Erector spinae plane block in pediatric surgery: a systematic review and meta-analysis

Citation:

Reysner M, Reysner T, Kowalski G, Janusz P, Mularski A, Daroszewski P, Kolasiński J, Wieczorowska-Tobis K. Erector spinae plane block in pediatric surgery: a systematic review and meta-analysis. BMC Anesthesiol. 2026; doi:10.1186/s12871-026-03617-1.

This systematic review and meta-analysis synthesised ten prospectively registered randomized controlled trials of erector spinae plane block in children undergoing cardiothoracic, abdominal, or hip surgery. Erector spinae plane block produced a small, high-certainty reduction in early postoperative pain compared with pooled alternatives and probably offers clear benefit over no regional block, but it showed little or no advantage over caudal, quadratus lumborum, or paravertebral blocks for later pain. Apparent large reductions in opioid use and longer time to first rescue analgesia are based on low-certainty, highly heterogeneous data, so erector spinae plane block should be seen as a reasonable option rather than a clearly superior new standard.

Study at a glance

- Design and setting: Systematic review and random-effects meta-analysis of 10 prospectively registered randomized controlled trials involving pediatric patients undergoing cardiothoracic, abdominal, or hip surgery, comparing ultrasound guided erector spinae plane block with no regional block or with other regional techniques (caudal, quadratus lumborum, or paravertebral blocks).

- Primary pain outcomes: Across eight randomized controlled trials (approximately 449 children), erector spinae plane block showed no important difference in pain within 2 hours after surgery (standardized mean difference minus 0.02, 95 percent confidence interval minus 0.24 to 0.21; moderate certainty). At about 4 hours, six trials (329 participants) demonstrated a small reduction in pain (standardized mean difference minus 0.27, 95 percent confidence interval minus 0.55 to 0.00; high certainty), largely driven by comparisons with no regional block. By 12 and 24 hours, high-certainty evidence showed little to no difference in pain (standardized mean difference minus 0.08 and minus 0.04 respectively, confidence intervals spanning no effect).

- Opioid-related outcomes: Nine trials (564 participants) suggested that erector spinae plane block may prolong time to first rescue analgesia and six trials (313 participants) suggested reduced total postoperative opioid consumption (pooled standardized mean difference minus 1.12, 95 percent confidence interval minus 2.12 to minus 0.13). However, both outcomes had very high heterogeneity (I squared around 93 to 96 percent), effects differed markedly by comparator (clearer benefit versus no block, little or none versus other blocks), and certainty was rated low. Intraoperative fentanyl consumption (three trials, 182 participants) showed no clear difference overall (standardized mean difference minus 0.21, 95 percent confidence interval minus 1.14 to 0.71; low certainty).

- Risk of bias and certainty of evidence: Individual randomized controlled trials were generally at low risk of bias, with Jadad scores of 5 in all 10 studies and mainly low risk assessments on standard domains, aside from one study with incomplete outcome data. For the meta-analyses, pain outcomes at 4, 12, and 24 hours were rated high certainty, while very early pain and 6-hour pain were moderate certainty. Time to first rescue analgesia, total postoperative opioid consumption, and intraoperative fentanyl use were all rated low certainty because of serious concerns about risk of bias and extreme heterogeneity. An independent methodological appraisal judged overall confidence in the review as low, citing incomplete searching and limited exploration of heterogeneity.

- Key limitations and clinical implications: The review pooled erector spinae plane block against clinically heterogeneous comparators, did not search trial registries or grey literature, and did not report a detailed table of excluded full texts. Very high and unexplained heterogeneity for opioid-related outcomes, plus an internal discrepancy in the effect size reported for time to first rescue analgesia (standardized mean difference 1.30 in the main text versus minus 1.08 in the GRADE table), further reduce confidence. Clinically, the evidence supports using erector spinae plane block as a reasonable component of multimodal analgesia, particularly when the alternative is no regional block, but it does not justify replacing well-established techniques such as caudal or quadratus lumborum block solely for better analgesia or opioid sparing.