EPOS

European Paediatric Ophthalmological Society

Abstract Preview

VALIDITY OF RETINAL DIGITAL IMAGING IN SCREENING PROGRAMMES FOR ROP

ZARRATEA HERREROS LAURA¹, TEJADA PALACIOS PILAR¹, DE LA CRUZ JAVIER¹, LORA DAVID¹, MORAL M. TERESA¹, CASERIO SONIA¹
¹HOSPITAL UNIVERSITARIO 12 DE OCTUBRE, MADRID, SPAIN

INTRODUCTION: Binocular indirect ophthalmoscopy (BIO) is the reference examination technique in screening programmes for retinopathy of prematurity (ROP) and it requires highly specialised training. RETCAM is a digital device that allows capturing, store and transferring retinal images. This study aims at estimating the validity of RETCAM for ROP screening. METHODS: Observational longitudinal study with prospective data collection. Eligible infants (n=86) for ROP screening (< 1250 g or < 31 weeks) were examined with both BIO and RETCAM (on the same examination, except for those infants (31%) participating in the pain assessment study where interval was 3 days), all in the ROP screening program at University Hospital 12 de Octubre, Madrid-Spain between November 2007 and May 2008. Data were analysed at two levels: 1) at the eye level: for each eye of each infant on each examination during the course of their screening the ophthalmologist determined worst ROP stage, zone and presence of plus disease. 2) At the child level: two standard definitions of severity were applied: close follow-up warranted ROP and referral-warranted ROP. Time to diagnosis was estimated together with Sensitivity (Se), Specificity (Sp) (95% confidence intervals), and other diagnostic validity performance parameters. RESULTS: 314 eyes were examined with both RETCAM and BIO. Quality data were available for 288 eyes. Any stage of ROP was present in 30% of the eyes; Se and Sp of RETCAM were respectively 71.6% (62.6-80.5) and 97.8% (94.4-99.1). Posterior Zone II was present in 30%, Se and Sp respectively 93.3% (86.1-96.9) and 77.4% (71.2- 82.6); both parameters were 100% for plus disease. Close follow-up warranted ROP was identified in 24.4% of children; Se of RETCAM was 84% and Sp 95%. Referral-warranted ROP in 7.7%; both Se and Sp were 100%. CONCLUSIONS: RETCAM always identifies plus disease. RETCAM was highly valid for identifying children warranting referral for ROP examination with BIO.