Development of a Risk Prediction Model for Infection After Kidney Transplantation Transmitted from Bacterial Contaminated Preservation Solution

Abstract

Background: The risk of transplant recipient infection is unknown when the preservation solution culture is positive.

Methods: We developed a prediction model to evaluate the infection in kidney transplant recipients within microbial contaminated preservation solution. Univariate logistic regression was utilized to identify risk factors for infection. Both stepwise selection with Akaike information criterion (AIC) was used to identify variables for multivariate logistic regression. Selected variables were incorporated in the nomograms to predict the probability of infection for kidney transplant recipients with microbial contaminated preservation solution.

Results: Age, preoperative creatinine, ESKAPE, PCT, hemofiltration, and sirolimus had a strongest association with infection risk, and a nomogram was established with an AUC value of 0.72 (95% confidence interval, 0.64-0.80) and Brier index 0.20 (95% confidence interval, 0.18-0.23). Finally, we found that when the infection probability was between 20% and 80%, the model oriented antibiotic strategy should have higher net benefits than the default strategy using decision curve analysis.

Conclusion: Our study developed and validated a risk prediction model for evaluating the infection of microbial contaminated preservation solutions in kidney transplant recipients and demonstrated good net benefits when the total infection probability was between 20% and 80%.

Keywords: kidney transplant; nomogram; risk factors; risk prediction model.