[Consistency and risk factor analysis of 2 growth chart in the diagnosis of extrauterine growth restriction in preterm infants with a gestational age at birth of less than 32 weeks].

Objective: To investigate the diagnostic consistency and the risk factors for extrauterine growth restriction (EUGR) assessed by the Fenton 2013 preterm growth charts (Fenton 2013) and the growth charts by International Fetal and Newborn Growth Consortium for the 21st Century (IG-21). Methods: This multicenter retrospective cohort study included 5 591 preterm infants with a gestational age (GA) at birth of less than 32 weeks admitted to 19 member hospitals of the Neonatal Perinatal Collaborative Network of Suxinyun from 2019 to 2024. Clinical data including baseline characteristics, complications, feeding practices and anthropometrics were processed and analyzed. EUGR was assessed using both the Fenton 2013 and the IG-21. A decrease in weight Z-score at discharge compared to admission by more than 1 was defined as longitudinal EUGR, and discharge weight below the P10 for the corresponding corrected GA was defined as cross-sectional EUGR. Diagnostic consistency was assessed using the Kappa coefficient between the 2 standards, and diagnostic performance of the 2 standards was compared using the McNemar test. Risk factors for EUGR under different definitions were analyzed using univariate analysis and multivariate Logistic regression analysis. Results: A total of 5 591 preterm infants were included, with a GA at birth of (29.7±1.6) weeks and a birth weight of (1 360±315) g and at discharge with a corrected GA of (36.3±2.0) weeks and weight of (2 246±370) g. Detection rates of cross-sectional and longitudinal EUGR diagnosed by Fenton 2013 were both higher than those by IG-21 (37.0% (2 214/5 991) vs. 23.7% (1 324/5 591), 61.1% (3 662/5 991) vs. 30.7% (1 714/5 591), χ2=326.77 and 1 358.05, both P<0.001). Using Fenton 2013 as a reference, IG-21 demonstrated superior diagnostic value and consistency in identifying cross-sectional EUGR compared with longitudinal EUGR (sensitivity of 100.0% (3 377/3 377) vs. 99.6% (1 922/1 929), specificity of 59.8% (1 324/2 214) vs. 46.6% (1 707/3 662), positive predictive value of 79.1% (3 377/4 267) vs. 49.6% (1 922/3 877), negative predictive value of 100.0% (1 324/1 324) vs. 99.6% (1 707/1 714), accuracy of 84.1% (4 701/5 591) vs. 64.9% (3 629/5 591), and Kappa 0.64 vs. 0.37, all P<0.001). In multivariate Logistic regression models, risk factors common to EUGR across both standards included smaller GA at birth, lower birth weight, boy, early-onset sepsis, late-onset sepsis and the elder age at full enteral feeding (all P<0.05). Hemodynamically significant patent ductus arteriosus remained an independent risk factor for longitudinal EUGR regardless of whether by the Fenton 2013 or IG-21 standard (adjust odds ratio (aOR)=1.25 and 1.27, 95%CI 1.09-1.42 and 1.11-1.45). In addition, under the IG-21 standard, severe bronchopulmonary dysplasia was an independent risk factor for cross-sectional EUGR (aOR=1.54, 95%CI 1.00-2.38), while severe necrotizing enterocolitis was an independent risk factor for longitudinal EUGR (aOR=2.18, 95%CI 1.01-4.73). Conclusions: IG-21 showed lower detection rates of both cross-sectional and longitudinal EUGR than Fenton 2013, suggesting greater clinical applicability of IG-21 by reducing overdiagnosis while maintaining sensitivity for predicting complications. Across both standards, cross-sectional EUGR facilitates early identification of growth restriction, whereas longitudinal EUGR better tracks dynamic growth patterns and complications of preterm infants.