Introduction: Not surprisingly given their multi-component nature, initiatives to improve integrated care often evolve to find the best way to bring about change. This paper provides an example of how an evaluation evolved alongside such an initiative designed to better integrate care across primary, community and hospital services in South Auckland, New Zealand.
Theory and methods: Using the explanatory power of a realist evaluative approach, theories of new ways of working that might be prompted by the initiative were explored in: (i) interviews with stakeholders in 2012 and 2015, (ii) online surveys of general practices and local care organisations, and (iii) a purposive sample of ten general practices.
Results: The results highlighted the institutional contexts that led to difficulties in implementing population health initiatives. They also revealed that changes in work practices focussed mostly on activities that improved the coordination of care for individuals at risk of hospital admissions.
Discussion: Multi-component complex interventions can vary in their delivery and be vulnerable to one or more components not being implemented as originally intended. In the case of this intervention, the move towards strengthening local relationships arose when contractual arrangements stalled. Realist evaluative approaches offer a logic that helps unpick the complexity of the relationships and politics in play, and uncover the assumptions made by those developing, implementing and assessing health service changes.
Conclusion: Given the multi-component and evolving nature of initiatives seeking to better integrate care, the realist evaluative emphasis on surfacing early the theories to explain how change is expected to occur helps overcome the challenge of evaluating “a moving target”.
In recognition that the projected demand for the local hospital was going to outstrip supply, a District Health Board in South Auckland, New Zealand introduced a major change programme focused on shifting the balance of care away from the hospital towards the community . Rather than a single intervention, the programme of change bundled together new financial incentives, new local decision-making networks and a new model of care to better manage long-term conditions. As different activities rose and fell in importance, along with the pattern of intermediate outcomes expected, the external evaluation needed an approach sensitive to the messier realities of health system change. Following the principles of realist evaluation , evaluative attention was paid to surfacing the theories expected to drive changes, testing these theories and then offering insights into what was working (or not working) for whom and in what contexts. Initial fieldwork revealed the contexts that meant local financial incentives made little progress while a new model of care made faster progress. Further fieldwork concentrated on exploring the impact of different local areas and practices in taking up a new model of care.
In 2012, in the context of a long standing interest in delivering care better designed around the needs of the local community, Counties Manukau District Health Board in South Auckland, New Zealand grouped a number of interventions together into a major change programme in order to deliver more care outside of the local hospital. An earlier initiative involving case management of patients who were high users of the hospital was extended to primary care to increase the management of patients in the community . Under the title of the ‘At Risk Individuals’ model of care, payments were provided to general practices to increase patient contact time and connect more with other services via a new information management system. At the same time four geographical localities were created and given responsibility for local planning, design and delivery . These new local networks (which became known as the Localities initiative) were expected to make the delivery of the At Risk Individuals model of care easier and motivate practice improvements by giving greater control over local budgets. Collectively it was hoped that a reduction in hospital bed days would be achieved allowing savings to then be invested in primary care.
Internationally, efforts to better integrate health care have taken many forms  often fueled by the belief that if care is better integrated for patients then avoidable hospital admissions can be reduced . The experience of those implementing integration programmes is that the concept itself takes a significant amount of time to clarify in context  and ambitious targets of reducing hospital admissions may not always be immediately realised, particularly with respect to the cost savings that can be achieved [8, 9]. However, there is good evidence that some initiatives do work to improve patient experience and the quality of care, particularly for those with long-term conditions [10, 11, 12].
The changes being put in place to increase the role of primary care by this District Health Board were informed by evidence that a greater focus on primary care is associated with better health, a more equitable distribution of health in populations, and lower health costs [13, 14, 15]. Further, that in order to manage the increasing numbers of the population with long-term conditions, strengthening population screening, monitoring and follow-up in primary care is valuable . Accordingly, the changes within the Localities initiative were informed both by new chronic care models of care [see for example, 17, 18], and tools such as primary care-led commissioning .
An independent evaluation was commissioned to track progress and evaluate the outcomes achieved from a Strategic Locality Partnership Agreement signed between Counties Manukau District Health Board and the five providers of primary health care within the district. In New Zealand, Primary Health Organisations represent the interests of groups of general practices that generally operate as independent businesses receiving a proportion of their funding from government capitation, and the remainder from patient co-payments. In this district, five Primary Health Organisations representing particular enrolled populations were in operation. In many cases these enrolled groupings cut across the geographical boundaries set up by the Localities initiative.
The Partnership between the five Primary Health Organisations and Counties Manukau District Health Board established a global budget for each locality centred on the expectation that if each locality was able to reduce demand across an agreed set of shared services then the savings would be directed towards innovative primary care services. The partnership agreement was the founding document of the Localities initiative.
