Predicting the risk of pulmonary infection after kidney transplantation using machine learning methods: a retrospective cohort study.

Purpose: Pulmonary infection is the most common and serious complication after kidney transplantation that affects the survival of the transplanted kidney and the quality of life of patients. This study aims to construct a machine learning model for predicting the risk of pulmonary infection after kidney transplantation.

Methods: We recruited 857 kidney transplant recipients from January 1, 2016, to December 31, 2021, in the Department of Nephrology, the First Affiliated Hospital of the University of Science and Technology of China. First, the distribution of baseline characteristics between patients with and without postoperative pulmonary infections was analyzed. Subsequently, six machine learning models were constructed to predict the risk of postoperative pulmonary infections. Finally, these models were subjected to external validation using an independent cohort. The performance of the models was evaluated by area under the receiver operating characteristic curve (AUC).

Results: Among kidney transplant recipients, a total of 186 individuals developed pneumonia, with 144 cases in the training cohort and 42 cases in the external validation cohort. The AUC range of the six machine learning models for predicting the risk of postoperative pulmonary infection was 0.758-0.822 for the training cohort and 0.642-0.795 for the testing cohort. Among the models assessed, the gradient boosting machine demonstrated the most favorable predictive accuracy.

Conclusions: Our study has developed a predictive model for assessing the risk of pulmonary infection after kidney transplantation, thereby providing a valuable foundation for the effective management of kidney transplant recipients.