Background
Transitions to the community from the hospital are often difficult for older adults. Because older adults are typically sicker than younger adults, they use more hospital days than the general population to accommodate extended recovery periods, or arrangement of home care or other long-term care services (Glasby et al., Reference Glasby, Littlechild, Le Mesurier, Thwaites, Oliver, Jones and Wilkinson2016). As a result of the potential for deterioration of functional status while in hospital, as well as a myriad of other factors such as multi-morbidity and limited support systems, older adults are vulnerable to post-hospital adverse events and hospital readmissions (Laugaland, Aase, & Barach, Reference Laugaland, Aase and Barach2012). In addition, older adults have reported difficulty (1) accessing health services following hospitalization (Jackson, Oelke, Besner, & Harrison, Reference Jackson, Oelke, Besner and Harrison2012; Lapum, Angus, Peter, & Watt-Watson, Reference Lapum, Angus, Peter and Watt-Watson2011), (2) being unsure of how to take care of themselves once at home (Knight, Thompson, Mathie, & Dickinson, Reference Knight, Thompson, Mathie and Dickinson2011), and (3) feeling rushed to make important decisions about their discharge destination (Lapum et al., Reference Lapum, Angus, Peter and Watt-Watson2011). Better ways to manage hospital-to-community transitions are needed, and integrated care approaches are a promising solution for enhancing hospital and community services coordination and collaboration.
Integrated Care
Integrated care is a health care approach recognized for providing high-quality care for older adults and/or people with chronic disease while maximizing health resource use (World Health Organization [WHO], 2015; Chappell & Hollander, Reference Chappell and Hollander2013). Integrated care approaches focus on enhancing coordination and collaboration within and between care sectors to reduce fragmentation (Kodner & Spreeuwenberg, Reference Kodner and Spreeuwenberg2002; Valentijn, Schepman, Opheij, & Bruijnzeels, Reference Valentijn, Schepman, Opheij and Bruijnzeels2013). Integrated care is believed to be most effective for populations with complex care needs who require care from multiple providers and in several settings. Thus, older adults may experience improvements in the quality and efficiency of care transitions with the application of integrated care (Maruthappu, Hasan, & Zeltner, Reference Maruthappu, Hasan and Zeltner2015).
Integrated care approaches may integrate care services along the vertical and/or horizontal continuum of care. Vertical integration refers to bringing together different levels of health care, such as primary and secondary care. Primary care includes basic health care services provided by professionals who are typically the “first contact” with patients (such as general practitioners [GPs]). Secondary care is specialist care, including basic hospital care (Gröne & Garcia-Barbero, Reference Gröne and Garcia-Barbero2001). Horizontal integration is the coordination and collaboration of services within one level of care and, in the context of this study, usually refers to integration of community-based social and health care services (Chappell & Hollander, Reference Chappell and Hollander2013; Gröne & Garcia-Barbero, Reference Gröne and Garcia-Barbero2001). Although the term social services is not clearly defined in the literature, it typically refers to all non-medical services, such as assistance with financial and housing needs, and provision of home support needs such as home maintenance, meal programs, daycare services, and/or transportation (Hollander & Prince, Reference Hollander and Prince2007).
The term integrated care describes a wide variety of strategies or approaches to improve care continuity and coordination; consequently, the implementation of integrated care is varied (Maruthappu et al., Reference Maruthappu, Hasan and Zeltner2015; Valentjin et al., Reference Valentijn, Schepman, Opheij and Bruijnzeels2013). This variation has made it challenging to compare integrated care initiatives (Valentjin et al., Reference Valentijn, Schepman, Opheij and Bruijnzeels2013). However, integrated care approaches that include vertical multidisciplinary interventions are efficacious for managing chronic diseases, managing care between primary and specialist care, and reducing hospital use (Martínez-González, Berchtold, Ullman, Busato, & Egger, Reference Martínez-González, Berchtold, Ullman, Busato and Egger2014; Mitchell et al., Reference Mitchell, Burridge, Zhang, Donald, Scott, Dart and Jackson2015). Less attention has been paid to integrated care practices in relation to care transitions between the hospital and community, despite the emphasis of integrated care as an approach that can reduce boundaries between levels of health care (Vedel et al., Reference Vedel, Monette, Beland, Monette and Bergman2011).