Reviews of new contracting arrangements designed to create savings and reduce utilisation have found mixed outcomes. Potential savings can be limited by the time taken to design and implement new contracts, pricing services and agreeing how performance will be assessed . Nevertheless, learning from other experiences seeking to set up global budgets has found such budgets can slow underlying growth in medical spending while improving quality of care . Reviews have found the effectiveness of financial incentives on changing health professional behaviours are likely to be context dependent, particularly as contractual approaches may vary . For these reasons, the evaluation team adopted a realist evaluative approach in order to explore the different contexts and contractual arrangements likely to influence new behaviours within each locality. The starting point of the realist evaluative approach is that the combination of resources offered by a programme (in this instance the Localities initiative) is directed towards altering people’s reasoning. Consequently, the evaluator’s task is to make explicit the theory of how this occurs and then to successively test that theory in order to conclude with an understanding of not only how a programme works, but the conditions that influence its success .
The evaluation team undertook an initial assessment of the outcomes expected to be attributed back to the Localities initiative and were struck by the variety of routes being put in place to achieve an outcome concentrated on reducing demand for secondary health care. These routes included new financial incentives via a risk/gain share contract,1 new administrative arrangements involving locality leadership teams, a new At Risk Individuals model of care to manage long-term conditions, and new collaborative activities tailored to the particular interests in each locality. Collectively these were expected to achieve a 20% reduction in the standardised rate of use of acute medical inpatient bed days and emergency department services over the next five years. Other outcomes included meeting or exceeding targets such as CVD risk assessments, immunisations and smoking cessation advice, as well as aims to establish primary care as the central focus and coordination mechanism of local health care.
Recognising that the Localities initiative was not a fixed thing involving an unchanging set of interventions amenable to a closed system investigation, the following evaluation questions were agreed with the specially created Localities Evaluation Advisory Group:
Using claims made in planning documents and governance meetings, as well as insights from relevant literature [12, 23, 24, 25, 26], three initial programme theories emerged to explain how the Localities initiative would work. The programme theories are couched as context-mechanism-outcome configurations reflecting realist guidance to pay attention to how and why the programme will lead to outcomes and in what kind of settings :
Following the articulation of these initial theories, twelve interviews were held with key individuals associated with the design and implementation of the Localities initiative to identify in more detail how the four localities would work. All interviews were held over the phone using a semi-structured interview guide and lasted up to 40 minutes. Following realist interview guidance to capture the story of the programme being evaluated , ten propositions were distilled describing the different ways the Localities initiative was expected to offer value to health providers (See Box One). Guidance on developing realist theories suggests that formal theories drawn solely from the literature can end up being too abstract , so this initial analysis helped ground the evaluation in changes that were being perceived on the ground, as opposed to concentrating solely on claims made in the planning documentation and the academic literature.
Box One: Propositions identifying ways Localities initiative would work
Eighteen months into the operation of the Localities initiative a second set of interviews were held with seventeen stakeholders. These interviews included some of the initial designers of the initiative, along with those now responsible for the implementation in each locality and a selection of those providers expected to be influenced by the initiative. Interviews were held over the phone using a semi-structured interview guide. In this guide interviewees were asked to rank and comment on the ten propositions in Box One.
Interviews were transcribed and uploaded into NVivo for management and analysis. The results revealed the most value from the Localities initiative was coming from improved collegial relationships, and improved knowledge about who to contact in the District Health Board. By contrast, the least value was coming from increased accountability for apportioning resources and the potential to receive more funding from the risk/gain share contract. The findings reflected the slow progress on the risk/gain share contract outlined in the Strategic Locality Partnership Agreement. While the District Health Board had hoped to create four entities that would be governing bodies in their own right, the Primary Health Organisations had stronger incentives to maintain what they described as “their own sovereignty”. The result was that the planned move from nominal budget holding (where results were tracked but no funding changed hands), to real budget holding (where funding could change hands) never happened.
At this stage, an interim report was delivered to the Localities Evaluation Advisory Group tracking the evolution of ideas about how the Localities initiative would achieve change. Displayed graphically in Figure 1, the change of emphasis away from new local budget holding networks to a more diverse set of activities led the Advisory Group to agree to a new set of evaluative research questions. These were:
In order to answer research questions a and b, two Qualtrics based2 on-line surveys were developed, drawing on recent scholarship evaluating integrated care initiatives . These surveys were sent to local general practices (41% response rate) and local care organisations (33.5% response rate). For the survey, local care organisations were defined as those organisations offering care and well-being services locally and were identified by locality managers. These involved, for example, home based services, pharmacies, health advocacy services and mental health services. Questions probed how providers were making sense of the broader integrated care agenda and which relationships were being improved and in what ways.
