Does calculating the textbook outcome based on its negative predictors enhance the transparency of intrahepatic cholangiocarcinoma surgery assessment?

Backgrounds/Aims: The definition of textbook outcome (TO) for liver surgery in patients with intrahepatic cholangiocarcinoma (ICC) varies depending on the data and TO model utilized in the study. This study aimed to clarify the separate estimation of TO in relation to its negative predictors, using two validated TO models for ICC surgery. Methods: We assessed the rates of achieving TO and failure of TO for both models in liver resection. Independent predictors of non-TO were identified through logistic regression and validated using AUC estimation. Results: TO was achieved in a similar proportion of patients across both models: 40% and 43%. The TO models did not differ in preoperative data affecting TO achievement. Independent predictors of non-TO for both models included tumor size > 10 cm,, open surgical approach, and biliary resection. TO achievement significantly differed among sub-groups that were homogeneous concerning the opposing values of the independent predictors, regardless of the TO model. The ratio of the frequency of negative predictors in the TO and non-TO groups, referred to as the TO coefficient, determines the likelihood of achieving TO given the presence of that predictor. Predictors were ranked by their negative impact on TO achievement, from the highest risk of TO failure (tumor size > 10 cm) to the lowest risk (open approach). Conclusions: Independent predictors of non-TO may improve the transparency of TO assessment and reduce biases related to unaccounted negative prognostic factors. The impact of these independent predictors on TO realization can be estimated and ranked using TO coefficient calculations.