Confounding and Adjustment Factors in the Effect of Intravenous Parecoxib on Shoulder Pain after Thoracotomy?

Abstract

We read with great interest the article by Pipanmekaporn et al. 1 in a recent issue of the journal. The authors performed a randomized trial evaluating 160 patients undergoing elective thoracotomy and concluded that intravenous parecoxib significantly reduced the incidence and severity of shoulder pain. The authors should be congratulated for performing a well-designed study on an important topic (eg, acute pain) in patients undergoing cardiothoracic surgery. 2,3 In addition, the renewed interest in the use of intravenous parecoxib to improve postoperative analgesia across many surgical procedures makes the topic timely in perioperative medicine. 4,5 Although the study of Pipanmekaporn et al. was conducted well, there are some concerns regarding the study that need to be clarified to determine the validity of the results. First, it is unclear if the authors standardized the intraoperative anesthetic strategy for these patients as this can affect some of the studied outcomes. 6 Second, it would be expected that postoperative pain would be similar among the groups as the postoperative analgesic treatments were used to keep pain scores acceptable to patients. Lastly, the authors evaluated pain scores across 7 time points (2, 6, 12, 24, 48, 72, and 96 hours) and therefore should have corrected their alpha values to avoid a type I error (Bonferroni correction, p o 0.007). This would have affected some of the authors' conclusions. We would welcome some comments by the authors as this would help to further substantiate the findings of this important randomized clinical trial.