Key result
Elevated homocysteine is linked to ~38% higher odds of postoperative AKI after cardiac surgery.
Why the study?
Acute kidney injury after cardiac surgery is common yet detected late, and most machine learning models for risk stratification either rely on postoperative data or lack a transparent, pre-specified evaluation.
Does a machine learning model using routine preoperative and intraoperative variables accurately predict acute kidney injury at the end of cardiac surgery in adult cardiac surgery patients?
Cohort (n=1,352)
No
Does a machine learning model using routine preoperative and intraoperative variables accurately predict acute kidney injury at the end of cardiac surgery in adult cardiac surgery patients?
Routine perioperative variables up to the end of cardiac surgery provide modest discrimination for predicting acute kidney injury (test AUC 0.722), with complex machine learning models offering no decisive advantage over parsimonious penalized linear models and preoperative homocysteine emerging as the top predictor.
Answers are grounded in this paper and cite their sources.
May support screening-oriented AKI prediction with routine variables; leaves open generalizability and utility pending external validation.
Background Acute kidney injury after cardiac surgery (CSA-AKI) is common yet detected late. Risk stratification at the end of surgery could enable timely renal protection, but most machine learning (ML) models either use postoperative data or lack a transparent, pre-specified evaluation. Methods In a retrospective single-center cohort of 1,352 adult cardiac surgery patients (211 AKI, 15.6%), we predicted AKI from 32 routine preoperative and intraoperative (up to end-of-surgery) variables. We benchmarked 210 combinations (five feature-selection strategies × seven algorithms × six class weights) by repeated stratified 5-fold cross-validation (10 repeats) on a fixed development set ( n = 1,081), selected the final model by a pre-specified two-stage rule, and evaluated it once on a fixed test set ( n = 271). A screening-oriented threshold (sensitivity ≥0.80) was pre-specified. We additionally tested the few-shot transformer TabPFN, stability selection, bootstrap optimism correction, and simple ensembling, with SHAP for interpretability. Results The selected random forest (positive-class weighting, isotonic calibration) achieved a test-set AUC of 0.722 (95% CI 0.634–0.799), sensitivity 0.762, and specificity 0.515. No model family was decisively superior; penalized linear models generalized comparably, and none of the small-sample strategies or ensembling improved discrimination. Bootstrap correction reduced the apparent AUC from 0.965 to a corrected 0.724. Homocysteine ranked first in SHAP importance and was among ten independent factors in multivariable logistic regression. Conclusion Routine perioperative variables support a screening-oriented, interpretable prediction of AKI at the end of cardiac surgery, with modest discrimination. The lack of model superiority argues for parsimony, and homocysteine merits prospective study; external validation is required before clinical use.
No takes yet. Share an insight, caveat, or question.
Li et al. (2026) conducted a cohort in Cardiac surgery-associated acute kidney injury (CSA-AKI) (n=1,352). Random forest machine learning model using 32 routine preoperative and intraoperative variables (with preoperative homocysteine as leading predictor per 5 μmol/L increase) vs. Penalized linear models and alternative machine learning combinations was evaluated on Creatinine-defined KDIGO acute kidney injury within 7 days after cardiac surgery (OR 1.38, 95% CI 1.20-1.59, p=<0.001). A random forest model using 32 routine perioperative variables predicted postoperative AKI with a test AUC of 0.722, with homocysteine as the top predictor.
Synapse has enriched 10 related papers. Consider them for comparative context: