Processes and challenges associated with establishing a linked national suicide database across the criminal justice system

INTRODUCTION: There is a wealth of data contained within healthcare and criminal justice system (CJS) datasets that, if successfully linked, could provide more information about this population, particularly those offenders who die in non-custodial CJS settings where, comparatively, much less is known. OBJECTIVES: This study aims to determine feasibility of conducting data linkage across key criminal justice datasets and outline the processes, methodological considerations and any other implications of setting up such a linkage. METHOD: Five CJS datasets were identified for potential inclusion for linkage with Office for National Statistics (ONS) mortality data. Respective data teams were contacted, and scoping discussions were held via email, telephone contact and in person. Information was sought on available data, quality and completeness, unique identifiers, processes for record matching, cost implications, estimated timescales, required approvals, data security considerations and quality of data. RESULTS: All five datasets were deemed important to include and responses from data teams suggest that the proposed linkage is both feasible and valuable, within a reasonable timeframe and with minimal associated costs. The discovery of an additional ‘spine’ dataset provides a more effective method of record matching by linking police identifiers to unique prison and probation identifiers. CONCLUSIONS: The proposed linkage could highlight key points across the criminal justice system at which to target suicide prevention strategies. A more comprehensive linkage, including healthcare services, would further extend the opportunity to target interventions.