Preoperative Nonopioid Analgesia Lowers Opioid Use After Arthroscopy

Key Takeaways

• Preoperative nonopioid medication was associated with a small overall reduction in opioid consumption during the first postoperative day.
• Statistically significant opioid-sparing effects were reported for COX-2 inhibitors and gabapentin, and COX-2 inhibitors were also linked to a small decrease in 24-hour pain scores.
• Nausea and vomiting differences were not significant, postoperative drowsiness was higher, and substantial heterogeneity tempered interpretation of the observed benefit.

Preoperative nonopioid analgesia before arthroscopic surgery was associated with lower postoperative opioid use in a systematic review and meta-analysis of randomized controlled trials, which found a 4.3 mg reduction in oral morphine equivalents at 24 hours versus placebo. The analysis included adults undergoing primary knee, hip, and shoulder arthroscopy. The clearest subgroup signals were seen with COX-2 inhibitors and gabapentin. The pooled effect reached statistical significance, but the magnitude was modest.

The review followed PRISMA guidance and a Cochrane framework, with searches in PubMed, Embase, and MEDLINE completed on December 12, 2024. It included 22 randomized trials of preoperative nonopioid medications in adults undergoing primary arthroscopic surgery. The primary outcome was postoperative opioid consumption, with pain scores and adverse events as secondary outcomes. Fifteen studies involved knee arthroscopy, four involved hip arthroscopy, and three involved shoulder arthroscopy. Medication classes included COX-2 inhibitors, gabapentin, pregabalin, dextromethorphan, dexamethasone, and duloxetine. Eighteen studies used placebo controls, 17 were blinded, and most pooled analyses focused on the first 24 to 48 postoperative hours.

Class-specific analyses showed that COX-2 inhibitors were associated with a 4.2 mg reduction in opioid consumption at 24 hours and a 4.8 mg reduction at 48 hours versus placebo. Gabapentin was associated with a 6.3 mg reduction in opioid consumption at 24 hours. COX-2 inhibitors were also linked to a 0.3 cm reduction in 24-hour visual analog scale pain scores. These subgroup findings remained statistically significant despite high between-study variability. The pain difference was significant, but it was small.

Procedure-level analyses showed higher postoperative opioid consumption after anterior cruciate ligament reconstruction than after general knee or shoulder arthroscopy. Pooled oral morphine equivalent values at 24 hours were 16.4 mg for anterior cruciate ligament reconstruction, 13.2 mg for knee arthroscopy, and 11.7 mg for shoulder arthroscopy. In four studies comparing the same COX-2 inhibitor given preoperatively vs postoperatively, postoperative opioid consumption was lower with preoperative dosing in all studies, with statistically significant differences in three. Two studies comparing different preoperative and postoperative medication strategies found no significant differences in opioid consumption or pain scores. Quantitative timing comparisons were limited by inconsistent reporting across trials. These procedure and timing differences added to the broader heterogeneity across the evidence base.

Safety findings were pooled across 18 studies that reported postoperative complications. There were no significant differences in postoperative nausea or in nausea and vomiting, but four studies involving 254 patients showed higher postoperative drowsiness with preoperative nonopioid analgesics. Three major adverse events were reported, including two pulmonary embolisms and one deep vein thrombosis, and these events were not directly correlated with medication use. Interpretation was constrained by variation in medication type, dose, timing, and procedure, as well as follow-up that usually did not extend beyond 48 hours. The investigators concluded that the literature was highly heterogeneous and that the observed reductions may not be clinically meaningful.