We thank Drs. Babul, Sloan, and Lipman for their interest in our meta-analysis,1which was primarily performed to study the effects of cyclooxygenase-selective and -nonselective inhibitors on morphine side effects. We agree with them (as we discussed in the article) that some concerns remain about the safety of short-term perioperative prescription of coxibs and that coxibs cannot be administrated to all surgical patients, especially patients with coronary artery disease or at risk of cerebral infarction. Two studies have indeed clearly demonstrated that the use of parecoxib and valdecoxib was associated with an increased risk of arterial thrombotic adverse events in patients scheduled to undergo coronary artery bypass surgery.2,3However, some evidence also suggests that short-term perioperative use of coxibs may offer benefits in comparison with nonsteroidal antiinflammatory drugs. For example, tonsillectomy is one of the most frequently performed ambulatory surgical procedures in children, and nonselective inhibition of cyclooxygenase by nonsteroidal antiinflammatory drugs increases significantly the rate of reoperation4,5but also decreases nausea and vomiting.5In that setting, celecoxib, has been demonstrated to relieve posttonsillectomy pain and to decrease bleeding risk in comparison with nonsteroidal antiinflammatory drugs.6Moreover, the risk of adverse cardiovascular events is extremely low in this population of young patients, as it is in patients devoid of arterial thrombotic pathology scheduled to undergo noncardiac surgery. Therefore, short-term perioperative use of coxibs could have a favorable risk–benefit ratio compared with nonsteroidal antiinflammatory drugs in patients without risk factors for arterial thrombotic events submitted to hemorrhagic surgical procedures.
*Tenon University Hospital, Paris, France. firstname.lastname@example.org