A total of 7656 papers were identified including 6426 from electronic databases, 1118 via Opengrey and 112 from websites and experts. After title and abstract screening, 809 articles underwent full-text screening, which resulted in a final 84 papers being included in the review. A flowchart of the study selection process, including the reasons for exclusion, is shown in Fig. 2.
Characteristics of included studies
The 84 included articles represented multiple locations across England, including regions in the south-west, northeast, north-west and Greater London. These were from a range of sectors; primary care, secondary care, social care, voluntary sectors, local government, local authority and public health. The most common sector was primary care (32%) and most frequent study design was qualitative (20%). There was substantial heterogeneity that included mixed-methods (14%), analysis/commentaries (14%), systematic/scoping/evidence reviews (12%), randomised controlled trials (11%), policy documents (8%), quantitative studies (8%), thesis (5%), editorials (2%) and books/book reviews (2%). Included articles were published between 1996 and 2020. The characteristics of included articles have been summarised in Fig. 3.
Level of integration
We counted the number of studies that considered integration at each level as set out by our conceptual framework (i.e. macro, meso or micro-level). There were 7% of studies that considered integration at the macro-level, 5% at meso-level and 30% at micro-level. Thirty-five per cent considered integration at all three levels. However, the combined number of studies focused on either one or both of micro and meso levels was 52%.
Summary of themes amongst included studies
We identified three themes from the analysis to summarise current research and progress on integrated primary care and social services for older adults with multimorbidity in England: (1) diverse focus on multi-level vs. multi-sector integration; (2) time needed for integration to embed; and (3) seeking structural integration while applying local flexibility. Each of these is described in turn below.
1. Multi-level and multi-sector integration
Several articles described previous research and clinical provision in primary care or social services for older adults with multimorbidity in England.[18, 19, 20 21]. These were often concerned with particular sectors (e.g. primary care) or scales of integration (e.g. clinical level), rather than whole-systems, as described by our conceptual framework.[22, 23, 24, 25, 26, 27, 28]. Studies focused on improving specific dimensions of integration such as leadership,[29, 30, 18] care models[23, 31, 32, 33, 34] or considered integrated working from the perspective of one or two levels of integration,[35, 36, 37, 38] most frequently the micro-scale or the micro/meso-scales together. Studies from 1996 onwards repeatedly stressed the need for more multi-level, systemic and comprehensive integration, [39, 40] although we found limited evidence of significant progress in achieving this ambition over the last two decades. A prevalent theme was the urgent requirement to mitigate or remove long-standing barriers such as incompatible record sharing systems and inadequate information sharing processes between sectors [41, 42, 43, 44, 45]; ‘siloed’ thinking in service provider organisations [46, 47, 48]; poor communication among health and social care professionals, both internally within their organisations and across sectoral boundaries.[49, 50, 51] There was an increasing emphasis on the need to tackle wider determinants of population health with suggestions that to achieve this, it is necessary to go beyond primary care and social services to include hospitals, GP community services, voluntary sectors and local government partners.[6, 7, 13, 34, 50] We observed a growing recognition in more recent literature that improving clinical care in one or two sectors may not be able to be as effective as simultaneously improving the organisation or design across services as one system of provision.[45] Solutions that were proposed emphasised the need for system-wide leadership across all scales, alongside a shared vision of integrated working across sectors.[52, 53, 54, 55] There was evidence highlighting the importance of the quality and style of organisational leadership, both in terms of delivering change and maintaining an integrated approach to service delivery.[12,29,56. We found very few examples of where this approach had led to individual and local successes, and widespread evaluation and evidence of application was very limited.[49, 57, 36, 33]
2. Time for integration to embed
A number of studies highlighted that integration requires time to allow new structures and relationships to develop and bed-in. Integrated care programmes take years to establish and need sufficient time to allow new care models to fully mature.[20, 58, 59, 60] Effective and enduring integration is ‘the result of a long-term process, facilitated by key local leaders, during which the capability and legitimacy of new ways of working is built up over time’.[6] The King's Fund report of the Vanguards concluded that the most successful models of integrated care are built on ‘trusting relationships and collaborative organisational cultures, often developed over time,’ which enable ‘clinical teams as well as key organisational leaders to work together effectively’ - where ‘success’ was defined in terms of perceptions of process.[31] Some studies suggest that the answer may lie in persistence and perseverance over several years to enable integrated care programmes to achieve their ‘objectives and become self-sustaining.’[53, 61, 62] This appeared to be influenced by the sustained commitment of key partners and the ‘longevity of the senior leadership.’[57] The challenge in the next phase of integrated care reform is ‘building clinical collaboration and system leadership in a statutory context’ that is ‘not designed for this purpose’,[31] alongside policymakers providing the necessary time for integrated care programmes to ‘evolve and mature’,[63] rather than moving onto the next new policy initiative.
3. Structure with flexibility
We identified inherent tensions between top-down and bottom-up driven approaches to integrated care, in particular, having in place a single comprehensive ‘whole-systems’ structure combined with local flexibilities. Studies suggested that integration should be implemented within a clear framework and a set of higher-level principles that allows for both macro-level systems-wide strategic management and oversight, combined with local autonomy and flexibility, described as ‘structured flexibility’.[64, 65, 66, 20 67] The benefit of holistic systems-wide approaches is that they ‘tend to be more strategic with clearer paths for scaling up, compared to ‘bottom-up’ approaches driven by highly motivated individuals at the micro-level.’[56] Nevertheless, a whole-systems strategy requires a twin-track approach [55], with ‘leadership from the bottom up’ driven by staff who are ‘empowered to integrate services where they see the need.’[53] Mechanisms for horizontal integration (mechanisms, structures and practices that connect care across the same level in the system)[5] were also seen as essential ‘at each organisational level (for example whole systems, community and individual levels). Vertical mechanisms (mechanisms, structures and practices which link together services up and down the different scales of the system) are also necessary ‘to integrate the various levels.’[39, 44] Successful examples of integrated care in the NHS indicate that when this is ‘pursued at all levels’, it could ‘overcome the risks of fragmentation, and of ‘service users falling between the cracks’ of care.[68] Critically however, the studies included in the review suggested that any programme of integrated care must be based on an understanding that ‘as barriers to integration are systemic in organisations designed for separation rather than integration and the historic paradigms of building bridges and tearing down walls is inherently flawed, and of limited effectiveness: a better metaphor is one of weaving integration into the fabric of organisational life.’[39]