For the survey sent to local general practices, 55% who replied were general practitioners, and 35% were practice nurses. Responses are reported as percentages as there were large differences in the number of people responding across the survey. Statistical hypotheses were not tested because the sample size was not large especially when broken down across the localities. For comparisons across the different localities, a pragmatic approach was taken in that an absolute difference of 10% was taken as a difference worth mentioning.
To answer research question c (i.e. understanding whether the most significant change project – the At Risk Individuals model of care – was achieving the changes expected), focus groups were held with practice staff in ten general practices. These practices were randomly selected to cover different Primary Health Organisations, different localities and different sized practices (i.e. small practices had less than 200 patients enrolled in the new model of care, medium practices had between 200–399 enrolled and large practices had 400 patients and over enrolled). The grid presented in Box Two displays the final sample based on those practices who agreed to participate. The integrated care team alongside PHO D in the grid below covered staff who were not organised in practices but as a wider team supporting walk-in clinics across the PHO. Not all practices in the initial lists from the Primary Health Organisations agreed to participate. In those cases, another practice that met the characteristics in the grid was randomly chosen by the evaluators.
Box Two: Focus group sample grid
The focus groups ranged in size from one through to seven members reflecting that smaller practices often only had one or two members of staff deeply engaged in the At Risk Individuals model of care. A total of 30 general practice staff participated in the groups. The majority of those who participated were practice nurses, though general practitioners, care coordinators and public health specialists were also included in some focus groups.
Three sets of propositions for how the At Risk Individuals model of care delivers value were developed from a review of the At Risk Individuals model of care set-up documents and the literature (particularly the literature on chronic care self-management programmes [26, 35, 36, 37, 38, 39, 40]). These propositions covered:
Focus group participants were asked about the extent to which these propositions resonated (or not) with their experience. The focus groups were audiotaped and then transcribed to create a qualitative data set that was entered into NVivo and coded against the different patient, practice and system contexts. We refined our understanding of each of our propositions by seeking both confirming and disconfirming evidence from our qualitative dataset. Emerging findings were cross checked within the research team with those who had experience of the day to day delivery of the Localities initiative. Consistent with the realist evaluative approach , the study as a whole relied on building a cumulative body of evidence involving different combinations of team members to ensure the trustworthiness of the emerging findings.
The study design was reviewed and approved by the Victoria University of Wellington Pipitea Campus Human Ethics Committee (# 0000021340). All participants had the research explained to them and gave informed consent.
The results presented in the following section provide findings from the surveys which explored how the Localities initiative was working as an ‘enabler’ of improved relationships. The next section focuses on one slice of the findings from the At Risk Individuals focus groups – that is, on what drives change for practices. Further results on the roll out of the At Risk Individuals model of care are presented elsewhere . These further results provide more detail on how being ‘at risk’ was being interpreted by staff. The final section summarises what was happening across the system over the period of the implementation of the Localities initiative.
Initial expectations were that within each of the localities the same project or event may unfold differently because of the different make-up of the populations served, the nature of the organisations within each locality and the different history of established services. Two on-line surveys revealed only small differences between each of the four localities. The lack of variability may be explained by the relatively high numbers of local care organisation respondents who did not align with one locality but were district wide (43%), and the small sample sizes which meant that only an absolute difference of 10% was taken as a difference worth mentioning. Nevertheless, given the considerable early thought that went into grouping providers into geographically distinct communities, the lack of significant differences between each of the localities in the survey results was notable.
In an environment where the central narrative of what the Localities initiative was expected to do shifted from a focus on risk/gain share contracts to local networks acting as “an incubator of integration” , providers made sense of the changes by testing whether they improved their knowledge around who to contact to arrange care for individual patients. For general practices, the relationships perceived to have strengthened most in the last three years were those with community pharmacy, home based carer support and mental health services. The relationships most likely to not have changed either positively or negatively (71%) were those with other social services (a category which included housing and welfare support as well as non-government organisations).
Local care organisations, despite already having connections with local general practices, were concerned that awareness of their services and information sharing were not as strong as they could be. Survey results revealed they were looking to Localities-linked activities to support them in their roles as an advocate and coordinator for their clients. Those that chose to provide comments on what influenced their judgements on relationships did refer to specific localities events and working groups, but a few also stressed the relevance of simply being in roles for some time. For example:
“I personally don’t feel my relationships have changed with other providers as I feel I have had good networks in these areas prior to the locality work getting up to speed…. I have witnessed others whose relationships have improved though …including awareness of other providers and networks. Some of this may be caused by the locality work but may also be caused by other factors happening.”
