Measuring integrated care at the interface between primary care and secondary care: a scoping review

Abstract

Purpose

In many healthcare systems, health policy has committed to delivering an integrated model of care to address the increasing burden of disease. The interface between primary and secondary care has been identified as a problem area. This paper aims to undertake a scoping review to gain a deeper understanding of the markers of integration across the primary–secondary interface.

Design/methodology/approach

A search was conducted of PubMed, SCOPUS, Cochrane Library and the grey literature for papers published in English using the framework described by Arksey and O'Malley. The search process was guided by the “Preferred Reporting Items for Systematic Reviews and Meta-Analyses” (PRISMA).

Findings

The initial database search identified 112 articles, which were screened by title and abstract. A total of 26 articles were selected for full-text review, after which nine articles were excluded as they were not relevant to the research question or the full text was not available. In total, 17 studies were included in the review. A range of study designs were identified including a systematic review (n = 3), mixed methods study (n = 5), qualitative (n = 6) and quantitative (n = 3). The included studies documented integration across the primary–secondary interface; integration measurement and factors affecting care coordination.

Originality/value

Many studies examine individual aspects of integration. However, this study is unique as it provides a comprehensive overview of the many perspectives and methodological approaches involved with evaluating integration within the primary–secondary care interface and primary care itself. Further research is required to establish valid reliable tools for measurement and implementation.

Citation

O'Farrell, A., McCombe, G., Broughan, J., Carroll, Á., Casey, M., Fawsitt, R. and Cullen, W. (2022), "Measuring integrated care at the interface between primary care and secondary care: a scoping review", Journal of Integrated Care, Vol. 30 No. 5, pp. 37-56. https://doi.org/10.1108/JICA-11-2020-0073

Introduction

Integrated care is a model of care within health systems and is considered a solution to the challenge of providing comprehensive, coherent and synergistic healthcare (Kodner and Spreeuwenberg, 2002; Valentijn et al., 2013; Goodwin, 2016). However, a lack of consistently applied definitions makes evaluating integrated care difficult, and there is a scarcity of “standardized, validated tools” used to evaluate integration outcomes (Armitage et al., 2009; Lyngsø et al., 2014; Strandberg-Larsen and Krasnik, 2009). Ambiguity and inconsistency around the terms, coupled with diverse outcome measures among integrated systems means uniform conclusions cannot be made about ideal integrated care model types and ways to evaluate each aspect of them (Lyngsø et al., 2014).

Nonetheless, past efforts have been made to develop effective integrated care assessment tools. For instance, with the aims of (1) identifying principles and factors facilitating effective care integration and (2) assessing the performance of integrated care models, in 2017 the European Commission's Expert Group on Health Systems Performance Assessment produced the “Blocks” report (Reynders, 2017). The report found that measuring integration is not the same as measuring integrated care performance. The report also establishes the term “building blocks” to monitor elements of integrated care. It describes the need to develop indicators that are specific to integrated care and stratify them to assess structures, processes and outcomes (Reynders, 2017). The Primary Health Care Impact, Performance and Capacity Tool (PHC-IMPACT) meanwhile is another integrated care assessment initiative. It uses numerous evidence based, mixed method indicators and pre-identified “Tracer conditions” to measure current integrated care structures, their performance and the effectiveness of primary healthcare in a region to inform its policy decision-making and aims to work towards global universal healthcare (Barbazza et al., 2019; Tello, 2019). Another initiative is the Scaling Integrated Care in Context (SCIROCCO) Project (Grooten et al., 2019). The project involved a study examining readiness for integration in health systems across 25 European Union sites. The “maturity” of healthcare systems and each site's ability to implement integrated care was assessed using a validated 12-dimensional tool (Grooten et al., 2019). The project has now concluded, and a new project is underway – “SCIROCCO Exchange”, which has refined the model for assessment and aims to support health systems in scaling-up integrated care (SCIROCCO Exchange) (Pavlickova, 2019).