Care Transitions for Older Adults
The challenges with transitions from hospital to community care are well documented. Post-hospital adverse events, emergency room visits, and readmission rates are high for older adults (Canadian Institute of Health Information, 2012; Forster et al., Reference Forster, Clark, Menard, Dupuis, Chernish, Chandok and van Walraven2004) due to multiple factors such as poor communication between the hospital and primary care services, lack of patient understanding of their diagnosis and self-care needs, and difficulties with accessing community care services (Jackson et al., Reference Jackson, Oelke, Besner and Harrison2012; Rennke et al., Reference Rennke, Nguyen, Shoeb, Magan, Wachter and Ranji2013).
The body of literature evaluating intervention programs designed to improve the quality and efficiency of care transitions from the hospital to the community is vast. It has been demonstrated that personalized discharge planning can bring small reductions in hospital length of stay and readmission rates for older medical patients (Gonçalves-Bradley, Lannin, Clemson, Cameron, & Shepperd, Reference Gonçalves-Bradley, Lannin, Clemson, Cameron and Shepperd2016). Studies on hospital-initiated interventions have identified two important features for reducing hospital readmissions: (1) the use of a dedicated, hospital-employed discharge transition provider or team, and (2) the use of a bridging strategy, whereby services are provided by the hospital beyond the discharge date (Rennke et al., Reference Rennke, Nguyen, Shoeb, Magan, Wachter and Ranji2013). A review by Leppin et al. (Reference Leppin, Gionfriddo, Kessler, Pablo Brito, Mair, Gallacher, Wang and Montori2014) of 47 trials using 18 types of discharge interventions found that more recent intervention studies have been less successful at improving health service use, even when implementing interventions that were successful in the past (Leppin et al., Reference Leppin, Gionfriddo, Kessler, Pablo Brito, Mair, Gallacher, Wang and Montori2014). Moreover, interventions that provide frequent and complex assistance to older adults in their own homes are most successful in reducing hospital readmissions (Leppin et al., Reference Leppin, Gionfriddo, Kessler, Pablo Brito, Mair, Gallacher, Wang and Montori2014). These findings signal the need for a new approach to care transitions that is consistent with integrated care principles such as reducing barriers within and between institutions, and meeting both social and health needs in the community.
The concept of integrated care seems promising for reducing fragmentation between the hospital and the community in order to improve care transitions, but has not yet been a focus of a systematic literature review. Because integrated care is not consistently defined or applied, it is important to understand how integrated care is currently being conceptualized in relation to hospital-to-community transitions. Scoping reviews are helpful for presenting a broad overview of the evidence to “map the concepts underpinning a research area” (Constand, MacDermid, Dal Bello-Haas, & Law, Reference Constand, MacDermid, Dal Bello-Haas and Law2014, p. 2). Scoping reviews also identify gaps in research, including determining the feasibility of a systematic review on the topic (Tricco et al., Reference Tricco, Lillie, Zarin, O’Brien, Colquhoun, Kastner and Straus2016). Our scoping review is unique and distinct from previous reviews on hospital-to-community transitions because of our explicit focus on integrated care approaches. Our overall aim was to determine research gaps and directions for future research on the topic of integrated care approaches for facilitating transitions from hospital to home for older adults. We did this by (a) outlining the size and scope of this body of literature, and (b) identifying how integrated care is being conceptualized in this body of literature.
Conceptual Framework for This Study
Two integrated care frameworks influenced our thinking while we conducted this study. The first is the enhanced continuing care framework (ECCF) (Hollander & Prince, Reference Hollander and Prince2007) because of its particular relevance to the topic of care transitions from hospitals to the community for older adults. The framework conceptualizes an ideal system for older adults with continuing care needs and specifically identifies linkages between the hospital and continuing care as important for meeting the needs of older adults. It emphasizes the importance of social as well as health care, promoting horizontal as well as vertical integration. The ECCF outlines the philosophical and policy prerequisites (e.g., patient-centred care; psychosocial model of care) that provide a base for the development and application of best practices of continuing care (e.g., coordinated administration, integrated information systems). Best practices then support the development of linkage mechanisms (e.g., staff that cross care boundaries, such as physician consultants in the community) between different layers of the health system and other sectors providing care services.
The second influence was a framework developed by Kodner and Spreeuwenberg (Reference Kodner and Spreeuwenberg2002) that adds to thinking about integrated care for adults transitioning from hospital in two ways. First, it includes integration of single institutions, such as integration of different departments and professionals within a hospital. Second, rather than providing a specific framework for an integrated system, it takes a more flexible approach by proposing that one or more integration strategies can be implemented in one or more domains depending on the needs of the care environment. Kodner and Spreeuwenberg (Reference Kodner and Spreeuwenberg2002) provided specific examples of strategies that can be implemented in five domains of integration that range from a macro level to a micro level: financial, administrative, organizational, service delivery, and clinical. For example, strategies for the financial domain include prepaid capitation and pooling of funds; strategies under clinical integration include decision support tools and regular patient contact.
