Affiliation:
1. The First Affiliated Hospital of Anhui Medical University
2. Dian Diagnostics Group Co., Ltd
3. Suzhou Hospital of Anhui Medical University
Abstract
Abstract
Background
Neonatal fundus disease lacks typical symptoms, thus early identification and prompt treatment are crucial for preventing visual dysfunction. This study aimed to identify the factors influencing the severity of retinal hemorrhage.
Methods
A total of 507 full-term newborns who underwent fundus disease screening at Suzhou Municipal Hospital from January 2023 to July 2023 were enrolled in this study. All full-term newborns underwent fundus examination using RetCam III shuttle within 7 days of birth, and clinical information on the newborns and their mothers was collected. Univariate and multivariate regression analyses were employed to identify independent risk factors and develop clinical risk nomogram. The risk nomogram was developed using the rms package and evaluated using decision curve analyses.
Results
Among the 507 newborns included in the study, a total of 210 cases exhibited fundus abnormalities. Retinal hemorrhage was identified as the most prevalent fundus abnormality (75.71%), followed by familial exudative vitreoretinopathy (14.76%) and retinopathy of prematurity (3.8%). Univariate and multivariate analysis demonstrated that vaginal delivery (OR = 4.875), binocular hemorrhage (OR = 6.434), and macular involvement (OR = 3.932) as independent risk factors for severe retinal hemorrhage. The C-index of the nomogram model was 0.804, with AUC of the ROC curve at 0.804 (95%CI, 0.733–0.875). The decision curve analysis indicated clinical risk nomogram has application value.
Conclusion
Vaginal delivery, binocular hemorrhage and macular involvement are risk factors for neonatal retinal hemorrhage. The clinical risk model can effectively predict the severity of neonatal retinal hemorrhage.
Publisher
Research Square Platform LLC