Development and validation of a comprehensive model to predict complications after hepatectomy

OBJECTIVE: Despite advances in perioperative care, hepatectomy remains associated with morbidity rates of up to 40%. Currently, available nomograms for predicting severe post-hepatectomy complications do not include early postoperative data. This retrospective observational study aimed to determine whether the parameters routinely measured in patients admitted to the Intensive Care Unit (ICU) after hepatectomy could represent risk factors for severe morbidity and to propose a nomogram scoring system to predict severe postoperative complications. PATIENTS AND METHODS: 411 adult patients who underwent elective hepatectomy at a high-volume tertiary care center for hepatic surgery from December 2016 to June 2022 were enrolled. The primary outcome was the assessment of predictors of 30-day severe postoperative complications following hepatectomy, defined as Clavien-Dindo grade 3a or higher. As a secondary outcome, we aimed to develop an easy-to-use scoring system to estimate the risk of severe postoperative complications. RESULTS: Severe complications occurred in 78 patients (19%). The final model included body mass index, preoperative bilirubin level, and ICU data (i.e., pH, lactate clearance, arterial lactate concentration 12 hours after ICU admission, need for packed red blood cell transfusions, and length of stay). Notably, the latter three variables were proven to be independent predictors of the outcomes. The model showed an overall good fit (C-index=0.754, corrected Dxy=0.692). A calibration plot using bootstrap internal validity resampling confirmed the stability of the model (mean absolute error=0.017, root mean square error of approximation=0.00051). CONCLUSIONS: We developed an accurate and practical scoring system based on preoperative and early postoperative data to predict poor outcomes after hepatectomy. Further external validation on larger series could lead to the integration of such a tool in the routine clinical practice to support patients’ management and early warning during ICU stay.