Ireland's healthcare system is currently in transition, as it endeavours to provide universal integrated healthcare, which is primary care centred with an emphasis on community care and an integrated system to cater for patients at all stages of life from disease prevention to diagnosis and disease management (Burke et al., 2018). Previous research in Ireland has identified the primary–secondary care interface as a problem area. Darker et al. reported that barriers to effective chronic disease management included difficulty in consulting hospital specialists and poor communication between primary care and hospitals teams (Darker et al., 2015). Further research reported that the relationship between primary and secondary care was considered “disconnected” and “fragmented” by almost half of the participants with some key issues relating to inadequate discharge summaries, communication difficulties with hospitals and difficulty accessing assessment units (Kennedy et al., 2016). A 2017 report “A Future Together” highlighted general practitioners (GPs)' concerns with inefficient communication systems, time consuming referral pathways and difficulty liaising with hospital staff (O'Dowd et al., 2017).

2020 sees the introduction of an “Integrated Care Programme for the Prevention and Management of Chronic Disease” (ICPCD) to replace the “diabetic” and “heartwatch” initiatives. This scheme will focus on increased formal general practice led care for a number of chronic diseases, which are a great burden for patients. Healthcare is increasingly being delivered through primary care, and there is an expected 46% rise in demand for primary care over the next 15 years (Health Service Executive (HSE), 2018).

Given the changing landscape of general practice in Ireland, it is timely and indeed necessary to evaluate the current relationship between primary and secondary care. As such, this review aims to examine the current literature to establish what information has been used to measure and assess integrated care at the interface between primary care and secondary care and thereby identify issues which may have an impact on future assessment of integrated practice at the primary–secondary care interface.

Methods

To outline the extant literature, its key concepts and the gaps in the research, we conducted a scoping review using the six-stage framework described by Arksey and O'Malley (Arksey and O'Malley, 2005).

• Stage 1: Identifying the research question

Our objective was to examine the interface between primary and secondary care to establish what markers could be used to evaluate integration between primary care and secondary care. The following research question was formulated: What information has been used from primary care to measure/assess integrated care at the interface between primary care and secondary care?

• Stage 2: Identifying relevant studies

A preliminary search of key databases was performed, and a reading list was generated. From this, medical subject heading (MeSH) terms were generated. Further adjustment of terms and inclusion of terms identified in the literature as international synonyms for integrated care were included in the search. A search of PubMed, SCOPUS and Cochrane was performed. The search terms were classified by category and results required reference to one or more search term in each category (See Figure 1). We chose not to limit the study search by year as research on integrated care assessment is limited, and we thus anticipated that a wide temporal focus would facilitate better inclusion of studies relevant to our research aims. Lastly, several additional articles of relevance were identified by “hand searching” for the grey literature on prominent health websites and databases using Google search functions.

• Stage 3: Selecting studies

Thereafter, a title and abstract review was conducted, followed by full-text reviews. The “Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)” flow diagram below (Figure 2) summarises the selection pathway. The literature was included irrespective of study design/methodology. This broad inclusion criterion facilitated the inclusion of a variety of study types and reviews. Endnote X9 software was used to track studies and manage citations. Studies were included if they were considered to examine the research question, and if they were published in the English language. Studies were excluded if they reported only patient perspective/satisfaction or focussed on specific individual conditions. All duplicate articles were excluded. Findings were reviewed by a second reviewer, and a finalised list of studies was agreed.

• Stage 4: Charting the data

To facilitate comparison and thematic analysis, the following data were extracted from the articles:

• Author(s), year of publication and title,

• Study population,

• Journal/Publication,

• Setting,

• Study aim/topic,

• Study design and

• Major findings.

• Stage 5: Collating, summarising and reporting results

An overview of the literature was detailed in a table summarising and charting the results (see Table 1).

Results

Studies identified

A total of 120 studies were examined; 112 were identified by data search and eight following review of key papers and journals. In total, eight duplicate papers were removed, leaving 112 articles to be screened. Following a title and abstract review, 86 studies were excluded as they were not relevant to the research question. The remaining studies underwent full text review and were analysed by a second reviewer. At this time, studies were removed based on unavailability (n = 1), language (n = 1) and lack of relevance (n = 7). A total of 17 studies were identified as relevant for analysis.

The 17 studies included ranged from 1993 to 2019 with the following geographical breakdown: USA (n = 7), United Kingdom (n = 4), Australia (n = 4), Denmark (n = 1) and Ireland (n = 1).

