index terms, was adapted for each included information source. Studies published from database inception to the date of the search, March 11, 2024, were eligible for inclusion. Search strategies for all included databases are available in Supplemental Material Appendix I.

Inclusion criteria

Participants

The population of interest for this review was people 60 years of age or older with on-going health and social care needs. Studies that identified the population as experiencing multiple chronic health conditions (e.g. diabetes, dementia, cardiac disease) or functional changes (e.g. cognitive decline, decreasing ability to perform ADLs, mobility challenges) and requiring continuous supportive care were included. Studies that focused on younger adults with disabilities or that focused on older people needing rehabilitation after an illness or injury were excluded.

Intervention

This review focused on multifactorial interventions that provided individualized support to older people with complex, ongoing care needs in the community. To be included, the studies must have been based in the community and incorporated more than one type of formal support or service based on individual needs. Excluded studies had assisted living or group home settings, programs aimed at rehabilitation, addressed acute care needs, provided short-term transitional care, or focused on palliative or end-of-life care.

Comparator(s)

Studies were included if they compared interventions to long-term residential care use or had no comparator.

Outcomes

This review focused on studies that measured quality of life and health system outcomes, such as hospital admission, emergency department visits, and primary care visits. Studies needed to include a quality of life (QoL) or health system outcome for inclusion. QoL outcomes were measured with tools such as the Health-Related Quality of Life Questionnaire [21] or EQ-5D [22].

Types of sources

For this review, all primary quantitative study designs including randomized controlled trials (RCTs), quasi-experimental, observational, case series and cohort studies were considered. There were no date or language restrictions on the studies.

Study selection

Citations identified in the search were uploaded into Covidence (Veritas Health Innovation, Melbourne, Australia) and duplicates removed. Two independent reviewers screened titles and abstracts against the inclusion criteria. Potentially relevant studies were retrieved in full and screened by two independent reviewers. Full-text studies that did not meet the inclusion criteria were excluded. Any disagreements that arose during screening between the reviewers were resolved through discussion or by a third reviewer. The results of the search and the study inclusion process are reported in full and presented in a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA: [23] flow diagram (see Fig. 1).

Fig. 1.

PRISMA Flow Chart

Assessment of methodological quality

Eligible studies were critically appraised by two independent reviewers at the study level for methodological quality using standardized critical appraisal instruments from JBI for experimental [24], quasi-experimental [25], cohort [26], and case series studies [27]. Conflicts between reviewers on methodological quality were resolved by a third reviewer or through discussion. All papers regardless of quality were included in the review.

Data extraction

Data was extracted by two independent reviewers using the data extraction tool, developed and tested by the authors, provided in Supplemental Material Appendix III. Extracted data included details about the population, country, intervention, comparator, and outcomes used to evaluate the effectiveness of the intervention. Disagreements that arose between the reviewers were resolved through discussion, or with a third reviewer.

Data synthesis

Findings are presented in narrative format, and in tables and figures to aid in data presentation where statistical pooling was not possible. Meta-analyses were conducted when three or more studies measured the same outcome at the same time to follow up, through the System for Unified Management, Assessment and Review of Information (SUMARI; JBI, Adelaide, Australia). Statistical analyses were performed using random effects, due to study heterogeneity, with effect sizes expressed as either odds ratios or weighted, and their 95% confidence intervals are reported. Heterogeneity and sensitivity analyses were conducted as per the priori protocol [20]. Subgroup analyses for sex and gender were not possible given the data available.

Results

Study inclusion

Searches were conducted initially in September 2022 and re-run March 2024 to locate the most up to date evidence. The searches produced a total of 45,149 potentially relevant studies. Duplicate records were removed and numbered 19,508. A total of 25,641 titles and abstracts were screened, and 25,388 sources were excluded at this stage. Full texts of 253 articles were retained for review. We were unable to locate four full text studies. A total of 249 full texts were assessed for inclusion and 194 were excluded. Reasons for exclusion were ineligible study design (n = 81); ineligible population (n = 67); ineligible intervention type (n = 23); ineligible outcomes (n = 20); and ineligible setting (n = 3). The total number of studies included in the review was 55. See Fig. 1 for the PRISMA flow chart (23).

Characteristics of included studies

See Table 1 for a summary of included studies. The research designs used in the included studies were RCTs (n = 24) [28–51], quasi-experimental (n = 23) [52–74], cohort (n = 5) [75–79] and case series (n = 3) [80–82]. The studies were conducted in 18 different countries, with the majority from the United States (n = 13) [32, 33, 36, 42, 55, 61, 65, 66, 68, 76, 79, 80, 82], Canada (n = 10) [28, 29, 37, 43–46, 56, 77, 78] and the Netherlands (n = 7) [50, 51, 58, 59, 67, 70, 71]. Other source countries were Italy (n = 4) [31, 57, 69, 74], China (n = 3) [40, 41, 72], Spain (n = 2) [60, 73], Sweden (n = 2) [30, 49], United Kingdom (n = 2) [53, 81], Australia (n = 2) [35, 39] and New Zealand (n = 2) [47, 48]. Countries with one study were Taiwan [34], Belgium [52], Switzerland [75], Korea [62], Germany [54], Finland [38], South Korea [63], and Argentina [64]. Studies were conducted between 1984 and 2024, with the median year being 2016. One study was written in French [29] and data was extracted and translated by native speakers and using google translate. All other studies were written in English.

Table 1.

