Definition, Types and Objectives of Home Care Services
The implicit definition of home care in the studies reviewed here appears to be any broadly defined health-related services provided in people’s homes, as well as the remote coordination of care provided to the recipient of such services 6-8. Content-wise, home care can include a vast array of services ranging from specialized medicine (for example, care provided by geriatricians, oncologists, psychiatrists, etc.) to services to help with daily living (meal preparation, house cleaning, etc.) and various types of care designed to promote health, gain or maintain autonomy, or cure ailments9. The diversity and breadth of what is described as home care in the scientific literature constitutes a significant challenge for integrative reviews such as ours.
Beyond that general definition (or lack thereof), the literature we analyzed can be characterized based on two main splits that we identified. First, there is a clear divide between fixed duration “reablement” interventions (also named restorative care in some jurisdictions) and indefinite-duration need-based interventions. The underlying causal hypothesis of reablement is that an intensive, multidisciplinary, time-limited (generally under 12 weeks) service can improve an older adult’s independence in daily activities so that no further services are needed afterwards 10-12. Such an approach has some good face value for specific groups (post-discharge patients for example). However, the literature analyzed often treats reablement as a substitute for indefinite-duration home care focused on supporting people whose functional independence is likely on a downward trajectory. For example, articles focused on assessing the effectiveness of reablement interventions tend to compare the outcomes of such interventions with “regular” long-term home care services10-17.
The second split has to do with the objectives of home care programs or interventions we reviewed 6. The first type of objectives are individual in focus and focus on the autonomy, satisfaction, and health of the older adults or their family/friend caregivers. Such objectives are frequently assessed using standardized instruments focused on health status (i.e. SF-12 and SF-36), mental health (i.e. 11-item CES-D Depression Scale, SPMSQ, Kessler-10 anxiety and depression scale), quality of life (i.e. EUROQOL, WHOQOL-Bref), functional capacities (i.e. OARS-IADL, Groningen Activity Restriction Scale-3, Barthel Index of Activities of Daily Living), caregivers’ burden and perceived unmet needs, or idiosyncratic indicators linked with specific interventions. The second type of objectives are societal in focus and focus on the system-level efficiency, appropriateness of resource allocation and the coherence between social values and available services. Indicators of system level objectives focus more on endpoint events (such as death or admission to hospital or nursing home), service use intensity measured through administrative databases (i.e. emergency room visits, number of hospitalizations, number of days hospitalized, number of visits, continued use of home care after a given time or number of visits), cost or cost benefits from those interventions as well as broader policy-level discussion of the alignment between practices and values.

Assessing Home Care Effectiveness

Objectives and indicators

Among the 113 reviewed papers, 45 were focused on measuring or synthesizing the effectiveness of one or multiple home care interventions. Our first challenge when trying to synthesize the evidence on home care intervention effectiveness revolved around the vague operational definition of the concept of home care itself, and the diversity in the objectives being pursued 9. The effectiveness-focused articles we reviewed each relied on given sets of indicators. Notably however, the rationale for those indicators were seldom discussed in relation to the ultimate goal of the intervention itself. For example, focusing on recipients’ quality of life versus focusing on rates of hospitalization or number of hospital visits suggest significantly different underlying objectives.

Shifting Goalposts

A second challenge concerns the research designs employed in the research. Most of the effectiveness-focused studies rely on randomized controlled trials (RCTs) and systematic reviews thereof. However, the nature of home care interventions makes it difficult to appropriately deploy such study designs. The vast majority of the RCTs we reviewed compare a given home care intervention to “usual care” but fail to define that baseline of “usual care.” In some studies, the “usual care” provided consisted of a generous basket of services, while in others it amounted to very little. Yet we only found three studies, all from the United States, comparing the outcomes of providing versus not providing home care services 18-20. Those studies all showed clear benefits from receiving home care services. There are obvious ethical considerations at play that might explain why few studies explicitly compare the outcome of providing versus not providing services to the same population. However, beyond those three studies, when looking at the whole field it is important to contrast, on the one hand, the significant efforts and resources invested in running RCTs measuring the precise level of marginal benefits between two models of home care and, on the other hand, the relative lack of interest in demonstrating the gain of providing home care in itself.

Unclear Causal Processes

Compounding the problem, most evaluations of a given home care intervention provided very little, if any, details on the nature of that intervention 17. The descriptions generally mentioned some elements, such as numbers of home visits, intervention durations, or types of professionals involved, but often failed to provide the level of detail needed to understand the process involved in the production of the outcomes being measured. This observation was also noted in other reviews of the field 21,22.
Because of the limitations discussed above, it is generally impossible to know whether an intervention which outperformed the baseline in a given context might prove beneficial in another. It also makes it impossible to determine the relative effectiveness of interventions with similar objectives. More generally, the policy value of most of the effectiveness studies we reviewed were limited to the time and jurisdiction of when and where the study took place.

