Telemedicine for corneal disease in rural Nepal.

We evaluated the diagnostic accuracy of slit-lamp photographs interpreted by telemedicine compared to a conventional clinical examination. A convenience sample of 21 patients with anterior segment disease was enrolled at the Lumbini Eye Institute (LEI) in Bhairahawa, Nepal. An ophthalmologist at the LEI examined each patient and assigned a diagnosis and follow-up interval; this represented the gold standard. An ophthalmic technician also obtained anterior segment photographs of each patient. Slit-lamp photographs and clinical histories were then uploaded to a server for access by three separate readers. These readers, one in Nepal and two in the US, provided a diagnosis and follow-up interval independently. The diagnostic agreement between the examiner and all readers was good (kappa = 0.75, P < 0.0001). The agreement regarding follow-up interval between the examiner and all readers was fair, with a kappa coefficient of 0.32 (P < 0.0001). However, the agreement was high when comparing the examiner with the reviewer in Nepal (kappa = 0.90) and was moderate when comparing the two US-based readers with each other (kappa = 0.45). In general, the ophthalmologists in Nepal recommended more rapid follow-up than their US-based counterparts. Our results suggest that the transmission of slit-lamp photographs from satellite clinics and eye health screening camps to the LEI and elsewhere for review and triage is an effective means of identifying anterior segment pathology.