Development and Validation of a 7-Day Postoperative Sleep Disturbance Predictive Nomogram in Non-Cardiovascular Surgical Patients: A Prospective Cohort Study of 3851 Adults

Jia-Li Mao, Shu-Hua Shu, Ling Hu, Xue-Feng Wang, Sheng Wang, Xiao-Qing Fan Department of Anesthesiology, The First Affiliated Hospital of USTC, Division of Life Sciences and Medicine, University of Science and Technology of China, Hefei City, Anhui Province, People’s Republic of ChinaCorrespondence: Shu-Hua Shu, Department of Anesthesiology, The First Affiliated Hospital of USTC, Division of Life Sciences and Medicine, University of Science and Technology of China, Hefei City, Anhui Province, People’s Republic of China, Email [email protected]: This prospective study developed and validated a nomogram to assess the likelihood of early postoperative sleep disturbance (PSD occurring within 7 days after surgery) following non-cardiovascular surgical procedures.Participants and Study Protocol: The study enrolled 3851 patients receiving non-cardiac procedures recruited in the First Affiliated Hospital of USTC from April 2024 to December 2024. These 3851 patients were randomly allocated into training cohort (n=2682, 70%) and validation cohorts (n=1169, 30%). Variable selection was performed using least absolute shrinkage and selection operator (LASSO) regression, followed by logistic regression analyses (univariate and multivariate) to identify independent risk factors. Based on these identified factors, a prognostic nomogram was developed and underwent comprehensive validation, including receiver operating characteristic (ROC) curve assessment, calibration plotting, and decision curve analysis (DCA) to evaluate its discriminative performance and clinical applicability.Results: About 37.7% of patients developed PSD within the first 7 postoperative days, with 12.1% persisting at 1 month after surgery. The analysis revealed ten independent PSD risk factors (all p < 0.05): female, higher ASA class (III), moderate-to-severe anemia, dissatisfaction with ward environment, anxiety, non-use of dexmedetomidine, older age, extended anesthesia time, lower sufentanil doses and higher postoperative NRS score. The nomogram incorporating these ten predictors demonstrated excellent discriminative performance, with AUC values of 0.826 (95% CI: 0.810– 0.843) in the training cohort and 0.822 (95% CI: 0.797– 0.847) in the validation cohort, complemented by optimal calibration.Conclusion: The prediction model incorporating ten routinely available clinical variables demonstrated excellent predictive accuracy and good calibration, highlighting its clinical utility in identifying short-term high-risk PSD patients (within 7 days postoperatively). This tool facilitates timely interventions to reduce PSD incidence and improve recovery outcomes.Keywords: risk factors, machine learning, dexmedetomidine, anesthesia, nomograms