Table 1 presents the results from the general practice surveys when asked about the outcomes that can be achieved from improved integrated care. Table 2 presents the results from the local care organisations. Not surprisingly general practices were most positive about the role of primary care as a coordinating mechanism whereas local care organisations prioritised the importance of local needs being met. Despite the attention paid to reducing the demand for hospital services, there was a high degree of uncertainty from general practice respondents about the potential to achieve this outcome.
|To what extent do you agree that integrated care has the potential to achieve the following outcomes?|
|n = 29–31/row percents (%)||Strongly agree*||Agree*||Neutral||Disagree||Strongly disagree||Not Relevant|
|Primary care will become the central focus and coordinating mechanism of healthcare.||32||35||16||10||3||3|
|Local needs will be better met.||19||55||16||6||3||0|
|The patient experience of care will be improved.||19||52||16||6||6||0|
|Inter-professional communication, satisfaction and relationships are improved.||17||53||17||3||10||0|
|Those in the health workforce will work more at the top of their scope of practice.||17||50||13||10||10||0|
|Health literacy and patient self-care will improve.||17||31||31||14||7||0|
|The health status of prioritised groups will improve at a faster rate.||14||41||24||14||7||0|
|Demand for hospital services will be reduced.||10||39||16||23||13||0|
|Health disparities for Māori and Pasifika will be reduced.||10||30||37||7||13||3|
|To what extent do you agree that integrated care has the potential to achieve the following outcomes?|
|n = 43–45/row percents (%)||Strongly Agree*||Agree||Neutral||Disagree||Strongly Disagree||Not Relevant|
|Local needs will be better met.||43||45||9||0||2||0|
|Primary care will become the central focus and coordinating mechanism of healthcare.||38||42||11||7||2||0|
|The health status of prioritised groups (for example those with chronic conditions) will improve at a faster rate.||33||35||30||0||2||0|
|The patient experience of care will be improved.||29||38||31||0||2||0|
|Inter-professional communication, satisfaction and relationships are improved.||29||53||13||0||2||2|
|Health literacy and patient self-care will improve.||25||41||30||2||2||0|
|Those in the health workforce will work more at the top of their scope of practice.||25||45||18||5||5||2|
|Demand for hospital services will be reduced.||20||41||32||5||2||0|
|Health disparities for Māori and Pasifika will be reduced.||20||36||34||0||5||5|
When participants from general practices provided comments about the concept of integration as whole, perspectives ranged from those who could clearly see the opportunities to achieve better outcomes for patients by being a more active hub of care, to those who had uncertainties over how a better working day could also be achieved within these new expectations. General practice respondents noted workload pressures as the most significant challenge to integrated working (50% strongly agreed and 26% agreed).
Despite workloads in general practice being noted as a significant challenge to integrated care, 97% of the general practices surveyed reported being involved in the At Risk Individuals model of care. Unlike the high percentage of general practices who noted workload pressures as the most significant challenge, local care organisations noted three joint challenges to working more closely together in their local area: (1) systems for sharing data across organisations, (2) differences in how primary, secondary and social care are funded, and (3) awareness of services available (for all three 81% strongly agreed or agreed).
This section examines the extent to which a shift towards proactive primary care was occurring based on focus groups in a purposive sample of practices. The At Risk Individuals model of care was implemented and managed by each of the five Primary Health Organisations with the support of the District Health Board, in the expectation that the plans and goals of each locality would be taken into account.
The focus groups revealed that those practices of whatever size that did not hold regular meetings to reflect on what had been learnt from implementing the At Risk Individuals model of care, that tried to absorb the model into their current style of operating and that left nurses to implement it in isolation, were less likely to demonstrate an increased capacity to make connections outside the practice. Those in small practices that tried to absorb the tasks involved in the model into business-as-usual described the struggle involved without having allocated time. For example, those staff who worked in a practice that continued to operate as a walk-in clinic found little support from the rest of the staff “… because it is another extra job in a very busy day”.
Another medium-size practice also reported struggling with incorporating the work into their normal routine, explaining that while they had training to highlight the difference between the At Risk Individuals model of care and an earlier chronic care management programme:
“… when you try and put the training in practice you just couldn’t as there was not time and you just couldn’t allow an hour every week to see a patient for six weeks and spend that time to get the goals and achieve them.” (Medium-size practice participant)
Other practices that had absorbed the At Risk Individuals model of care into business-as-usual were more likely to agree they were resourced for the time involved, often because they had already been doing this work and were now being reimbursed. What characterised these three practices, despite their small size, was the strength of team dynamics they could draw on, built from past experiences delivering care to those with long-term health conditions.
