Skip to main content

Table 1 Steps in the realist synthesis – a summary of the study approach

From: Modelling successful primary care for multimorbidity: a realist synthesis of successes and failures in concurrent learning and healthcare delivery

Step

Summary of approach

1 Focusing the review

The social interactions between GPs, patients and trainees where chosen as a focus within the context of multimorbidity in primary care because (i) multimorbidity is an increasing clinical and educational challenge, (ii) workplace-based learning has to occur concurrently with healthcare delivery to ensure future doctors are equipped to meet the needs of patients with multimorbidity

2 Developing a theory:

(a) Initial rough theory – we theorised that as social interactions are known to shape learning (meaning-making and knowledge construction) it was likely that social mechanisms influenced concepts of success and failure in the absence of cure and hence understanding this was essential to understanding mechanisms that would lead to relationship-centred needs-based learning and care delivery

(b) Review of evidence – an extensive systematically conducted database search with citation follow-up was conducted as described in this paper and our protocol.

(c) Refined theory – the model presented below represents the mechanisms which, if triggered, are most likely to lead to constructive transformations and learning for GPs, patients and trainees

3 Search strategies:

These are detailed fully in the protocol, with the Medline search provided in additional file 2 as an exemplar.

4 Selection and appraisal of documents

As described in the main text citations were selected according to relevance and rigour.

5 Applying realist principles in analysis

The data extraction sheet provided a framework for ensure that data was pulled from each citation to inform understanding of social interactions, complexity, concurrency, success or failure in multimorbidity, learning and service provision (version 1 and 2 – a more focused version for later rounds of data extraction can be found in additional files 4 & 5). The use of this framework which incorporated the elements of the VICTORE model [8] as well as members of the study team being required to keep note of how each citation informed ‘what works , for whom, in what circumstances and why?’ with respect to the study aims ensured that data extraction included the seeking of explanations (how in a given context did a mechanism generate an outcome), comparison of interventions, aligning evidence to theory taking a bi-directional approach (allowing the evidence to refine our theory as well as theoretically informed searching for evidence) and iterative development of the proposed model

  1. Further details on the design of the review can be found in our published protocol.