Validation of the “TWO-ROP” Algorithm at a Multi-Neonatal Intensive Care Unit Tertiary Referral Center

Abstract

PurposeWe aim to validate the previously published TWO-ROP algorithm on an external dataset. DesignRetrospective consecutive study. SubjectsInfants screened for retinopathy of prematurity (ROP) between January 2013 and August 2023 at a tertiary referral multi-site. MethodsInfants with higher birth weight (BW) and longer gestational age (GA) were included and stratified into three groups as follows: group 1 (BW<1500 g, GA≥30 weeks), group 2 (BW≥1500g, GA< 30 weeks), and group 3 (BW≥1500g, GA≥30 weeks). Main Outcome MeasuresThe rate of ROP, treatment-warranted ROP (TW-ROP), and number of inpatient examinations were evaluated in the three groups. Results1,095 (33.8%) patients met the inclusion criteria. The number of patients in groups 1, 2, and 3 were 837 (76.4%), 72 (6.6%), and 186 (17.0%), respectively. ROP was detected in 120 (11.0%) patients; the rate was 9.8% in group 1, 20.8% in group 2, and 12.4% in group 3 (p=0.013). The overall mean number of inpatient examinations for patients undergoing traditional, TWO-ROP 36-week, and TWO-ROP 40-week screening systems were 1.95, 1.43, and 0.99, respectively (p<0.001). Stage 3 was found in 9 eyes of 5 patients (0.5%, all zone II). Three eyes of 2 patients (0.2%) had plus disease. Two patients had bilateral laser treatment at 44 and 39.4 weeks postconceptional age; 3 out of 4 of these eyes met Type 1 treatment criteria. Overall, the ROP screening burden saved was 9.0% and 16.7% for the TWO-ROP 36-week and 40-week systems, respectively. The sensitivity for TW-ROP was 100% for TWO-ROP 36-week system and 99.4% for TWO-ROP 40-week system. ConclusionThe TWO-ROP algorithm can reduce the number of inpatient examinations while maintaining safety. To ensure timely management, we recommend that the single first ROP examination occurs at 38-39 weeks postconceptional age.