Background and objective: The aim of this systematic review was to indirectly compare the efficacy of any intervention, administered perioperatively, on acute and persistent pain after breast surgery. Databases and data treatment: We searched for randomized trials comparing analgesic interventions with placebo or no treatment in patients undergoing breast surgery under general anaesthesia. Primary outcome was intensity of acute pain (up to 6 hr postoperatively). Secondary outcomes were cumulative 24-hr morphine consumption, incidence of postoperative nausea and vomiting (PONV), and chronic pain. We used an original three-step approach. First, meta-analyses were performed when data from at least three trials could be combined; secondly, trial sequential analyses were used to separate conclusive from unclear evidence. And thirdly, the quality of evidence was rated with GRADE. Results: Seventy-three trials (5,512 patients) tested loco-regional blocks (paravertebral, pectoralis), local anaesthetic infiltrations, oral gabapentinoids or intravenous administration of glucocorticoids, lidocaine, N-methyl-D-aspartate antagonists or alpha2 agonists. With paravertebral blocks, pectoralis blocks and glucocorticoids, there was conclusive evidence of a clinically relevant reduction in acute pain (visual analogue scale > 1.0 cm). With pectoralis blocks, and gabapentinoids, there was conclusive evidence of a reduction in the cumulative 24-hr morphine consumption (> 30%). With paravertebral blocks and glucocorticoids, there was conclusive evidence of a relative reduction in the incidence of PONV of 70%. For chronic pain, insufficient data were available. Conclusions: Mainly with loco-regional blocks, there is conclusive evidence of a reduction in acute pain intensity, morphine consumption and PONV incidence after breast surgery. For rational decision making, data on chronic pain are needed. Significance: This quantitative systematic review compares eight interventions, published across 73 trials, to prevent pain after breast surgery, and grades their degree of efficacy. The most efficient interventions are paravertebral blocks, pectoralis blocks and glucocorticoids, with moderate to low evidence for the blocks. Intravenous lidocaine and alpha2 agonists are efficacious to a lesser extent, but with a higher level of evidence. Data for chronic pain are lacking.