Summary of included studies

Study				Aim	Intervention			Control
Primary Author	Year	Country	Design		(n)	Age (mean)	Chronic Conditions (mean)	(n)	Age (mean)	Chronic Conditions (mean)	Frailty Measure	Intervention title	Summary of findings
Beland (27)	2006a	Canada	RCT	“Assess a transformation of the organization and delivery of health and social services with intensified community-based interventions for frail elderly persons.” (p.367)	606	82	4.9	624	82	5	Nagi scale	SIPA [French acronym for System of Integrated Care for Older Persons]	Integrated systems appear to be feasible and have the potential to reduce hospital and nursing home utilization.
Beland (28)	2006b	Canada (French)	RCT	To examine differences in utilization and costs of health and social services between an experimental group receiving the SIPA Intervention and a control group	650	82	4.9	580	82	4.99	Functional Autonomy Measure	SIPA [French acronym for System of Integrated Care for Older Persons]	“…the results indicate that it is possible to expect integrated service systems for frail older persons to reduce the use and costs of hospital services and nursing homes without increasing overall health care costs, reducing quality of care, or increasing the burden on older persons and their relatives.” (p.39).
Berglund (29)	2015	Sweden	RCT	“Analyse effects of a comprehensive continuum of care persons’ life satisfaction as compared to those receiving usual care.” (p.1080)	85	20 (65–79), 65 (80+)	-	76	18 (65–79), 58 (80+)	-	Frailty was assessed as having three or more indicators, including weakness, fatigue, weight loss, reduced physical activity, impaired balance, reduced gait speed, visual impairment and impaired cognition.	Comprehensive continuum of care	“There was a positive effect of the continuum of care intervention on older persons’ life satisfaction. The older persons receiving the intervention were about two to three times more likely to improve or maintain satisfaction than those who received usual care…” (p.1086)
Bernabei (30)	1998	Italy	RCT	To evaluate the impact of an integrated social and medical services programme on admissions to institutions, use and cost of health services, and functional decline among frail elderly people living in the community	99	80.7	4.7	100	81.3	4.8	-	Integrated Care Program	An integrated community care programme implemented by an interdisciplinary team reduced the risk of hospital admission and length of stay in either hospital or nursing home.
Boult (31)	2011	United States	cRCT	“…report the effects of guided care on multimorbid older patients’ use of 6 health services through an additional 12 months… “ (p.461)	446	77.1	4.3	404	77.8	4.3	Used Hierarchical Condition Category predictive model	Guided care	Compared with usual care, patients receiving guided care used similar amounts of health care services. The only significant difference between the two groups was a 29.7% reduction in home health care episodes for the intervention group. (p.463)
Boult (32)	2013	United States	cRCT	Report the final results of a 32-month study designed to test the effect of Guided Care teams to usual care teams on functional health and quality of care for their patients (p.613)	485	77.2	4.3	419	78.1	4.3	Used Hierarchical Condition Category predictive model	Guided care	The data do support the hypotheses that Guided Care improves patients’ perceptions of the quality of their health care, their access to telephone advice, and significantly reduces such patients’ use of home health care. (Page 617).
Chen (33)	2021	Taiwan	cRCT	The aim of this study was to evaluate the effectiveness of the High-Need Community-Dwelling Older Adults Care Delivery Model.	71	77.38	1.92	74	78.53	2.16	High-Need Community Dwelling Older Adult Screening Scale	High-Need Community Dwelling Older Adults Care Delivery Model	This model had contributed to improved functional ability and quality of life, reduced depressive symptoms, greater access to adequate health care and social services, and more satisfaction with care delivery and service linkage among high-need community-dwelling older adults.
Fairhall (34)	2015	Australia	RCT	To compare the costs and cost-effectiveness of a multifactorial interdisciplinary intervention versus usual care for older people who are frail.	120	83.4	7.44	121	83.2	7.37	Cardiovascular Health Study frailty criteria.	Frailty Intervention Trial	Integrated systems appear to be feasible and have the potential to reduce hospital and nursing home utilization.
Federman (35)	2022	United States	RCT	To compare the effectiveness of HBPC versus office-based primary care for improving outcomes among home-bound older adults with chronic conditions and functional limitations.	114	82	-	112	81	-	Edmonton Symptom Assessment System	Home-based Primary Care (HBPC)	“In conclusion, HBPC improved satisfaction with care for homebound older adults and reduced hospitalization rates, but it was also associated with more deaths, a finding that warrants additional evaluation.” (p.453)
Fisher (36)	2020	Canada	RCT	“To determine the effectiveness of a 6-month, community-based, multimorbidity intervention for community-dwelling older adults with multimorbidity that were newly referred to and receiving home care services.” (abstract)	30	63% 75+	8.63	29	72% 75+	8.72	-	Multi-morbidity intervention	“While the intervention was cost neutral in comparison to usual care, it was not found to improve the primary outcome or the secondary health outcomes (mental functioning, mental health, self-efficacy).” (p.15)
Kari (37)	2022	Finland	RCT	“The aim of this study was to evaluate real-world health outcomes, quality of life and physical performance, and cost-utility of a people centred care model in primary care compared with that of usual care.” (p.3005)	151	81	-	126	81.4	-	Short Performance Physical Battery. Scale 0–12, worst to best.	People-centred Care Model	“While no statistically or clinically significant differences were observed … in their QoL, at the two-year follow-up, the physical performance in the usual care group was significantly (p = 0.03) lower than in the intervention group.” (p.3008)
Kinchin (38)	2022	Australia	step-wedge cRCT	“To assess the cost of implementation, delivery and cost-effectiveness of a flagship community-based integrated model against usual primary care” (abstract)	80 (split into 3 groups)	80.7	-	-	-	-	Functional Independence Measure	OPEN ARCH	“OPEN ARCH was associated with a favourable Bayesian CE profile in improving functional status and dependency levels, avoiding or reducing inpatient stay compared with usual primary care in the Australian context.” (abstract)
Leung (39)	2004a	China	RCT	Explore the impact of a case management project for older persons in Hong Kong.	130	74.4	2.7	130	75.3	2.9	Minimum Data Set for Home Care	case management project for community- dwelling frail elderly patients discharged from hospitals	“Our findings showed that timely and appropriate intervention using a case management approach is effective and reduces health care costs. Further- more, participants in the intervention group showed less decline in mental functioning and continence than those in the control group.” (p. 77)
Leung (40)	2004b	China	RCT	“… the aim of the project was to provide continuous support to this group of patients and their caregivers to reduce the utilization of hospital services.” (p.80)	45	75.5	3 + 51%	47	75.5	3 + 61.7%	Minimum Data Set for Home Care	Case Management	“The study demonstrated that utilisation of hospital services could be significantly reduced when a group of elderly patients and their caregivers received timely interventions and appropriate services through case management services.” (abstract)
Levine (41)	2012	United States	RCT	To assess the efficacy of a home care program designed to improve access to medical care for older adults with multiple chronic conditions at risk for hospitalization.	156	81.1	2.51	140	80.6	2.39	Electronic risk assessment screening	Choices for Healthy Aging program	“In addition, findings of significant reductions in hospital days and the decreased probability of hospitalization for CHA intervention patients suggests that home care and support may reduce use of costly acute medical service.” (p.275)
Markle-Reid (42)	2006	Canada	RCT	Evaluate the comparative effects and costs of a proactive nursing health promotion intervention in addition to usual home care for older people compared with usual home care services alone. (p.381)	120	83.37	60 (1 condition), 60 (2 conditions)	122	84.25	55 (1 condition), 67 (2 conditions)	-	Nursing health promotion intervention	“Our results suggest that with modest reorganization of the delivery of existing home care services, giving greater priority to nursing health promotion, frail older home care clients can experience improved quality of life at no additional cost to society as a whole.” (p. 392)
Markle-Reid (43)	2018	Canada	RCT	To compare the effect of a 6-month community-based intervention with that of usual care on quality of life, depressive symptoms, anxiety, self-efficacy, self-management, and healthcare costs in older adults with type 2 diabetes mellitus and 2 or more comorbidities.	80	26 (65–69), 32 (70–74), 22 (75+)	8.4	79	24 (65–69), 31 (70–74), 24 (75+)	8.3	-	Community Program	Participation in a 6-month community-based intervention improved quality of life and self-management and reduced depressive symptoms in older adults with type 2 diabetes mellitus and comorbidity without increasing total healthcare costs.
Miklavcic (44)	2020	Canada	RCT	To compare the effect of a community-based intervention versus usual care on physical functioning in older adults with type 2 diabetes mellitus and 2 or more comorbidities, and compare the effects on mental health, depressive symptoms, anxiety, self efficacy, self-management and healthcare costs.	70	44% 75+	81.4% had 6 or more	62	41.9% 75+	90.3% had 6 or more	-	The Aging, Community and Health Research Unit Community Partnership Program	“Although the overall benefits of the intervention were inconclusive for physical and mental functioning, the program shows the potential for significant improvements in mental functioning in participants with lower baseline scores…” (p.9)
Montgomery (45)	2003	Canada	RCT	To determine the impact of enhances case management, as compared to usual home care service provision, on health care utilization, emergency or hospital utilization and use of long-term care services (home care and facility-based).	78	81.4	-	74	81.4	-	-	South Winnipeg Integrated Geriatric Program (SWING)	The principle positive finding of this program was the possible reduction in the need for long-term care placement for the intervention group. (p.279–280)
Parsons (47)	2017	New Zealand	RCT	To establish the effectiveness of a restorative home support service on institutional-free survival in frail older people referred for needs assessment.	56	82.7	-	57	83.5	-	Need for help with everyday activities	Community Flexible Integrated Responsive Support Team (FIRST).	“The trial showed a trend towards improved survival and lower rates of placement in residential care for the Community FIRST model over usual care.” (p.31)
Parsons (46)	2012	New Zealand	RCT	“The primary aim of the research was to determine the effect of COSE on residential care placement and death, health-related quality of life, and caregiver burden…” (p.87)	169	80.8	-	182	81	-	Need for help with everyday activities	Coordinator of Services for Elderly (COSE)	“This trial showed that a primary care integrated model of care management, which was designed to facilitate independent living, when tested in a randomized controlled trial, decreased residential care placement and death.” (p.89)
Sandberg (48)	2015	Sweden	RCT	The aim of this study was to evaluate the effects of a case management intervention for frail older people by costs and utility.	80	81.4	3	73	81.6	4	Dependency in number of ADLs	Case management intervention	The intervention was cost neutral and does not seem to have affected health-related quality of life, however, it seems to have reduced hours and cost of informal care and help required with instrumental activities of daily living.
van Hout (49)	2010	Netherlands	RCT	To discover if preventive home visits by nurses prevent functional decline, institutionalization and mortality through multidimensional assessments and individualized care plans.	331	81.3	2.1	320	81.5	2	Groningen Activity Restriction Scale	Preventive home visits	Preventive home visits by nurses did not demonstrate any preventive effects for older people in primary care. (p.739)
van Leeuwen (50)	2015	Netherlands	Step-wedge cRCT	“To evaluate the cost-effectiveness of the Geriatric Care Model (GCM), an integrated care model for frail older adults based on the Chronic Care Model, with that of usual care.” P. 2494	1147 (split into 4 groups)	80.5	-	-	-	-	Frailty Index	Geriatric care model	“There were no statistically significant differences in primary and secondary outcomes between intervention phases and usual care phases (Table 2).” (p.2498)
de Almedia Mello (51)	2016	Belgium	Quasi	To examine the effects of home care interventions for frail older people in delaying permanent institutionalization during 6 months of follow-up (p.2252)	4607 (mild 1999; mod 2608)	80.9/81.6 (mild/mod)	-	3633 (mild 1871; mod 1762)	82.9/84.9 (mild/mod)	-	Edmonton Frail Scale or Katz Scale	Home Care Interventions for frail older people at risk of institutionalization	Occupational therapy and multi-component interventions were the most effective types of interventions in delaying institutionalization. Older people with moderate to severe impairment who received night support at home with full supervision had a higher risk of institutionalization and death.
Chapman (52)	2019	United Kingdom	Quasi	“To assess the ‘Okay to Stay’ plan to investigate if this reduces visits to emergency departments, unplanned admissions and elective admission to hospital in elderly patients with long-term health conditions.” (abstract)	50	77.5	all reported 1+	-	-	-	-	Okay to Stay	“The Okay to Stay programme has demonstrated a significant reduction in numbers of annual visits to emergency departments, and in the number of unplanned emergency admissions… ” (p.5)
Hasemann (53)	2022	Germany	Quasi	Evaluate the effectiveness of this multi-component community-based care approach for older people with functional impairments in terms of progression in need for care and self- reported health-related quality of life.	873	80.13	-	1797	80.32	-	LUCAS-FI/long-term care grade	Innovative Care Approach (NWGA)	The analyses did not reveal considerable health effects of the community-based intervention as applied in this study.
Hughes (54)	1984	United States	Quasi	Explore the impact of the Five Hospital Homebound Elderly Program, a model long-term, comprehensive, coordinated home care program in Chicago.	162	80.4	3.1	167	77.6	3.4	The Duke/OARS Multidimensional Functional Assessment Questionnaire (OMFAQ)	The Five Hospital Homebound Elderly Program (FHHEP)	Major findings include a significant reduction in the nursing home admissions and nursing home days of experimental group clients. The reported analyses also show an increase in experimental clients’ sense of physical health well-being and a decrease in their number of previously unmet needs for community services.