Design and Methods
This study was guided by a six-stage methodology proposed by Levac, Colquhoun, and O’Brien (Reference Levac, Colquhoun and O’Brien2010), who refined methodology developed by Arksey and O’Malley (Reference Arksey and O’Malley2005). The first five stages outline considerations for (1) identification of the research question, (2) identification and (3) selection of relevant studies, (4) charting of data, and (5) summarizing and reporting results. As it is optional, we did not implement the sixth stage, which recommends consultation of stakeholders to gain additional perspectives on preliminary findings. We also followed recommendations as outlined in The Joanna Briggs Institute Reviewers’ Manual: Methodology for JBI Scoping Reviews (Joanna Briggs Institute; 2015) that provides detailed recommendations on protocol development, data abstraction procedures, and required components in the reporting of results.
Eligibility Criteria
Consistent with scoping review methodology, this study was broad in its inclusion of different types of literature and did not evaluate the quality of the studies (Levac et al., Reference Levac, Colquhoun and O’Brien2010). Both peer-reviewed and grey literature was searched with no methodological requirement for study inclusion. Protocols were included since they provided information that was not otherwise captured, either because the studies were not yet published, or in the case of one study, the protocol had been altered during study implementation. We excluded conference abstracts and Microsoft PowerPoint presentations as they did not provide enough detail to determine document eligibility and/or extract data accurately. Eligibility criteria for content were developed according to the JBI guidelines (2015) that suggest the use of the mnemonic PCC (population, concept, and context) to target the desired focus and scope for the review:
1. Population: The document needed to focus on hospitalized older adults (defined as age 65 or older). Documents presenting research on participants with a mean age of 65 were included, even if some of the participants were under age 65.
2. Concept: The document needed to explicitly explain, evaluate, describe, or propose an integrated care initiative. As an objective of this review was to understand how the concept of integrated care is being used in research on care transitions, we included all types of integrated care.
2. Context: The document needed to focus on transitions of care from an acute care environment to other care environments (including, but not limited to, intermediate care, nursing home, home with or without home care services). Interventions applied in the community to prevent hospitalization readmission immediately following hospitalization were included, but interventions aimed at preventing hospitalization for community-dwelling older adults were excluded.
Information Sources and Search Strategy
The search was comprehensive and designed to locate both published and unpublished literature within the field of health and health services. The search strategy was developed with the advice of an experienced academic health librarian. We adapted the strategy following a preliminary search of some potential keywords to ensure search specificity. For example, a preliminary search using the term “care transition” did not provide adequate breadth of results, and thus was replaced by the terms “discharge OR hospital”. The search strategy included the terms (or related terms) “integrated care”; gerontology (geriatrics, older adults, elderly); “discharge OR hospital” (with Boolean operators AND and * where relevant). We tailored the search plan to suit each database or search source. Details of the search strategies are available from the first author.
We searched the following electronic databases from January 1, 2005 to December 31, 2016 for documents written in French or English: Scopus (includes PubMed and EMBase), Proquest and EBSCOhost Full Text (includes 3,000 periodicals including AgeLine and CINAHL), and the University of Manitoba library one-stop search feature (includes all physical items in the University of Manitoba libraries, subscribed e-books, documents from over 200 databases, course reserve materials, and subject guides). The one-stop search feature was included to ensure breadth of the search, since it searches all library database materials, not just health and social databases. The Journal of Integrated Care and the International Journal of Integrated Care were searched separately due to the high potential of relevant documents. For further grey literature searching, we used both Google Advanced Search and the Canadian Health Research Collection. Once we had chosen documents for inclusion in the study, we reviewed their reference lists to search for additional studies.
Study Selection
The selection of documents was completed in three main stages (see Figure 1). In stage 1, the first author screened the document titles to eliminate French versions of documents for which we had copies in English and to eliminate documents that clearly did not meet the inclusion criteria (e.g., pediatric population). In stage 2, we independently determined the eligibility of documents based on the title and abstract, and then met to discuss and reconcile any differences. For grey literature with no abstract, we accessed the executive summaries or table of contents of the documents. In stage 3, we independently reviewed the full text of documents that were potentially eligible following stage 2, and then again met to discuss and reconcile any differences regarding eligibility, which left 48 documents.
