Problem with non-inferiority margin: a letter to the editor

Abstract

To the editor We read with intense interest the recently published article by Lee et al in regard to the comparison of postoperative analgesic effects between posterior quadratus lumborum block (QLB) and intrathecal morphine (ITM) in laparoscopic donor hepatectomy. We highly appreciate the efforts of the authors in conducting this outstanding study. Nevertheless, we feel that there may be some uncertainties in this study. In this study, intravenous patientcontrolled analgesia with fentanyl was programmed to deliver a 15 μg bolus (1 mL) dose and a lockout interval of 15 min in the postanesthetic care unit (PACU). When Numeric Rating Scale (NRS) >4, some additional pain remedies would be implemented, as intravenous pethidine (25 mg) was injected as firstline therapy. Intravenous fentanyl (25–50 μg) would injected when the target analgesic effect was not achieved after 15 min. And at surgical wards, all participants received intravenous ibuprofen or oral Mypol capsule for pain management. When NRS>4, intravenous hydromorphone (2 mg) was injected. The primary outcome in the study was the resting pain score at 24 hours postsurgery. The results showed that, in both groups, the median pain score at admission to PACU was beyond 6, with a maximum of 10. And a lot of patients in both QLB and ITM group still had pain scores greater than 4 at 24 and 48 hours postsurgery. It indicated that there were many patients who failed to achieve analgesic goal after surgery, making us doubt whether the general analgesic program in this paper is appropriate. What’s more, the mean differences of resting pain score at 24 hours postsurgery between QLB and ITM groups was 1.11, which is lower than the recommended minimal clinically important difference 1.5. It indicated that the difference in resting pain score is of no clinical value and meets the definition of noninferiority. We think that the predetermined noninferiority margin (δ) was too small to acquire the noninferiority result. We would be interested in hearing the author’s thoughts on our concerns.