Retinopathy of prematurity: associated factors and prediction

Abstract

Introduction  Retinopathy of prematurity (ROP) is a vasoproliferative disease that affects the retina of prematurely born children. Therefore, this study proposed to identify associated factors and to predict retinopathy in prematurity.

Method  A cross-sectional study including preterm newborns divided into two groups: with and without retinopathy. The variables were categorized using the likelihood ratio test, and the variables with p-values less than 0.2 were analyzed by a multiple logistic model. The model used the construction of a Receiver Operating Characteristic (ROC) curve as predicting tool.

Results  Of the 296 newborns, 32 (10.81%) had retinopathy. The following variables represented a risk factor according to the logistic model: gestational age ≤ 30 weeks (p-value = 0.0214), birth weight ≤ 1250 grams (p-value = 0.0013) and Apgar score at 1 minute <7 (p-value = 0.0352). The ROC curve with sensitivity of 0.771 and specificity of 0.774 was a precise model, showing accuracy of 0.8448 and a probability point of 0.79.

Conclusions  This model can be used to predict retinopathy in premature newborns with a risk factor of 0.79, emphasizing that the need for additional attention.

Keywords:  retinopathy of prematurity; risk factors; logistic models