Figure 1: Document identification, screening, and selection flow
Data Extraction
For documents eligible for inclusion, data was abstracted by the first author and reviewed by the second author. First, basic information (e.g., publication date, document objective, document type, etc.) was abstracted from each included document. To accomplish this, the first author developed an extraction file based on a list of key information recommended for extraction by the Joanna Briggs Institute (2015), and both authors met to revise the file after it was piloted on five documents. Next, we looked at the integrated care initiatives described in the documents. Several of the grey literature reports described multiple integrated care initiatives, but not all the initiatives met the inclusion criteria for this review. For example, documents may have described initiatives that were developed for different populations or that were not specific to care transitions. Thus, we reviewed all the initiatives reported within the included grey literature using our PCC inclusion criteria. The first author then extracted information on each initiative meeting the inclusion criteria so that the diversity of initiatives as well as literature types could be conveyed in the results.
In the final phase of data extraction, we used a qualitative content analysis approach to further address the study objective of determining how integrated care is being conceptualized in the literature from the perspectives of the documents’ authors (Vaismoradi, Turunen, & Bondas, Reference Vaismoradi, Turunen and Bondas2013). Content analysis allows for reporting of common trends and patterns in large amounts of text with a low level of interpretation and can allow for the quantification of qualitative information (Vaismoradi et al., Reference Vaismoradi, Turunen and Bondas2013). Information about the features of each integrated care initiative was extracted, grouping text that was similar in meaning. We used an inductive approach, rather than attempting to fit the data into current frameworks of integrated care. However, knowledge of the Hollander and Prince (Reference Hollander and Prince2007) and Kodner and Spreeuwenberg (Reference Kodner and Spreeuwenberg2002) frameworks ensured that we considered multiple domains of integrated care (financial, administrative, clinical, etc.) and were referenced to assist with developing category labels.
When extracting data, we included all features of the initiative that would be (or were) in place during the implementation of the integrated care initiative. In some cases, this included features of the system or initiative that were already in place and were being kept in place during implementation of the integrated care initiative. As new information was extracted from different documents, similar information was grouped together. Following extraction of approximately 10 initiatives at a time, we reviewed the data as a whole, categorized it, and then applied labels to grouped data. This iterative process continued, altering labels and categories as new data were incorporated into the emerging framework. Once data had been extracted from all the documents, we continued to collapse similar categories until we were left with three main groups of data containing categories and subcategories that summarize the features of integrated care in this body of literature. Finally, we counted how many articles had information on each of the category features.
Results
Types of Documents
The number of peer-reviewed articles totaled 25 (19 with empirical results, four protocols, an editorial, and a program description), 22 grey literature documents (19 reports, 2 dissertations, and 1 clinical care guideline), and one non-peer-reviewed journal article. These 48 documents taken together described 45 different integrated care initiatives to support older adults transitioning from the hospital to the community. The initiatives were in various stages of implementation; 32 (71.1%) had been implemented, 3 (6.7%) were partially implemented, and 9 (20.0%) were proposals or visioning documents (see tables 1 and 2 for details of the included documents).
Table 1: Peer-reviewed research reporting results
a All population groups were age 65 and older, or the population mean was 65 years of age or older.
b Only the results related specifically to acute to non-acute transitions are reported here.
ED = Emergency Department; GP = general practitioner; IC = integrated care; LoS = length of stay; QoL = quality of life
Table 2: Characteristics of other documents included
ADL = activities of daily living; GP = general practitioner; IC = integrated care; QoL = quality of life; RCT = randomized controlled trial
The peer-reviewed research (n = 19) was heterogeneous in terms of research design. There were six randomized controlled trials (RCTs), five qualitative studies, five prospective cohort studies, two mixed method studies, and one quasi-experimental study. Of the four protocols, two were from the same study, with one protocol describing the intervention and the other the research design (quasi-experimental with nested RCT). The remaining two protocols described an RCT and a prospective matched control before and after study.
Of the grey literature (n = 22), 19 were reports describing planned or implemented integrated care initiatives. Only one of the reports had an empirical approach and data reporting (mixed methods); the rest of the reports were descriptive in nature. The reports were most commonly authored by health authorities, health associations, or the health government branch (n = 9). For the two dissertations, one was a process evaluation and the other, a quasi-experimental study (see Figure 2).
