Leiv Arne Rosseland1. 1. Department of Anesthesiology, Rikshospitalet University Hospital, Oslo, Norway. lei.arne.rosseland@rikshospitalet.no
Abstract
BACKGROUND AND OBJECTIVES: Intra-articular (IA) injection of morphine has been the subject of many randomized clinical trials (RCTs). Both negative and positive results have been obtained in trials with a preemptive design, and the question of efficacy remains unresolved. Recent RCTs on patients whose inclusion was delayed until a baseline pain of at least moderate intensity was documented have illuminated the pitfalls of IA analgesic trials. Previously published systematic reviews may have included flawed RCTs in the analyses. METHODS: A systematic, qualitative review of RCTs on the analgesic efficacy of IA morphine after knee arthroscopic surgery. RESULTS: Of the 67 screened RCTs, 46 RCTs (43 publications) of IA morphine were included. Thirty-six trials were placebo controlled. Twenty-three of these RCTs were of low scientific quality; randomization and blinding were not adequately described or the method used for statistical analysis of repeated measurements was unsound. Among the 13 RCTs with usable information, 4 of the positive outcomes may be explained by the uneven distribution of patients whose natural course was low postoperative pain intensity. Uneven sex distribution may be a confounding factor in one of these trials. Seven negative and 2 positive trials had reliable information. The only RCT with documented control over baseline pain intensity was negative. Most positive trials had small sample size. Publication bias favors the reporting and publication of positive trials more often than negative ones. CONCLUSIONS: There are few well-controlled RCTs on IA morphine, and the negative trials of higher quality counter the evidence from the numerous positive ones of lower quality. The quality of most published trials is poor, and performing meta-analysis on these data is not meaningful. Properly controlled trials, in which early postoperative pain intensity is documented, suggest that there is no added analgesic effect of IA morphine compared with saline alone.
BACKGROUND AND OBJECTIVES:Intra-articular (IA) injection of morphine has been the subject of many randomized clinical trials (RCTs). Both negative and positive results have been obtained in trials with a preemptive design, and the question of efficacy remains unresolved. Recent RCTs on patients whose inclusion was delayed until a baseline pain of at least moderate intensity was documented have illuminated the pitfalls of IA analgesic trials. Previously published systematic reviews may have included flawed RCTs in the analyses. METHODS: A systematic, qualitative review of RCTs on the analgesic efficacy of IA morphine after knee arthroscopic surgery. RESULTS: Of the 67 screened RCTs, 46 RCTs (43 publications) of IA morphine were included. Thirty-six trials were placebo controlled. Twenty-three of these RCTs were of low scientific quality; randomization and blinding were not adequately described or the method used for statistical analysis of repeated measurements was unsound. Among the 13 RCTs with usable information, 4 of the positive outcomes may be explained by the uneven distribution of patients whose natural course was low postoperative pain intensity. Uneven sex distribution may be a confounding factor in one of these trials. Seven negative and 2 positive trials had reliable information. The only RCT with documented control over baseline pain intensity was negative. Most positive trials had small sample size. Publication bias favors the reporting and publication of positive trials more often than negative ones. CONCLUSIONS: There are few well-controlled RCTs on IA morphine, and the negative trials of higher quality counter the evidence from the numerous positive ones of lower quality. The quality of most published trials is poor, and performing meta-analysis on these data is not meaningful. Properly controlled trials, in which early postoperative pain intensity is documented, suggest that there is no added analgesic effect of IA morphine compared with saline alone.