G579(P) Can the online winrop algorithm predict the develoment of severe retinopathy of prematurity in a multi-ethnic population of preterm babies?

Résumé fourni par la source

Retinopathy of prematurity (ROP), a retinal condition of the preterm retina, can cause blindness if undetected and untreated. Babies at risk (<32w GA and/or <1501g BW) undergo screening to detect ROP requiring treatment (ROP-T). Over 95% of at risk babies do not develop ROP-T1 but are screened and utilise skilled resources. The online algorithm WINROP uses GA, BW and serial weights to identify early babies who subsequently develop ROP-T. It has the potential to improve ROP screening service delivery but it was developed and has been validated in mainly White populations. We investigated WINROP’s ability to identify babies from a multi-ethnic population who subsequently developed ROP-T. This service delivery improvement project was performed in a single NICU using retrospective data. Babies at risk of ROP, born between 01/09/2015 and 30/06/2016, and admitted to the NICU at <48 hours of age were identified. Data on GA, BW, and serial weights within the first 6 weeks of life were entered into WINROP. Its sensitivity and specificity for identifying babies who subsequently developed ROP-T was calculated. 107 seven infants were included. Self-reported maternal ethnicity of the cohort was 37% Black, 29% White, 15% Asian, 15% Other, 3% Mixed, 1% Unknown. WINROP’s sensitivity and specificity for identifying babies who subsequently developed ROP-T were 100% and 54% respectively. The positive predictive and negative predictive values were 17% and 100%, respectively. In this small, multi-ethnic cohort, WINROP identified all babies who subsequently developed ROP-T. WINROP might have a role in improving ROP screening service delivery but further work is required to confidently validate its performance. References Adams GGW, Bunce C, Xing W, et al. Treatment trends for retinopathy of prematurity in the UK: active surveillance study of infants at risk. BMJ Open 2017;7:e013366.doi:10.1136/bmjopen-2016-013366.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.