Trends in Retinopathy of Prematurity over 10 years in a Colorado Cohort

Abstract

Purpose : To determine trends in retinopathy of prematurity (ROP) in infants born at risk for ROP between January 2006 and April 2017.

Methods : A retrospective cohort study was conducted using records from an ROP registry. All infants fulfilled the American Academy of Pediatrics 2013 screening criteria for ROP and were delivered at the University of Colorado Hospital or Children’s Hospital Colorado (tertiary care facilities) between January 2006 and April 2017. ROP was graded using the International Classification of ROP. The main outcome was development of ROP with maximum grade of retinopathy defined as the highest stage and lowest zone of ROP in the worse eye. ROP was categorized as type 1 or type 2 ROP versus low grade or no ROP. Other variables included in the trend analysis were gestational age and birth weight at delivery, and infant mortality. Univariate spline models were fit to assess the linearity of trends over time by birth date (in days) as a continuous variable for each of the variables.

Results : The cohort contained 1462 total infants. 1336 infants survived long enough to undergo ROP screening, of which 134 (10%) infants were diagnosed with type 1 or type 2 ROP (71 type 1 [5.3%] and 63 type 2 [4.7%]), 308 (23%) had low grade ROP, and 894 (67%) had no ROP. There was no overall trend in ROP rates (shown for type 1 and type 2 ROP in Figure 1), infant deaths or gestational age over the observation period (p= 0.88, p = 0.14, and p = 0.30, respectively). However, for the entire study period, there was a marginal decreasing linear trend over time for birth weight (p = 0.07).

Conclusions : We found relatively low rates of type 1 and type 2 ROP in this cohort and no significant trend in the incidence of ROP during the 10-year study period. Although it is difficult to assess the impact of potential changes in neonatal care, our data demonstrates the validity of combining patients treated over the last decade into larger cohorts for analysis.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.