A U-shaped threshold in the monocyte-to-lymphocyte ratio predicts adverse outcomes after pediatric cardiac surgery: a retrospective cohort study

Children undergoing congenital heart disease (CHD) surgery with cardiopulmonary bypass (CPB) face substantial postoperative morbidity, but cost-effective preoperative risk-stratification tools remain limited. Inflammatory indices derived from routine complete blood counts predict outcomes in adults, but their utility in children—whose immune physiology differs markedly—remains unknown. Moreover, most studies assume linearity, potentially overlooking clinically meaningful nonlinear patterns. We therefore evaluated the association between six preoperative inflammatory indices and adverse postoperative outcomes in this population. This retrospective cohort study included 776 children (< 18 year) who underwent CHD surgery with CPB at a single tertiary pediatric hospital between 2018 and 2024. Six preoperative inflammatory indices—systemic immune-inflammation index (SII), systemic inflammation response index (SIRI), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), monocyte-to-lymphocyte ratio (MLR), and pan-immune-inflammation value (PIV)—were natural-log transformed and examined using multivariable logistic regression with the false discovery rate (FDR) controlled by the Benjamini–Hochberg method. Nonlinear associations were explored using generalized additive models and two-segment piecewise regression. The composite primary endpoint was in-hospital mortality, prolonged mechanical ventilation (> 72 h), unplanned reintubation, or prolonged hospitalization (> 90th percentile). Of 776 children, 153 (19.7%) met the composite endpoint. All six indices were independently associated with adverse outcomes after full adjustment and FDR correction (all q < 0.05). ln‑MLR suggested a U-shaped threshold association with an inflection point at − 2.86 (corresponding to raw MLR = 0.057). Above this threshold, each log-unit increase nearly doubled the odds ratio (OR) of adverse outcome (OR = 1.99; 95% confidence interval (CI), 1.41–2.82; P < 0.001); below it, the association was not significant (OR = 0.15; P = 0.16). MLR achieved the highest area under the curve (AUC) (0.676; sensitivity 63.4%, specificity 63.7%). Anemia status significantly modified this association (non-anemic: OR = 2.18; anemic: OR = 0.69; P for interaction = 0.006). Although preoperative ln‑MLR showed a statistically significant U‑shaped association with postoperative adverse events, its poor discriminative performance (AUC = 0.676) indicates that MLR cannot serve as an independent clinical risk‑stratification tool. None of the six routine inflammatory indices achieved adequate standalone predictive capacity in this pediatric cardiac cohort. The observed U‑shaped threshold effect and anemia‑related effect modification remain hypothesis‑generating findings that require external prospective validation. Anemia status critically modifies the inflammatory index‑outcome association and should be considered during preoperative risk interpretation. Prospective multicenter validation is warranted before any clinical application of MLR‑based preoperative risk evaluation. ChiCTR2400091749 (registered 20 December 2024).