Kelly (55)	2015	Canada	Quasi	The goal of the present study was to examine the effectiveness of the Fraser Health surveillance nurse telephone support initiative.	930	83.2	-	930 (after matching)	83.1	-	MAPLe Method for Applying Priority Levels	Surveillance nurse telephone support intervention	Those in the treatment condition experienced better outcomes than their matched control counterparts, in terms of rate of emergency room registrations, hospital admissions and days in hospital per 100 days in the home care program.
Landi (56)	1999	Italy	Quasi	To examine the impact on hospital use/cost of a specific home care program based on comprehensive geriatric assessment and case management.	115	77.5	3 to 4	-	-	-	Minimum Data Set for Home Care	Integrated Home Care Program	Home care services based on full integration and case management and on the use of an innovative and comprehensive assessment instrument can achieve cost savings in long-term care of frail older individuals.
Landi (73)	2001	Italy	Quasi	“…To examine the effect of a home care program based on comprehensive geriatric assessment— Minimum Data Set for Home Care—and case management on hospital use/cost of frail elderly individuals”. (p.968)	1204	77.4	3.5	-	-	-	Minimum Data Set for Home Care	Integrated Home Care Program	“It appears that home care services based on full integration and case management as well as on the use of an innovative and comprehensive assessment instrument … can achieve cost savings in long-term care of frail elderly individuals.” (p.970)
Looman (57)	2016	Netherlands	Quasi	To examine the effect of the Walcheren Integrated Care Model on health outcomes, functional abilities and quality of life of community-dwelling frail older people.	184	81.8	3.8	193	82.3	3.9	Groningen Frailty Indicator (GFI)	Walcheren Integrated Care Model	“Our study shows that the Walcheren Integrated Care Model has a positive effect on the ability to receive love and friendship, and the WICM moderately preserves the general quality of life of frail older people. The WICM is not effective in terms of health outcomes and functional abilities.” (p.6)
Looman (58)	2014	Netherlands	Quasi	Explore the short-term value of integrated care for the frail elderly by evaluating the effects of the Walcheren Integrated Care Model on health, quality of life, health care use and satisfaction with care.	205	82	-	212	82.46	-	Groningen Frailty Indicator (GFI)	Walcheren Integrated Care Model	The main conclusion is that the Walcheren Integrated Care Model had only a small overall effect. The model had a positive effect on attachment, a dimension of quality of life, however, health care use was not affected by the integrated care intervention.
Mateo-Abad (59)	2020	Spain	Quasi	Evaluate the impact of CareWell integrated care model for older patients with multimorbidity in the Basque Country.	101	79.6	9.6 Charleston Comorbidity index	99	79.2	9.9 Charleston Comorbidity index	Basque population-based risk stratification.	CareWell integrated care model	The study shows different utilization of health resources in the intervention and control groups. Patients in the intervention group had fewer hospital admissions and made fewer emergency visits, whereas they had more contacts with their GPs, and more face-to-face visits with their PC nurse.
Moreno (60)	2021	United States	Quasi	Objectives were to reduce unnecessary healthcare utilization, enhance provider and patient satisfaction, and improve communication between patient and the healthcare team. (p.1628)	420	74.4	-	700	75	-	Medical group risk score	Connecting Provider to Home	“… we found that a home-based social worker and CHW intervention was associated with significant reductions in utilization rates for ED visits and hospitalizations in the 12 months after program enrollment as compared with the 12 months before enrollment.” (p.1632-3)
Noh (61)	2021	Korea	Quasi	“In this study, we aimed to apply the Community-Based Integration Service model among older adults living alone and evaluate its effectiveness.” (p.1489)	331	< 80 375; >80 171	4.44	546	< 80 239; >80 92	3.9	Korean Frailty Index	Community-Based Integration Service Model	“The Community-Based Integration Service model improved frailty, loneliness and QoL in community-dwelling older adults living alone.” (p.1495)
Oh (62)	2021	South Korea	Quasi	To compare physical function and institutionalisation-free survival time over 30 months between intervention group and comparison group.	187	76.3	1.5	196	76.3	1.5	Deficit-accumulation frailty index with 47 items as part of comprehensive geriatric assessment (0–1)	A multicomponent intervention	The 24-week multicomponent intervention showed sustained improvement in physical function, temporary reduction in frailty and longer institutionalisation-free survival over 30 months.
Perman (63)	2021	Argentina	Quasi	The objective was to evaluate if a new health and social integration programme for frail older adults decreases the hospital admission rate of the participants compared to the best standard of care. (p.29)	121	86.2	-	121	84.5	-	Definition of frailty – “A decline in intrinsic capacity (composite of all the physical, psychological and mental capacities of an individual).”	Health and Social Care Integration Programme	We found that those who received the health and social care integration programme intervention had about half the risk of hospital admissions, better quality of life and a non-statistically significant trend towards lower hospital deaths. (p.32–33)
Popejoy (64)	2015	United States	Quasi	The goal of this study was to compare utilization and cost outcomes of patients who received long-term care coordination in an Aging in Place program to patients who received care coordination as a routine service in home health care.	213	78.8	-	585	75.4	-	Not a frailty measure but OASIS (Outcomes and Assessment Information Set) was used to assess functional status, behavioral health, and pain.	Aging in Place Program	“This study adds to the growing body of evidence about the effectiveness of nurse care coordination. Care coordination managed by RNs can influence utilization and cost outcomes, and impact health and functional abilities.” (p.312–313).
Prior (65)	2012	United States	Quasi	“To investigate and report on the impact of a community-based senior outreach program designed to positively impact the lives of elders who had been repeatedly hospitalized and decrease rates of hospital readmissions and emergency department visits of those clients.” (p.348)	193	69.5	-	-	-	-	Participants had multiple ED visits and were enrolled in senior outreach program	Senior Outreach Program	“…research thus far suggests positive outcomes for the program in decreasing emergency department visits and hospital readmissions and improving clients’ quality of life in a number of areas.” (p.354)
Ruikes (66)	2016	Netherlands	Quasi	“We evaluated the effectiveness of a general practitioner–led extensive, multicomponent program integrating cure, care, and welfare for the prevention of functional decline.” (abstract)	287	83.1	-	249	80.5	-	The frailty index was defined as the proportion of accumulated deficits.	CareWell primary care program	“We found no effects of the CareWell primary care program on functioning, quality of life, mental health, health-related social functioning, institutionalization, hospitalization, and mortality among community-dwelling frail elderly people in Dutch primary care.” (p.213)
Schubert (67)	2016	United States	Quasi	“This article describes the lessons learned and outcomes observed during a practical clinical trial of the implimentation of GRACE…” (p.1504)	179	78	-	77	77	-	Charlson Comorbidity Index	Geriatric Resources for Assessment and Care of Elders (GRACE)	The enrollment in the GRACE program was associated with significantly fewer emergency department visits, readmissions, hospitalizations, and total bed days of care per 100 veterans. The GRACE program can effectively improve care quality and reduce healthcare costs for high-risk older veterans.
Tiozzo (68)	2019	Italy	Quasi	To describe the impact of a Care Management Program developed for patients affected by chronic heart failure and multimorbidity.	244	78.4	4.3	244	79.1	4.5	-	Care Management Program	“Our study shows that a care management program can be successful in reducing hospitalization rates and ER visits in patients with multimorbidity, chronic heart failure and complex health care needs.” (p. 245)
van Dijk (69)	2016	Netherlands	Quasi	Evaluation of the effects of Integrated Neighborhood Approaches on older people’s health-related quality of life	186	81.6	98% had 1+	186	79.8	93% had 1+	Tilburg Frailty Indicator (TFI)	Integrated Neighborhood Approaches	“The Integrated Neighborhood Approaches was found to have no substantial effect; the control group showed slightly better well-being and physical functioning than the intention to treat group, but these differences were not clinically relevant.” (p.6)
Vestjens (70)	2019	Netherlands	Quasi	To report on the “cost-effectiveness of the proactive, integrated primary care program Finding and Follow-up of Frail older persons compared with usual primary care for community-dwelling frail older persons…” (p.1 – abstract).	232	82.45	92.6% 2+	232	82.41	89.7% 2+	Tilburg Frailty Indicator (TFI)	Finding and Follow-up of Frail older persons (FFF)	“The results of our economic evaluation indicate that proactive, integrated care for community-dwelling frail older persons as provided in the FFF program is most likely not a cost-effective initiative compared with usual primary care in the Netherlands, in terms of wellbeing and QALYs over a 12-month period” (p.9).
Vila (72)	2015	Spain	Quasi	To improve the efficiency and effectiveness of care, guarantee continuity of care, and optimize healthcare resources, a home healthcare program was created for individuals with multiple chronic conditions.	261	84.4	90% had 2+	32	?	all had 2+	PROgnostic Model and FUNctional Prediction Developed (PROFUND) for Pluripathologic Patients in Spain Indexes	home healthcare program	The intervention reduced the number of hospital admissions and length of stay, resulting in good patient satisfaction and lower costs. The main features that helped reduce admissions were multidisciplinary teamwork, electronic medical records, and the ability to deliver complex care in participants’ homes.
Yu (71)	2020	China	Quasi	“…to examine the effectiveness of an integrated care model supported by frailty assessment, personalized care plan and coordinated care services as arranged by community centres for older people in pre-frail and frail older people.” (p.1049)	183	76.3	-	270	75.9	-	FRAIL scale (includes fatigue, resistance, ambulation, illness and weight loss). Scores range from 0–5 where 3–5 represents frail.	Integrated Intervention	This model of care was found to reduce frailty scores in both pre-frail and frail community-dwelling older adults. However, no significant improvements were found for health services use.
Di Pollina (74)	2017	Switzerland	Cohort	Testing the efficacy of providing integrated care at home to reduce unnecessary hospitalizations, emergency room visits, institutionalization, and mortality in community dwelling frail and dependent older adults.	122	81.8	-	179	81.9	-	Presence of at least one indicator of frailty (impaired cognition, falls, social isolation, or frailty of the informal caregiver support) was required for inclusion	Formal care coordination	This trial showed that formally coordinating existing resources of the public and private sectors for frail and dependent-community dwelling older adults reduced the rate of hospitalizations, decreased unnecessary hospitalizations, lowered the rate of emergency room visits, and increased the proportion of patients dying at home.
Edwards (75)	2023	United States	Cohort	Description of a cohort of Veterans Affairs HBPC patients, their care patterns, clinical outcomes and care trajectories over one year.	10,571	77.7	-	-	-	-	Jen Frailty Index	Veterans Affairs HBPC	“After enrollment, HBPC patients have reductions in acute care use, spend a large majority of time in non-institutional settings, and most HBPC decedents receive hospice care.” (p.88)
McGregor (76)	2018	Canada	Cohort	“To assess the influence of a multidisciplinary HBPC program on acute hospital use through ED visits and hospital admissions, and examine the same outcomes for community-dwelling older adults using home care over the same time.” (p.2)	246	85.2	-	492	84.1	-	CHESS (Hospital and Community Outcome Measures); MAPLe Method for Applying Priority Levels	ViVE (Home-Based Primary Care)	“Although both populations were substantially different, results suggest that expansion of HBPC to individuals at an earlier stage of their frailty trajectory may be an opportunity to “bend the curve” of increasing hospital use seen in the home care cohort, thereby improving care and reducing cost.” (p.9)
Rosenberg (77)	2012	Canada	Cohort	“To evaluate the effect of medical Primary Integrated Interdisciplinary Elder Care at Home on acute hospital use and mortality in a frail elderly population.” (abstract)	198 (active)	86.7	-	50 (discharged)	89.2	-	Canadian Study on Health and Aging Clinical Frailty Scale	Primary Interdisciplinary Elder Care at Home	“This article demonstrates that this model of care may also reduce some of the costs in the acute hospital care system and that, for active patients, there can be a significant reduction in acute hospital admissions and hospital days, as well as a smaller reduction in ED use.” (p.1344)
Valluru (78)	2019	United States	Cohort	To determine the effect of home-based primary care (HBPC) for frail older adults, operating under Independence at Home incentive alignment on long-term institutionalization. (pg 1495)	570	49% 85+	-	w/out HBPC 573; with HBPC 82	44.9% 75+/44% 85+	-	Jen Frailty Index	HBPC integrated with Long-Term Services and Supports	“We found in our study that providing HBPC integrated with long-term support services in the community, as part of a program with aligned Medicare incentives, improves community survival free of long-term institutionalization for frail, medically complex Medicare beneficiaries without increasing the costs of community supports.” (p.1500)
Kling (79)	2023	United States	Case-control	“We conducted a longitudinal, matched case-control study to evaluate the impact of the program on healthcare service utilization” (p.213)	117	85.2	-	328	84.4	-	Hospital Frailty Rist Score and modified Charlson Comorbidity Index	Home-based primary care	“This HBPC program increased the amount of primary care provided to patients but did not change utilization of specialty of hospital services relative to matched controls” (p.219)
Lewis (80)	2017	United Kingdom	Case Series	“The primary aim of the study was to determine if unplanned hospital admission and ED presentations could be reduced through a CVW [community virtual ward] model.” (p.988)	54	81.6	all reported 2+	-	-	-	Rockwood Clinical Frailty Index Scale	Community Virtual Ward	“This study demonstrates that a CVW model for the care of older persons with complex health care and social care needs can assist in reducing the number of ED presentations and unplanned hospital admissions.” (p.990)
Wajnberg (81)	2010	United States	Case Series	To describe the sociodemographic, clinical, and functional characteristics of patients enrolled in House Calls Program and to determine whether enrollment in House Calls Program is associated with fewer hospital and long-term care admissions.	179	79	All participants identified as homebound	-	-	-	seven-item activity of daily living scale (score/14)	House Calls Program Model of Care	A House Calls Program may be associated with fewer hospitalizations and long-term care placements.

