Background
In the face of global population ageing and individuals living longer with chronic health conditions, governments, healthcare advocates and older adults themselves are tasked with finding the balance between rising healthcare expenditures and improving older people’s health and wellbeing. Many countries have opted to shift the site of care provision from hospitals and long-term care facilities to home and community settings, with the “expectation of cost savings, or at the very least, that such substitution might lead to more efficient use of nursing home and hospital beds” [
1,
2] (p.108). The World Health Organization’s Global Strategy and Action Plan on Ageing and Health also highlights the importance of delivering home and community-based care to enable older people to 'age in place' with dignity [
3,
4]. As a result, home care is becoming a national priority in many countries.
There is no universal definition of home care. This paper defines home care as “services [that] help people to receive care at home, rather than in a hospital or long-term care facility, and to live as independently as possible in the community.” [
5] The focus of this paper is on formal (paid) home care rather than informal home care (unpaid care provided by family, friends or neighbours). Formal home care encompasses both home health care (HHC) services and home support services (HSS) [
6]. The former refers to nursing or rehabilitative care and other services delivered by licensed health professionals. The latter refers to personal care, housework, meal preparation, and/or respite care delivered by personal support workers or volunteer agencies. The services encompassed under the term home care are often arbitrary. Similarly, depending on context, home care is often characterized under broader services in the healthcare and social care sectors such a long-term care services, community aged-care services, etc. Nevertheless, the goals of all services are to maintain or improve quality of life and functional abilities to promote greater independence and satisfaction while living at home or in the community.
The emphasis on home care reflects an emergent widespread preference; older adults who require assistance prefer to 'age in place' in familiar settings rather than receive care for dementia, other chronic health conditions or ageing needs in an institutionalized setting [
7‐
10]. For the individual, home care has been associated with decreased mortality, reduced hospitalisations, delayed institutionalisation and improved quality of life [
11‐
13]. At a health system level, in comparison to long-term care facilities or hospitalisations, home care in the community has been associated with significantly lower overall healthcare costs [
14]. The appeal of home care is that it can produce health outcomes comparable to those achieved in institutionalised settings, respond to the call for personalised care for individuals to live in the comfort of their own home and improve cost-efficiency [
1].
For home care to serve the above purposes, it is important to establish the characteristics of the individuals who use home care services. This provides “the fundamental information necessary to develop the most cost-effective services for each patient group… This information will also allow healthcare providers and policymakers to prepare and provide services” for patients [
15](p.512). As the population ages, governments and policymakers need to know the current drivers of service use, which will allow them to predict how to best allocate resources and manage the expected increase in demand on the home care system [
16]. Significant modifications and reforms to the home care system become more feasible when we understand what drives utilization.
Unfortunately, the predictors of home care utilization are unevenly understood; while the health determinants impacting home care use are well established by previous reviews [
17,
18], there is less literature looking at how social factors collectively impact use of home care services. Health factors such as cognitive impairment, mobility issues and number of chronic conditions (collectively known as frailty) frequently increase use of home care services [
15,
17,
19]; this trend is in keeping with general patterns that healthcare services, not specifically home care, are positively correlated with frailty [
20,
21]. But the nature of home care service provision relies on resources in the community, beyond individual health status. When two individuals have the same degree of health problems, the difference between being able to remain independent at home with assistance and having to find alternate living accommodations can be explained by social circumstances [
22].
Social factors (or social determinants, terms are used interchangeably in this paper) are defined as the conditions in which people live, work and age, influencing their health and care needs [
23]. Though potentially relevant for people of all ages, the protective or deleterious accumulation of social factors is particularly relevant in older age [
22,
24]. Older adults using home care services are increasingly reliant on social supports (i.e. caregivers and community resources), but at the same time, have dwindling social networks (due to death of friends and family, or greater difficulty participating in social activities due to health and functional impairments) [
22]. And while there have been studies investigating social determinants that affect home care use, most fail to look at the whole picture by focusing on a single or a limited number of determinants [
22]. Therefore, drawing upon a socio-ecological framework, the overarching purpose of this paper is to provide a broad and comprehensive overview of all the social factors influencing home care utilization, which is a current gap in the literature; to our knowledge, a review bringing together the whole range of social determinants has not previously been conducted.
