Clinical prediction rules of postoperative reintubation within 24 hours after general anesthesia: a retrospective case-control study
5
Issued Date
2025-12-01
Resource Type
eISSN
14712253
Scopus ID
2-s2.0-105026349050
Pubmed ID
41291422
Journal Title
BMC Anesthesiology
Volume
25
Issue
1
Rights Holder(s)
SCOPUS
Bibliographic Citation
BMC Anesthesiology Vol.25 No.1 (2025)
Suggested Citation
Morakul S., Charernboon T., Patumanond J., Sombatthaveekul P., Eowsakul N. Clinical prediction rules of postoperative reintubation within 24 hours after general anesthesia: a retrospective case-control study. BMC Anesthesiology Vol.25 No.1 (2025). doi:10.1186/s12871-025-03508-x Retrieved from: https://repository.li.mahidol.ac.th/handle/123456789/114013
Title
Clinical prediction rules of postoperative reintubation within 24 hours after general anesthesia: a retrospective case-control study
Corresponding Author(s)
Other Contributor(s)
Abstract
Background: Reintubation after planned extubation (RAP) following general anesthesia is a serious complication associated with intensive care unit admission, prolonged hospitalization, and increased mortality. Despite its clinical significance, no routinely validated clinical scoring system currently exists for predicting RAP. This study aimed to develop a clinical prediction rule for reintubation within 24 h after general anesthesia. Methods: This retrospective case-control study included 657 patients (235 cases and 422 controls) who underwent general anesthesia at Ramathibodi Hospital between 2014 and 2018. Cases were defined as patients reintubated within 24 h after planned extubation, and controls were randomly selected from those with successful extubation on the same operative day. Multivariable logistic regression was used to identify predictive factors, and significant predictors were transformed into a point-based risk score. Results: Significant predictors of reintubation included age < 1 or > 65 years, ASA classification ≥ III, emergency surgery, neurosurgical or thoracic procedures, vasopressor or inotrope use, positive fluid balance ≥ 40 mL/kg, and failure to follow commands after anesthesia. The score-based model demonstrated strong discrimination with an area under the receiver operating characteristic curve (AUROC) of 0.831 (95% CI: 0.795–0.868). Hosmer–Lemeshow goodness-of-fit test using 9 groups: χ²(df = 7) = 10.67, p = 0.154. Bootstrap validation confirmed consistent performance, with an optimism-adjusted AUROC of 0.831 (95% CI: 0.798–0.870). Based on total score ranges, patients were stratified into two risk categories. Those with a score of 0–9 was classified as low risk with a positive likelihood ratio (LHR+) of 0.693 (95% CI: 0.526–0.913, p = 0.004), and scores of 9.5–20 were considered high risk with an LHR + of 11.363 (95% CI: 5.611–25.306, p < 0.001). Conclusion: The RAP prediction score is a validated clinical prediction tool with good discrimination of postoperative RAP. It effectively stratifies postoperative patients into distinct risk categories and may guide for recognition and decision making for extubation during postoperative period.