The Correlates of Recoarctation After Repair in Pediatric Patients with Aortic Coarctation: A Real-World Descriptive Study

Background/Objectives: This retrospective descriptive real-world study aimed to characterize clinical features and factors associated with recoarctation in children with coarctation of the aorta (CoA) and develop a composite metric to quantify postoperative residual aortic obstruction burden. Methods: We retrospectively enrolled 126 pediatric CoA patients (98 surgical, 28 interventional) from January 2012 to December 2025. Baseline, perioperative and follow-up data were analyzed descriptively only, without therapeutic comparison. Univariate Cox and 24-month logistic regression screened for potential correlates of recurrent obstruction. Multicollinearity among hemodynamic indicators was assessed, and the Composite Residual Obstruction Score (CROS) derived from principal component analysis (PCA) was developed to mitigate collinearity bias. Results: Baseline characteristics differed substantially between treatment groups. Rates of postoperative hypertension and recoarctation were similar. Patients with recoarctation exhibited heavier preoperative and postoperative hemodynamic burden and higher hypertension prevalence. Recoarctation rates varied by CoA subtype, though intergroup differences were not statistically significant (p = 0.1518). Hypertension and abnormal hemodynamic indices were associated with higher recoarctation risk. The CROS was generated from four discharge hemodynamic variables; the first principal component explained 49% of variance. The CROS showed strong discriminative ability for 24-month recoarctation (AUC = 0.927; 40.0% for CROS > 28 versus 1.4% for CROS ≤ 28). Conclusions: Postoperative hypertension and residual obstruction measured by the CROS were associated with elevated recoarctation risk in children with CoA. As an internally derived exploratory tool from a single-center cohort with limited recoarctation events (n = 20), the CROS integrates collinear hemodynamic indices to facilitate risk evaluation and reduce multicollinearity bias. The CROS cut-off value of 28 is population-specific for the present cohort and should not be interpreted as a well-established universal clinical threshold.