Little is known about the role played by emergency departments in cases of partner violence. To find out what happens and what works, for whom and in what circumstances, this research employs a mixed-method case series approach to explore pathways through emergency care. One element of the study is a retrospective medical record case review, which will be used to compare health service management of emergency department contacts arising from partner violence. However, conducting partner violence research in health systems in England is methodologically challenging. This poster focuses on the problems of sample design to ‘find cases’ for a medical record case review of a largely ‘hidden’ population.
Medical record case identifiers or clinical codes for partner violence are not normally applied in emergency department health systems. The poster illustrates the clustered, multi-phase, random sample design which has been developed for the retrospective medical record review of cases of partner violence in emergency departments in Lancashire. There are, however, limitations in the sample design that stem from health data originating from individual pathology (physical injury from assault) and incidence (episodes of pathology), which collectively may obscure prevalence and wider health consequences of partner violence. Furthermore, this may perpetuate bias towards particular characteristics of ‘cases’ in health, limiting the view to extreme partner violence ‘types’.