Wide-field digital retinal imaging versus binocular indirect ophthalmoscopy for retinopathy of prematurity screening: a two-observer prospective, randomised comparison

Aim: To compare the diagnostic accuracy of wide-field digital retinal imaging (WFDRI) with the current "gold standard'' of binocular indirect ophthalmoscopy (BIO) for retinopathy of prematurity (ROP) screening examinations. Methods: A consecutive series of premature infants undergoing ROP screening at Edinburgh Royal Infirmary were eligible for recruitment into this prospective, randomised, comparative study. Infants were screened using both WFDRI (Retcam II with neonatal lens) and BIO by two paediatric ophthalmologists who were randomised to the examination technique. Both examiners documented their clinical findings and management plans in a masked fashion. WFDRI eye findings were compared with those of BIO. Results: A total of 81 infants were recruited, and information from 245 eye examinations was analysed. The sensitivity of WFDRI in detecting any stage of ROP, stage 3 ROP and "plus'' disease was 60%, 57% and 80%, respectively, and specificity 91%, 98% and 98%, respectively. The proportional agreement between WFDRI and BIO was 0.96 for detecting stage 3 disease and 0.97 for detecting "plus'' disease. There was very good agreement on management decisions (kappa 0.85). Conclusion: When used in a routine ROP screening setting, a randomised comparison of WFDRI and BIO, WFDRI showed relatively poor sensitivity in detecting mild forms of ROP in the retinal periphery. This resulted in difficulty in making decisions to discharge infants from the screening programme. Sensitivity was better for more severe forms of ROP, but at present WFDRI should be regarded as an adjunct to, rather than a replacement for, BIO in routine ROP screening.