Background: effectiveness reviews of complex interventions may not provide conclusive answers. Evidence syntheses are complicated by heterogeneity in interventions, outcomes and who receives the interventions, under what circumstances. As an example, an existing Cochrane effectiveness review shows that advocacy interventions for domestic violence may help abused women to stop or reduce repeat victimisation and overcome its consequences. Yet the review cannot make clear recommendations for practice.
Realist reviews are increasingly used for a deeper understanding of the mechanisms underlying such interventions and what works for whom and under what contexts. We thus worked with the Cochrane Developmental, Psychosocial and Learning Problems Group on an exemplar realist review of advocacy interventions for abused women, to augment an existing effectiveness review.
Objectives: our aims were 1) to develop an exemplar for Cochrane realist reviews; 2) to develop a programme theory for advocacy interventions for abused women, encapsulating what mechanisms work, where, in what circumstances, for whom and how.
Methods: we undertook three main search phases: scoping for candidate theories; iterative recursive searches to deepen and fill gaps in our understanding of context-mechanism-outcome (CMO) combinations and develop a rounded programme theory; systematic search for studies to test, confirm or refute our developed theory. We searched 13 databases and relevant websites, using the CLUSTER approach and did not exclude any study type.
Results: we extracted data from 139 articles (95 studies). Following synthesis, and review by a stakeholder panel, we constructed a conceptual model from eight core principles based on study CMO combination patterns, which we will present. There was panel consensus on the importance of considering women’s vulnerabilities and intersectionalities, and the trade-offs of abuse-related decisions in the contexts of the individual women’s lives (e.g. their dependencies on their abusers or their community, diversity-related factors). There were links between setting, theory underpinning interventions, type, duration and intensity of advocacy, advocate discipline and outcomes. A good therapeutic alliance was important. Significant challenges for advocates included inter-organisational working, vicarious traumatization, and lack of clarity on how much to support women, given that advocacy is tailored to the individual.
Conclusions: realist reviews offer important insights into complex interventions that complement effectiveness evidence syntheses. This exemplar review highlights the complex way that factors combine and interact for effective advocacy and facilitates clear recommendations for future practice and research.
Patient/healthcare consumer involvement: our stakeholder panel included abused women, who contributed significant changes in our understandings of the nuances of, and wordings used to describe, our theoretical concepts.