If It Were Easy, It Would Succeed More Often

One observation that stands out from the quantitative evaluations of home care interventions we reviewed is this - designing an effective intervention is not easy. Of the 45 evaluations of home care effectiveness, 22 were able to demonstrate the success of a given home care model or intervention while 23 offered unconvincing or negative results. Other reviews in the field reported similar results23.
For example, Bouman, van Rossum, Nelemans, et al.24 systematic review of seven RCTs on preventive home visits found that those programs had no effect. Berger, Escher, Mengle, et al. 25 systematic review of 38 articles on occupational therapy-based home care services found some benefits for participants’ occupational performance, mixed results for quality-of-life indicators and no effect on health care utilization. Our analysis also included one primary study 26 covering similar ground that found no meaningful effect from the restorative occupational therapy intervention studied.
Turning to reablement-focused interventions in particular, one 2013 Cochrane systematic review of this approach 10 failed to identify any high quality evidence showing that such interventions worked. The literature we reviewed included one primary RCT13 that found a reablement intervention to be superior to regular care, yet others 27,28 which found no meaningful benefits. Moreover, the primary studies of reablement services we reviewed (n=10) tended to rely on methodological choices - such as outcome measures focused on short-term physical autonomy gains – that likely overemphasize the benefits of such approaches12.
Luker, Worley, Stanley, et al. 29 conducted a systematic review of 31 RCTs of home care interventions aimed at preventing the institutionalization of older adults, and found no meaningful benefit overall. The authors then conducted a sub-analysis of what they described as ”complex multifactorial interventions” which suggested that this subgroup of interventions do in fact lower institutionalization. However, the definition and boundaries of that subgroup of intereventions is unclear and this approach could be interpreted as data dredging (data dredging describe the aggressive slicing of data and use of sub-group analysis to find statistically significant results). Four of the studies labelled by Luker, Worley, Stanley, et al. 29 as ”complex multifactorial interventions” are part of our own data set. Two of those30,31 found no meaningful effects and two32,33 showed positive effects. Other primary RCTs reviewed here 34 that focused explicitly on multifactorial interventions found no effect as compared to the control group. Aggressive sub-analysis and potential data dredging is also present in some primary studies focused on similar interventions35-37
We also reviewed various nurse-led assessment and coordination of care interventions in different high-risk populations 38-40that failed to measure a meaningful effect. One study combining three different RCTs 33 found positive impacts on mental health in one study and no impact in the two other RCTs. Another study41 of nurse-based home care found some effects on secondary variables (such as the incidence of falls) but no improvement in the primary outcome measure, which was related to quality of life.
The RCTs we reviewed with interventions aimed at improving post discharge coordination between hospitals and home care services or providing additional home care post discharge for frail elderly patients both failed to measure any meaningful effects 30,42.
Low, Fletcher, Gresham, et al. 43 found no convincing link between the need for home care services, the wait time required to obtain said services, or even the amount of services allocated. They also found no link between the services received and outcomes measured, such as quality of life indicators.
This list of negative results provides half the portrait; as we stated earlier, nearly half of the effectiveness-focused literature we reviewed found significantly better outcomes for the intervention under study. What we want to emphasize with this summary of non-effects is that one should not underestimate the challenges related to the design and implementation of an effective and efficient home care intervention.

Insights on What Seems to Work

Three characteristics of home care interventions stand out in the subsample of research studies which found positive results: intersectoral integration, care coordination and relational continuity. Regarding integration, Bernabei, Landi, Gambassi, et al.44 RCT of a multidisciplinary, integrated social and health care intervention, among frail older adults in northern Italy, found it improved functionality and mental health and lowered hospital or nursing home admission and costs per recipient. Likewise, Landi, Gambassi, Pola, et al. 45 quasi-experimental study on the effectiveness of another multidisciplinary, integrated social and health care intervention (also in in northern Italy), found it lowered hospital admissions and costs. Beland and colleagues32,46 led the development and pilot testing of SIPA, an integrated care intervention for vulnerable community-dwelling older adults in Québec (Canada). The SIPA model involves community-based multidisciplinary teams responsible for the full spectrum of care provision, and led to improved accessibility, satisfaction and lower per-recipient costs. Stewart, Georgiou, Westbrook 47systematic summary of the results from the broader PRISMA research program, a different model also from Québec (Canada), identified 45 journal articles and two books authored or co-authored by the team that, overall, support the effectiveness of integrated models of home care on client satisfaction and limiting functional decline.
In the same way, multiple studies on the effectiveness of care coordination in home care showed positive results. Onder, Liperoti, Soldato, et al. 48 large-scale retrospective cohort study of multiple European case-management programs found those to effectively lower nursing home admissions. Scharlach, Graham, Berridge 19 evaluation of an integrated assessment and care coordination model suggests an improvement in clients’ satisfaction and autonomy. Markle‐Reid, Weir, Browne, et al.49 RCT on the benefit of proactively providing older people with home care nursing, health promotion and care coordination in Ontario (Canada), found positive effects on mental health for similar costs. Melis, van Eijken, Teerenstra, et al. 9ran a RCT in the Netherlands demonstrating improved client mental health associated with a nurse-centred home care assessment and management intervention. Morales-Asencio, Gonzalo-Jiménez, Martin-Santos, et al. 50 Spanish, quasi-experimental study showed that their nurse-led case management intervention had a positive impact on functional capacity and caregiver burden. Bass, Judge, Lynn Snow, et al. 51 surmised that in the United States, a Veteran Health Administration’s care-coordination program focused on caregiver needs lowered their level of unmet needs, strain and depression. Parsons, Senior, Kerse, et al. 21’s meta-analysis of RCTs of three different care management models in New Zealand highlighted an association between those models and lower long-term care admissions and the incidence of death. Lastly, Shapiro, Taylor 18’s US-based RCT of a nurse-led assessment and case-management intervention also found improved satisfaction, lowered institutionalization and deaths.
The third characteristic we want to discuss here is relational continuity. Relational continuity is an ongoing therapeutic relationship between a service recipient and a professional 52. In comparison with integration and care coordination, there is far less hard evidence that links continuity with positive outcomes in the documents we reviewed. Three effectiveness-focused studies identified continuity as a characteristic associated with desirable outcomes49,53,54 while two other more analytical articles55,56 discuss its importance in high-performing home care models. However, 34 of the documents we reviewed (30%) mention continuity – most often relational continuity, with informational continuity as a more secondary focus – as a desirable feature of home care interventions and a core process leading to high quality services.
The next section on the main insights and policy-level lessons that can be learned from the 113 documents analyzed here.