Editor,—I have some comments on the recently published article by Choi et al 1 dealing with long term refractive outcome and oculometry variables in Korean children of very preterm delivery. As for the sample under study (n=65) there are certain points to suggest a marked deviation in composition from the usually analysed preterm cohorts. The material appears highly selective; over a 6 year period, from two university clinics, only 10–11 preterm infants have been included per year. Screening limits were 1500 g birth weight and 28 weeks' gestational age. Exclusion of a great number of preterms appears likely, but criteria are not specified or discussed.

Eighty three per cent acquired active ROP of at least stage 3. If unselected, this is the highest figure of advanced (and of any) ROP ever …