The following is a summary of “Machine learning models to predict systemic inflammatory response syndrome after percutaneous nephrolithotomy,” published in the July 2024 issue of Urology by Zhang et al.
The objective of this study was to develop and assess the efficacy of machine learning models for predicting the likelihood of systemic inflammatory response syndrome (SIRS) following percutaneous nephrolithotomy (PCNL). Researchers conducted a retrospective analysis of clinical data from 337 patients who underwent PCNL between May 2020 and June 2022. The dataset was divided into a training set comprising 80% of the data and a testing set containing 20%. About 6 distinct machine-learning algorithms were employed to construct predictive models using the training dataset. The performance of each model was evaluated based on key metrics, including the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity, all calculated using the testing set.
Additionally, the study group utilized regression coefficients to analyze the contribution of each variable to the model’s predictive performance. Among the six algorithms evaluated, the support vector machine (SVM) demonstrated the highest performance, achieving an accuracy of 0.868 and an AUC of 0.942 (95% CI 0.890–0.994) in the testing set. Further investigation with the SVM model revealed that prealbumin levels were the most significant predictor of SIRS, followed by factors such as preoperative urine culture results, systemic immune-inflammation (SII), neutrophil to lymphocyte ratio (NLR), presence of staghorn stones, fibrinogen levels, operation duration, preoperative urine white blood cell (WBC) count, preoperative urea nitrogen levels, hydronephrosis status, stone burden, patient sex, and preoperative lymphocyte count.
These findings indicate that machine learning-based predictive models can effectively forecast the occurrence of SIRS after PCNL by leveraging patient clinical data, providing valuable insights that could assist surgeons in clinical decision-making. Implementing these predictive models in clinical practice may ultimately enhance patient outcomes by facilitating early identification and management of potential complications associated with PCNL.
Source: bmcurol.biomedcentral.com/articles/10.1186/s12894-024-01529-1
The post Predicting Systemic Inflammatory Response Syndrome Following PCNL Using Machine Learning Models first appeared on Physician's Weekly.