Bleeding statistics.

摘要

I read with interest the recent work by Chandra et al and commend them for collecting such a group of patients, but I wish for clarification on a few points. They described the present work as a retrospective case-control study; I feel that this is incorrect. The patients studied were selected on the basis of an exposure of interest, in this case to warfarin pharmacotherapy and matched with respect to age, sex and presenting complaint. The participants were followed (retrospectively) to record outcomes of pars plana vitrectomy as well as presenting features. Therefore, this study describes a retrospective (or historical) matched cohort study, not a retrospective case-control study. When performing the matching was there a reason that other patient-specific variables, both systemic (blood pressure, liver disease, clotting disorders, medication) and ocular, which could potentially influence the outcome were not controlled for?