Development and Validation of a Multivariate Nomogram for Predicting Retinopathy of Prematurity in Infants with Gestational Age ≤ 34 Weeks

Leilei ShenRuixue ZhengXiaodong SunSheng Chen^*

• Department of Pediatrics, First Affiliated Hospital of Army Medical University, Chongqing, China

Purpose: To delineate risk factors and develop a predictive nomogram for retinopathy of prematurity (ROP) in infants with gestational age (GA)≤34 weeks.We conducted a comprehensive retrospective analysis of infants with GA ≤34 weeks, divided into ROP and non-ROP groups based on fundus screening results. Clinical and laboratory data were collected to identify risk factors associated with ROP. Multivariable logistic regression was performed to identify independent predictors, and a nomogram was developed to predict the occurrence of ROP in infants with GA ≤34 weeks.Results: Our analysis identified five independent risk factors for ROP in infants with GA ≤34 weeks: hypertensive disorders of pregnancy (HDP), number of blood transfusions, oxygen therapy time (OTT), oxygen therapy concentration (OTC) >50%, and blood glucose spikes in 2 the first postnatal week. These predictors were incorporated into a nomogram to estimate individual ROP risk. The predictive model achieved a C-index of 0.923 (95% CI: 0.888-0.959), indicating high predictive accuracy. Internal validation of the nomogram demonstrated excellent calibration and practical utility for clinical decision-making. Conclusions: The established predictive model, incorporating five key clinical parameters, offers clinicians a practical instrument to stratify ROP risk in neonates born at ≤34 weeks' gestation. This clinical tool demonstrates significant utility in guiding intervention protocols, potentially enhancing patient outcomes through early identification and optimized management strategies. Strengths and limitations of this study: Strengths 1. Identifies key clinical and laboratory risk factors associated with ROP, providing valuable insights for early diagnosis and intervention. 2. Develops and validates a robust predictive nomogram with high accuracy (C-index: 0.923) and clinical utility for neonatal care.3. Provides a practical and user-friendly predictive tool to aid clinical decision-making and optimize management strategies for ROP risk stratification.1. The study is limited to a single center, which may reduce the generalizability of the findings to broader neonatal populations.2. The predictive model has not been externally validated, necessitating further testing in diverse clinical environments to ensure robustness and applicability.