Methods
Objectives
1)
To provide (by identifying and describing) an overview of reported social factors influencing home care utilization, exploring the breadth (comprehensiveness), rather than depth (details) of available evidence to answer the research question.
2)
To evaluate the influence of each discrete social factor on patterns of home care utilization.
Search strategy
A scoping review was conducted using the Arksey and O’Malley (2005) framework refined by Levac, Colquohoun and O’Brien (2010) [
25,
26]. A search of primary studies was conducted in MEDLINE, EMBASE, SCOPUS, Social Science Citation Index (SSCI), the National Health Service Economic Evaluation Database (NHS EED) and Cochrane Library. The last search was conducted July 12, 2020. The search was developed in MEDLINE after consultation with a health sciences librarian, and then translated to the subsequent databases (full search available in Additional File
1). To ensure the findings of the review were relevant to social circumstances in high-income countries, a validated low-and-middle-income country (LMIC) search filter was applied to exclude studies conducted in these countries as defined by the World Bank [
27]. Additionally, documents from the websites of two international bodies (World Health Organisation’s Ageing and Life Course section and the International Home Care Nurses Organisation) and reference lists of included studies and previous literature reviews were hand-searched for records [
28,
29].
Citations were imported into Covidence (2020), a systematic review web platform that removes duplicates and facilitates screening of titles and abstracts, full text retrieval and eligibility assessment [
30]. The first author was responsible for screening of titles and abstracts using Covidence. She also retrieved and screened the full-text articles using the same platform. All authors contributed to establishing the scoping review protocol at the beginning of the study and were consulted throughout the process of data screening and retrieval to ensure adherence to the planned review protocol. The scoping review protocol was not registered but this paper does follow PRISMA-ScR guidelines (see Additional File
2).
Eligibility criteria
Studies were included when:
-
Study participants were community-dwelling older adults greater or equal to 60 years of age living in high-income countries (population). Studies were included if > 50% of the sample met this criteria.
-
Studies examined a social factor (intervention or exposure) in relation to home care services. Social factors included in this review had to meet the definition of at least one social determinant of health (income and social status, employment and working conditions, education and literacy, childhood experience, physical environments, social supports and coping skills, healthy behaviours, access to health services, biology and genetic endowment, gender, culture, race/racism) [
23,
31].
-
Study findings included the use of formal home care services (outcome). Utilization (or use) refers to whether clients have actually received home care services. Therefore, need and access are not outcomes included in this review. Informed by previous literature, outcomes domains of home care service utilization can be divided into:
-
1) Intensity of home care use - defined as the amount of services (hours of service, number of services, costs of home care, etc.); and
-
2) Propensity home care use – defined as a dichotomized outcome of having received or not received any home care [
1,
32].
-
Studies were quantitative (type of study); quantitative studies were felt by the authors to better answer the second objective of the scoping review (to evaluate the discrete influence of each social factor on propensity and intensity of home care use), although we acknowledge that qualitative studies have the potential to add a richness to the understanding of the issue and thus consideration of results from qualitative literature represents an important area for further exploration.
Non-English studies and studies prior to 2010 were excluded. The authors feel that the approach to caring for ageing populations in their own home is best reflected in more recent literature. By limiting the search to the last ten years, this scoping review enables assessment of relevant practices and challenges in home care, which may have more policy relevance in the current political and healthcare climate. Studies looking primarily at long-term care homes or hospital based care programs (i.e. day hospital programs, which can be considered part of home care in some countries) were excluded. Studies focusing on populations of older adults with dementia or diagnosed palliative conditions were also excluded; the former due to a recent scoping review in patients with dementia and the latter because a palliative population often has access to additional care services [
18].
