On the Use of Sampling Weights for Retrospective Medical Record Reviews

Retrospective medical record review is often used to answer the "why" questions that statistical modeling cannot. In addition to its utility as an explanatory tool, it can be used to generate hypotheses using available retrospective data and so is a convenient guide for developing future prospective studies. A recent review of articles that used the retrospective medical record review method listed 10 best practices that ought to be followed. However, an issue that is not listed is the use of sampling weights, which are important when one can only conduct retrospective medical record review for a sample of the target population. Although that review acknowledged the importance of carefully selecting a sampling strategy for such a scenario and indeed had outlined 3 commonly used sampling methods (convenience, simple random, and systematic), the authors say nothing of the use of sampling information at the data analysis stage. This article aims to fill that gap and to demonstrate why the use of sample weights ought to be another best practice to add to the list by reviewing well-known theoretical details and some published data analysis examples.