Discussions in one small practice noted that it could be logistically difficult to pass on patients to the one-to-two nurses in the practice whose time was increasingly booked up. Nevertheless, the practice found a way to split responsibilities between the general practitioners who initiated the process and the nurses who followed up, agree specific goals, and develop a care plan. As they explained, “… when it started it was one of the huge issues in our head where were we going to get the time to do all this but it has worked”, [as now]:
“As a doctor I can say OK book 30 minutes with me and I spend that time with the patient and then pass on to the practice nurse … it doesn’t reduce the time I spend having the nurse involved but it increases the quality of what patients receive” (Small-size practice participant).
Larger practices reported adapting to the At Risk Individuals model of care demands by appointing nurse leads, sharing caseloads across nurses and experimenting with different ways of “pulling nurses off the floor” to cover the extra time involved. These staff were also able to draw on a team-based approach to delivering care in partnership with the doctors in the practice. While the At Risk Individuals model of care did involve more work for the nurses, with the right team dynamic they were able to argue for changes to the way the practice allocated time:
“When it first started we were all lost … we did not know what to do … but now as we are going ahead it is more work for us but we are getting more time allocated to do that work as well. We are getting set times. As it is going when we do need more time we are speaking out and we are getting an allocation” (Large-size practice participant).
As being part of a new locality network was expected to influence practices to work in a more planned proactive way for those with long-term conditions, the focus group results revealed different locality contexts had much less influence than the team dynamic within each practice. Those practices prepared to change their organisational processes in order to support nurses to confidently take on new responsibilities for those with long-term conditions, were those most likely to embrace the opportunities to work in a planned proactive way.
The changes put in place by the Localities initiative were collectively expected to shift the balance of care from the hospital to the community by better integrating services between primary and secondary care, moving from a more reactive to proactive model of care and encouraging the provision of more services in primary care settings. However, an assessment of the high-level trends in secondary care demand across the Auckland region found no evidence of change that could be confidently attributed back to the Localities initiative.
Confirming that integrated care “is not a fixed thing but instead is a fluid state that requires constant amending and adapting” , the Localities initiative adapted and evolved and required corresponding shifts in the initiative’s evaluation. Between 2012 and 2015, the Localities initiative shifted from an emphasis on four new local networks charged with reshaping services across a local population to four new networks operating as an “enabler of change” and an “incubator of integration”. Rather than charging new networks with clinical governance decisions, time was spent in local peer-to-peer networks discussing improvements for individual patients and adapting to a new model of care for those with long-term conditions. This shift had implications for how the success of the overall initiative was judged, and the type of new behaviours and engagement expected from different groups of health professionals.
The original Strategic Locality Partnership Agreement was yoked to an ambitious target of a 20% reduction in the projected standardised use of acute medical in-patient bed days and emergency department services over five years. When the Strategic Locality Partnership Agreement was reshaped into a series of enabling activities across the four localities, the policy target remained. However, using the theory-driven realist approach the evaluation sought to assess progress through the intermediate measures of: (i) whether primary care practices were working more proactively for those with long-term conditions; and (ii) the ways in which relationships were improving in order to deliver more integrated care. It is worth reflecting that this smaller scale of experimentation involving a series of enabling activities was forced on the District Health Board by the different power dynamics between the state (in this case the District Health Board) and the non-state (the Primary Health Organisations).
This experience adds weight to broader calls to acknowledge the power dynamics within integrated care  and the call for deeper understanding of the facets involved in governing between organisations (collaborative governance) as opposed to within organisations . The fact that the localities’ geographic boundaries rarely matched the boundaries in place around the enrolled population for each Primary Health Organisation, further complicated governance arrangements. Concerns over how the risk/gain share contract would acknowledge the different contributions across five Primary Health Organisations differentially linked to four local communities proved a major stumbling block. In a similar vein, Accountable Care Organisations in the United States are also co-located in geographic areas where other Accountable Care Organisations exist, creating a potential free rider problem if benefit is given to those who improve the health for all people in a given population .
From the beginning, the Localities initiative encompassed a bundle of interventions. Box One outlined ten original propositions of how the initiative was supposed to work. These propositions included both actions that reshaped services for different local populations and actions that improved the coordination of care for individual patients. As time progressed, the Localities initiative became distinguished by the attention paid to the introduction of a new case management programme (i.e. the At Risk Individuals model of care) and the opportunities to build new working relationships. This finding suggests that of all the propositions in play the ones gaining the most traction were those aiming for improvements for individual patients. Smith’s 2011 analysis of New Zealand’s Primary Health Care directions pointed to unresolved tension between population and patient perspectives within New Zealand primary care. Her assessment was that reform emphasis was placed on the population health gains expected from capitated funding, and less management and policy attention was paid to the development of more integrated primary care services for individual patients . Adaptions of the Localities initiative provide some evidence of an orientation back towards improvements for individual patients.
