Abstract
In Reply.—We are glad Dr Gumprecht appreciates the need for postmarketing surveillance studies, and we appreciate the opportunity to respond to his concerns. Unfortunately, we may not have communicated as clearly as we could have, leading to some misinterpretations.First, our study did not show a low risk of operative site bleeding associated with this drug, but in fact a high risk (39.6%; Table 1). However, this risk was not significantly higher than that in a roughly comparable control group (38.6%); even after adjusting for possible confounding, the relative risk was not elevated. A higher risk was observed as the dose of ketorolac increased (P<.001) and in the elderly (P<.02). In contrast, the risk of clinically serious operative site bleeding was much lower (1.55% vs 1.8%, respectively).Second, the study1 to which Gumprecht refers was not a double-blind trial—it was a retrospective cohort study like ours and less