A range of study designs were identified including a systematic review (n = 3), mixed methods study (n = 5), qualitative (n = 6) and quantitative (n = 3).

Study populations included primary care physicians (PCPs) alone (n = 9), six studies examined the views of PCPs and others including practice staff (n = 1); eHealth IT specialists (n = 1); parents of patients (n = 1); hospital management (n = 1); physician specialists (n = 1); PCPs with patients and physician specialists (n = 1); one review included studies across a variety of domains and two study populations included health care management staff.

Studies examining integration across the primary–secondary interface

In total, eight studies assessed integration across the primary–secondary care interface: three were mixed methods studies, four qualitative studies and one systematic review. The studies focused on characteristics of successful integration, including communication, attitudes and education.

A systematic review by Mitchell et al. examined outcomes of models that integrate primary and secondary care (Mitchell et al., 2015). This review examines the effectiveness of these models. Except for disease control, limited advances were reported in terms of patients' clinical outcomes compared with usual care. However, substantial improvements were noted in service-related process outcomes.

In their qualitative study, Bouamrane et al. reported that eReferral substantially improved communication between general practice and secondary care and noted that instant transfer of referral and the availability of an electronic audit trail were two key advantages over paper-based systems (Bouamrane and Mair, 2014). Interviews with 25 GPs reported benefits including the system being more user friendly (n = 11), referral transfers being more immediate (n = 9), clinical advice and referral guidance functions (n = 5), improved organisational work processes and patient management through the health service (n = 8) and sharing of electronic patient information across the health service (n = 5).

Murphy et al. conducted a mixed method study examining GP satisfaction with electronic discharge summaries and accuracy of ICD-10 coding by non-consultant hospital doctors (Murphy et al., 2017). Overall satisfaction level with electronic discharge summaries was high (91–100%). List of diagnoses, treatments, procedures, GP information and follow up and discharge medications were all noted to be of key value to GPs. All were satisfied with electronic prescriptions and all found information regarding patients' medications that were stopped/ held useful.

Lee et al. examined PCP perceptions of electronic consult systems in relation to workflow, specialist access and patient care (Lee et al., 2018). Many physicians reported that the systems resulted in timelier speciality input, improved scheduling, educational benefits and a positive change in relationship between specialists and physicians.

Isaac et al. also evaluated the interface through their mixed-method study reviewing attitudes of 350 GPs to tertiary teaching hospitals (Isaac et al., 1997). In total, 93% were keen to see an extension of shared care. Ongoing concerns were communication and time. They found that 84% were not informed of patient admission or change in patient well-being, including death (87%) and discharge (75%). An area of concern was early discharges: where GPs were concerned for patient well-being (65%) and felt discussion was required prior to taking over responsibility for the patient.

The relationship between primary and secondary care was examined in a qualitative study by Brousseau et al. (2011). Parents' and PCPs' feelings regarding direct emergency department (ED) attendances were reviewed. In general, physicians approved of parents' decisions attend second level care directly. Physicians understood the potential reasons for attending ED, and neither PCPs nor parents felt that these non-urgent ED attendances were a “significant enough” breach in continuity of care to warrant changes in physician care practices (e.g. integrative initiatives directing such patients away from ED towards primary care).

A mixed methods study by Tuzzio et al. examined the impact of education at the interface in the form of “peer coaching” on specialty referrals (Tuzzio et al., 2017). All participants reported benefit of peer discussion on patient care. All reported that they reflected on their referral decision- making and considered new approaches to referral and for managing patient expectations following the meetings. Time constraints were noted to be a barrier for optimising referrals.

In a quantitative study by Southern et al. participating GPs noted that they felt coordination between GPs and hospitals regarding patient management was sub-optimal. GPs also mentioned that hospital involvement in patient care was insufficient (Southern et al., 2001). In this study, only 41% of GPs claimed that they were involved in an admitted patient's care, 18% reported being involved in discharge planning and a third mentioned receiving information about patient's hospital medication. A third of rural GPs were involved with hospital committees vs 8.4% of urban practitioners. Only 28% of GPs were linked to other healthcare services by computer technology. Remuneration was cited as a barrier to integration by 22% of study participants.

Integration measurement

In total, two studies examined integration in general terms: one a quantitative study and one was a systematic review.