ADL, activities of daily living; cRCT, cluster randomized controlled trial; CVW, community virtual ward; ED, emergency department; GFI, Groningen Frailty Indicator; GRACE, Geriatric Resources for Assessment and Care of Elders; HBPC, home-based primary care; Quasi: quasi-experimental; RCT, randomized controlled trial; SIPA, French acronym for System of Integrated Care for Older Persons; TFI, Tilburg Frailty Indicator

Intervention characteristics

There were 51 different interventions reported in the 55 included studies. See Supplemental Material Appendix IV for a more detailed description of the interventions. While the interventions were all community-based, they were delivered via several models, including through primary care (n = 24), community organizations (n = 11), home care (n = 12), and mobile interprofessional geriatric teams (n = 4). The most common intervention element was case management (n = 47), which involved personalized support from case managers who tended to be nurses or social workers, with some interventions using primary care providers, physical therapists, or community health workers to provide case management. Care planning was an element of 38 interventions, and other common intervention elements included a comprehensive initial assessment (n = 40) (e.g. comprehensive geriatric assessment), in-home visits (n = 34), a multi-disciplinary team (n = 34), medication review (n = 13), and caregiver resources (n = 24). See Table 2 below for a description of the interventions and a summary of included elements. Subgroup analyses were conducted to examine the impact of the model of delivery and elements of the interventions and no statistically significant results were found.

Table 2.

Intervention description and included elements

Author/year	Intervention Name	Intervention Description	CM	CA	CP	HV	MD	+CG	PC	Social Care	Med Rev
de Almedia Mello (2016)(51)	Home Care Interventions for frail older people at risk of institutionalization	“Interventions were identified as single component (occupational therapy (OT), psychological support, night care, day care) or multicomponent. The latter included case management in combination with OT and psychological support or physiotherapy, with rehabilitation services or with OT alone.” (abstract)	X				X
Beland (2006a)(27) Beland (2006b)(28)	SIPA [French acronym for System of Integrated Care for Older Persons]	“…community-based multidisciplinary teams with full clinical responsibility for delivering integrated care through the provision of community health and social services and the coordination of hospital and NH (nursing home) care…” (p.368)	X	X			X	X		X
Berglund (2015)(29)	Comprehensive Continuum of Care	“The intervention included geriatric assessment, case management, interprofessional collaboration, support for relatives and organising of care-planning meetings in older persons’ own homes.” (abstract)	X	X	X	X		X
Bernabei (1998)(30)	Integrated Care Program	“[intervention group] received case management and care planning by the community geriatric evaluation unit and general practitioners. All the services considered necessary were provided in an integrated fashion…” (p.1348)	X	X	X	X	X
Boult (2011)(31) Boult (2013)(32)	Guided Care	“A registered nurse collaborated with two to five primary care physicians in providing eight services to participants: comprehensive assessment, evidence-based care planning, proactive monitoring, care coordination, transitional care, coaching for self-management, caregiver support, and access to community-based services” (abstract)	X	X	X	X	X	X	X
Chapman (2019)(52)	Okay to Stay	Plan is developed through stages: “Community Matrons support the patient to complete the plan (with family members or carers involved); the plan is completed and saved on the Electronic Patient Record; the plan is shared with and discussed with the GP; the plan is emailed to the GP collaborative…; several copies of the plan are printed for the patient and family members to keep…; the Community Matron provides any indicated follow-up, such as referrals on or changes to care; the plan is reviewed every three months or when indicated by significant change.” (p.2)			X			X
Chen (2021)(33)	High-Need Community Dwelling Older Adults Care Delivery Model	“…involved case screening, comprehensive assessment and care coordination…Based on the results of the assessments and the types of the high-need older adults, the care coordinator used a specifically designed list of categorized services to formulate an individualized care plan…” (p.300)	X	X	X	X		X
Di Pollina (2017)(74)	Formal Care Coordination	In addition to usual care by their primary care physician and nursing team, “the intervention group received an additional home evaluation by a community geriatrics unit with access to a call service and coordinated follow-up” (abstract)	X	X		X	X				X
Edwards (2023)(75)	Veterans Affairs Home-Based Primary Care (HBPC)	“…comprehensive longitudinal care to patients with complex, chronic disabling disease” (abstract)	X			X	X
Fairhall (2015)(34)	Frailty Intervention Trial (FIT)	“A 12-month multifactorial, interdisciplinary intervention targeting identified frailty characteristics” (abstract)	X	X	X		X				X
Federman (2023)(35)	HBPC	“At the initial visit, the physician completed a medical history and physical exam and developed a plan of care. Additional assessments included symptoms, activities of daily living, and fall risk if indicated…Follow-up visits occurred at the team’s discretion.” (p.445)	X	X	X	X	X			X
Fisher (2020)(36)	Multimorbidity Intervention	Intervention was delivered by an interprofessional team through four main components: in-home visits, monthly case conferences, case management and usual home care services. (p.4–5)	X		X	X	X	X	X
Hasemann (2022)(53)	Innovative Care Approach (NWGA - (‘NetzWerk GesundAktiv’)	“The innovative care approach includes a geriatric assessment, a case and network management as well as digital supporting tools…” (abstract)	X	X	X			X		X
Hughes (1984)(54)	The Five Hospital Homebound Elderly Program (FHHEP)	“…a model long-term, comprehensive, coordinated home care program…” (abstract)	X	X	X			X		X
Kari (2022)(37)	People-Centred Care Model (PCCM)	“[intervention] comprised: an at-home patient interview by a named nurse and a pharmacist; completing health (the named nurse) and clinical medication (a pharmacist) reviews; and agreeing on the care and medication plan based on the patient’s care targets and needs…During the two years of follow-up, care coordination and health support were provided by the named nurse.” (p.3006)	X	X	X	X	X		X		X
Kelly (2015)(55)	Surveillance Nurse Telephone Support	“The intervention consisted of regularly scheduled telephone calls from a surveillance nurse to proactively assess the individual’s well-being, care plan status, use of and need for services (home support, adult day program, physiotherapy, etc.) and home environment (e.g. informal caregiver support).” (abstract)	X		X			X		X
Kinchin (2022)(38)	OPEN ARCH	“…intervention delivered integrated care for community-dwelling frail people through systemised integration of primary and secondary care. This involved a prevention model of comprehensive assessment, coordination and management by co-located community-facing specialist geriatric and primary care services.” (p.3)	X	X	X		X		X
Kling (2023)(79)	HBPC	“The HBPC program included elements of successful home-based primary care models, such as team-based home visits by medical providers and social workers, interdisciplinary team meetings once a week to discuss newly enrolled patients and ongoing patient concerns, and after-hours care to support patients and their caregivers.” (p.214)				X	X		X
Landi (2001)(73) Landi (1999)(56)	Integrated Home Care Program	“Each patient was assessed with the Minimum Data Set for Home Care, and, subsequently, a case manager and a multidisciplinary team delivered social and health care services as indicated.” (abstract)	X	X	X		X
Leung (2004a)(39)	Case Management Project for community- dwelling frail elderly patients discharged from hospitals	Services were delivered by a case manager (social worker plus registered nurse). Including: regular home visits and telephone consultations; comprehensive geriatric assessment; care plans; links to formal health and social services; on-site or phone health and psychosocial counseling; health educational programs and support groups. (p.72)	X	X	X	X		X		X
Leung (2004b)(40)	Case Management	“…intervention group received case management services through assigned case managers… services included: regular monitoring of subjects’ health status…; availability for phone assistance…; home visits, if needed; prescribing of community-based supportive services, including community nursing services; and access to the case geriatrician…” (p.80)	X			X	X
Levine (2012)(41)	Choices for Healthy Aging (CHA) Program	“Goals of the CHA program… (1) early identification and treatment of exacerbation of the illness; (2) patient-specific health education; (3) self-management or caregiver management of the disease, and (4) advance care planning and other psychosocial issues.” (p.e270)	X	X	X	X	X	X			X
Lewis (2017)(80)	Community Virtual Ward	“The model was set up to work within existing resources within primary and secondary care through an integrated approach to older persons’ care overseen by a clinical case manager… This was key in providing timely service interventions and follow-up, as well as access to the day hospital…” (p.987)	X	X		X	X	X		X
Looman (2016)(57) Looman (2014)(58)	Walcheren Integrated Care Model	“Geriatric nurse practitioners and secondary care geriatric nursing specialists were assigned as case managers and co-ordinated the care agreed upon in a multidisciplinary meeting. The general practitioner practice functions as a single-entry point and supervises co-ordination of care.” (abstract)	X	X	X	X	X	X
Markle-Reid (2006)(42)	Nursing Health Promotion Intervention	“In addition to usual home care, the nursing group received a health assessment combined with regular home visits or telephone contacts, health education about management of illness, coordination of community services, and use of empowerment strategies to enhance independence.” (abstract)	X	X	X	X		X			X
Markle-Reid (2018)(43)	Community Program	“Client-driven, customized self-management program with up to 3 in-home visits from a registered nurse or registered dietician, a monthly group wellness program, monthly provider team case conferences and care coordination and system navigation.” (abstract)	X		X	X	X	X	X
Mateo-Abad (2020)(59)	CareWell Integrated Care Model	“…based on the coordination between health providers, home-based care and patient empowerment, supported by information and communication technology tools.” (abstract)	X	X	X		X		X
McGregor (2018)(76)	ViVE (Home-Based Primary Care)	“The program includes family physicians and nurse practitioners… HBPC services include planned regular home visits, responsive day-time and after-hours care for emergencies, and nursing, physical and occupational rehab services as needed.” (p.2)	X	X	X	X	X
Miklavcic (2020)(44)	The Aging, Community and Health Research Unit Community Partnership Program (ACHRU-CPP)	“The intervention involved up to three in -home visits, a monthly group wellness program, monthly case conferencing, and care coordination.” (abstract)	X			X	X	X	X	X
Montgomery (2003)(45)	South Winnipeg Integrated Geriatric Program (SWING)	“Intervention subjects received a multidimensional assessment as soon as possible by a specially trained coordinator, who had enhanced access to geriatric medical and day-hospital services; intervention patients were case managed for a 3-month period.” (abstract)	X	X	X		X
Moreno (2021)(60)	Connecting Provider to Home	“Community-based healthcare program delivered by a social worker and community health worker for older adults with complex medical and social needs.” (abstract)	X	X		X	X			X
Noh (2021)(61)	Community-based Integration Service Model	“The model comprises eight healthcare services, five social care services and tailored case management.” (p.1489)	X	X		X	X		X	X
Oh (2021)(62)	Multicomponent Intervention	“The 24-week multicomponent intervention programme consisted of group exercise, nutritional supplementation, depression management, medication review and home hazard reduction.” (p.2159)		X		X					X
Parsons (2012)(46)	Coordinator of Services for Elderly (COSE)	“After a standardized comprehensive assessment, a package of required support services was contracted, and the same COSE worker maintained continuation of care from referral until discharge…” (p.87–88)	X	X	X			X
Parsons (2017)(47)	Community Flexible Integrated Responsive Support Team (FIRST).	“A registered nurse case manager coordinator undertakes a comprehensive geriatric assessment and sets mutually agreed and meaningful short and long-term goals for the older person…” (p.28)	X	X	X	X		X	X	X
Perman (2021)(63)	Health and Social Care Integration Programme	“…a health and social care councillor that systematically reviewed the social and biological situation following a structured process, evaluating: functionality, nutrition, mobility, pain, cognition, medication reconciliation and adherence, need for care, quality of care and environmental safety.” (abstract)	X	X	X			X			X
Popejoy (2015)(64)	Aging in Place Program	“…nurse care coordinators working with advanced practice registered nurse expert to manage a comprehensive care plan that coordinated physicians, nurses, and other professionals’ interventions to improve or support older adults’ medical conditions, physical functioning, medication management and supervision of health and social services necessary to maintain older adults in their homes.” (p.307–308)	X		X						X
Prior (2012)(65)	Senior Outreach Program	“The program provides comprehensive in-home case management services to clients to manage chronic illness through the development of an individual case plan based on the assessment of eight domains…” (p.348)	X	X	X	X	X	X	X	X
Rosenberg (2012)(77)	Primary Interdisciplinary Elder Care at Home (PIECH)	“Primary geriatric care provided by a physician, nurse and physiotherapist in participants’ homes.” (abstract)	X	X	X		X
Ruikes (2016)(66)	CareWell Primary Care Program	“…consisted of proactive care planning, case management, medication reviews and multidisciplinary team meetings with a general practitioner, practice and/or community nurse, elderly care physician and social worker.” (abstract)	X	X	X		X	X			X
Sandberg (2015)(48)	Case Management Intervention	“The case management intervention consisted of four dimensions: traditional case management tasks (assessment, care plans, care coordination, home visits, telephone calls and advocacy), general information (about the healthcare system, social activities, nutrition, exercise), specific information (related to the respondent’s specific health status, individual needs and medication) and aspects related to safety…” (p.3)	X	X	X	X				X
Schubert (2016)(67)	Geriatric Resources for Assessment and Care of Elders (GRACE)	“…is a model of care that works in collaboration with primary care providers and patient-centered medical homes to provide home-based geriatric care management focusing on geriatric syndromes and psychosocial problems commonly found in older adults.” (p1504)	X	X	X	X	X		X
Tiozzo (2019)(68)	Care Management Program (CMP)	“The CMP for multimorbid complex patients puts the patient at the center of care, granting care coordination, care planning, medicine review, adverse even monitoring and treatment adherence.” (p.242)	X	X	X	X	X	X	X		X
Valluru (2019)(78)	HBPC Integrated with long-term services and supports (LTSS)	“Each site has HBPC proactively integrated care coordination with community supports based on availability of resources and local relationships. Support to patients ranged from adult day healthcare and home health aide-provided personal care services to transportation, meals and equipment (e.g. stair glides).” (p.1498)	X			X	X
van Dijk (2016)(69)	Integrated Neighborhood Approaches (INA)	“…professionals and residents are asked to watch over neighbors and report manifestations of frailty to INA community workers… Community workers visited older people at home and mapped their social and physical needs and capabilities in respect to factors such as housing, mobility issues and social activities…Together with older people, they sought appropriate solutions to identified problems or needs and composed individualized support plans.” (p.3)		X	X	X			X	X
van Hout (2010)(49)	Preventive Home Visits	“…proactive home visits by trained community nurses. The nurses (a) assessed the care needs… (b) determined care priorities together with the person; © designed and executed individually tailored interventions; and (d) monitored participants by telephone and on average three home visits.” (abstract)	X	X	X	X			X		X
van Leeuwen (2015)(50)	Geriatric Care Model	“[intervention] consisted of the following components: a regularly scheduled in-home comprehensive geriatric assessment b a practice nurse followed by a customized care plan, management and training of practice nurses by a geriatric expert team, and coordination of care through community network meetings and multidisciplinary team consultations of individuals with complex care needs.” (abstract)	X	X	X	X	X		X
Vestjens (2019)(70)	Finding and Follow-up of Frail older persons (FFF)	“This approach combines several interrelated components with the aim of providing high-quality proactive, integrated primary care for frail community-dwelling persons… The FFF approach is implemented in general practitioner practices and led by general practitioners.” (p.3)	X	X	X	X	X	X	X		X
Vila (2015)(72)	Home Healthcare Program	“The program’s medical coordinator evaluated all individuals, determined whether the criteria for inclusion in the program were met, and assigned a primary care physician and a nurse to attend the person while his or her condition remained stable (follow-up team).” (p.1018)	X	X	X	X	X	X
Wajnberg (2010)(81)	House Calls Program (HCP) Model of Care	“…consisted of two primary care physician (PCP)-nurse practitioner (NP) teams and one social worker…The assigned PCP performs a comprehensive initial visit … [which] consists of a complete medical history; physical examination; and assessment of function, disability and cognition. After the initial visit, the NP makes monthly visits…” (p.1145)	X	X		X	X
Yu (2020)(71)	Integrated Intervention	“Participants in the intervention group received an integrated intervention consisting of in-depth assessment, personalised care plans and coordinated care.” (p. 1049)	X	X	X					X	X