Figure 2: Types of documents included in the scoping review of integrated care initiatives
The documents came from 14 different countries on four continents (Europe, North America, Asia, and Australia) with the majority originating from Europe (31; 64.6%). In order of most to least documents produced, documents were from England (14; 29.2%), Canada (7; 14.6%), the Netherlands (6; 12.5%), Norway, (4; 8.3%), Hong Kong (3; 6.3%), and Singapore (3; 6.3%); additionally, two documents each (4.2%) were from Australia, Sweden, Scotland, and the United States, and one document each (2.1%) from Spain, Denmark, and Ireland. The initiatives described in the documents were generally from the same 14 countries, although some documents reported on initiatives in other countries, and some initiatives were described in more than one document. For example, there were six documents and two initiatives from the Netherlands, and two documents and four initiatives from the United States.
According to the publication year, this topic is becoming more popular, with the number of documents increasing over the past 10 years. Nine of the documents were released between 2005 and 2010 (18.8%), while the remaining 39 were released between 2011 and 2015 (81.3%). The year with the most documents released was 2015 (13; 27.1%).
Outcomes Measured
In the quantitative research from peer-reviewed publications (n = 14; 8 RCTs and 6 non-RCTs), the most commonly measured outcome was health utilization, with the most frequent indicators being hospital readmissions (n = 9 studies), and hospital length of stay (n = 7 studies). Other health utilization measures included emergency department visits and outpatient visits (n = 7 studies). Health and function indicators were used in seven studies; health care costs in five studies; patient satisfaction of care in five studies; quality of life of the patient or informal caregivers in four studies; service provider perspectives of care, and health care costs, in three studies. See Figure 3 for a summary of outcomes reported from quantitative instruments. The qualitative studies involved thematic analysis of interviews, focus groups, or narrative documents, and thus outcomes were not extracted for these studies and are not included in Figure 3. All six of the studies that analysed qualitative data explored perceptions of staff on the implementation of an integrated care initiative, and two of the seven also explored patient perceptions of their care transition experience.
Figure 3: Outcomes measured in quantitative peer-reviewed literature
Patient Populations
The patient populations in the peer-reviewed quantitative research were also variable. Of those reporting specific enrolment criteria (n = 11 of 14 studies), some studies included only older adults with specific health conditions (COPD, n = 2; hip fracture, n = 1; chronic cardiorespiratory disease, n = 1), whereas others included all hospitalized older adults (n = 4). Some studies focused on older adults considered to be vulnerable to hospitalization due to existing disability (n = 2), or older adults at risk of readmission according to screening tools (n = 2).
Conceptualization of Integrated Care
To determine how integrated care was conceptualized in this body of literature, we first tracked the terminology related to the word integrate in all of the documents to explore consistency in the use of this term. In the 48 documents, there were 37 different uses of language or terminology for integrated care or integration. The four most common were (1) integrated health and social care (or services) (11.1% of the tracked terms); (2) integrated care (9.9%); (3) integrated care model (7.4%); and (4) service integration (6.2%).
The next step in conceptualizing integrated care was to track how the documents defined integrated care. Ten of the 48 documents provided explicit definitions of integrated care (20.8%). All of the definitions referred to improving partnerships or coordination of care, with most referring specifically to coordination and partnerships across different health and/or social care settings and/or services. Thirteen (27.1%) of the documents did not provide an explicit definition, but the characterization of integrated care was evident from the description of the goals of care. Seven of these 13 documents characterized integrated care by describing locally developed or adopted models of integrated care, whereas six of these 13 documents were aiming to reduce the absence of integration (such as system fragmentation, a lack of coordination, or siloed care). Finally, more than half (25; 52.1%) of the documents provided no definition or characterization of integrated care.
Another approach we took to understanding how integrated care is being conceptualized was to track the breadth of the initiatives. Of three levels of breadth reported in these documents, the narrowest application was within-hospital integration, whereby the goal was to improve coordination, communication, and care pathways within the hospital setting. This was the least common approach, with only two articles describing this type of integrated care (British Columbia Provincial Seniors Hospital Care Working Group, 2012; Joint Improvement Team, 2015). Vertical integration of health services was the second most common type of integrated care described, with 17 of the 45 initiatives (37.8%) aiming to integrate inpatient hospital services and at least one other level of care. Of these initiatives, the most common approach to integration was to integrate inpatient hospital care and primary care (10 initiatives; 22.2%). Other approaches included linking inpatient hospital services with inpatient rehabilitation initiatives, the Emergency Department, GPs, and/or nursing homes. Initiatives that aimed for both horizontal integration of community health and social initiatives, and vertical health services integration were most common (26 of 45 initiatives; 57.8%). These initiatives had the most breadth of service inclusion – for example, including palliative care, ambulance services, geriatric day centres, and mental health services in addition to hospital care and health and social community care. The broadest vision of integrated care was a proposal by National Health Service (NHS) Forth Valley (2012) for whole-system integration of the health, social, housing, education, and voluntary sectors.
