Abstract
<title>Abstract</title> <p>Background Ultrasound-guided brachial plexus block (USG-BPB) is widely used for upper limb surgery, but failure remains clinically important. This study aimed to identify predictors of USG-BPB failure and develop a clinical risk score. Methods We performed a retrospective cohort study of patients undergoing upper limb surgery with USG-BPB at a university hospital from 2018 to 2023. Block failure was defined as the need for supplemental block, additional multimple dose of analgesics or sedatives, deep sedation, or conversion to general anesthesia. Demographic, procedural, and surgical variables were analyzed using multivariable logistic regression. Model discrimination was assessed by the area under the curve (AUC). A point-based clinical prediction score was derived from the regression coefficients. Results Of 657 cases reviewed, 606 were included in the analysis. Block failure occurred in 112 cases (18.5%). Significant predictors of failure were liver disease (OR = 4.81, p = 0.031), humerus/elbow or forearm surgery compared with hand surgery (p = 0.019), surgical duration > 120 minutes (OR = 3.54, p < 0.001), block-to-incision interval > 30 minutes (OR = 1.80, p = 0.012), single local anesthetic solution versus mixed solution (OR = 3.04, p = 0.002), and block performed by first- or second-year residents compared with regional anesthesiologists (OR = 5.84, p < 0.001). Diabetes mellitus (OR = 0.43, p = 0.004) and larger local anesthetic volume (OR = 0.93, p = 0.012) were protective factors. The final model showed good discrimination (AUC = 0.75, 95% CI 0.70–0.79). The derived clinical risk score also showed good discrimination (C-statistic = 0.75, 95% CI 0.70–0.80). Conclusion USG-BPB failure was associated with operator inexperience, liver disease, surgery at the humerus/elbow or forearm, prolonged duration, delayed incision, and single-agent local anesthetic use. Diabetes mellitus and higher block volume were protective. The point-based clinical risk score showed good discrimination and may help identify patients at risk of block failure. Trial registration : Not applicable</p>