The socio-ecological model
The complex, interconnected relationships between individuals, their social circumstances and their environment present a challenge for researchers who study social determinants of health [
33]. In particular, the patterns of how social circumstances affect health behaviours are not easily explained by considering each social factor in isolation. To illustrate, educational attainment is a property of individuals; yet, ability to go to school is associated with household income and both are measures of neighbourhood deprivation and unemployment. The socio-ecological model provides a way of conceptualizing and disentangling this complex interplay while still acknowledging the contribution of all factors. Based on ecological systems theory by Bronfenbrenner, the model assumes that health behaviours are shaped by the individual’s relationships with caregivers (micro and meso systems) the community (exosystem) and broader society (macrosystem) as opposed to health behaviours solely being the product of individual characteristics and choices [
34].
In applying the socio-ecological model to this review, it is therefore assumed that home care use (the health behaviour) is influenced by factors at each level of influence to various degrees. The social factors identified by this review will be organised into four levels of influence: individual (microsystem), relationship (meso-system), community (exosystem) and societal (macrosystem). The individual level refers to the intrapersonal characteristics of the individual who requires home care. The relationship level examines the interpersonal relationship between the individual and their caregivers. The community level explores the assets or deficits in the community that impact home care. The societal level refers to the public policy and cultural norms that shape the availability of home care services. These factors impacting home care utilization interact across all levels to impact behaviour change. The social-ecological model is best understood as a dynamic model for the purposes of this review, allowing for factors to be added or removed from categories of influence (e.g. household income can be a relationship factor for married adults, but an individual factor for single older adults).
A data extraction form was created to collect information from each paper. This included: (1) General information (country and type of home care system, ethics, funding sources, conflicts of interest); (2) Study methods (aim, design, start/end dates, data source, model/theory); (3) Participants (inclusion/exclusion criteria, method of recruitment, characteristics); (4) Social factors and exposures; (5) Outcome data (relevant to social factors); (6) Key author conclusions relevant to the review question.
The high heterogeneity of study populations, designs, data sources and statistical adjustment techniques (in addition to the variable ways of measuring social factors and home care outcomes and inconsistent types of effect measures) precluded a statistical combination of results through meta-analytic techniques. Instead, in line with the scoping review framework, the following process of collating, summarizing and reporting occurred to produce a narrative rather than statistical summary.
First, all studies were investigated using a qualitative content analysis approach including reduction, explication and restructuring [
18]. After the content from each study was reduced to only the social determinants influencing home care, the material was organised in relation to the socio-ecological model and grouped according to outcome (intensity or propensity). Since multiple studies provided results for more than one social factor, studies may appear in multiple areas of the framework. To complement the narrative synthesis, vote counting per social factor occurred based on direction of effect and statistical significance. Vote counting synthesizes information to answer the question: is there any evidence of effect? This question is particularly relevant for the objectives of this scoping review [
35]. Where appropriate (for at least 2 studies per social factor), harvest plots were created to visualize vote counting per social factor and outcome domains, accounting for study type (colour) and study measures (labelled on x-axis) [
35‐
37]. Harvest plots group studies based on direction of effect (i.e. positive association, null association, negative association) and each study is represented by a bar positioned according to its categorization [
35].
Second, an evaluation of aggregated findings per social factor was conducted to determine consistency across findings. This evaluation process has been endorsed by previous narrative reviews on healthcare utilization [
17,
21,
38].
-
Step 1: If the majority (≥60%) of the studies indicated that a social factor was associated (either positively or negatively) with home care utilization, then that social factor was evaluated as likely “associated” across all studies. If 40–59% of the studies showed an association (either positively or negatively), then the outcome of the evaluation for this social factor was deemed “uncertain”. If < 40% of the studies supported any association (either positively or negatively), the consensus is that the social factor is likely “not associated” with home care utilization across all studies.
-
Step 2: For social factors that were deemed “associated” in Step 1, the same evaluation took place using all significant studies (positively and negatively associated) as the new denominator. The same cut offs above were used to determine the direction/pattern of the association: “positive/more”, “negative/less”, or “equivocal”.
