J Thorac Cardiovasc Surg 2004;127:605
© 2004 The American Association for Thoracic Surgery
Letter to the editor
300 Elliot St, Newton, MA 02464, USA
To the Editor:
Ott and colleagues,1 and the editorial staff of the Journal,deserve much credit for carrying out and publishing a study on the use of cyclooxygenase 2 inhibitors for postoperative pain after coronary artery bypass grafting. However, the study report is beset with serious deficiencies in the presentation of the study results, which should be noted to avoid similar deficiencies in what I hope will be future publications of studies on postoperative pain relief.
Ott and colleagues1 commendably sought to provide a risk-benefit assessment of the use of parecoxib and valdecoxib in the postcoronary artery bypass grafting setting by reporting on both the degree of pain relief and differences in adverse effects associated with the study drugs relative to the control group. The first issue is the choice of primary outcome measure: amount of reduction in morphine consumption administered by a patient-controlled analgesia (PCA) pump. Ott and colleagues1 reported an overall reduction of morphine consumption of approximately 20% in the parecoxib/valdecoxib group relative to the placebo group. Although this may indicate a statistically significant analgesic effect, especially in light of the hints provided later about the secondary analgesic efficacy measures, the article did not discuss whether this degree of opioid sparing was clinically meaningful in this population. As has been discussed at length in multiple recent US Food and Drug Administration Advisory Committee meetings, opioid sparing alone does not necessarily imply clinical benefit. Clinical benefit of opioid sparing must be demonstrated directly, for example by showing a reduction in opioid-related side effects. In this study Ott and colleagues1 appropriately reported relative side effects in the two treatment groups. One would hope to see a reduction in typical opioid side effects, such as nausea, vomiting, dizziness, sedation, fatigue, and constipation; however, these were numerically about the same in the two groups, with the possible exceptions of dizziness (higher in the control group) and nausea (higher in the cyclooxygenase 2 inhibitor group). Thus one cannot conclude that there was clinically meaningful benefit of the study drugs according to the primary outcome measure of the study.
Of course, in studying an analgesic the critical issue is whether pain control is improved. Although some have argued that it is unrealistic to expect reductions in pain intensity when both treatment and control groups have access to morphine PCA, in fact multiple published studies of nonopioid analgesics versus placebo in the setting of PCA have succeeded in demonstrating pain reduction. Unfortunately, Ott and colleagues1 did not provide any interpretable pain data, which is inexcusable in a study of an analgesic for postoperative pain. They did provide data on the "peak pain intensity difference," defined as the "difference between maximum daily sternotomy pain and pretreatment sternotomy pain" calculated for each day of treatment. Given that the study drugs were administered twice a day, and that the pretreatment pain scores are not presented, the meaning of this outcome measure is unclear. Furthermore, the single figure in which these data are presented suggests that the difference between active drug and placebo was less than 1 unit on a 4-point pain intensity scale, a difference of uncertain clinical meaningfulness.
Ott and colleagues1 did obtain data on various subscales of the Brief Pain Inventory, but they chose not to present the data. Instead, they indicated that differences in several of these subscales between groups had reached statistical significance, which does not inform the reader about the magnitude of potential clinical benefit. Although the small differences in patient and physician global assessments of study drug went in the hoped-for direction, global assessments are not a direct measure of pain.
This study report presented a thorough discussion of the disturbing safety issues associated with the cyclooxygenase 2 inhibitors; however, the data presented do not justify the conclusion that "the parecoxib/valdecoxib regimen demonstrated superiority for pain relief over an aggressive therapeutic regimen supplemented with PCA." Future reports of analgesic clinical trials should include the specific pain data obtained, not just P values, so that readers can decide the clinical meaningfulness of any claimed benefits. This point should be noted by both authors and by journals, who should both be encouraged to continue to foster research on postoperative analgesia.
This article has been cited by other articles:
M. W. Khalil, A. Chaterjee, G. MacBryde, P. K. Sarkar, and R. R. D. Marks
Single dose parecoxib significantly improves ventilatory function in early extubation coronary artery bypass surgery: a prospective randomized double blind placebo controlled trial
Br. J. Anaesth., February 1, 2006; 96(2): 171 - 178.
[Abstract] [Full Text] [PDF]
|HOME||HELP||FEEDBACK||SUBSCRIPTIONS||ARCHIVE||SEARCH||TABLE OF CONTENTS|
|ANN THORAC SURG||ASIAN CARDIOVASC THORAC ANN||EUR J CARDIOTHORAC SURG|
|J THORAC CARDIOVASC SURG||ICVTS||ALL CTSNet JOURNALS|