Epidemiology of Retinopathy of Prematurity in the US From 2003 to 2019

Abstract

ImportanceRetinopathy of prematurity (ROP) is a potentially blinding retinal disease with poorly defined epidemiology. Understanding of which infants are most at risk for developing ROP may foster targeted detection and prevention efforts.ObjectiveTo identify changes in ROP incidence in the US from 2003 to 2019.Design, Setting, and ParticipantsThis retrospective database cohort study used the Healthcare Cost and Utilization Project Kids’ Inpatient Databases. These nationwide databases are produced every 3 years, include data from over 4000 hospitals, and are designed to generate national estimates of health care trends in the US. Participants included pediatric newborns at risk for ROP development between 2003 and 2019. Data were analyzed from September 30, 2021, to January 13, 2022.ExposuresPremature or low-birth-weight infants with relevant International Classification of Diseases, Ninth Revision or Tenth Revision codes were considered ROP candidates. Infants with ROP were identified using relevant codes.Main Outcomes and MeasuresROP incidence in selected subpopulations (based on database-reported race and ethnicity, sex, location, income) was measured. To determine whether incidences varied across time or subpopulations, χ2 tests of independence were used.ResultsThis study included 125 212 ROP discharges (64 715 male infants [51.7%]) from 23 187 683 births. The proportion of premature infants diagnosed with ROP increased from 4.4% (11 720 of 265 650) in 2003 to 8.1% (27 160 of 336 117) in 2019. Premature infants from the lowest median household income quartile had the greatest proportional increase of ROP diagnoses from 4.9% (3244 of 66 871) to 9.0% (9386 of 104 235; P < .001). Premature Black infants experienced the largest increase from 5.8% (2124 of 36 476) to 11.6% (7430 of 63 925; P < .001) relative to other groups (2.71%; 95% CI, 2.56%-2.87%; P < .001). Hispanic infants experienced the second largest increase from 4.6% (1796 of 39 106) to 8.2% (4675 of 57 298; P < .001) relative to other groups (−0.16%; 95% CI, −0.29% to −0.03%; P = .02). The Southern US experienced the greatest proportional growth of ROP diagnoses, increasing from 3.7% (3930 of 106 772) to 8.3% (11 952 of 144 013; P < .001) relative to other groups (1.61%; 95% CI, 1.51%-1.71%; P < .001). ROP diagnoses proportionally increased in urban areas and decreased in rural areas.Conclusions and RelevanceThis cohort study found that ROP incidence among premature infants increased from 2003 to 2019, especially among Black and Hispanic infants. Infants from the lowest-income areas persistently had the highest proportional incidence of ROP, and all regions experienced a significant increase in ROP incidence with the most drastic changes occurring in the South. These trends suggest that ROP is a growing problem in the US and may be disproportionately affecting historically marginalized groups.

Epidemiology of Retinopathy of Prematurity in the US From 2003 to 2019

“…Despite our alternative explanation for the observed trend reported by Bhatnagar et al, it is crucial to recognize that ROP remains a very serious and blinding condition; our comments should not be misinterpreted as suggesting anything to the contrary. However, in the absence of manual record reviews or prospective observational trials, we would like to offer that the terminology “incidence of coding for ROP” replace “incidence of ROP” in any retrospective reports that collect data solely on the basis of diagnosis and procedure codes.…”

“…To the Editor The recent article by Bhatnagar et al described a relatively constant incidence of retinopathy of prematurity (ROP) from 2003 until about 2006, followed by a relatively linear rise until 2012, followed again by a relatively constant incidence (Figure 1). This increase in ROP over time is surprising given the relative stability in factors that correlate with ROP; furthermore, this type of incidence curve does not appear to be physiologic.…”

Nationwide Increase in the Incidence of Retinopathy of Prematurity in the US—A Growing Problem?

“…Although both studies used the same databases from somewhat different time periods, the reported ROP incidence rates are substantially different, likely due to the differences in defining the target population (eg, different denominator for incidence rate calculation). The study by Bhatnagar et al defined the eligible study population using birth weight 1500 g or less or gestational age 36 weeks or less, while the study by Ludwig et al used premature infants with length of stay in hospital longer than 28 days. A recent study by Thangamathesvaran et al analyzed the National Inpatient Sample databases for ROP incidence in the US between 2009 and 2018 and found that the ROP incidence among premature infants who met US ROP screening criteria (defined as birth weight 1500 g or less or gestational age 30 weeks or less) increased from 11% in 2009 to 15% in 2018.…”

“…In this issue of JAMA Ophthalmology , Bhatnagar et al report the change of ROP incidence in the US between 2003 to 2019. Based on their retrospective analysis of the National Healthcare Cost and Utilization Projects Kids’ Inpatient databases that included more than 125 000 ROP discharges from over 23 million births, they found that ROP incidence rate among premature infants increased from 4.4% in 2003 to 8.1% in 2019 (86% relative increase) nationwide with the largest increase found in Black infants (from 5.8% to 11.6%) and Hispanic infants (4.6% to 8.2%).…”

Epidemiology of Retinopathy of Prematurity in the US From 2003 to 2019—Reply

“…Groth and Donahue offer that the change in screening guidelines should have decreased the observed incidence. This would hold true if our study varied the definition of a ROP candidate from year to year to reflect the changes that occurred in 2006 and 2013. As explained in the Discussion, the criteria used to define an ROP candidate were held constant over the study period and were more inclusive than the most inclusive guidelines.…”