ICROP Third-Edition Phenotypes of Retinopathy of Prematurity and Their Association with Gestational Age in Indonesian Preterm Infants: A Retrospective Study

Abstract

Background: Retinopathy of prematurity (ROP) is a leading cause of preventable childhood blindness, and middle-income countries face a third epidemic as neonatal survival outpaces screening capacity. Indonesian data using the third ICROP edition are scarce.

Objective: To describe the ICROP third-edition phenotype distribution among Indonesian infants with ROP and evaluate the associations of severity with gestational age and birth weight using exact inference and partial identification.

Methods: This retrospective STROBE-compliant study characterised the ICROP phenotype of 39 infants with ROP at a tertiary referral hospital in Denpasar between January 2020 and December 2022, and examined severity against gestational age and birth weight. Proportions carry 95% Wilson intervals; the gestational-age-by-stage cross-classification was reconstructed under the published marginals, tested by exact methods, and bounded across every reconstruction.

Results: Mean gestational age was 28.7 ± 2.7 weeks and mean birth weight was 1091 ± 320 g; 94.9% were of very low birth weight. Stage was documented for 38 infants: stage 1, 68.4%; stage 2, 28.9%; and stage 3, 2.6%. Pre-plus and plus disease each affected 7.7%, and aggressive ROP affected 2.6%. Stage ≥2 occurred in 70.0% of infants born before 28 weeks versus 17.9% of those born later (odds ratio 9.16, 95% CI 1.90–44.10; exact P = 0.005; Cramér’s V = 0.370), with a monotone trend across strata (exact P = 0.004). Across all 30 admissible reconstructions, the exact P ranged from 0.004 to 0.115 and the odds ratio from 0.03 to 9.16. Birth weight carried no usable information because 94.9% of the cohort occupied a single stratum.

Conclusion: Gestational age, not birth weight, should anchor ROP surveillance intensity in Indonesian tertiary units, while a broad birth-weight entry threshold is retained.