RÉSUMÉ
OBJECTIVE: To detect the screening efficiency of general ophthalmologists (ophthalmic residents) as well as non-ophthalmologists (pediatric residents and nurses posted in neonatal intensive care unit) in screening (ROP) retinopathy of prematurity on the basis of posterior pole vascular changes. METHODS: Prospective consecutive review in a tertiary care hospital setting. Five groups (each comprising of one ophthalmic resident, one pediatric resident and a nurse) examined the posterior pole vessels of 200 eyes of ROP with a direct ophthalmoscope and compared with an ROP specialist using indirect ophthalmoscope. SPSS (Statistical Package for the Social Science), version 10.0 was used for the analysis. RESULTS: Ophthalmic residents findings were: (sensitivity 95.68%, specificity 92.85%, positive predictive value 94.81%, negative predictive value 93.97%; pediatric residents findings were : (sensitivity 92.24%, specificity 88.09%, positive predictive value 91.45%, negative predictive value 89.15%); and nurses, finding were: (sensitivity 88.79%, specificity 85.71%, positive predictive value 89.56%, and negative predictive value 84.70%). The results had no statistically significant difference in diagnostic reliability. Kappa agreement analysis was significant for ophthalmic residents (0.887), pediatric residents (0.805) and nurses (0.744) compared with the ROP specialist. None of the children diagnosed with pre-threshold or threshold ROP was thought to have normal posterior pole vessels by the trainees. CONCLUSIONS: Given adequate training, general ophthalmologists and non-ophthalmologists (pediatricians and nurse practitioners) are independently reliable in detecting posterior pole changes in ROP babies using direct ophthalmoscope and can be provided with a screening protocol.