Finally, we categorized the features of the integrated care initiatives to determine commonalities. See Table 3 for feature definitions and Table 4 for results. We identified three overarching categories of features: philosophy and policy, clinical features, and administrative and operational features. The category of philosophy and policy was defined as information that conveys the underlying beliefs or attitudes embodied in the initiative. All the initiatives stated their goals.
Table 3: Definitions of features of integrated care programs categories
Table 4: Features of integrated care initiatives (n = 45 initiatives)
a Proportions for the features in the Administrative and Operational Features category are derived from the total number of programs that provided description of administrative and organizational features (n = 34).
The most common goal was to make the most efficient use of health care resources by minimizing hospital lengths of stay, reducing hospital readmissions, and supporting older adults in the community (66.7% of initiatives). The second, third, and fourth most commonly reported goals had similar frequency rates: enhancing coordination (42.2%), being patient-centred (40.0%), and improving health outcomes (35.6%). Almost a quarter (24.4%) of the initiatives identified stakeholder engagement as an important part of their development and feedback mechanisms, reflecting a dedication to sustainable decision-making. Another quarter of the initiatives (24.4%) identified the importance of the development of a shared inter-agency culture or model of care to promote high-quality care provision. These categories of goals were not mutually exclusive. For example, two initiatives used fictitious older adult exemplars to engage administration and staff across health settings and services in the development of solutions to care barriers (Cunnane, Reference Cunnane2013; Thistlethwaite, Reference Thistlethwaite2011). The pseudopatient was given a name, and depicted with a written case story and image, to help health professionals “look through the eyes of the patient” in developing patient-centered care solutions that transcended care boundaries.
The second overarching category was clinical features of the initiatives, defined as elements of the initiative that are related to direct clinical care. All 45 initiatives provided at least one detail about their clinical components. The descriptions were typically comprehensive and detailed allowing for the development of several subcategories. Consistent with the finding that most of the initiatives were aiming for horizontal, or horizontal and vertical, integration, 71.1 per cent of the initiatives described specific strategies that were in use, or were planned, to ensure care coordination across care settings. The other clinical features for which there was the most consensus were (1) individual care coordination (51.1% of initiatives), (2) multidisciplinary teams (51.1% of initiatives), (3) post-hospitalization care support (48.9% of initiatives), and (4) individualized comprehensive assessment (44.4%).
The final overarching category was administrative and operational features of the initiatives, defined as elements that were not directly related to clinical care. Details on these features were more difficult to extract than the clinical features because the administrative structure was often not described. When possible, we used supplementary information from health region websites to determine the administrative and operational structure of organizations named in the articles. For example, several articles from England identified the use of clinical commissioning groups that organize the delivery of several levels of health care. Three quarters (75.6% or 34) of the initiatives identified at least one administrative or operational feature of the initiative. For the remaining one quarter, the description did not include the operation of the initiative or health system, making it unclear whether or not these initiatives have autonomous administrative structures. One exception was an initiative that was deliberately avoiding macro-level integration (Cunnane, Reference Cunnane2013). Since it was the only article describing this approach, we did not include it in the article count for Table 4.
The most common features for administrative and operational features were administrative or organizational integration (20 of the 34 initiatives; 58.8%), financial integration (55.8%), and specific strategies to promote organizational integration (47.1%). Administrative integration was most typically in the form of system-wide horizontal or vertical mergers. In fewer cases, the administrative structure was developed for a specific initiative. For example, for the implementation of a transitional care bridge model in the United States, an administrator worked across sites to implement the initiative (Altfeld, Pavle, Rosenberg, & Shure, Reference Altfeld, Pavle, Rosenberg and Shure2012). Financial integration was most typically described as some type of fund sharing structure, either vertically or horizontally, although other types of fund sharing were described, such as agencies working together to secure grant funding.
Discussion
The results of this scoping review indicate a heterogeneous literature base for the topic of integrated care approaches to care transitions for older adults. The heterogenous nature was both in regard to the types of documents, as well as how the documents applied the concept of integrated care. The literature ranged from unpublished clinical guidelines to randomized controlled trial results. The integrated care initiatives ranged from efforts to coordinate services within the hospital to comprehensive systems of vertically and horizontally integrated social and health care. That there were numerous grey literature documents indicates that the published peer-reviewed literature provides a narrow view of international developments in integrated care initiatives for care transitions.
