Construction of a nomogram risk prediction model for prolonged mechanical ventilation in patients following surgery...

This study aims to analyze the risk factors associated with prolonged mechanical ventilation (PMV) in patients following surgical treatment for acute type A aortic dissection (ATAAD). The objectives include constructing a predictive model for risk assessment and validating its predictive efficacy.

A total of 452 patients diagnosed with ATAAD and undergoing surgical procedures at a tertiary hospital in Nanjing between January 2021 and April 2023 were selected using a convenience sampling method. Patients were categorized into two groups: PMV group (n = 132) and non-PMV group (n = 320) based on the occurrence of prolonged mechanical ventilation (PMV), and their clinical data were compared. The data were randomly divided into a modeling set and a validation set in a 7:3 ratio. Risk factors for PMV were identified in the modeling group using logistic regression analysis. A risk prediction model was constructed using R 4.1.3 software, visualized via a column chart. Receiver Operating Characteristic (ROC) curves were generated using the validation set to assess model differentiation. Calibration curves were plotted to evaluate accuracy and consistency, and Decision Curve Analysis (DCA) was applied to evaluate clinical utility.

The logistic regression analysis identified age, body mass index, preoperative white blood cell count, preoperative creatinine, preoperative cerebral hypoperfusion, and cardiopulmonary bypass time as significant risk factors for postoperative PMV in patients with ATAAD. The area under the curve (AUC) for the validation set ROC curve was 0.856, 95% confidence interval (0.805-0.907), indicating good discrimination. Calibration curves revealed strong alignment with the ideal curve, and the Hosmer-Lemeshow goodness-of-fit test indicated a well-fitted model (P = 0.892). The DCA curve demonstrated a high net benefit value, highlighting the model's strong clinical utility.

The risk prediction model developed in this study for PMV in patients undergoing surgery for ATAAD exhibits robust predictive performance. It provides valuable insights for healthcare practitioners in predicting the likelihood of PMV and devising timely and personalized intervention strategies.