CM, case management; CA, comprehensive assessment; CP, care planning; HV, home visits; MD, multidisciplinary; +CG, caregiver involvement; PC, primary care; Med Rev, medication review; GP, general practitioner; HBPC, home-based primary care

Participants

Sample sizes ranged from 50 to 10,571 participants. The average age of participants across studies was approximately 80 years. Just over half of the studies reported on the number of chronic conditions (n = 29, 53%), with the average being four conditions. Comorbidities included conditions such as hypertension, heart failure, stroke, chronic obstructive pulmonary disease, diabetes mellitus and cognitive decline such as dementia. Almost all studies identified participants as frail or functionally impaired using pre-determined criteria, such as the Tillberg Frailty Indicator or the Groningen Frailty Indicator (n = 46, 84%). Four studies analyzed their findings by level of frailty or functional ability [35, 52, 54, 72].

Methodological quality

The overall quality of the included studies was assessed as moderate. The RCTs demonstrated moderate methodological quality with an average score of 66% or 8.6 out of 13. Due to the nature of the interventions, many of the studies did not include assessor or participant blinding, and randomization was often clustered. The quality of the quasi-experimental studies was moderate to high, with three out of 23 studies demonstrating high quality with a score of 9 out of a possible 9. Four studies had lower scores because they used a pre/post design and did not include control groups. The three case series studies were assessed to have overall moderate quality with an average score of seven out of a possible ten. However, only one study included cases consecutively and participants were included using enrollment date or a convenience sample. Finally, the five included cohort studies demonstrated overall moderate quality with an average score of eight out of a possible 11. Many of the studies did not adequately address confounding factors, such as differences in baseline characteristics, and did not address incomplete follow-up of participants. See Appendix II in the Supplemental Material for full description of methodological quality appraisal results.

Main findings

Quality of life

Twenty-one studies reported on quality of life (QoL) [30, 33–39, 43–45, 49, 51, 54, 58, 59, 62, 64, 67, 70, 71]. The studies were either RCTs (n = 13) or quasi-experimental (n = 8). Quality of life measures focused on dimensions such as mobility, usual activities, physical pain or limitations and overall mental health. Nine studies [35, 39, 49, 58, 59, 62, 64, 67, 70] measured health-related quality of life using the EQ-5D, and two studies used the EQ-5D with an added cognitive component [59, 67]. Another frequently used tool was the Short-Form Health Survey (n = 8; [33, 37, 38, 43–45, 51, 54]. Four versions of the tool were used, including the SF-12, SF-36 (versions 1 and 2), and the SF-6D. Five other tools were used to measure QoL: EQ-VAS [62], QoL in Alzheimer’s disease [36], the LiSat-11 [30], WHOQOL-BREF (Taiwan version; [34]), and ICECAP [58, 59].

Five of the 21 (24%) studies that measured QoL outcomes reported a statistically significant increase in QoL following the intervention: High-Need Community Dwelling Care Delivery Model [34], Community Program [44], Nursing Health Promotion Intervention [43], Community-based Integration Service Model [62], and Health and Social Care Integration Program [64]. Two studies reported QoL as an improvement in the mental health component of the SF-12 and SF-36 [43, 44]. Two other studies reported that the intervention group had significantly higher WHOQOL-BREF and EQ-5D scores [34, 64], indicating a positive effect of the intervention on QoL. Finally, one study [62] reported statistically significant improvement in QoL scores for the intervention group with the EQ-VAS, but not the EQ-5D-5 L.

Meta-analysis

Due to the heterogeneity of study design and time to follow up, meta-analysis was limited to nine of the 13 studies that measured QoL at 12 months post intervention [35, 36, 38, 49, 51, 58, 64, 67, 70]. Of the nine studies, only one found a statistically significant positive effect of the intervention on quality of life scores [64]. A comparative meta-analysis of studies was performed using standardized mean difference (SMD) to account for the different QoL measures. Results show a low effect (Cohen’s d = 0.09) that was not significant (P = 0.477). There was considerable heterogeneity of the studies with a I² of 93 that indicates much of the variance was accounted for in study heterogeneity. A forest plot is included as Fig. 2.

Fig. 2.

Meta-analysis of effects of intervention on QoL at 12 months

Health system outcomes

See Fig. 3 for a summary of the number of studies reporting on each of the health system outcomes included here.

Fig. 3.