Lyngsø et al., in 2014 published a systematic review examining instruments to assess integrated care (Lyngsø et al., 2014). They found no generally agreed measurement instrument. A diverse combination of methods was found to have been used. Most studies looked at structural and process aspects of integration with only four studies examining all six criteria defined as central for a measurement tool. These criteria include a defined construct, theoretical framework, defined level of analysis, structural aspects, process aspects and cultural aspects. The three elements most commonly examined were the following: IT, information transfer, commitment and incentives and clinical care.

Gillies et al. looked at measuring integration in their quantitative study (Gillies et al., 1993). They reviewed 12 organised delivery systems. Focus was put on the “perceptions of integration” based on the thought that improvements must first deal with the current zeitgeist. They report moderate integration at a functional level but at low levels of physician-system integration or clinical integration. There is a link reported between perceived integration and perceived effectiveness: that the better the coordination, the more effectual system is.

Factors affecting care coordination

The remaining seven studies identified several themes pertinent to care coordination including time, finances, resources and the value of GPs as stakeholders. They consisted of three qualitative studies, two mixed methods studies, one systematic review and one was a quantitative study.

A qualitative study by Chaudry et al. reports increased paperwork and administrative work associated with managed care (Chaudry et al., 2003). Poor patient understanding was thought to contribute to the inappropriate use of services. Communication, complex needs and reimbursement were key concerns voiced by participants.

A time study by Chen et al. reports that 20% of a physician's workday was spent on activities outside of office visits (AOVs) that in turn adversely impact care coordination (Chen et al., 2011). They found that 38% of this time was spent on visit specific tasks (i.e. completing tasks generated during a consultation), and 62% were non-visit specific AOVs (phone calls 26%, follow-up diagnostics 22% and prescriptions 12%). It was thought that 15% of these tasks could have been completed by support staff.

Fowler Davis et al. examined GP views on Enhanced Primary Care Programmes (Fowler Davis et al., 2018). Capacity to Support Integrated Care teams was one of the main themes. Many felt that the schemes did not enhance the workings of the multidisciplinary team (MDT). GPs were selective in their implementation to benefit their practice demand, without increasing their own workload.

A qualitative study by Blakeman et al. examines perceived barriers associated with delivering coordinated care (Blakeman et al., 2001). They highlight the importance of time, organisation, communication, education and available resources. Barriers included poor links with MDT/secondary care, including delays and inadequate discharges documents, difficulty with contact, poor knowledge of services available, lack of understanding regarding roles and inadequate community services. It highlights that care coordination relies on the effectiveness of other forms of integration in order to achieved desired outcomes.

Directing resources towards coordinated care was the aim of the Somerset Practice Quality Scheme reported by Close et al., in 2019 (Close et al., 2019). Ultimately, time savings and MDT improvements were recorded, and decreased administrative work was appreciated by disincentivising quality and outcome framework targets (QOFs) and redirecting resources to target complex patients with multi-morbidities.

Gosden et al. conducted a systematic review in 2016 examining payment methods of physicians and the affects that this may have (Gosden et al., 2000). It concluded that fee for service (FFS) resulted in greater compliance with suggested attendances; more attendances at primary care, specialist care, diagnostic services but fewer secondary referrals and repeat prescriptions when compared with capitation. Greater continuity of care and improved compliance were reported among FFS participants.

Financial incentives were the focus of the quantitative study by Grumbach et al. (1998). They reported on the behaviour of physicians to tailor their management based on incentives. Of the 766 physicians involved in managed care programmes, 38% received an incentive/ bonus. Pressure to limit referrals was reported by 58%, where 17% reported that this compromised care. Pressure to see more patients was reported by 75%, where 24% felt that this compromised care. Physician satisfaction was reported as lower when incentives were linked to productivity vs physicians for whom incentives were linked to quality of care.

Discussion

This study sought to develop understanding of how primary care has informed the measurement and assessment of integrated care at the primary–secondary care interface. It is clear from the literature that the measurement and assessment of integration needs to take into account several elements, dimensions and points of view. These include perspectives on primary–secondary care interactions and issues concerning management of primary care time, financial and human resources. Further, diversity of perspective is also evidenced by the fact that the studies examined in this review used a wide variety of methods including surveys, interviews, questionnaires, data analyses, literature reviews and observational techniques to assess integration. The methodological diversity used in this review's included studies shows that no single approach covers all aspects of integration but many cover individual elements of integration.

