Abstract
BACKGROUND: Non-steroidal anti-inflammatory drugs (NSAIDs) are recommended for perioperative analgesic treatment. However, NSAID treatment may be associated with increased risk of adverse events. We investigated the risks of serious adverse events with perioperative NSAID treatment in patients undergoing orthopaedic surgery. METHODS: We conducted a systematic review (PROSPERO: CRD42022342003) of randomised clinical trials assessing the harmful effects associated with the use of NSAIDs versus placebo, usual care, or no intervention in patients undergoing orthopaedic surgery. The primary outcome was the incidence of serious adverse events. We systematically searched for eligible trials up to October 13, 2025. We performed risk of bias assessments to account for systematic errors, Trial Sequential Analysis to account for the risks of random errors, performed meta-analyses using R, and used The Grading of Recommendations Assessment, Development and Evaluation framework to assess the certainty of the evidence. RESULTS: We included 38 trials enrolling 6593 patients for our primary outcome. Most trials were of high risk of bias. Meta-analysis showed no statistically significant difference between NSAID and placebo for risk of serious adverse events, RR 0.83; 95% CI, 0.61 to 1.14; p = 0.26; very low certainty of evidence. Post hoc beta-binomial regression sensitivity analyses, including trials with zero events, indicated no difference in risks of serious adverse events between NSAID and placebo (OR 0.89; 95% CI, 0.76 to 1.06; p = 0.19). Trial Sequential Analysis showed that insufficient information was obtained to confirm or reject a relative risk change of 25%. CONCLUSION: In adult patients undergoing orthopaedic surgery, we found no evidence of a difference between NSAID and placebo regarding risk of serious adverse events. However, the available data were insufficient to draw firm conclusions, and the certainty of evidence was generally very low. EDITORIAL COMMENT: This systematic review with meta-analysis was aimed at assessing non-steroidal anti-inflammatory drug treatments for reported serious adverse effects in post-operative analgesia trials in orthopedic surgical cohorts. These drugs are routinely part of modern multimodal or pre-emptive analgesia treatment routines. The published evidence was generelly assessed to have a high risk of bias or low certainty for this specific question. No clear difference in reported event risk between this type of drug and placebo was found, though it is not clear that this is conclusive.