Number of studies reporting on each health system outcome. LTC: long-term care; QoL: quality of life; ED: emergency department

Hospital admissions

Description of studies that measured effect of intervention on hospital admissions

Thirty-seven studies reported on the effect of interventions on future hospital admissions [28, 29, 31–33, 35–37, 39–42, 46, 50, 53, 55–57, 59–61, 64–69, 72–78, 80–82]. Of those, 14 were RCTs [28, 29, 31–33, 35–37, 39–42, 46, 50], seven were quasi experimental [56, 60, 64, 65, 67, 69, 72], eight were pre-post design with a control group [55, 59, 61, 66, 68, 73, 75, 77], and eight were pre-post with no control group [53, 57, 66, 74, 76, 78, 81, 82]. A summary of the studies that evaluated hospital admission following the intervention by comparison with a control group is included in Table 3, and those that used pre/post designs in Table 4.

Table 3.

Impact of intervention on hospital admission for studies with control groups

Study author/year	Intervention Name	What was reported	Intervention	Control	P value
Bernabei (1998)(30)	Integrated Care Program	Number admitted to hospital at least once over 12 months	n = 36	n = 51	0.05*
Federman (2022)(35)	Home-Based Primary Care (HBPC)	Percent change 12 months pre/post intervention	-35.4%	-17.5%	0.001*
Levine (2012)(41)	Choices for Healthy Aging	Percent with one or more inpatient days during 12-month study period	25.6%	37.1%	0.02*
Leung (2004a)(39)	Case Management Project	Percent decrease of unplanned hospital admissions	36.8%	20.4%	0.011*
Leung (2004b)(40)	Case Management	Total episodes of hospital admission (mean) post intervention	2.3	2.7	0.05* difference for IG
Mateo-Abad (2020)(59)	CareWell Integrated Care Model	Rate per year of hospitalizations	0.5	0.8	0.026*
Moreno (2021)(60)	Connecting Provider to Home	Difference in no hospitalizations pre/post intervention (percentage)	10.9%	5.3%	< 0.05*
Perman (2021)(63)	Health and Social Care Integration Program	Crude hazard ratio of proportion of intervention to control without hospital admission after a year	0.622	Not reported	0.013*
Tiozzo (2019)(68)	Care Management Program	Reduction in hospitalization rates (percentage)	39%	No change reported	Not reported*
Vila (2015)(72)	Home Healthcare Program	Number of admissions adjusted for participant and month	0.36	0.19	< 0.001*
Beland (2006a)(27)	System of Integrated Care for Older Persons [SIPA fr.]	Percent increase in inpatient care utilization	-5%	Not reported	Not reported
Beland (2006b)(28)	System of Integrated Care for Older Persons [SIPA fr.]	Percent access to institutional services - hospitalizations	51%	50%	Not reported
Boult (2013)(32)	Guided Care	Overall treatment effect (7 pods) of adjusted mean annual per capita rate ratio guided care: usual care	0.94	Not reported	Not reported
Boult (2011)(31)	Guided Care	Mean Annual per Capita Use	0.7	0.72	Not reported
Di Pollina (2017)(74)	Formal Care Coordination	Percent hospitalized at least once while in study (5–41 months)	59%	60%	Not reported
Fairhall (2015)(34)	Frailty Intervention Trial	Number of ED visits over 12-month study period	152	135	0.79
Fisher (2020)(36)	Multimorbidity Intervention	Number of participants hospitalized over 6-month period of intervention	n = 1	n = 4	Not reported
Hughes (1984)(54)	The Five Hospital Homebound Elderly Program (FHHEP)	Posttest hospitalization rate (percent)	35%	53%	Not reported
Kelly (2015)(55)	Surveillance Nurse Telephone Support	Mean hospital admissions (after matching)	0.76	0.86	Not reported
Kinchin (2022)(38)	OPEN ARCH	Mean inpatient stays	0.301	0.232	-0.089 (adjusted)
Kling (2023)(79)	HBPC	Mean adjust pre/post difference for inpatient hospitalizations	0.5	0.4	0.75
Looman (2014)(58)	Walcheren Integrated Care Model	Percent of participants with hospital admission over 3 months	9%	9.5%	Not reported
Montgomery (2003)(45)	South Winnipeg Integrated Geriatric Program (SWING)	Likelihood of accessing hospital services	Not reported	Not reported	Not reported
Popejoy (2015)(64)	Aging in Place Program	Percent with one hospitalization during study period	28%	30%	Not reported
Ruikes (2016)(66)	CareWell Primary Care Program	Percent change during follow-up	18.1%	22.9%	0.17
Schubert (2016)(67)	Geriatric Resources for Assessment and Care of Elders (GRACE)	Percent decrease in hospitalizations	37.9%	Not reported	0.14
van Hout (2010)(49)	Preventive Home Visits	Percent admitted to the hospital at least once	49.2%	55.9%	0.19
Yu (2020)(71)	Integrated Intervention	Mean number of hospitalizations after 12 months	51	67	Between group comparison not reported

* Significantly fewer admissions in Intervention group; HBPC: home based primary care

Table 4.

Impact of intervention on hospital admission for pre/post studies

Study author/year	Intervention Name	What was reported	Pre	Post	P value
Chapman (2019)(52)	Okay to Stay	Pre/post emergency hospital admissions (mean)	2.64	1.3	0.015*
Edwards (2023)(75)	Veteran Affairs HBPC	Percentage of hospitalizations 6 months pre/post entry	40.6%	25%	Not reported*
Landi (1999)(56)	Integrated Home Care Program	Percentage of persons admitted to hospital at least once 6 months pre/post intervention	56%	46%	< 0.001*
Landi (2001)(73)	Integrated Home Care Program	Percentage of persons admitted to hospital at least once 12 months pre/post intervention	44.5%	26.3%	< 0.001*
Lewis (2017)(80)	Community Virtual Ward	Median number of unplanned hospital admissions	1	0	0.001*
Prior (2012)(65)	Senior Outreach Program	Mean monthly hospitalizations during study period, 24-months or follow-up	0.21	0.13	0.02*
Rosenberg (2012)(77)	Primary Interdisciplinary Elder Care at Home (PIECH)	Number of admissions pre/post program entry (active)	n = 84	n = 34	0.001*
Wajnberg (2010)(81)	House Calls Program (HCP) Model of Care	Percent of participants with one or more hospitalizations	61%	38%	< 0.001*
McGregor (2018)(76)	ViVE (Home-Based Primary Care)	Mean hospital admission rate per 1000 patient days (Intervention group)	2.3	2.2	0.726

* Significantly fewer admissions after the intervention; HBPC: home-based primary care

Eighteen studies (49%) reported a significant reduction in hospital admissions following the intervention [31, 36, 40–42, 53, 57, 60, 61, 64, 66, 69, 73, 74, 76, 78, 81, 82]. Of the studies reporting a significant reduction in admissions, 10 included a control group and eight were pre/post studies with no control group.

Meta-analysis

A meta-analysis was completed for studies that reported the number of hospital admissions 12-months after the intervention. Sixteen studies were ineligible for inclusion in meta-analysis due to their design [53, 55, 57, 59, 61, 66, 68, 73–78, 80–82]. Of the 21 studies that reported on hospital admissions at 12 months, two were cohort studies [74, 78], one case-control [80] and one was quasi experimental before/after [53] and were therefore excluded from the meta-analysis. An additional seven studies reported hospital admission data as a mean (average number of admissions per person) or as a proportion of the sample without admissions and were also excluded [32, 33, 41, 56, 64, 68, 75]. The ten studies that reported odds ratios of the probability of admission to hospital after intervention compared to a control group were included in a comparative meta-analysis. The result showed a clear trend in the direction of the intervention but was not statistically significant (p = 0.153). The results are displayed in the forest plot below (See Fig. 4).

Fig. 4.

Meta-analysis of intervention effect on hospital admission

Emergency department (ED) visits

Description of studies that measured effect of intervention on ED visits

Twenty-six studies reported on the effect of interventions on the number of visits to the ED [28, 31–33, 36, 37, 39–42, 50, 53, 56, 60, 61, 65, 66, 68, 69, 73, 75–78, 80, 81]. Eleven were RCTs [28, 31–33, 36, 37, 39–42, 50], six were quasi experimental [56, 60, 61, 65, 68, 69], three were pre-post design [53, 66, 73], four were cohort studies [75–78], and one each were case series or case control [80, 81]. See Table 5 for a summary of the findings from studies with control groups reporting on ED visits and Table 6 for a summary of findings related to ED visits from studies that used designs without a control group.

Table 5.

ED visits for studies with control groups

Study author/year	Intervention	Results reported	Intervention	Control	P value
Bernabei (1998)(30)	Integrated Care Program	Relative risk of being admitted to ER	6	17	< 0.025*
Di Pollina (2017)(74)	Formal Care Coordination	Percent with at least one ER visit	6.6% (n = 8)	14.5% (n = 26)	0.04*
Mateo-Abad (2020)(59)	CareWell Integrated Care Model	Rate per year (mean)	0.3	1.3	< 0.001*
Moreno (2021)(60)	Connecting Provider to Home	No ED visits difference (percentage) pre/post intervention	12.3%	1.9%	< 0.05*
Tiozzo (2019)(68)	Care Management Program	Percent reduction	33%	0	Not reported*
Beland (2006a)(27)	SIPA [French acronym for System of Integrated Care for Older Persons]	Percent increase in utilization of emergency department (visits)	-11%	Not reported	Not reported
Boult (2013)(32)	Guided Care	Overall treatment effect (7 pods) of adjusted mean annual per capita rate ratio guided care: usual care	1.02	Not reported	Not reported
Boult (2011)(31)	Guided Care	Mean Annual per Capita Use	0.44	0.44	Not reported
Federman (2022)(35)	HBPC	Number of ED visits during intervention period	18	17	Not reported
Fisher (2020)(36)	Multimorbidity Intervention	Number of participants with ED visits over 6 months	6	4	Not reported
Kelly (2015)(55)	Surveillance Nurse Telephone Support	Mean ED visits (after matching)	1.56	1.66	Not reported
Kinchin (2022)(38)	OPEN ARCH	Mean number of ED visits	0.250	0.226	0.577
Kling (2023)(79)	HBPC	Mean difference pre/post intervention	0.3	0.3	0.84
Leung (2004a)(39)	Case Management Project for community- dwelling frail elderly patients discharged from hospitals	Mean number of ED visits after 6 months	0.4	0.2	Not reported
Leung (2006b)(40)	Case Management	Total episodes of ED visits (mean) post intervention	0.3	0.8	Not reported
Levine (2012)(41)	Choices for Health Aging Program	Percent with one or more ED visits during 12-month study period	16.7%	21.4%	0.19
Popejoy (2015)(64)	Aging in Place Program	Percent with no ED visits	39%	47%	Not reported
Schubert (2016)(67)	Geriatric Resources for Assessment and Care of Elders (GRACE)	Annual emergency department visits per 100 veterans	252.9	282.7	0.59
van Hout (2010)(49)	Preventive Home Visits	Percent with at least one ED visit	32.3%	28.8%	Not reported
Vila (2015)(72)	Home Healthcare Program	Mean number of ED visits difference (intervention/control) pre/post intervention	0.3	0.2	Not reported

ED, emergency department; * indicates significant decrease; HBPC, home-based primary care

Table 6.

