Medical Record Review: Methodological Considerations

Purpose Medical chart or record review (MRR) is commonly used in epidemiological studies. However, there is a dearth of literature summarizing the most methodologically rigorous way to conduct this type of retrospective research. The purpose of this paper is to present a) guidelines concerning the development and implementation of a MRR based on a thorough literature review and b) methodological issues related to MRR within the context of a work-related traumatic brain injury (wrTBI) surveillance and prevention study involving MRR. Methods A comprehensive literature review regarding MRR methodological considerations was conducted. Key words including medical, chart, record, retrospective, review, method(s) and methodology were used within various databases such as Ovid, CINAHL, scholars-portal, PubMed, and Medline. The principles obtained from the literature were applied to a MRR reliability study concerning workers who sustained work related injuries, primarily wrTBI, and who were assessed at a multi-disciplinary neurology service in Ontario, Canada within the years 1998 & 1999. This MRR was designed to abstract variables related to pre-injury and incident related person, occupation and environment factors, and post-injury sequelae related to wrTBI to inform injury prevention strategies. Results The authors developed and utilized a model of MRR methodology, the MRR-Conduction Model, which outlines the relationships between different entities involved in a MRR. These entities include the investigator, the data abstractor(s), the data source, the abstraction tool, and the abstraction manual. Within each entity, guidelines were found that aided in creating a more feasible, effective and reliable MRR. These guidelines and their application to improved MRR reliability will be presented. Conclusion MRR is often perceived as an easy and potentially expedient data collection method however, creating an effective MRR takes planning and time. The MRR-Conduction Model facilitates understanding of the entities and relationships involved in using MRR methodologies.