-
The evaluation was only performed for social factors with at least 2 studies. If studies reported both univariate and multivariate analyses, only the multivariate results were included.
All simple statistics (i.e. frequencies, proportions) were conducted with STATA-IC 16.1. All figures were created using Microsoft Excel or Lucidchart [
39].
Discussion
Main Findings & How They Compare with Other Reviews
By providing an overview of social factors influencing home care utilization, this paper has brought together a wider range of social determinants in relation to home care utilization than those included in previous literature reviews. As summarized in Tables
2, 35 discrete social determinants were identified. This review also distinguished the level of influence at which these social factors were studied. Most studied social factors were properties of individuals or related to family units or peer groups. Social factors influencing home care were also identified at community levels (e.g. neighbourhood racial composition) and at societal levels (e.g. comparing national entitlement to home care systems), but these were less common.
There was a high degree of heterogeneity in the studies reviewed. Despite the differences with respect to patients sampled, countries of study, methods and analyses used, there were some findings that were similar across all studies. The factors consistently showing any association (positive, negative or equivocal in pattern) with home care propensity were: age, ethnicity/race, self-assessed health, insurance, housing ownership, housing problems, marital status, household income, children, informal caregiving, social networks and urban/rural area. Across all studies, social factors consistently showing any association (positive, negative or equivocal in pattern) with home care intensity were: age, personal finances, housing ownership and living arrangements.
For the social factors in commonality with previous reviews, these findings also support the findings of Kadushin (2004) and Johnson and colleagues (2018). The former concluded that living alone, a low level of informal support and Medicaid coverage have a relationship with higher use of home care services; it also noted most studies were from the USA [
17]. The latter, a Canadian review, found that age, gender and location of residence influenced home care utilization [
19]. Together, these reviews do not invalidate the message from other literature that links health status to the utilization of home care. Rather, independent of health-related needs, the reviews add to collective evidence supporting that social determinants influence home care utilization. In comparison to a review of formal community care in persons with dementia, not all factors were the same [
18]. Ethnicity, living situation, region of residence and gender were compatible with this review but viewing formal care as a threat to independence and effect of previous experiences with home care were not. A number of explanations may explain these differences. By including qualitative studies, Bieber and colleagues (2019) were able to capture studies investigating attitudes towards home care and experiences with services (which were outside the scope of this review). Due to their population of interest (persons with dementia), the review by Bieber and colleagues (2019) strongly reflects the voice of the caregivers rather than the care recipients and it is well established that these two populations can have differing needs [
18].
Limitations
The findings of this review must be interpreted with caution. A number of limitations arose, especially the findings related to objective two (evaluating the influence of each discrete social factors on patterns of home care utilization). Generalising across studies is problematic given that 23 different countries and 64 data sources were included in the findings, each with different home care systems, cultural norms, study populations and designs, ways of measuring social factors, etc. For example, although one social factor (i.e. insurance) may be influential in one setting (i.e. USA), that same social factor may not apply in another country with different context (e.g. Netherlands which provides universal home care). Seeing the same social determinants repeated across multiple studies gives some confidence that there is a persistent association between these factors and home care and raises additional research questions regarding why these factors are consistent across many contexts; but each social factor is still best understood within its proper context. A key difference between many countries is the extent to which home care is organized. For example, nursing led versus social work led. In the United Kingdom (UK), it is social work led and nursing care is an entirely different service [
49]. In Austria (part of SHARE), it is nursing led with a multi-disciplinary team involved in home care provision [
82]. This would change the eligibility criteria for services, particularly as health systems tend to be universal in Europe, but social care services are often linked to social assistance programmes that are means-tested. In some countries like Germany there is also a further pillar – the long-term care system. As the social factors found by this review move from the individual to the societal levels in the model, the findings are increasingly variable. Even individual or relationship level factors that appear consistently associated across studies (i.e. house ownership or living arrangements) can be highly dependent on those variable community/societal factors (i.e. comprehensiveness of national social security programs). How these factors come together within a country to influence the direction of association is likely to be fairly unique.