The finding that included studies examined integration from a wide variety of perspectives using a multitude of research techniques is not surprising as previous research has also demonstrated that this is often the case (Barbazza et al., 2019; Tello, 2019; Burke et al., 2018; Darker et al., 2015; Pavlickova, 2019). However, this study makes a valuable contribution to knowledge in the sense that it sheds new light on the diversity of perspectives and approaches within research examining integration in the primary care sector and the primary–secondary care interface.

The included studies' findings also have implications for understanding of how integrated care systems may be better evaluated in healthcare systems both in Ireland and internationally. One of the included studies was conducted in Ireland (Murphy et al., 2017), and several studies were conducted in countries with socioeconomic dynamics, cultural backgrounds and healthcare systems like those in Ireland. Thus, this review's findings will likely prove useful with regards to answering questions posed by existing integrated care policy documents and initiatives in various countries (Burke et al., 2018; Darker et al., 2015; Health, 2018). Based on this review's findings, it is recommended that policymakers take the time to account for the multitude of professional perspectives within healthcare systems before implementing policy reform. Further, we recommend that policy focused evaluations standardise integration assessment tools as much as is possible to avoid the confusion resulting from methodological ambiguity evident among peer reviewed studies to date.

This study has several methodological strengths and limitations. Our adoption of Arksey and O'Malley's framework for instance was beneficial, as it facilitated greater rigour and transparency in the research process. Also, based on a review of the literature, a comprehensive set of search terms were gathered. Further, we feel the decision to not limit our literature search by year was justified as it facilitated inclusion of several valuable studies published prior to 2010 (Chaudry et al., 2003; Blakeman et al., 2001; Southern et al., 2001; Isaac et al., 1997; Gosden et al., 2000; Grumbach et al., 1998; Gillies et al., 1993). Our search, however, did not include all databases, which may have resulted in omission of some relevant studies, and we did not evaluate the study quality of the included literature. We also only included the literature published in English, which may have excluded other relevant studies.

Conclusion

It is clear that measurement and assessment of integration within the primary–secondary care interface and primary care itself is complex and involves giving voice to multiple perspectives. Further, understanding of these complexities may benefit from the application of standardisation within integrated care evaluation processes. Thus, the challenge ahead for Irish and international clinicians, researchers and policymakers lies in establishing valid reliable tools for assessment and then implementing them.

