Answers are grounded in this paper and cite their sources.
Conventional static anatomical assessments inadequately predict velopharyngeal surgery outcomes in obstructive sleep apnea (OSA). This study aimed to develop and validate a prediction model integrating awake structural burden, sleep-state morphology, and pharmacological reversibility. In this prospective cohort study, 106 consecutive OSA patients (modeling cohort) underwent velopharyngeal surgery following standardized evaluation including the TCM score, polysomnography, and drug-induced sleep endoscopy (DISE) with dexmedetomidine. Dynamic reversibility was quantified by the proportional change in apnea-hypopnea index following nasopharyngeal tube insertion (ΔAHI-DISE) and oropharyngeal lateral wall collapse patterns. An independent cohort ( n = 28) served for temporal validation. Multivariable logistic regression was used to construct a nomogram, with performance assessed by discrimination (AUC), calibration (Hosmer-Lemeshow test, Brier score), and decision curve analysis. Static anatomical metrics (TCM score) showed modest discrimination (AUC 0.667, 95%CI 0.568–0.765). Dynamic parameters demonstrated superior predictive value: ΔAHI-DISE achieved an AUC of 0.769 (95%CI 0.674–0.864), and oropharyngeal lateral wall response to nasopharyngeal tube insertion remained independently predictive (odds ratio 0.254, 95%CI 0.089–0.688, P = 0.008). The integrated nomogram (TCM + dynamic metrics) achieved robust discrimination (AUC 0.822, 95%CI 0.740–0.904), significantly outperforming TCM alone (DeLong test, Z = 3.42, P < 0.001). Calibration was excellent (intercept 0.000, slope 1.000, Brier score 0.167). Temporal validation showed acceptable discrimination (AUC 0.728, 95% CI 0.535–0.921) but limited calibration (slope 0.524, intercept 0.112), indicating a need for recalibration rather than confirmed transportability; the optimal cutoff shifted from 0.545 to 0.697 in the lower-BMI validation cohort. Dynamic functional reversibility may outperform static anatomical burden in predicting velopharyngeal surgery success. The proposed nomogram may provide preliminary individualized risk stratification, but prospective multicenter validation and recalibration remain warranted.
No takes yet. Share an insight, caveat, or question.
Xu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: