Early Warning Model for Gallbladder Polyps in a Nurse Population: A Retrospective Cohort Study.

PURPOSE Gallbladder polyps pose a health concern among nurses, and early identification is essential for timely intervention and risk management. This study aimed to identify key risk factors and validate an early warning model for GBP specifically tailored to nurses using a retrospective cohort design.

METHODS A cohort of 3,507 nurses was recruited from a tertiary care hospital in northern China, with 3,181 participants after exclusions. The cohort was divided into training (2,407 cases) and validation (774 cases) subsets. Logistic regression analysis of demographic, lifestyle, and biochemical factors was used to develop a prediction model for GBP. Model performance was assessed using sensitivity, specificity, ROC curves, calibration curves, and the Hosmer-Lemeshow test.

RESULTS The prevalence of GBP increased from 1.45% in 2013 to 12.7% in 2022. Key predictors included age, gender, walking time, fried food consumption, and total bilirubin levels. The model demonstrated moderate accuracy with an area under the curve of 0.64 in the training cohort and 0.65 in the validation cohort. Sensitivity reached 60%, specificity was 64-65%, and the negative predictive rate was 92%. Internal validation using bootstrap resampling confirmed model stability.

CONCLUSIONS The early warning model for GBP among nurses demonstrated fair discriminative performance and reliable calibration, indicating potential clinical utility. The model can facilitate early intervention and strengthen occupational health management. Further research should aim to refine the model and validate it across diverse populations to enhance generalizability and accuracy.