Figures

Figure 1

Search strategy

Figure 2

Summary of papers identified for study

Table 1

Included studies

	Author(s)	Year	Study title	Study population	Journal/ Publication	Setting	Study aim/Topic	Study design	Major findings
1	Murphy et al. [37]	2017	Electronic discharge summary and prescription: improving communication between hospital and primary care	GPs (n = 13) and Chart data (n = 90)	Irish Journal of Medical Science	Ireland	To evaluate the effect of electronic discharge on GP satisfaction and accuracy of diagnosis	Mixed methods, qualitative interview 13 semi- structured interviews and quantitative chart data	EDS has led to improved timelines and GP satisfaction with communication between hospital and primary care
							Communication		Coding is inaccurate
2	Close et al. [49]	2019	Longitudinal evaluation of a countrywide alternative to the quality and outcome framework in UK general practice aimed at improving person- centred, coordinated care	Practices involved (n = 55) patients, healthcare professionals practice managers and staff	British Medical Journal	The United Kingdom	To evaluate the deincentivisation of QOFs	Mixed method study. Tools such as quantitative P3C-EQ, patient experiences P3C, practitioner P3C-OCT, organisational data, hospital statistics, emergency admission and data- interrupted time series Qualitative component: semi- structured interviews (not dealt with in this paper)	Initiative saved time Time was freed during consultation Time was freed for admin staff and GPs outside of consultation time Practice data showed a significant increase in organisational processes, stronger federation links and informal networks Increased MDT working, relocation of resources, changes to structure and timings of appointments No disbenefits were detected in admission data
3	Lee et al. [38]	2018	Primary care practitioners' perceptions of electronic consult systems: A qualitative analysis	Primary care practitioners (n = 40) 20 DHS v 20	JAMA Internal Medicine	USA	To understand PCP perceptions of the results of eConsult initiation on PCP workflow, specialist access and patient care	Qualitative interviews	Four themes Access and timeliness of specialist care- variable shift of speciality care to PCP Relationship with specialist: all variable in response, both positive and negative
				Non-DHS (Dept. of Health Services) (12 internists 17 family practice practitioners 11 advanced practice clinicians (n = 40)
4	Fowler Davis et al. [47]	2018	Factors affecting decisions to extend access to primary care: results of a qualitative evaluation of general practitioners' views	GPs (n = 24)	British Medical Journal	The United Kingdom	To report GPs' views and experiences of an Enhanced Primary Care Programme (EPCP)	Qualitative interviews	Four main themes: receptivity to the aims of the EPCP; capacity to support integrated care teams; capacity to manage urgent care; value of schemes to enhance locality-based primary care’s variable responses GPs were selective to benefit local demand without increasing their own workload
5	Bouamrane and Mair [36]	2014	A qualitative evaluation of general practitioners' views on protocol-driven eReferral in Scotland	GPs, 25 semi- structured interviews, one focus group with members of the Scottish Electronic Patient Record programme and one interview with a senior architect of the Scottish Care Information national eReferral system	BMC Medical Informatics and Decision making	UK	GP perspective on information management processes	Qualitative interviews/focus group. They conducted semi-structured interviews, and they analysed data using a	eReferral streamlines communication, improvements over paper: immediate transfer and electronic audit trail. Some felt templates cumbersome, while some felt were useful GPs considered any benefits of eReferral greatly outweighed any disbenefits
								framework based on
								DeLone and McLean's model of
								quality in information systems [53]
6	Brousseau et al. [40]	2011	Nonurgent emergency- department care: analysis of parent and primary physician perspectives	Parents of children (26) and primary care physicians (20)	Pediatrics	USA	To better	Qualitative interviews 26 interviews of 21 female and 5 male parents were completed	Parents felt they acted appropriately Physicians neither approved their decision nor felt the need to change
							understand parental decisions to seek care for their children and
							physician perceptions of parents' decisions to seek non-urgent emergencydepartment care
7	Chen et al. [46]	2011	Patient care outside of office visits: a primary care physician time study	Primary care physicians (n = 33)	Journal of General Internal Medicine	USA	To describe primary care physicians' ambulatory patient care activities outside of office visits (AOVs) and their perceptions of the extent these AOVs substitute for visits and could be performed by support staff	Mixed methods, cross- sectional, direct observational study and Qualitative questionnaire	It was found that 20% of physicians’ workday was spent on AOVs AOVs can substitute for some visits, which would otherwise occur approx. five visits per day Some tasks could be delegated to another staff member (15%) Policies needed to save costs, time and improve care
8	Chaudry et al. [45]	2003	Caring for patients under medicaid mandatory managed care: perspectives of primary care physicians	14 physicians,	Qualitative Health Research	USA	Care coordination	Qualitative interviews	Four themes: provider hassle, complex needs improved access to care under managed care and individual providers disconnect from policy and evaluation
				7 individual interviews, 7 participated in focus groups and CEOs of 3 health maintenance organisations
				(HMO)
9	Blakeman et al. [48]	2001	Evaluating general practitioners' views about the implementation of the enhanced primary care medicare items	GP (n = 30)	Medical Journal of Australia	Australia	Measuring barriers: use of EPC items, difficulties with implementation and suggestions for improving implementation	Qualitative interview	Five main topics: Time, Organisation, Communication, Education and Resources
