To the Editor:

Retinopathy of prematurity (ROP) is a leading cause of preventable childhood blindness among premature infants worldwide [1]. The development of neonatal intensive care units increased the survival rate of preterm infants, and consequently the incidence of ROP and its long-term sequelae [1].

Infants with ROP can develop early and late visual disabilities that may lead to poor neurodevelopmental (ND) outcomes [1]. Independent of visual deficits, premature infants are at increased risk for ND disabilities due to its associated brain damage [1,2,3]. Whether the severity of ROP, its treatment, its visual sequelae, or the prematurity-associated comorbidities is the reason for the resultant poorer ND outcomes in these children is unknown. While some studies demonstrated that ROP severity is a marker for subsequent functional disability, particularly in the presence of unfavourable visual acuity [4], others concluded that neither ROP severity nor its treatment was related to ND outcomes [1, 5].

To further examine ND outcomes in infants with ROP, we performed a retrospective analysis of premature infants between 2007 and 2016 using the IBM MarketScan database. Conditions and outcomes were identified using International Classification of Disease 9th and 10th editions diagnosis codes. Five major ND outcomes were compared between infants with ROP (±treatment) and premature infants without ROP using univariate and multivariate logistic regression analyses. The Stanford University Institutional Review Board ruled this analysis of de-identified administrative data exempt from approval.

The study population included 79,382 premature infants (Table 1). More infants with treated ROP had extremely low birth weight (<1000 g) and a gestational age of <30 weeks compared with infants with untreated ROP and premature infants without ROP. In addition, a larger proportion of these infants had multiple comorbidities, intellectual disabilities (75.2%), psychiatric and behavioural disorders (43.2%), speech and language impairment (32.4%), motor deficits (18%), and hearing loss (23.4%) at 1 and 2 years of age compared with the other groups. Multivariate logistic regression analysis revealed that both infants with treated and untreated ROP had increased odds of intellectual disabilities (OR 2.83, 95% CI 1.94–4.12; OR 1.70, 95% CI 1.57–1.83), psychiatric and behavioural disorders (OR 1.62, 95% CI 1.17–2.23; OR 1.36, 95% CI 1.23–1.49), speech and language impairment (OR 1.91, 95% CI 1.40–2.66; OR 1.34, 95% CI 1.22–1.48), motor deficits (OR 1.38, 95% CI 0.93–2.03; OR 1.73, 95% CI 1.53–1.96), and hearing loss (OR 1.51, 95% CI 1.07–2.14; OR 1.41, 95% CI 1.26–1.57) (Table 2).

Table 1 Baseline demographic and clinical characteristics of premature infants with treatment-requiring retinopathy of prematurity (treated ROP), no treatment-requiring ROP (untreated ROP), and no ROP in the IBM MarketScan database 2007–2016.
Full size table
Table 2 Multinomial logistic regression model for odd of adverse neurodevelopmental outcomes within 1 and 2 years of age in premature infants with treatment-requiring retinopathy of prematurity (treated ROP) and no treatment-requiring ROP (untreated ROP) in the IBM MarketScan database 2007–2016.
Full size table

Our large cohort study confirms that infants with ROP have worse ND outcomes compared with premature infants without ROP. We demonstrate that severity of ROP, as reflected by treatment-requiring disease, is likely associated with worse outcomes. The limitations of this study include lack of visual acuity data and reliance on diagnostic and procedural codes. An inability to account for coding or billing errors may have resulted in a falsely low number of treated ROP patients. Finally, infants who are covered by Medicaid (which may be up to 40% of insured infants) are not represented in this study. In conclusion, our findings emphasize the importance of ND assessment and monitoring along with ophthalmic monitoring.