Background: Despite the increasing implementation of selective histopathologic policies for post-cholecystectomy evaluation of gallbladder specimens in low-incidence countries, the fear of missing incidental gallbladder cancer (GBC) persists. This study aimed to develop a diagnostic prediction model for selecting gallbladders that require additional histopathological examination after cholecystectomy.
Methods: A registration-based retrospective cohort study of nine Dutch hospitals was conducted between January 2004 and December 2014. Data were collected using a secure linkage of three patient databases, and potential clinical predictors of gallbladder cancer were selected. The prediction model was validated internally by using bootstrapping. Its discriminative capacity and accuracy were tested by assessing the area under the receiver operating characteristic curve (AUC), Nagelkerke's pseudo-R2, and Brier score.
Results: Using a cohort of 22,025 gallbladders, including 75 GBC cases, a prediction model with the following variables was developed: age, sex, urgency, type of surgery, and indication for surgery. After correction for optimism, Nagelkerke's R2 and Brier score were 0.32 and 88%, respectively, indicating a moderate model fit. The AUC was 90.3% (95% confidence interval, 86.2%-94.4%), indicating good discriminative ability.
Conclusion: We developed a good clinical prediction model for selecting gallbladder specimens for histopathologic examination after cholecystectomy to rule out GBC.
Copyright © 2022. Published by Elsevier Ltd.