ED visits for pre/post studies

No Control	Intervention		Pre	Post	P value
Chapman (2019)(52)	Okay to Stay	Pre/post ED visits (mean)	3.25	1.3	0.009*
Lewis (2017)(80)	Community Virtual Ward	Median number of ED visits	1	0	< 0.001*
Prior (2012)(65)	Senior Outreach Program	Mean monthly ED visits during study period, 24-months or follow-up	0.17	0.11	0.007*
Edwards (2023)(75)	HBPC	Percent of ED visits 6 months pre/post intervention	24.6%	18%	Not reported
McGregor (2018)(76)	ViVE (HBPC)	Mean number of ED visits pre/post intervention per 1000 patient days	4.1	3.7	0.332
Rosenberg (2012)(77)	Primary Interdisciplinary Elder Care at Home (PIECH)	Number of ED visits pre/post program entry (active)	90	82	0.66

ED, emergency department; * indicates significant decrease; HBPC, home-based primary care

Eight studies (31%) reported statistically significant differences in the direction of the intervention for the number of ED visits (see Table 4) [31, 53, 60, 61, 66, 69, 75, 81]. Of those studies, five included a control group and three were pre/post.

Meta-analysis

A meta-analysis was performed using studies that reported ED visits at 12-months following the intervention. Twenty one of the 26 studies that reported on ED visits were excluded from the meta-analysis due to study design (n = 9; [53, 66, 73, 75–78, 80, 81], type of data reported (n = 7; [31–33, 41, 56, 60, 68] or time of measurement (n = 5; [28, 37, 39, 40, 50]. A comparative meta-analysis was performed with findings from five studies and are reported in Fig. 5. The results are inconclusive and not statistically significant (p = 0.664).

Fig. 5.

Meta-analysis of ED visits at 12 months

Long term care (LTC) use

Description of studies that measured effect of intervention on LTC use

Seventeen studies reported on the effect of interventions on LTC use [28, 29, 31, 32, 35, 46–48, 50, 52, 55, 56, 63, 67, 76, 79, 82]. Of those, nine were RCTs [28, 29, 31, 32, 35, 46–48, 50], five were quasi-experimental [52, 55, 56, 63, 67], two cohort studies [76, 79], and one case series [82].

The operational definition of LTC use differed among the studies. For example, some studies identified the occurrence of LTC use as admission to facilities as a permanent outcome, while others included short stay use of LTC such as for rehabilitation after a hospital admission or illness. Table 6 summarizes how LTC use was reported. Of the seventeen studies that reported on LTC use after a complex intervention, 41% (n = 7) found a significant decrease in LTC use after the intervention [46, 47, 52, 55, 56, 63, 82]. In Table 7 we summarize the findings of the included studies by reporting how LTC use was measured and whether the study found statistically significant reductions in use.

Table 7.

Effect of interventions on LTC use

Study author/year	Intervention	How results were reported	Significantly less LTC use
de Almedia Mello (2016)(51)	Home Care Interventions for frail older people at risk of institutionalization	Relative risk for institutionalization	Y
Hughes (1984)(54)	The Five Hospital Homebound Elderly Program (FHHEP)	Overall institutionalization rate (%)	Y
Montgomery (2003)(45)	South Winnipeg Integrated Geriatric Program (SWING)	Number designated for LTI after randomization	Y
Oh (2021)(62)	A Multicomponent Intervention	30-month mean institutionalization-free survival	Y
Parsons (2012)(46)	Coordinator of Services for Elderly (COSE)	Combined death and LTI	Y
Wajnberg (2010)(81)	House Calls Program (HCP) Model of Care	Number of Patients ≥ 1 skilled nursing facility admissions	Y
Beland (2006a)(27)	SIPA [French acronym for System of Integrated Care for Older Persons]	Admission to skilled nursing home	NS
Beland (2006b)(28)	SIPA [French acronym for System of Integrated Care for Older Persons]	Length of stay (days)	NS
Bernabei (1998)(30)	Integrated Care Program	Number admitted to nursing home or hospital	NS
Boult (2011)(31)	Guided Care	Skilled nursing home admissions and days spent in care	NS
Edwards (2023)(75)	Veterans Affairs HBPC	Percentage with a nursing home admission	Not reported
Fairhall (2015)(34)	Frailty Intervention Trial (FIT)	Number of users of permanent high-level residential care	NS
Kelly (2015)(55)	Surveillance Nurse Telephone Support	Nursing home admission (%)	Significance not reported for LTC use alone
Parsons (2017)(47)	Community Flexible Integrated Responsive Support Team (FIRST)	Number of people entering permanent residential care	NS
Ruikes (2016)(66)	CareWell Primary Care Program	Number of residential and nursing home admissions	NS
Valluru (2019)(78)	HBPC Integrated with long-term services and supports (LTSS)	Long-term institutionalization rate at 36 months (%)	NS
van Hout (2010)(49)	Preventive Home Visits	Number institutionalized	NS

LTI, long-term institutionalization; Y, yes; NS: not significant; HBPC, home-based primary care

Meta-analysis

Of the seven studies that reported at this time point, one was a cohort study [79]. There were six studies with control groups that measured LTC admission at 12 months post intervention eligible for inclusion in a meta-analysis. Two of these studies reported data as a mean or hazard ratio [32, 63]. Data from the remaining four studies were used in the meta-analysis [31, 35, 56, 67]. The results demonstrate a strong trend favoring the intervention with less LTC use at 12 months for those in the experimental groups, but the results were not conclusive with large confidence intervals that crossed the line of no effect (Fig. 6).

Fig. 6.

Meta-analysis of studies reporting LTC admission at 12 months

Primary care visits

Description of studies that measured effect of intervention on primary care visits

Twelve studies reported on the effect of an intervention on the number of primary care visits [32, 33, 35, 42, 48, 55, 59, 60, 65, 71, 72, 80]. Of those studies, five were RCTs [32, 33, 35, 42, 48], six were quasi-experimental [55, 59, 60, 65, 71, 72] and one a case-control [80].

Three of the twelve studies (25%) reported a significant effect of the intervention on the number of primary care visits [55, 60, 80]. Hughes et al. (1984) reported that the mean number of self-reported visits to primary care significantly decreased for the intervention group [55]. Kling et al. (2023), in their study of home-based primary care (HBPC), reported a significant decrease of non-HBPC in-clinic primary care visits, however when combined with HBPC medical provider visits, there was a significant increase in primary care visits for the intervention group [80]. Mateo-Abad et al. (2020), also reported a significant effect on primary care visits, however, the mean use of primary care per year increased for the intervention group [60]. In contrast, the authors noted that hospital admissions decreased significantly in this study [60]. The remaining studies reported non-significant results for the number of primary care visits after a complex intervention.

Meta-analysis

A meta-analysis was completed on studies that reported primary care visits at 12 months (n = 4) (Fig. 7). Of the nine studies that reported primary visits at 12 months, four reported the results as a mean number of visits. The remaining five studies were excluded due to type of data reported [32, 48, 65, 72] or an incompatible study design [80]. The results showed no significant effect of the intervention, favoring the direction of the control groups.

Fig. 7.

Meta-analysis of studies reporting primary care visits at 12 months

Hospital days

Description of studies that measured effect of intervention on number of hospital days

Fifteen studies included the effect of the intervention on number of days in hospital [32, 35, 39–41, 46, 55, 57, 68, 71, 73, 74, 77, 78, 81]. Six studies were RCTS [32, 35, 39–41, 46], six were quasi-experimental [55, 57, 68, 71, 73, 74], two were cohort [77, 78] and one case series [81]. 53% (53%; n = 8) of the studies reported a significant reduction in hospital days for the intervention group compared to control groups [40, 41, 57, 68, 73, 74, 78, 81]. However, the heterogeneity of the studies prevented meta-analysis. A description of how hospital days were reported, including significance, is shown in Table 8.

Table 8.

Effect of intervention on hospitalization days

Study author/year	Intervention	What was reported	P value
Landi (1999)(56)	Integrated Home Care Program	Total number of hospital days pre/post	0.0018*
Landi (2001)(73)	Integrated Home Care Program	Total number of hospital days pre/post	0.001*
Leung (2004a)(39)	Case Management Project for community- dwelling frail elderly patients discharged from hospitals	Total number of bed-days	0.006*
Lewis (2017)(80)	Community Virtual Ward	Number of hospital bed-days	0.001*
Rosenberg (2012)(77)	Primary Interdisciplinary Elder Care at Home (PIECH)	Number of bed days for active participants (pre/post)	0.004*
Schubert (2016)(67)	Geriatric Resources for Assessment and Care of Elders (GRACE)	Annual bed days per 100 veterans	0.01*
Vila (2015)(72)	Home Healthcare Program	Average length of stay	< 0.001*
Boult (2011)(31)	Guided Care	Mean annual per capita use	Not reported
Fairhall (2015)(34)	Frailty Intervention Trial (FIT)	Number of hospital days	0.79
Hughes (1984)(54)	The Five Hospital Homebound Elderly Program (FHHEP)	Mean number of days	Not reported
Kinchin (2022)(38)	OPEN ARCH	Mean average length of stay (days)	-0.415
Leung (2004b)(40)	Case Management	Mean number of hospital bed-days (significant increase for CG)	0.05
McGregor (2018)(76)	ViVE (HBPC)	Hospital days	Not reported
Montgomery (2003)(45)	South Winnipeg Integrated Geriatric Program (SWING)	Number of bed days	Not reported
Vestjens (2019)(70)	Finding and Follow-up of Frail older persons (FFF)	Mean hospital days (increased for both IG/CG)	Not reported

IG, intervention group; CG, control group; * significantly fewer hospital days recorded; HBPC, home-based primary care

Mortality

Thirteen studies reported the effect of the intervention on mortality [36, 47, 48, 50, 52, 54, 55, 63, 64, 67, 73, 76, 80]. Four of the studies were RCTs [36, 47, 48, 50], seven were quasi-experimental [52, 54, 55, 63, 64, 67, 73], and one each of cohort [76] and case-control [80]. As reported in Table 9, seven of the studies found no difference in mortality between intervention and control groups after participation in the intervention [48, 50, 52, 55, 64, 67, 73]. Three of the studies (n = 3) reported a statistically significant higher mortality rate for the intervention group compared to the control group [36, 54, 80]. Additionally, there were a few studies that combined mortality and institutionalization as institutional free survival time [47, 48, 63]. Two of these studies reported significant improvements for the intervention groups [47, 63].

