(138) Comparing Study Designs in Case-Crossover Studies of Chronic Medications Versus Acute Medications in a Systematic Review

Publication Number: 469

Background: The case-crossover design is a case-based method that is used to compare exposure immediately preceding an event occurrence with exposure in earlier control periods. The design is most suitable for acute exposures as a recent simulation study from Kubota et al showed concerns for within-subject exposure dependency when using the case-crossover design for chronic exposures with multiple control windows.

Objectives: Our goal was to compare design and analysis choices in case-crossover designs with chronic versus acute medication exposures in a systematic review.

Methods: We conducted a PubMed search to identify recent case-crossover, case-time-control, and case-case-time-control studies focused on medication exposures. Articles noting these study designs that published between January 2015 and December 2021 in English language were identified. Reviews, methodological studies, commentaries, articles without medications as the exposure of interest, and articles with no available full text were excluded. The remaining articles were reviewed by two independent reviewers. Information including study design, outcome, risk window, control window, sample size calculation, and sensitivity analysis were summarized overall and by acute versus chronic medication.

Results: Of the 914 articles initially identified, 100 articles were eligible for inclusion. From these articles, the case-crossover was the most common (n= 87), followed by the case-time-control (n=18), and case-case-time-control (n= 4). Fifty-five articles included only acute medications, 32 articles included only chronic medications, and 13 articles included both. From 2015 to 2021, the proportion of articles evaluated chronic medications went from 29% to 46%. Overall, 58 articles used a single control window. Among articles included only acute medications or only chronic medications, 58% and 63% of articles used a single control window, respectively. Among the 45 articles that evaluated at least one chronic medication, 35% used multiple control windows. The proportion of articles that used multiple control windows for chronic medication increased from 28% to 63% from 2015 to 2021.

Conclusions: We found an increase in the proportion of studies using a case-crossover design for chronic medications from 2015 through 2021. Importantly, there was a simultaneous increase in the proportion of those studies focusing on chronic medications and using multiple control windows which has the potential to bias the observed effect estimates. Researcher should consider the impact of the potential bias when opting to use as case-crossover design with chronic medication exposures.