One of the objectives of this scoping review was to determine if there is an adequate literature base to conduct a systematic review or meta-analysis (Levac et al., Reference Levac, Colquhoun and O’Brien2010). This review included 11 quantitative studies, but the heterogeneity in population and outcomes would make them difficult to synthesize. Particularly notable was that the studies’ authors used different approaches to measuring outcomes. Integrated care goals are often two-pronged, addressing both patient and system efficiency outcomes; it is therefore not surprising that the outcomes in these documents are variable in approach and perspective. However, the lack of a gold standard for measuring the success of integrated care in improving care transitions is perpetuating variability in outcome measurement. Most articles in this study focused on reducing institutional health care use as a desired outcome of improved care transitions, with health and function outcomes being less commonly measured. Outcome measurement in this body of literature had little emphasis on the patient perspective compared to the service perspective, which is inconsistent with the goals purported by these initiatives.
Another challenge for interpreting the quantitative literature is the lack of appraisal of whether the integration goals were achieved, making it difficult to determine the extent of the influence of integration on outcomes. The qualitative literature included in our review revealed challenges with the implementation of integrated care initiatives. For example, Hjelmar, Hendriksen, and Hansen (Reference Hjelmar, Hendriksen and Hansen2011) described how an attempted collaboration between hospital and primary care had difficulty eliciting cooperation for the initiative from the community general practitioners. Further, documenting the success of initiatives in promoting integration is important because integrated care success may be more dependent on the consistency between the macro and micro environment (administration and front-line staff) than on the type of initiative (Calciolari & Ilinca, Reference Calciolari and Ilinca2011).
Previous authors have pointed out the lack of standardized, validated tools for evaluating to what extent care is integrated. A systematic review of integrated care delivery and services found only three tools that measure the extent of integration (Armitage, Suter, Oelke, & Adair, Reference Armitage, Suter, Oelke and Adair2009), and a recently published protocol also aimed to address this gap by means of an inventory of indicators for assessing achievement of an integrated system (Oelke, Suter, da Silva Lima, & Van Vliet-Brown, Reference Oelke, Suter, Da Silva Lima and Van Vliet-Brown2015). The proposed scoping review by Oelke et al. (Reference Oelke, Suter, Da Silva Lima and Van Vliet-Brown2015) addresses a sorely needed area of research. We also recommend an inventory of tools that evaluate the patients’ perception of care integration. The patient perspective is notably absent from the inventory compiled by Armitage et al. (Reference Armitage, Suter, Oelke and Adair2009) and is not explicitly addressed in the protocol by Oelke et al. (Reference Oelke, Suter, Da Silva Lima and Van Vliet-Brown2015). Incorporating the patient voice into this research is important as, despite the lack of consistency in the definition of integrated care, there is consistency in the view that integrated care requires a patient-centred focus (Hollander & Prince, Reference Hollander and Prince2007; Kodner & Spreeuwenberg, Reference Kodner and Spreeuwenberg2002; WHO, 2015). The National Voices in England developed a document that provides narrative statements that could be used for developing integrated care indicators. For example, statements that specifically relate to transitions include “When I use a new service, my care plan is known in advance and respected” (National Voices, n.d.).
The second objective of this review was to determine how the concept of integrated care is being applied in the literature on care transitions. Although at least half of the documents in this review did not describe their conceptualization of integrated care, analysis of the initiatives determined that there were three broad applications of integrated care: (1) integration of services within the hospital, (2) vertical integration of health services, and (3) vertical health as well as community horizontal integration. In most of the initiatives included in this review, integrated care was a strategy used across organizational boundaries rather than being applied only within an organization (only two articles focused on within-hospital integration). More than half of the initiatives aimed to integrate community health and social services, indicating that robust community supports beyond health issues are believed to be important for supporting older adults’ post-hospitalization.
In terms of the initiatives’ features, many were consistent with integrated care frameworks, such as having shared values, integrated information systems, jointly managed programs, and co-location (Hollander & Prince, Reference Hollander and Prince2007; Kodner & Spreeuwenberg, Reference Kodner and Spreeuwenberg2002). However, we expected more documents to describe patient and family involvement due to the consistent focus of patient-centred care in integrated care frameworks (Hollander & Prince, Reference Hollander and Prince2007; Kodner & Spreeuwenberg, Reference Kodner and Spreeuwenberg2002). Some of the features were more consistent with care coordination rather than integration – for example, scheduling follow-up appointments prior to hospital discharge, or ensuring that written information is provided to the community by the hospital in a timely and standardized fashion (Leutz, Reference Leutz1999). Further, some features may or may not have been in the spirit of integrated care depending on how they were operationalized. For example, when developing the category of patient and family involvement, we included all references to family and patient communication and support without judging the quality of the proposed interactions. Some of the interactions were consistent with the spirit of integrated care in that they involved patient and family input into decision-making and goal-setting, whereas for other initiatives it was simply stated that information about decisions would be provided to families and patients. It is well recognized that depending on the setting and context, coordination or linkage may be a more realistic goal than full integration (Leutz, Reference Leutz1999); however, none of the documents in this review identified the extent of integration that the initiatives aimed to achieve.
