Can complexity dynamics be harnessed to improve integration of care? The implementation of the Health Links in Ontario, Canada

Abstract

Introduction: Attempts to understand the emergence of integrated care using traditional approaches provide limited insights. Increasingly, scholars are applying complex adaptive systems (CAS) theory to explain the implementation of integrated care policies and interventions. A CAS perspective suggests that integrated care efforts designed to support sensemaking, which allows staff to give meaning to their experience and efforts, self-organizing, and adaptive approaches are more likely to be successful than efforts which are more ‘mechanistic’ and ‘linear’. We examined the implementation of the ‘Health Links’ (HLs), a provincial integrated care initiative in Ontario, Canada. Launched in 2012, there are currently 82 operational HLs, consisting of voluntary partnerships among organizations aimed at improving care for the top 5% of health system users in terms of costs. The HLs were introduced in a ‘low rules’ policy framework to stimulate grass-roots networks and approaches. In this study, we explored the extent to which this approach has generated interconnections, sensemaking, self-organization, emergence and coevolution, key aspects of CAS. Theory/Methods: We used key terms from CAS theory to deductively code transcripts from 37 interviews conducted in 2014/2015 with managers and clinicians from 24 HLs in all 14 regional health authorities in Ontario. Two reviewers coded a subset of interview data, in duplicate, that had been coded as addressing CAS themes in an earlier study. Results: Health Links generated novel interconnections between organizations and professionals across health and social services. These interconnections stimulated sensemaking, as individuals came together to understand, interpret, and adapt the initiative within their own contexts. Several self-organizing processes were identified as staff designed, tested and refined their structures and processes, and at the regional level where some regions standardized HLs within their borders. Despite enthusiasm for the ‘low-rules’ environment, some participants struggled with the lack of guidance or consistency from the province, and all agreed that some standardization was necessary. Early signs of emergence and co-evolution were identified, however, many HLs were still in early stages of implementation. Discussions: Our results show that a flexible policy environment encouraged interconnections, sensemaking, and self-organization among actors in the system. However, there were different capacities and preferences for self-organizing, and limits to what could be achieved through self-organization. Successful implementation required leadership and some operational consistency across HLs. Barriers to self-organization persisted in the form of pre-existing organizational and regional regulations. Conclusions: A ‘low-rules’ integrated care policy can generate novel interconnections, adaptive sensemaking, and productive self-organization among actors. Lessons learned: Operationalizing CAS theory domains is challenging and requires clarification of definitions and the level of analysis. Limitations: As a qualitative study, the results are limited in their generalizability. Interviews were conducted within three years of HLs launch and are limited to the initial implementation. Suggestions for future research: Future research should develop clear and applied definitions of CAS domains to support empirical studies of complexity dynamics in integrated care.