Preoperative Serum Raman Spectroscopy: A Rapid, Label-Free Predictor of Postoperative Pulmonary Infection After Off-Pump Coronary Artery Bypass Grafting

Objective: This study aimed to develop and validate a preoperative risk-stratification tool based on serum Raman spectroscopy to identify patients at high risk for postoperative pulmonary infection (PPI) following off-pump coronary artery bypass grafting (OPCABG). Design: A prospective diagnostic cohort study. Setting: The study was conducted at a single academic medical center. Participants: A total of 276 adult patients undergoing OPCABG were enrolled. Preoperative serum samples were collected from all participants. Interventions: Preoperative serum was analyzed using 532 nm confocal Raman spectroscopy (400^-1,800 cm^-1). Participants were randomly split 7:3 into training and test sets for model development and validation. Measurements and Main Results: PPI within 72 hours was adjudicated by Centers for Disease Control and Prevention/National Healthcare Safety Network criteria. Fifty-seven patients (20.7%) developed early PPI. After feature selection via 2-sample t-test, LASSO regularization, and multivariable logistic regression, a 2-peak Raman signature (664 cm^-1 and 842 cm^-1) was identified. The final logistic model demonstrated strong performance on the test set, with an area under the receiver operating characteristic curve of 0.794 (95% confidence interval: 0.676-0.912), sensitivity of 70.6%, specificity of 72.7%, and a Brier score of 0.133. SHapley Additive exPlanations analysis confirmed that elevated intensities at these peaks were associated with increased early PPI risk. Decision-curve analysis indicated net clinical benefit across threshold probabilities of 15% to 35%. Conclusions: A rapid reagent-free serum Raman signature comprising 2 peaks (664 cm^-1 and 842 cm^-1) provides an interpretable and effective preoperative prediction of early PPI risk in OPCABG patients, as demonstrated by internal validation.