Development and validation of a prognostic model for postoperative hypotension in patients receiving epidural analgesia
Year: 2025
Type: article
Study design: Observational Study
Abstract: Background and Aims: Postoperative hypotension is common in adults receiving epidural analgesia. Although risk factors have been reported in the literature, prognostic models have not been developed or validated. We aimed to develop and validate a multivariable, prognostic model for postoperative hypotension in patients receiving epidural analgesia. Material and Methods: candidate exposures. The primary outcome was postoperative hypotension during the first 72 hours after surgery. Multiple logistic regression was performed to evaluate a multivariable model. Exposures identified as statistically significant were used for logistic regression, linear discriminant analysis, and decision-tree model of random forest. Classification error was used to compare models, and variable importance was used for random forest analysis. Results: In total, 829 participants were included. The incidence of postoperative hypotension was 38.8%. Multivariable analysis identified the following independent prognostic factors: male sex, white race, body mass index, intraoperative hypotension, use of arterial line, bupivacaine concentration of 0.125% (vs. lower concentrations), and anesthesia duration. The error misclassification rate was 67% for multiple logistic regression, 27% for linear discriminant analysis, and 33.4% for random forest model. Conclusion: Using retrospective cohort data, a prognostic model of hypotension produced the best performance results using linear discriminant analysis, with an error misclassification rate of 27%. Further studies are required to perform model optimization for future clinical use.
Authors: Carlos E. Guerra-Londono, Erika Taco Vasquez, Efrain Riveros, Ehsan Noori, David Greiver, Srikanth Pillai, Theodore Schiff, James Soetedjo, Maylyn Wu, Jaime Garzon Serrano
Language: English
Subfield: Surgery
Field: Medicine
Domain: Health Sciences
Open Access status: diamond
OpenAlex ID: W4408766928