Heterogenic control groups in randomized, controlled, analgesic trials of total hip and knee arthroplasty.

ABSTRACT

INTRODUCTION: Postoperative analgesic interventions are often tested adjunct to basic non-opioid analgesics in randomized controlled trials (RCTs). Consequently, treatment in control groups, and possible assay sensitivity, differs between trials. We hypothesized that postoperative opioid requirements and pain intensities vary between different control groups in analgesic trials. EVIDENCE ACQUISITION Control groups from RCTs investigating analgesic interventions after total hip and knee arthroplasty were categorized based on standardized basic analgesic treatment. Morphine consumption 0 to 24 hours postoperatively, and resting pain scores at 6 and 24 hours for subgroups of basic treatments, were compared with ANOVA. In an additional analysis, we compared pain and opioid requirements in trials where a non-steroidal anti-inflammatory drug (NSAID) was administered as an intervention with trial where NSAID was administered in a control group. EVIDENCE SYNTHESIS: We included 171 RCTs employing 28 different control groups with large variability in pain scores and opioid requirements. Four types of control groups (comprising 78 trials) were eligible for subgroup comparisons. These subgroups received "opioid" alone, "NSAID + opioid", "acetaminophen + opioid", or "NSAID + acetaminophen + opioid", respectively. Morphine consumption and pain scores varied substantially between these groups, with no consistent superior efficacy in any subgroup. Additionally, trials administering NSAID as an intervention demonstrated lower pain scores and opioid requirements than trials where NSAID was administered in a control group. CONCLUSIONS: Analgesic treatment in RCT control groups varies considerably. Control groups receiving various combinations of opioid, NSAID and acetaminophen did not differ consistently in pain and opioid requirements. Pain and opioid requirements were lower in trials administering NSAID as an intervention compared with trials administering NSAID in a control group.