Predictors and Prevalence of Prolonged Mechanical Ventilation in Pediatric Patients Undergoing Congenital Heart Disease Surgery: A Comprehensive Analysis
Loading...
Date
Journal Title
Journal ISSN
Volume Title
Publisher
Brieflands
Abstract
Background: Prolonged mechanical ventilation (PMV) after congenital heart disease (CHD) surgery in children is associated with morbidity and increased resource use; however, definitions and predictors vary across studies. Objectives: We evaluated perioperative factors, including the vasoactive-inotropic score (VIS) and inflammatory indices, that were associated with PMV in a contemporary pediatric cohort. Methods: We conducted a retrospective observational study of consecutive patients aged < 18 years who underwent CHD surgery at a tertiary center. Prolonged mechanical ventilation was defined a priori as postoperative mechanical ventilation lasting ≥ 48 hours. Candidate predictors included age, sex, weight, underlying diagnosis, Risk Adjustment for Congenital Heart Surgery (RACHS-1) category, cardiopulmonary bypass time, VIS on postoperative days 1 - 2, and neutrophil-to-lymphocyte (N/L) ratios on postoperative days 1 - 2. Variables with P < 0.05 in univariable analyses or strong clinical relevance were included in a multivariable logistic regression model. Results: Among 268 children (mean age, 30.96 ± 35.41 months; mean weight, 10.81 ± 8.28 kg), 42 (15.7%) experienced PMV. In univariable analyses, younger age, lower weight, higher RACHS-1 scores, higher VIS, and specific diagnoses, including dextro-transposition of the great arteries, coarctation of the aorta (CoA), and total anomalous pulmonary venous connection (TAPVC), were associated with PMV. In the adjusted model, lower weight remained independently associated with a higher risk of PMV (P < 0.001), as did higher surgical complexity (RACHS-1: odds ratio [OR], 2.18; 95% confidence interval [CI], 1.21 - 3.94; P = 0.010). Higher POD2 VIS was associated with increased odds of PMV (OR, 1.13; 95% CI, 1.08 - 1.19; P < 0.001), whereas POD1 VIS was not retained in the final parsimonious model. Neither N/L ratio was retained in the primary three-predictor model. The POD2 N/L ratio was not independently significant in the adjusted sensitivity analysis. The diagnoses d-TGA, CoA, and TAPVC were more frequent among PMV cases in univariable analyses. Conclusions: In pediatric CHD surgery, PMV (≥48 hours) was common and independently associated with lower weight, higher RACHS-1 scores, and greater vasoactive requirements on POD2. Diagnosis-specific associations, including those for d-TGA, CoA, and TAPVC, were observed in univariable analyses. Early risk stratification incorporating patient size, surgical complexity, and POD2 hemodynamic burden may help to target preventive strategies; however, prospective multicenter validation is necessary.