The analysis of initiative features in this review confirms that, similar to the larger body of literature on integrated care, micro-level clinical and service delivery integration is typically targeted rather than macro-level integration of funding, administration, and/or organization (Kodner & Spreeuwenberg, Reference Kodner and Spreeuwenberg2002). Because there was a higher level of description of programs’ clinical features, we were able to ascertain more consensus as to what clinical-level integration entails for older adults’ care transitions than for administrative integration. There were three specific clinical features included in more than 50 per cent of the initiatives, but for macro strategies, we were unable to achieve this level of specificity.
A broader description of macro levels of integration is required for integrated care comparison. It may be that the focus on micro clinical integration in this literature reflected the scope of the integration initiatives. However, different health regions and different countries will have varying levels of integration inherent in their health system infrastructure. Thus, in order to compare integrated care initiatives across regions and countries, a full understanding of integration across both the macro and micro domains is needed to truly understand the context of a given initiative. This is particularly important considering the international scope of this literature. Further, it has previously been noted that integration at a macro level does not automatically lead to integration at the micro level (Baillie et al., Reference Baillie, Gallini, Corser, Elworthy, Scotcher and Barrand2014; Vedel et al., Reference Vedel, Monette, Beland, Monette and Bergman2011), and that context is particularly important for integration initiatives, as new initiatives need to match local needs (Armitage et al., Reference Armitage, Suter, Oelke and Adair2009). A useful way to describe integrated care is in terms of the extent of integration (autonomous, coordinated, or integrated; Leutz, Reference Leutz1999) for each of five integration domains (financial, administrative, organizational, professional, clinical; Kodner and Spreeuwenberg, Reference Kodner and Spreeuwenberg2002).
In summary, the main implication for research from this study is the need for improved consistency and standardization in describing and evaluating integrated care initiatives aiming to improve care transitions. The grey literature provides depth and breadth on possible approaches to integrated care for health policy decision-makers, but few of the initiatives included evaluation. For both grey and empirical literature, description of the macro- as well as micro-level context of integrated care initiatives is important to promote comparison and assist decision-makers, such as health authorities, to determine if the initiatives would be practical in their own contexts. In addition, evaluation of both the achievement of integration and the outcomes of integration is needed to help in understanding how the integration initiative is contributing to outcomes. Of particular importance for the topic of care transitions is the inclusion of the patient perspective when evaluating outcomes. Care transitions research has identified a lack of patient inclusion and satisfaction in care transition decisions and processes (Jackson et al., Reference Jackson, Oelke, Besner and Harrison2012; Lapum et al., Reference Lapum, Angus, Peter and Watt-Watson2011), and integrated care research has indicated that, in some cases, front-line health professionals’ perceptions of improvements in care delivery is not always matched by those of patients and families (Holstege et al., Reference Holstege, Caljouw, Zekveld, Van Balen, De Groot, Van Haastregt and Achterberg2015). Therefore, consideration of both system and patient perceptions is important for determining the impact of integration initiatives on care transitions.
A limitation of this study was that we included only literature that was very explicit in language regarding integration. There is a large body of research on the topic of care transitions for older adults in which many different care frameworks are explicitly and implicitly used. Frameworks that may have used an approach to enhance integration but did not explicitly use this terminology would not have been captured in this review. This review also did not include documents that focused on preventing hospital admission by providing community-oriented integrated care, as we were interested specifically in older adults who had been hospitalized. The use of content analysis required the authors to interpret the meaning of text based on their own knowledge of clinical and administrative health systems and practices. The lack of use of the “consultation” step in the scoping review process may have limited the relevance of our findings to health decision-makers and health services researchers (Levac et al., Reference Levac, Colquhoun and O’Brien2010).
In conclusion, the literature on using integrated care approaches for transitions from hospital into the community among older adults is diverse in terms of literature type as well as scope and extent of integration. There are opportunities for increasing knowledge on this topic by evaluating integrated pilots and initiatives currently planned or under way. However, to learn how to apply these initiatives in other contexts, increased reporting of how integrated care is conceptualized and whether integration was achieved, as well as a description of the macro-level integration context, is needed.
