Challenges in interpretation of randomised clinical trials included in systematic reviews and clinical guidelines.

Abstract

Although systematic reviews, meta-analyses, and network meta-analyses are central to evidence-based medicine, most of those assessing the effectiveness of perioperative analgesic interventions have critical limitations that may lead to flawed conclusions, making them clinically uncertain: major concerns include the lack of rigorous inclusion and exclusion criteria, failure to critically assess the quality of the studies, heterogeneity of included studies, and poor quality of included studies (bias, use of basic analgesics, lack of dynamic pain scores, inadequate follow-up, fragility, spin, inaccurate interpretation, etc.). Importantly, clinical guidelines based on such meta-analyses may be inaccurate, despite the use of well-recognised sophisticated complex statistical analyses, which may not compensate for the inherent flaws or study heterogeneity. A properly conducted meta-analysis would have a meticulous systematic review with inclusion of homogenous RCTs, and the included studies would be assessed critically for current clinical relevance including well-defined early mobilisation, the use of basic analgesics, assessment of functional pain and recovery, and clinically relevant reductions in pain and opioid use. Overall, critical appraisal, careful interpretation, and judicious application of meta-analyses are necessary before using them in guidelines and day-to-day practice.