Comparative efficacy of local anaesthetic adjuvants for the erector spinae plane block on postoperative analgesia: a systematic review and dose-stratified network meta-analysis

Background The erector spinae plane block (ESPB) provides opioid-sparing analgesia, but single-injection blocks are short-lived, prompting the addition of local-anaesthetic adjuvants. Their comparative effectiveness is undefined, and existing syntheses are single-agent, pairwise and disregard dose. We compared and ranked ESPB adjuvants and characterised the dexmedetomidine dose-response. Methods We searched PubMed, Embase, the Cochrane Library and Web of Science to 1 June 2026 for randomised trials of adults receiving an ESPB with local anaesthetic plus an adjuvant. A frequentist network meta-analysis used the ratio of means (ROM) for time to first rescue analgesia (primary outcome) and 24-hour opioid consumption, and odds ratios for postoperative nausea and vomiting (PONV). The dexmedetomidine node was split by dose (0.5 versus 1.0 µg/kg). Results Fifteen trials (1,127 participants) evaluated dexmedetomidine, dexamethasone, ketamine, magnesium sulfate and nalbuphine. Magnesium sulfate, 1.0 µg/kg dexmedetomidine, ketamine and nalbuphine each prolonged analgesia and reduced opioid consumption versus plain local anaesthetic, whereas dexamethasone and 0.5 µg/kg dexmedetomidine did not. Nalbuphine ranked first for opioid sparing. A coherent dexmedetomidine dose-response was found: 1.0 µg/kg was superior to 0.5 µg/kg for time to first rescue (ROM 1.75, 95% CI 1.04–2.97), and only the higher dose separated from plain local anaesthetic. Pooled dexmedetomidine reduced PONV, but this benefit dissolved on dose stratification. Conclusions Local-anaesthetic adjuvants differ substantially in ESPB efficacy. Magnesium sulfate, ketamine, nalbuphine and 1.0 µg/kg dexmedetomidine prolonged analgesia and spared opioids, nalbuphine ranking first for opioid sparing, whereas perineural dexamethasone was ineffective and dexmedetomidine's effect was dose-dependent. Several rankings, from sparse data, need head-to-head confirmation.