Table 9.

Effect of interventions on mortality

Study	Intervention	What was reported	IG	CG	P value
de Almedia Mello (2016)(51)	Home Care Interventions for frail older people at risk of institutionalization	Relative risk of death (mild impairment)	Not reported	Not reported	Not reported
		Relative risk of death (moderate to severe impairment)	0.3 for day care use	Not reported	Not reported
Edwards (2023)(75)	Veterans Affairs HBPC	Number of deaths	2026 (19.2%)	n/a	Not reported
Federman (2022)(35)	HBPC	Number of deaths	24 (21.1%)*	12 (10.7%)	p < 0.0001
Hasemann (2022)(53)	Innovative Care Approach (NWGA - (‘NetzWerk GesundAktiv’)	Percent of participants who died	4.12%	2.89%	0.001a
Hughes (1984)(54)	The Five Hospital Homebound Elderly Program (FHHEP)	Number of deaths	14 (11.5%)	13 (10.6%)	Not reported
Kling (2023)(79)	HBPC	Number of deaths	35 (30%)*	33 (10%)	p < 0.0001
Oh (2021)(62)	A Multicomponent Intervention	30-month mean institutionalization-free survival time at 30-months (percent)	87%	64.9%	0.001
Parsons (2017)(47)	Community Flexible Integrated Responsive Support Team (FIRST)	Number of participants who died during study period	13	16	Not reported
Parsons (2012)(46)	Coordinator of Services for Elderly (COSE)	Number of participants who died during study period	17	21	Not reported
Perman (2021)(63)	Health and Social Care Integration Programme	Number of deaths	18 (14.9%)	16 (13.2%)	0.711
Ruikes (2016)(66)	CareWell Primary Care Program	Number of deaths	31 (10.8%)	21 (8.4%)	0.7
van Hout (2010)(49)	Preventive Home Visits	Number of deaths	20 (6.4%)	23 (6.9%)	Not reported
Vila (2015)(72)	Home Healthcare Program	Number of deaths	103 (40%)	18 (56%)	Not reported

NS, not significant; Sig, significant; HBPC, home-based primary care; * significant increase; ^a unreliable estimation

Discussion

We synthesized the evidence on the effectiveness of complex, multifactorial interventions for frail older adults in the community. We found 55 relevant studies that reported on 51 unique interventions. The quality of the included studies was moderate. Interventions were delivered via primary care, home care, community organizations and mobile geriatric health teams. Primary care interventions included both traditional clinic-based models and models of home visits. The complex interventions commonly included care planning, case management and comprehensive assessment but also included multidisciplinary teams, medication review, caregiver involvement and patient centered approaches. Methodological differences posed challenges to meta-analyses. While the studies often reported on similar outcomes, there were differences in the measures used and the time periods at which data were collected. We completed five meta-analyses of outcomes that were measured 12 months after the intervention, namely quality of life, number of hospital admissions, number of ED visits, primary care use and LTC use. The meta-analyses of studies reporting on the effectiveness of complex interventions on the number of hospital admissions and LTC use showed effects toward intervention groups, but results were not statistically significant. Meta-analyses on the findings related to ED visits, and primary care use showed effects in the direction of controls, however, were also not statistically significant. The studies reporting on quality of life showed a high degree of heterogeneity.

As acute hospital care is one of the most costly and detrimental aspects of the healthcare system for older people, the findings of this review related to hospitals are particularly important. The body of evidence related to the effect of complex interventions for community-dwelling frail people on admission to hospital shows fairly strong evidence of effectiveness. Eighteen of the 37 studies (49%) that reported on hospital admissions showed a statistically significant lower number of hospital admissions for the intervention group. Meta-analysis of the nine studies that reported odds ratios of the probably of admission to hospital after the intervention showed pooled effectiveness in the direction of the intervention groups. Also, the effectiveness of complex interventions on the number of days in hospital, while reported differently across studies precluding metanalysis, was statistically significant in eight of fifteen studies (53%). Together, these findings show that complex community-based interventions for older people with complex needs may be well suited to preventing hospital use.

The review supports what has been found in other reviews about the heterogeneity of design and evaluation of studies examining the interventions for community-dwelling older people. For example, Ho et al., 2023 [15] noted the variability in outcomes that were measured and tools to measure them and Sadler et al., 2023 [83] also cited that there were discrepancies in outcomes measured, and the time points when measurements were made. For the studies included in this review, it was common to measure outcomes at 12 months post intervention, however just under half used other time frames (n = 23, 42%). The inconsistency of outcome reporting is particularly evident in the reports on health system outcomes in this review, with outcomes such as hospital use being reported using so many different measures, we were unable to synthesize the findings. For future research in this area, we recommend authors include the findings in formats that can be used in meta-analysis, whether that is mean scores or odds ratios, and measuring outcomes up to at least 12 months.

Our review found that case management was a common component of complex interventions. Indeed, the role of case management for supporting older people with complex needs in the community has been explored elsewhere [83, 84]. A review of this literature highlighted four key elements of case management that have been associated with positive outcomes for older people: assessment of needs, patient/family participation in decision making, evidenced-informed care using multiple approaches and an interdisciplinary team [85]. These components are similar to components of complex interventions described in this study thus suggesting there is conceptual ambiguity about the essential components of both case management and complex interventions. While guidelines such as the Medical Research Council guide to developing and evaluating complex interventions are likely useful, they are not specific to interventions for the needs of frail older people [86]. There needs to be further development of a taxonomy of terms used in community-based complex intervention development so that evidence on the aspects of interventions that work well to support frail older people can be better understood. This would also support further work such as that done by Crocker et al. (2024) [87] that uses network analysis to examine the various components of interventions on outcomes.

The interpretation of the results of this review is complicated by the jurisdictional differences of each of the included studies [88, 89]. The studies included in this review took place in 18 different countries, with 13 from the United States alone, which itself has various state-specific models for supporting frail older people in the community. Such jurisdictional differences related to complex interventions for older people include differences in public and private care, structure of primary care, availability of multidisciplinary teams, role of informal caregivers and payment structures for medical services [90]. Jurisdictional differences are also seen in how data is collected (e.g. structure of forms, health records) and how health conditions are defined (e.g. frailty) [91]. There are also cultural elements of the local context, particularly related to familial responsibility and/or the importance of informal caregiving [88]. Such differences often go unexamined in effectiveness studies and are thus not included in discussions of knowledge synthesis. However, they have implications for the integration of services and supports and what is or is not included in the interventions, as well as out-of-pocket and public costs [8, 92, 93].

The growing interest in, and use of, implementation science to support the success of complex interventions provides further support for exploring jurisdictional differences in relation to complex interventions to support frail older people in the community. Such literature highlights the need to understand and account for context in relation to intervention design and suggests that interventions be tailored to the specific context where it is to be implemented [86]. This suggests an integration of jurisdictional and policy analysis and community-level interventions is important. This could take the form of including information on contextual elements in the reporting of individual studies or using research methods that account for differences in context in their design. For example, case study research that identifies specific jurisdictions that have the desired outcomes, such as low levels of admissions to hospitals for frail older people or high levels of quality of life among frail older people, and exploring differences in how community supports and services are offered and provided to such populations. One example of a comprehensive description of jurisdictional differences that would support an examination of context in relation to interventions to support people with dementia in the community is Peckam and colleagues (2022) [94] who compared five jurisdictions in Canada and the United States. Research that includes an examination of context with evaluation of effectiveness could contribute to narrowing the focus of interventions to elements that have been shown to affect population level outcomes.

This review was informed by the perspective of older people themselves, and therefore we were particularly attuned to individual level indicators of intervention effectiveness. One of the primary outcomes of interest for this review was QoL. Only five of the 21 studies reporting on QoL outcomes found statistically significant outcomes, and a meta-analysis on findings from similar studies showed very little effect and high heterogeneity. We question whether there are other measures from the perspective of older people that could be used to determine how effective complex interventions are in meeting the needs of individuals. Beswick et al., 2008 [95] noted that outcomes such as autonomy, independence, self-esteem and self-confidence relating to the older adult receiving the intervention may be more accurate indicators of effectiveness. More research is needed to connect patient perspectives on frailty care in the community with the evaluation of interventions [96]. For example, it may also be relevant to include measures of community connection and engagement, or continuity of care as these have been shown to be important to older people and their caregivers.

Limitations

This systematic review was limited by the operationalization of frailty and complex care that we used. Others have noted the challenges of using definitions with a biomedical or social foundation and their limits [97]. With the growing use of frailty measures in the literature, it may be possible to better refine the search parameters in the future. Also, we did not include cost as an outcome in evaluating the interventions. This should be the focus of further research and should be conducted with consideration for both out-of-pocket and societal level costs in relation to jurisdictional differences. While we did not limit our search to any one language, our search terms and databases were English based. This may have resulted in relevant studies written in languages other than English being missed.

Conclusions

While there is a large body of evidence related to supporting older people with complex needs in the community, there is little agreement of the effectiveness of complex interventions on quality of life and health system outcomes. Jurisdictional differences, including how terms are defined, eligibility criteria and cultural expectations may make the synthesis of literature reporting on such interventions particularly difficult. There is an ongoing need to understand what helps older people with complex needs live well in the community and what level of engagement with the health care system is most appropriate.

Abbreviations

ADL

activities of daily living

CG

control group

cRCT

cluster randomized controlled trial

ED

emergency department

GFI

Groningen Frailty Indicator

GRACE

Geriatric Resources for Assessment and Care of Elders

HBPC

home-based primary care

IG

intervention group

LTC

long-term care

LTI

long-term institutionalization

NS

not significant

Quasi

quasi-experimental

QoL

quality of life

RCT

randomized controlled trial

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Sig.

significant

SIPA

French acronym for System of Integrated Care for Older Persons

SMD

standardized mean difference

SUMARI

System for Unified Management, Assessment and Review of Information

TFI

Tilburg Frailty Indicator

Y

yes

Authors’ contributions

Authors EM and EPJ came up with the review concept and design. Data was acquisitioned, analyzed and interpreted by authors HM, MM, LEW, RA, EPJ, CM and AM. Drafting of the manuscript was completed by EM, HM and AM. Authors LEW, MM and RA completed a critical revision of the manuscript. Search strategy and methods were completed by EL. EM supervised this project. All authors read and approved the final manuscript.

Funding

This study received funding from CIHR through the SPOR Evidence Alliance.

Data availability

Availability of data and materials: Datasets generated and analyzed during this study are included in this published article and their supplementary information files.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.