Multi-component complex interventions can vary in their delivery and be vulnerable to one or more components not being implemented as originally intended . Recent work summing up the emerging achievements of new vanguard models of care in England draws attention to the benefits of building relationships before new ways of working are formalized through organisational or contractual changes . In the Localities case, the move towards the benefits of strengthening local relationships arose when contractual arrangements stalled. Realist evaluative approaches offer a logic that helps unpick the complexity of the relationships and politics in play  and uncover the assumptions made by those developing, implementing and assessing health service changes . In the case of the Localities evaluation, building an understanding of the behavioural landscape of professionals , offered value, despite the major shifts in what was happening to drive change. The key recommendations from the evaluation were to:
The strength of this evaluation centres on the application of the realist approach. Recognising that the Localities initiative is a social programme involving human decisions and actions, attention was paid to identifying how change is expected to occur, and then testing the extent to which these theories were borne out. A recent realist evaluation of two integrated care initiatives tried to retrospectively identify the context, mechanism and outcomes in play but struggled to identify the relationships between them . This evaluation has made greater progress in surfacing testable theories of change, but confidence in the findings needs to be tempered by the partial nature of the evidence collected with respect to (i) the survey response rates and (ii) the sheer diversity of indicators where we might expect to see change. The smallness of the sample size for each locality may also have masked some local differences.
The Localities initiative demonstrated an agility in adapting to the complex web of incentives around primary health care in New Zealand, but the consequence was that attributing what was done to earlier defined metrics of success was difficult. The realist logic of enquiry regards programmes as “theories incarnate” . Surfacing these incarnate theories early helped identify which components of the Localities initiative were rising or falling in importance and was vital to ensuring the evaluation evolved to concentrate on where results could add the most value.
1In risk/gain share contracts, gains and losses are calculated according to the expected costs of delivering care to a defined population with any difference in the actual costs being shared with the contracted health providers. The assumption is that any gain in costs that has occurred comes from efforts by the provider to keep their population healthy. In the Localities initiative, the expectation was that funds that would have been used for extra hospital staff/resources would now be invested in primary and community settings to buy additional services.
Thanks are owed to staff from Counties Manukau District Health Board, and in particular Dr Kathryn DeLuc, who greatly assisted this research. We are also immensely grateful to Dr Claire O’Loughlin who provided invaluable editorial support for this paper.
Peter Crampton, Professor of Public Health, Kōhatu, Centre for Hauora Māori, University of Otago, New Zealand.
Alexandra Hulst, PhD, LMFT, Integrated Behavioral Health Advisor, Rocky Mountain Health Plans, Colorado, USA.
HR is employed by the organisation delivering the Localities initiative, and had direct involvement in aspects of the At Risk Individuals programme. The other authors – LM, MP and JC – were commissioned to deliver an independent evaluation funded by Counties Manukau District Health Board. Approval was sought for the publication of these results as part of the contract for services.
Rea, H, Kenealy, T, Wellingham, J, Moffitt, A, Sinclair, G, McAuley, S, Goodman, M and Arcus, K. Chronic care management evolves towards integrated care in Counties Manukau, New Zealand. The New Zealand Medical Journal (Online), 2007; 120(1252). http://www.nzma.org.nz/journal/120-1252/2489/.
Bardsley, M, Steventon, A, Smith, J and Dixon, J. Evaluating integrated and community-based care: How do we know what works? Research report. London: Nuffield Trust; 2013. https://www.nuffieldtrust.org.uk/research/evaluating-integrated-and-community-based-care-how-do-we-know-what-works.
Goodwin, N. Understanding integrated care. International Journal of Integrated Care, 2016; 16(4): 6. DOI: https://doi.org/10.5334/ijic.2530
Damery, S, Flanagan, S and Combes, G. Does integrated care reduce hospital activity for patients with chronic diseases? An umbrella review of systematic reviews. BMJ Open, 2016; 6(11): e011952. DOI: https://doi.org/10.1136/bmjopen-2016-011952
Imison, C, Curry, N, Holder, H, Castle-Clarke, S, Nimmons, D, Appleby, J, et al. Shifting the balance of care: Great expectations. Research report. London: Nuffield Trust; 2017. https://www.nuffieldtrust.org.uk/research/shifting-the-balance-of-care-great-expectations.
Curry, N and Ham, C. Clinical and service integration: The route to improved outcomes. London: The King’s Fund; 2010. https://www.kingsfund.org.uk/publications/clinical-and-service-integration.
