A practical guide to conducting and reporting medical record review studies, including instruction and insights on topics ranging from idea conception to manuscript submission, is provided.Summary
Research based on the retrospective collection and analysis of data from medical records is often the first type of research conducted by new investigators; such research can be particularly useful in high-acuity populations and is often more feasible and cost-effective than prospective studies. The research process begins with formulating a valid hypothesis-driven or exploratory research question and refining it into one that is testable; this can be facilitated by incorporating information from various sources (e.g., relevant literature, consultation with experts, observations of practice). Accurate sample size calculations are important in ensuring study feasibility (i.e., the estimated sample sizes required can be obtained). Appropriate data collection methods are particularly crucial in medical record review studies; key considerations include pertinent variables, systematic subject identification, standardized data collection instruments, training of data collectors to minimize interrater variability, processes to ensure the accuracy and reliability of data, and blinded collection of the outcome variable. Checklists are provided to guide investigators in retrospective data collection, manuscript preparation, and decisions regarding authorship eligibility and sequencing.Conclusion
The quality of medical record review studies and the likelihood of their acceptance for publication can be improved through adherence to recommended standards of research design, data collection and analysis, ethical authorship, and manuscript preparation.