Fletcher A
Dalkin S
Anderson R
Baker RM
Donaldson C
Hibberd V
Kumar MB
Shenton F
Westhorp G
Wong G
Wright J
Bate A
Soc Sci Med. 2026;393119050
BACKGROUND: In the evaluation of complex interventions, economic evaluations aim to determine the relative cost-effectiveness of interventions but generate little explanation of how or why contexts and underlying causal mechanisms impact this. Conversely, realist approaches aim to explain 'what works, for whom, in which circumstances and why' but rarely capture the economic costs and consequences of interventions. As a result, many evaluations remain partial. OBJECTIVE: To identify past attempts to integrate realist and economic evaluation approaches and summarise the recent developments in realist and economic evaluation approaches in the evaluation of complex health and social care interventions. METHODS: We conducted a series of scoping reviews using online academic databases, personal libraries and expert stakeholder workshops, to identify the theoretical, methodological, and practical challenges and developments in bringing together realist and economic evaluation approaches. FINDINGS AND RECOMMENDATIONS: Although increasing, there remain relatively few examples of evaluations that have attempted to integrate realist and economic evaluation approaches, and challenges for their integration mean that further guidance is required. The wider literature indicated challenges in the theoretical (e.g. ontology, causality), methodological (e.g. accounting for context, study design, mixing methods) and practical (e.g. terminology, scale and scope) domains, for which we have developed recommendations. CONCLUSION: To deliver services that are both effective and efficient, evaluations must synthesise relevant explanatory evidence with cost and outcome data to enable policymakers and commissioners to make informed decisions. Findings and recommendations from this review were used to inform the development of guidance for the integration of realist and economic evaluation approaches.
