Long-Term Retinopathy of Prematurity Outcomes with Mandated Screening Interval and Grader Continuity at a High-Volume Quaternary Center

Abstract

Objective: To summarize screening patterns and treatment rates from a quaternary teaching hospital and regional retinopathy of prematurity (ROP) referral center, highlighting mandated screening interval protocols and grader continuity.

Design: Retrospective cohort study.

Methods, intervention, or testing: A retrospective review and descriptive analysis of all patients who underwent inpatient ROP screening within a quaternary teaching hospital from 2013 to 2024. The screening model employed here used continuous screening by assigned graders without handoff and mandated screening interval with binocular indirect ophthalmoscopy (BIO). Variables compared included estimated gestational age, birth weight, weight at each screening, inborn vs outborn status, screening date and intervals, treatment, risk category, and location. ANOVA testing was performed for multiple comparisons, and two-sample T-testing was performed for binary comparisons.

Main outcome measures: Incidence of Treatment-Warranted ROP and rate of retinal detachment.

Results: Within the study period, 15,798 examinations occurred for 1,860 patients. 1,651 (88.8%) patients were only examined by one of four screeners. For each of the four screeners, between 98.1% and 99.3% of examination intervals were weekly or more often. Patients came from 34 different hospitals and 334 (18.0%) were outborn. Outborn (vs inborn) patients had significantly (P < .005) younger mean (±SD) EGA (27.8 ± 3.0 weeks vs 29.0 ± 3.0 weeks) and lower mean (±SD) birthweight (1066.9 ± 408 grams vs 1181.4 ± 416 grams). The screener who accounted for the most treatments also saw more high risk (nano- and micro-premature) infants, with a 19.4% treatment rate of these infants. Volume increased over time, but no seasonal trends were identified (P = .51). A hundred (5.4%) required treatment, more often those who were outborn (4.6% inborn vs 9.0% outborn, P = .002). None developed a retinal detachment.

Conclusion: Twelve years of data from a BIO ROP screening model which relies on single screeners for continuity of care and mandated weekly screening demonstrated excellent outcomes. No infants succumbed to blindness from ROP within this screening model. Grader continuity may have contributed to lower intervention rates.