Additionally, the method of evaluating outcomes across studies (see data synthesis section) was chosen as a well-documented way of comparing consistency across heterogeneous studies; this method has been used by other reviews on healthcare utilization [
17,
21,
38]. However, this method produces results using a chain of procedural decisions. Evaluations using different tools or cut-offs may well lead to different findings. For social factors considered by small numbers of studies, this evaluation method may lead to erroneous conclusions.
Only studies in English were included in the inclusion criteria, again for feasibility reasons. Although this review found substantial numbers of included studies from Europe and Asia, additional literature written in non-English languages is likely missing. While this review followed rigorously defined scoping review methodology, a single reviewer conducted the screening and data extraction, which may potentially have introduced personal bias in these steps. Ideally, a double screening and extraction approach would have been carried out to increase internal validity and reliability, and reduce the chance of random error or personal bias. Since there was significant consultation between the authors to ensure adherence to the planned review protocol and to establish consensus for any difficult screening questions, it is unlikely that a double screening approach would change the overall findings of this paper.
Another important limitation of this review is the lack of quality appraisal of individual studies and grading of evidence for each outcome domain. Even though quality appraisal of studies in scoping reviews is controversial, its exclusion is often seen as a limitation to applying scoping reviews to policy and practice [
26]. Only a quarter of scoping reviews published in 2014 included a quality appraisal step [
18,
26]; hence, quality appraisal is not part of the traditional scoping methodology [
25]. The lack of assessment of bias and lack of review of methodological quality may explain some of the inconsistencies in the results section. An assessment of methodological quality could be seen as especially important because most of the study designs included in this review were observational studies or secondary analyses; there was a noticeable absence of the traditional higher levels of evidence (i.e. RCTs).
In addition, the scope of this research question looked at home care use, rather than need or access. Access to, or need for, home care is a much more difficult concept to define and measure and would merit further research in this area. Ideally, use of home care correlates with need for care, but it is likely that this review fails to capture populations of older adults who lack the means to access home care altogether; hence a more vulnerable older adult population who could be adversely affected by social determinants of health.
Policy implications
Overall, from this review, the take home messages for policymakers are clear: 1. Social factors influence home care use and should be accounted for in policy decisions; 2. These social factors should be understood as interdependent and not isolated from one another; 3. Enhancing home care infrastructure for older adults is inadequate so long as social circumstances make it difficult for individuals to remain at home. Rather, a multi-level approach can help create social circumstances and policies that make it safe, attractive and economical for individuals to use homecare; ideally helping to promote individuals living in the comfort and familiarity of their own home, shifting site of care away from institutions while improving cost-efficiency of the healthcare system. It may seem attractive to focus on individual or caregiver factors, but intervening at societal level (i.e. reducing co-payments or changing eligibility criteria) has potential to make population level shifts towards using more home care services; and 4. For policymakers who are interested in examining home care from a holistic perspective, there is a substantial body of literature that exists looking at factors influencing home care from all levels of influence in the socio-ecological model.
Future research
This review addresses questions about
what evidence on social factors and home care utilization is available and
which social factors appeared consistently in the literature (breadth of the topic). But this paper does not answer
why and
how these social factors influence home care patterns. Understanding mechanisms is required for future research in this area. Understanding the complex, causal pathways by which social circumstances impact home care utilization requires recognition that some social factors are more relevant to specific contexts than others. For researchers who wish to look into the social factors influencing home care utilization in their communities/countries, Additional File
3 has compiled examples of data sources from each included country that would be conducive for this purpose. The results of this scoping review can also provide the basis for conducting a more focused systematic review (e.g. to establish the quality of evidence or answer questions regarding the effectiveness of social interventions). Finally, whether these social factors are impacting home care need or access (rather than home care use) is an area where future research can build on the limitations of this review.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.