Sir,

We read with interest the article ‘The effect of erythropoietin on the severity of retinopathy of prematurity’.1 Even if the results are interesting, we have a few concerns and comments. The use of sedatives to screen all the babies for ROP is surprising and can be risky for these vulnerable infants.2 The babies with stage 3 ROP were treated when they reached threshold stage despite ETROP recommendation3 to treat babies, at high risk, prethreshold, and beyond. This questions the efficiency of the treatment strategy followed in the study. The images of the infants needing treatment were reviewed by another examiner before treatment probably implies that the findings of the screening physicians required reconfirmation. In such a case, babies labelled with severe ROP after erythropoietin injection should also have been reconfirmed by a retina specialist. Although the indication to give erythropoietin was under the discretion of the paediatrician, some details like baseline haemoglobin concentration and platelet count at the time of erythropoietin injection would have been more informative as they are known to affect the severity of ROP independent of erythropoietin. Reports say that effect of erythropoietin on ROP depends whether it was given at early or late post-natal life.4 So a similar division in the study could have given extra information.

The cumulative dose of erythropoietin received by neonate with mild ROP (3200 units) was more than those with severe disease (2750 units).To establish erythropoietin as a contributory cause, it is important for the cause to not only have statistical significance but also alter the effect on altering the cause with a proven dose–response relationship, which the study fails to show.