image: (A) ROC of the training set. (B) ROC of the validation set. (C) PR curve of the training set. (D) PR curve of the validation set. (E) Calibration curve of the training set. (F) Calibration curve of the validation set. (G) DCA of the training set. (H) DCA of the validation set. AUROC and AUPRC are expressed as the point estimates and 95% CI. An AUROC > 0.8 is considered to give good discriminatory accuracy for a clinical prediction model. An AUPRC > 0.7 can be considered evidence that the model has good performance and can reach a certain level of the comprehensive performance in precision and recall. It has certain application value for clinical prediction and other tasks, and can effectively identify and predict target events to a certain extent. Calibration curves are used to evaluate the consistency between the predicted probability of the model and the actual probability of occurrence. An ideal model has pairs of observed and predicted probabilities that lie on the 45◦ line, and the P-value of the Hosmer–Lemeshow test should be greater than 0.05. DCA is used to evaluate the clinical net benefit of the prediction model under different decision thresholds, taking into account true positives, false positives, true negatives, and false negatives, as well as the potential benefits and risks of different decisions. Net benefit represents true cases of postoperative infection that would be treated preventatively. Abbreviations: CI, confidence interval; AUROC, area under the receiver operating characteristic curve; AUPRC, area under the precision-recall curve.
Credit: hLife
Postoperative infections remain a serious threat to transplant recipients, with incidence rates as high as 30%–80% within the first month. While previous models have focused on clinical and immunological factors, a new study introduces a more comprehensive approach by incorporating lifestyle and psychological variables.
Led by Professor Ning-Ning Liu from Shanghai Jiao Tong University School of Medicine, the research team conducted a prospective observational study across six major transplant centers in China. They collected standardized data from 615 liver and kidney transplant patients, including dietary habits, psychological status, and clinical indicators.
Using machine learning and multivariate regression, the team identified several novel risk factors. For example, regular tea consumption was associated with a 57% lower risk of infection, while higher guilt scores on the Transplant Effects Questionnaire (TxEQ) were linked to a 12-fold increase in infection risk. Preoperative serum creatinine levels also played a significant role.
The resulting prediction model—validated using ROC curves, precision-recall analysis, and decision-curve analysis—showed robust performance in stratifying patients into high- and low-risk groups. Although performance dipped in the validation set, the model still provided clinically useful thresholds for intervention.
The authors suggest that future studies include more transplant types and pathogen-specific data to refine the model further.
About Author: (Briefly outline the corresponding author's academic achievements and research contributions) (One author only)
Dr. Liu obtained his Ph.D. from Shanghai Institutes of Biological Sciences, Chinese Academy of Sciences in 2012. After graduation, he continued his research in Harvard Medical School/Boston Children’s Hospital as a post-doctoral researcher. In 2017, he joined Shanghai Jiao Tong University School of Medicine as a faculty member until now. Dr. Liu’s research has been focused on the mechanistic investigation of fungi-host interaction during pathogenesis. He has published more than 40 original papers in prestigious journals (Cancer Cell, Cell Host & Microbe, Nature Microbiology, Nature Communications, Nature Protocols, PNAS, Cell Reports etc.) and has been licensed 4 patents for invention.
Digital Object Identifier (DOI): 10.1016/j.hlife.2025.09.002
Journal
hLife
Article Title
Risk prediction model of postoperative infection after transplantation
Article Publication Date
11-Sep-2025