Davy, C, Bleasel, J, Liu, H, Chan, MT, Ponniah, S and Brown, A. Effectiveness of chronic care models: Opportunities for improving healthcare practice and health outcomes: A systematic review. BMC Health Services Research, 2015; 15(1): 194. DOI: https://doi.org/10.1186/s12913-015-0854-8
Goodwin, N, Dixon, A, Anderson, G and Wodchis, W. Providing integrated care for older people with complex needs: Lessons from seven international case studies. London: The King’s Fund; 2014. https://www.kingsfund.org.uk/publications/providing-integrated-care-older-people-complex-needs.
Starfield, B. Is primary care essential? Lancet, 1994; 344(8930): 1129–33. DOI: https://doi.org/10.1016/S0140-6736(94)90634-3
Starfield, B, Shi, L and Macinko, J. Contribution of primary care to health systems and health. The Milbank Quarterly, 2005; 83(3): 457–502. DOI: https://doi.org/10.1111/j.1468-0009.2005.00409.x
Wagner, EH, Glasgow, RE, Davis, C, Bonomi, AE, Provost, L, McCulloch, D, et al. Quality improvement in chronic illness care: A collaborative approach. The Joint Commission Journal on Quality and Patient Safety, 2001; 27(2): 63–80. DOI: https://doi.org/10.1016/S1070-3241(01)27007-2
Addicott, R. Commissioning and contracting for integrated care. London: The King’s Fund; 2014. https://www.kingsfund.org.uk/sites/default/files/kings-fund-commissioning-contracting-integrated-care-nov14.pdf.
Song, Z, Safran, DG, Landon, BE, Landrum, MB, He, Y, Mechanic, RE, et al. The ‘Alternative Quality Contract,’ based on a global budget, lowered medical spending and improved quality. Health Affairs, 2012; 31(8): 1885–1894. DOI: https://doi.org/10.1377/hlthaff.2012.0327
Flodgren, G, Eccles, MP, Shepperd, S, Scott, A, Parmelli, E and Beyer, FR. An overview of reviews evaluating the effectiveness of financial incentives in changing healthcare professional behaviours and patient outcomes. Cochrane Database of Systematic Reviews, 2011; 7, CD009255. DOI: https://doi.org/10.1002/14651858.CD009255
Pawson, R. Evidence-based policy: A realist perspective. London: Sage Publications; 2006. DOI: https://doi.org/10.4135/9781849209120
Provan, KG and Kenis, P. Modes of network governance: Structure, management, and effectiveness. Journal of Public Administration Research and Theory, 2008; 18(2): 229–252. DOI: https://doi.org/10.1093/jopart/mum015
De Silva, D. Helping people help themselves: A review of the evidence considering whether it is worthwhile to support self-management. London: The Health Foundation; 2011. https://www.health.org.uk/sites/health/files/HelpingPeopleHelpThemselves.pdf.
Pawson, R. The science of evaluation: A realist manifesto. London: Sage; 2013. DOI: https://doi.org/10.4135/9781473913820
McDermott, I, Checkland, K, Coleman, A, Osipovic, D, Petsoulas, C and Perkins, N. Engaging GPs in commissioning: Realist evaluation of the early experiences of Clinical Commissioning Groups in the English NHS. Journal of Health Services Research & Policy, 2017; 22(1): 4–11. DOI: https://doi.org/10.1177/1355819616648352
Rea, HH, Kenealy, TW, Rolland, T-M, Naumann, C, Hou, T and Sheridan, N. Enabling Integrated Health and Care. International Journal of Integrated Care, 2016; 16(6): A126, 1–8. DOI: https://doi.org/10.5334/ijic.2674
Battersby, M, Harris, M, Smith, D, Reed, R and Woodman, R. A pragmatic randomized controlled trial of the Flinders Program of chronic condition management in community health care services. Patient Education and Counseling, 2015; 98(11): 1367–75. https://www.sciencedirect.com/science/article/pii/S0738399115002761?via%3Dihub. DOI: https://doi.org/10.1016/j.pec.2015.06.003
Penm, J, MacKinnon, NJ, Strakowski, SM, Ying, J and Doty, MM. Minding the gap: Factors associated with primary care coordination of adults in 11 countries. Annals of Family Medicine, 2017; 15(2): 113–19. http://www.annfammed.org/content/15/2/113.full. DOI: https://doi.org/10.1370/afm.2028
Manzano, A. The craft of interviewing in realist evaluation. Evaluation, 2016; 22(3): 342–360. DOI: https://doi.org/10.1177/1356389016638615
RAMSES II Project Team. Developing realist theories: The RAMSES II project. NIHR; 2017. http://www.ramesesproject.org/media/RAMESES_II_Developing_realist_programme_theories.pdf.
