Retinopathy of the premature (ROP)

New algorithms for early screening identified

The retinopathy of the premature (ROP) is a disease of the retina caused by neovascularisation resulting from various conditions related to prematurity. Although it is a treatable condition, early diagnosis is important as theclinical evolution of this pathology is rapid and can lead to retinal detachment and blindness.

To date, there are still several "unmet needs'. with regard to ROP screening and the only effective intervention is to routinely subject preterm babies to fundus examinations.

A recent retrospective study published in Acta Ophtalmologica by an Italian team coordinated across three centres (Udine, Padua and Verona) and with transversal expertise, assessed the sensitivity and specificity of three predictive algorithms, named WINROP, ROPScore and CHOP ROP, for ROP screening. The analysis of the effectiveness of the three algorithms involved 445 preterm births mainly with gestational age ? 30 weeks and/or with weight ? 1500 g. The results, which were promising and unique, indicated that two out of three algorithms, namely ROPScore and CHOP ROP, identified with 100% sensitivity the presence of pathology and that this was present in both eyes in 26.1% of the cases. As expected, the gestation age was significantly lower in the group of ROP births. Furthermore, the authors, analysing the area subtended by the curve ROC (Receiver Operating Characteristic - Area Under Curve) for detection of severe ROP reported how the two predictive algorithms performed particularly well in terms of sensitivity. also showed excellent performance in this respect. The tmedian time of 'alarm to treatment' was 5.1 weeks for ROPScore and 9.1 weeks for CHOP ROP, respectively.

Taken together, these data indicate that predictive algorithms could constitute a reliable tool for the early identification of infants requiring referral to an ophthalmologist and to reduce the use of stressful procedures for preterm infants.

Bibliography

Piermarocchi S et al. Predictive algorithms for early detection of retinopathy of prematurity. Acta Ophthalmol. 2017 Mar;95(2):158-164.

Dr. Carmelo Chines
Direttore responsabile