Efficacy of Computer-Aided Three-Dimensional Ultrasound Guidance for Neuraxial Anesthesia in Adult Patients: A Systematic Review and Meta-Analysis.

Artificial intelligence for ultrasound scanning in regional anesthesia is a rapidly developing interdisciplinary field. This study aimed to evaluate the efficacy of computer-aided three-dimensional ultrasound (C-aided US) guidance for neuraxial anesthesia in adult patients. We searched all randomized controlled trials (RCTs) of adult patients who required neuraxial anesthesia in the MEDLINE, CENTRAL, Embase, International Clinical Trials Registry Platform (ICTRP), and ClinicalTrials.gov databases on June 19, 2023. The primary outcomes were first-pass success, procedure time, and incidence of procedure-related adverse events. We used the Risk of Bias 2 to evaluate the risk of bias for each outcome, a random-effects model to conduct a meta-analysis, and the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach to evaluate the certainty of evidence. Seven RCTs (594 patients) were included. The C-aided US guidance results were as follows: first-pass success (risk ratio = 1.39, 95% confidence interval (CI) = 0.89 to 2.16; low certainty) and total procedure time (mean difference = 0.85 minutes, 95% CI = -0.81 to 2.5; low certainty). Four RCTs reported procedure-related adverse events (e.g., paresthesia, back pain, hemorrhagic events) in both groups (low certainty). The updated meta-analysis showed that there might be no differences in the first-pass success rates and total procedure times between C-aided US guidance and anatomical landmark guidance for neuraxial punctures.