Purpose: This study aimed to inform a basis for nursing intervention through investigating the clinical risk factors for prolonged length of stay (PLOS) in patients with chronic acalculous cholecystitis (CAC) after cholecystectomy and constructing a predictive model.
Methods: The prediction model for the PLOS in CAC patients after cholecystectomy was developed using a retrospective research design. Data were extracted from Medical Information Mart for Intensive Care IV (MIMIC-IV). Logistic regression was used to explore the influencing factors of PLOS using the interpolated data. A variety of statistical methods, such as receiver operating characteristic (ROC) analysis, decision curve analysis (DCA) analysis, 5-fold cross-validation method, and SHapley Additive exPlanations (SHAP), were conducted to assess, validate, and interpret the nomogram model.
Results: The study included 204 CAC patients who underwent cholecystectomy. The average age was 60.23 ± 14.30 years, with 65.2% experiencing a PLOS. Multivariable logistic regression analysis showed that age (OR=1.04, 95%CI: 1.01-1.08), calcium (OR= 0.48, 95%CI: 0.24-0.93), ciprofloxacin (OR= 3.08, 95%CI: 1.14-9.91), fentanyl (OR= 4.08, 95%CI: 1.36-14.53), and mild liver disease (MLD) (OR= 0.30, 95%CI: 0.11-0.76) may be the influencing factors for PLOS. The nomogram model based on the five variables demonstrated moderate discrimination capacity according to ROC and DCA analyses. The ROC results based on the 5-fold cross-validation method revealed that the average AUC was .75 (.67-.83) in the training set, .71 (.54-.88) in the validation set, and .72 (.53-.90) in the test set. SHAP analysis identified age as the most influential factor, followed by calcium, fentanyl, ciprofloxacin, and MLD.
Conclusion: The nomogram model, based on age, calcium, fentanyl, ciprofloxacin, and MLD, demonstrated moderate discrimination capacity in predictive PLOS in CAC patients after cholecystectomy. These findings may support the identification of at-risk patients and inform early nursing interventions.
扫码关注我们
求助内容:
应助结果提醒方式:
