Fleck et al. have noted international differences in inter-observer variation in ROP grading and management between UK and Australian/New Zealand (ANZ) ophthalmologists [1, 2].

Eight consultant ophthalmologist ROP screeners in Scotland graded RetCam (Natus Medical, Pleasanton, CA, USA) images of 25 eyes of 25 infants regarding the presence and severity of plus disease, and stated their management decision (‘treatment’ or ‘no treatment’). Clinical information on gestational age at birth, birth weight, and post-gestational age at the time of imaging was provided. Graders viewed the revised standard ICROP photograph for plus disease prior to commencing the task [3]. Thirteen of the 25 eyes were treated for ROP.

Inter-observer agreement (Fleiss-Kappa, where 0.21–0.40 is considered ‘fair’) was 0.34 (95% CI = 0.28–0.39) for the presence of plus and 0.40 (95% CI = 0.33–0.48) for management decision, slightly greater than that reported by Fleck et al. for UK ophthalmologists (0.2 and 0.33, respectively) [2].

Our data supports the poorer degree of inter-observer agreement found for UK ophthalmologists compared with those of the ANZ group, who had lower treatment rates, although comparable visual outcomes, in the BOOST II trials [4]. The ANZ grading group self-certified using a training website. Given that a possible tendency towards over-treating ROP could result in unnecessary morbidity, the development of a similar online tool to improve standardisation and management decisions would be of potential benefit to UK ophthalmologists.