Prediction of massive intraoperative blood transfusion requirement in lung transplantation.

BACKGROUND

Massive transfusion (MT) during lung transplantation (LTx) is associated with primary graft dysfunction and increased risk of post-LTx mortality and morbidity. To date, predictive factors for MT during LTx are unknown. This study aimed to establish a predictive model for intraoperative MT in patients undergoing LTx to provide personalized risk prediction.

METHODS

This retrospective cohort study included adult patients who underwent LTx at the Second Affiliated Hospital of the Zhejiang University School of Medicine between 2020 and 2023. MT was defined as an intraoperative transfusion of > 5 units of packed red blood cells. The dataset was randomly split into a training cohort (70%) and a test cohort (30%). Three algorithms were applied for variable selection in the training cohort and logistic models were used to establish a predictive model. Model accuracy was assessed using goodness-of-fit tests, receiver operating characteristic analyses, and bootstrap resampling.

RESULTS

In total, 390 patients were included. The MT incidence was 25.9% in the training cohort (n = 274) and 25.0% in the test cohort (n = 116). The multivariable model for predicting MT incorporated primary diagnosis, hemoglobin level, patient entry route, surgical type, and approach, activated partial thromboplastin time (APTT), and cold ischemia time (CIT). The model demonstrated an AUC of 0.830 (95% CI: 0.775-0.884) in the training cohort and 0.759 (95% CI: 0.657-0.861) in the test set for predicting MT.

CONCLUSION

A nomogram combining seven predictive factors may help predict the need for intraoperative MT in patients undergoing LTx to optimize individual transfusion strategies.