Severe retinopathy of prematurity and its association with different ratesof survival in infants of less than 1251 g birth weight

Background-There is controversy over whether improved survival of preterm i nfants has resulted in a higher incidence of severe (grade 3 or greater) re tinopathy of prematurity (ROP). Aim-To compare survival rates and rates of greater than or equal to stage 3 ROP-that is, with a high risk of sequelae-in preterm infants in five Engli sh cities where, anecdotally, the incidence of ROP is reported to show cons iderable variation. Methods-All infants of birth weight < 1500 g and or gestational age < 32 we eks, born in 1994 in one of the cities or transferred in within 48 hours, w ere studied. The populations were adjusted for case mix variation using CRI B (clinical risk index for babies, a disease severity scoring system). The incidence of severe ROP, the actual death rate, and that adjusted for disea se severity were determined. Results-The rate of severe ROP per 1000 births was higher in city 1 than in all the other cities. This increase in comparison with city 2 and city 4 w as significant (city 1, 167 (95% confidence interval (CI) 96 to 260); city 2, 24 (6 to 59); city 4, 16 (1 to 84)). A significant difference was not se en between city 1 and cities 3 (23 (1 to 120)) and 5 (74 (21 to 79)). The r elative risk of developing severe ROP in city 1 compared with all the other cities was 5.5 (2.5 to 11.9). The actual death rate per 1000 births in cit y 1 was significantly lower than that predicted by modelling death against CRIB score (city 1: actual 270; predicted 385 (95% CI 339 to 431)). In cont rast, the other cities had actual death rates as predicted, or worse than p redicted, by CRIB. Interpretation-A significantly higher incidence of severe ROP was identifie d in one of the five cities studied. Variation in survival rates among high risk infants may explain this observation.