# Abnormal coagulation after hepatectomy in patients with normal preoperative coagulation function

## TL;DR

This study identifies risk factors for abnormal coagulation after liver surgery in patients with normal preoperative coagulation and builds a predictive model.

## Contribution

A novel three-factor predictive model for postoperative coagulation abnormalities after hepatectomy is developed and validated.

## Key findings

- PAC occurred in 54.8% of patients with normal preoperative coagulation after hepatectomy.
- Preoperative INR, intraoperative succinyl gelatin infusion, and major hepatectomy were identified as independent risk factors for PAC.
- A predictive model with an ROC of 0.742 in the training set and 0.711 in the validation set was successfully developed.

## Abstract

To explore the risk factors for postoperative abnormal coagulation (PAC) and establish a predictive model for patients with normal preoperative coagulation function who underwent hepatectomy.

A total of 661 patients with normal preoperative coagulation function who underwent hepatectomy between January 2015 and December 2021 at the First Affiliated Hospital of Sun Yat-sen University were divided into two groups: the postoperative abnormal coagulation group (PAC group, n = 362) and the normal coagulation group (non-PAC group, n = 299). Univariate and multivariate logistic analyses were used to identify the risk factors for PAC.

The incidence of PAC in 661 patients who underwent hepatectomy was 54.8% (362/661). The least absolute shrinkage and selection operator (LASSO) method was used for multivariate logistic regression analysis. The preoperative international normalized ratio (INR), intraoperative succinyl gelatin infusion and major hepatectomy were found to be independent risk factors for PAC. A nomogram for predicting the PAC after hepatectomy was constructed. The model presented a receiver operating characteristic (ROC) curve of 0.742 (95% confidence interval (CI): 0.697–0.786) in the training cohort. The validation set demonstrated a promising ROC of 0.711 (95% CI: 0.639–0.783), and the calibration curve closely approximated the true incidence. Decision curve analysis (DCA) was performed to assess the clinical usefulness of the predictive model. The risk of PAC increased when the preoperative international normalized ratio (INR) was greater than 1.025 and the volume of intraoperative succinyl gelatin infusion was greater than 1500 ml.

The PAC is closely related to the preoperative INR, intraoperative succinyl gelatin infusion and major hepatectomy. A three-factor prediction model was successfully established for predicting the PAC after hepatectomy.