First Things First: How to Elicit the Initial Program Theory for a Realist Evaluation of Complex Integrated Care Programs.

Policy Points Realist evaluation (RE) is an emerging and promising research approach for evaluating integrated care, addressing what works, how, for whom, and in what circumstances. The rich philosophical foundation of RE, critical realism, can help to systematically unravel an integrated care program's initial theory prior to implementation, as a first step within RE. RE can be considered a robust methodological asset in integrated care research by facilitating a deeper level of insight into program functioning than traditional forms of evaluation do and by shaping a realist-informed monitoring and evaluation process. CONTEXT: The complexity of integrated care and the need for transferable evaluation insights ask for a suitable evaluation paradigm. Realist evaluation (RE), underpinned by the philosophy of critical realism, is a theory-driven approach that addresses what works, how, for whom, and in what circumstances. The current study illustrates the process needed for RE's first step: eliciting the initial program theory (IPT). The TARGET program, a Dutch primary care initiative to facilitate more integrated care for chronically ill patients, i.e., care that is efficient, tailored, and holistic,  was taken as a real-world case.
METHODS: An RE approach informed the phased IPT elicitation: (1) identifying an abstract theory framework; (2) formulating the preliminary IPT, building on the abstract theory and informed by previous scientific studies that underpin TARGET; and (3) refining the preliminary IPT, informed by RE expert interviews (n = 7). An RE heuristic tool, specifying the interplay between intervention-context-actors-mechanisms-outcomes (ICAMO) and retroductive reasoning, was applied to synthesize the underlying theory of individual TARGET components into TARGET's IPT.
FINDINGS: Separate but related IPTs were identified for the two main types of actors involved in TARGET: primary care professionals (PCPs) and patients. For both actors, two sorts of mechanisms are assumed to be activated by TARGET, which-via instrumental outcomes-contribute to long-term quadruple aim targets. The first is confidence to enhance PCPs' person-centered conversational skills and to increase patients' active engagement in TARGET. The second is mutual trust, between PCPs and patients and between PCPs and their network partners. A supportive context is assumed crucial for activating these mechanisms-for example, sufficient resources to invest in integrated care.
CONCLUSIONS: Although the IPT elicitation process is time intensive and requires a mind shift, it facilitates a deeper insight into program functioning than accommodated by the prevailing experimental designs in integrated care. Furthermore, the design of a realist-informed evaluation process can be informed by the IPT.