Can We Optimize Retinopathy of Prematurity Screening by Combined Risk Score Analysis?

Abstract

To evaluate a composite metric incorporating the DIGIROP-Birth (DRB) and Neonatal Research Network Bronchopulmonary Dysplasia Outcome Estimator (NRN-BPD) for predicting severe retinopathy of prematurity requiring treatment (TR-ROP). This was a retrospective cohort study of 990 infants born prematurely undergoing dilated eye examinations between January 2010 and August 2023. We performed a chart review to assess a primary outcome of TR-ROP and secondary outcome of stage 2 or greater ROP. DRB and 14-day NRN-BPD scores were quantified for each infant, and optimal thresholds for predicting TR-ROP were analyzed using receiver operator characteristic curves. Sensitivity and specificity for TR-ROP were assessed. Of the 990 infants, 364 (36.8%) had stage 2 or greater ROP and 68 (6.9%) had TR-ROP. Receiver operator characteristic analysis with (95% CI) showed areas under the curve of 0.867 (0.829-0.906) for DRB and 0.845 (0.809-0.881) for NRN-BPD. Optimal cutoff scores were 1.7 for DRB and 40% for NRN-BPD with respective sensitivities of 97% and 96%. Composite screening for babies meeting either cutoff allowed 100% sensitivity for predicting TR-ROP while decreasing number of infants qualifying for screening from 990 to 562 (43% reduction). A composite metric using DigiROP-Birth and NRN-BPD Outcome Estimator scores may allow an early, simple approach to predict TR-ROP with 100% sensitivity yet allow significant reduction in the number of infants requiring screening eye examinations. Additional large validation studies are needed.