									Difficulty incorporating items into daily practice without support EPCs need implementation and depend on other aspects of integration to succeed
10	Southern et al. [42]	2001	Integration from the Australian GP's perspective	GPs (n = 208)	Australian Family Physician	Australia	GPs' perceptions about their role in relation to activities that support integration and what they are doing	Quantitative survey- based on the agreed focus group statements	GP responses relating to holistic individualised care were positive Statements about care-coordination scored low Rural GPs had more involvement with secondary care Many obstacles to integrated care were identified: policy and attitudinal, and financial incentives required infrastructure to support education
11	Isaac et al. [39]	1997	The GP hospital interface: attitudes of general practitioners to tertiary teaching hospitals	GP 350	Medical Journal of Australia	Australia	To assess GP perceptions of liaison with two local tertiary teaching hospitals	Mixed methods survey,	GP dissatisfaction with communication GPs not notified of admissions: 84% Changes in patient condition: 87% Discharge: 75% Poor access to results Changes in organisation and attitudes needed
								quantitative, qualitative open- and closed- questions were conducted as room for comments and questionnaire-based survey was conducted with a five point Likert scale
12	Gosden et al. [50]	2000	Capitation, salary, fee-for- service and mixed systems of payment: effects on the behaviour of	Four studies, 640 primary care physicians and more than 6,400 patients	Cochrane Database Systematic Review	The United Kingdom	Impact of diff methods of	Systematic review	FFS resulted in increased GP visits, visits to specialist/ diagnostics/curative services but fewer hospital referrals and repeat prescriptions compared with capitation
			primary care physicians				payment on clinical behaviour of GPs	2 RCTs 2 before and after designs	Compliance with visit numbers was higher and continuity of care was better with FFS
13	Tuzzio et al. [41]	2017	Design and implementation of a physician coaching pilot to promote value-based referrals to specialty care	Four primary care physicians and four coaches	The Permanente Journal	USA	To assess feasibility and acceptability of a coaching/mentoring programme to evaluate specialty referral decisions	Mixed methods, qualitative evaluation, single-arm observational pilot study	(1) Peer-to-peer dialogue relieved isolation and was a vehicle to learn from each other
								with four	(2) Reflection and acquiring new skills improved knowledge and decision-making capacity and main reasons for ref.
								dyads of qualitative and quantitative evaluation (interviews)	were clinical uncertainty and patient request. New strategies were developed for use
									(3) Lack of time was reported as a barrier to programme participation and to optimizing referrals
									(4) There was support for sustainability if supported
14	Grumbach et al. [51]	1998	Primary care physicians' experience of financial incentives in managed care systems	766 primary care physicians	New England Journal of Medicine	USA	Types of incentives for PCP in managed care systems	Quantitative questionnaire	pressure to limit referral, incentive via bonus
									If bonus, then more pressure was not to limit referral: this was felt compromised care
									Physicians with incentive based on productivity felt pressure to see more patients and felt this compromised care
15		2014	Instruments to assess integrated care: a systematic review	Systematic review of 23 articles, patients, health professionals, healthcare systems, organisational delivery systems and hospitals	International Journal Of Integrated Care	Copenhagen Denmark	Systematic review of instruments to assess integrated care	Systematic review	Eight organisational elements found IT, information transfer/communication an access Commitment and incentives to deliver integrated care Clinical care Organisational culture and leadership Education Financial incentives Patient focus Quality improvement/ performance measure
	Lyngsø et al. [8]
16	Mitchell and Burridge	2015	Systematic review of integrated	Systematic review	Australian Journal of Primary	Australia	To identify outcomes of different models	Systematic review	Few improvements in clinical outcomes
					Health		That integrate
	Zhang et al. [35]		Models of health care delivered at the primary– secondary interface: how effective is it and what determines effectiveness?	Patients with chronic complex illness, primary care doctors specialists and doctors			Specialist and primary care practitioners and characteristics of models that delivered favourable clinical outcomes		Improvement in process outcomes regarding disease control and service delivery increased cost of primary- secondary integration No negative effects compared with usual care Six elements identified that were common to the models in integrated primary–secondary care Interdisciplinary teamwork Communication/information exchange Shared care guidelines or pathways Training and education access and acceptability for patients viable funding model
17	Gillies et al. [43]	1993	Conceptualizing and measuring integration: findings from the health systems integration study	12 systems examined healthcare admin, managers and hospital group board members	Hospital and health services administration	USA	Looks at 12 organised delivery systems	Existing literature review, quantitative measures of perceived functional integration, physician- system integration and clinical integration both horizontal and vertical were obtained using a 54 point Likert scale self-administered questionnaire	Some functional integration areas are positively associated with both physician-system and clinical integration that in turn are positively related to each other Perceived integration was found to be positively associated with perceived effectiveness