Wistow, G, Gaskins, M, Holder, H and Smith J. Putting integrated care into practice: The North West London experience. London: Nuffield Trust and London School of Economics and Political Science: Personal Social Services Research Unit; 2015. https://www.nuffieldtrust.org.uk/research/putting-integrated-care-into-practice-the-north-west-london-experience.
Brady, TJ, Murphy, L, O’Colmain, BJ, Beauchesne, D, Daniels, B, Greenberg, M, et al. A meta-analysis of health status, health behaviors, and health care utilization outcomes of the Chronic Disease Self-Management Program. Preventing Chronic Disease, 2013; 10. DOI: https://doi.org/10.5888/pcd10.120112
Harvey, J, Dopson, S, McManus, RJ and Powell, J. Factors influencing the adoption of self-management solutions: An interpretive synthesis of the literature on stakeholder experiences. Implementation Science, 2015; 10(1): 159. DOI: https://doi.org/10.1186/s13012-015-0350-x
Kennedy, A, Bower, P, Reeves, D, Blakeman, T, Bowen, R, Chew-Graham, C, et al. Implementation of self management support for long term conditions in routine primary care settings: Cluster randomised controlled trial. BMJ, 2013; 346: f2882. DOI: https://doi.org/10.1136/bmj.f2882
Nolte, S and Osborne, RH. A systematic review of outcomes of chronic disease self-management interventions. Quality of Life Research, 2013; 22(7): 1805–1816. DOI: https://doi.org/10.1007/s11136-012-0302-8
Packer, TL, Boldy, D, Ghahari, S, Melling, L, Parsons, R and Osborne, RH. Self-management programs conducted within a practice setting: Who participates, who benefits and what can be learned? Patient Education and Counseling, 2012; 87(1): 93–100. DOI: https://doi.org/10.1016/j.pec.2011.09.007
Panagioti, M, Richardson, G, Small, N, Murray, E, Rogers, A, Kennedy, A, et al. Self-management support interventions to reduce health care utilisation without compromising outcomes: A systematic review and meta-analysis. BMC Health Services Research, 2014; 14(1): 356. DOI: https://doi.org/10.1186/1472-6963-14-356
Janse van Rensburg, A, Rau, A, Fourie, P and Bracke, P. Power and integrated health care: Shifting from governance to governmentality. International Journal of Integrated Care, 2016; 16(3): 17. DOI: https://doi.org/10.5334/ijic.2480
Minkman, MMN. Longing for integrated care: The importance of effective governance. International Journal of Integrated Care, 2017; 17(4): 10. DOI: https://doi.org/10.5334/ijic.3510
Casalino, LP, Erb, N, Joshi, MS and Shortell, SM. Accountable care organizations and population health organizations. Journal of Health Politics, Policy and Law, 2015; 40(4): 821–837. DOI: https://doi.org/10.1215/03616878-3150074
Craig, P, Dieppe, P, Macintyre, S, Michie, S, Nazareth, I and Petticrew, M. Developing and evaluating complex interventions: The new Medical Research Council guidance. BMJ, 2008; 337: a1655. DOI: https://doi.org/10.1136/bmj.a1655
Naylor, C, Curry, N, Holder, H, Ross, S and Marshall, L. Clinical commissioning groups: Supporting improvement in general practice. Research report. London: Nuffield Trust and King’s Fund; 2013. https://www.nuffieldtrust.org.uk/research/clinical-commissioning-groups-supporting-improvement-in-general-practice. DOI: https://doi.org/10.1016/j.socscimed.2018.04.006
Dainty, KN, Golden, BR, Hannam, R, Webster, F, Browne, G, Mittmann, N, et al. A realist evaluation of value-based care delivery in home care: The influence of actors, autonomy and accountability. Social Science & Medicine, 2018; 206: 100–109. DOI: https://doi.org/10.1016/j.socscimed.2018.04.006
Kaehne, A. Integration as a scientific paradigm. Journal of Integrated Care, 2017; 25(4): 271–279 DOI: https://doi.org/10.1108/JICA-07-2017-0023
Khaldi, A. A question of behaviours: Why delivering care integration and managing acute demand depends as much on changing behaviours as new systems and structures. London: iMPOWER Consulting Ltd; 2015.
Busetto, L, Luijkx, K and Vrijhoef, HJM. Development of the COMIC Model for the comprehensive evaluation of integrated care interventions. International Journal of Care Coordination, 2016; 19(1–2): 47–58. http://journals.sagepub.com/doi/abs/10.1177/2053434516661700. DOI: https://doi.org/10.1177/2053434516661700