Study design of principal analysis: case-case-time-control. Two concomitant case-crossover (CCO) analyses were conducted: 1 among the “cases” and 1 among “future cases.” Cases were patients with VOE (stripped bar, index date) during the study period. Corticosteroids exposure is expected to vary by seasonal periods, resulting in a variable probability of exposure over time. To best control for this exposure trend, using an adjustment cohort with a probability of exposure similar to cases is recommended. This adjustment cohort is made of “future cases.” They were patients with VOE the year after the index date (VOE of the case), matched with cases on age, sex, and residence place. In each CCO analysis, the frequency of exposure during the case period (28 days before VOE) was compared with the frequency of exposure during the control period (also with a duration of 28 days, separated from the case period by a 28-day washout period). The ratio (OR in cases/OR in future cases) gives the OR associated with the effect of the exposure on the event adjusted for the exposure trend.