A teleophthalmology system for the diagnosis of ocular urgency in remote areas of Brazil

• Autor(a) principal: Ribeiro,Anna Giselle
• Data de Publicação: 2014
• Outros Autores: Rodrigues,Renan Albert Mendonça, Guerreiro,Ana Maria, Regatieri,Caio Vinicius Saito
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Arquivos brasileiros de oftalmologia (Online)
• Texto Completo: http://old.scielo.br/scielo.php?script=sci_arttext&pid=S0004-27492014000400214
• topic: Telemedicine; Remote consultation; Ophthalmology; Diffusion of innovation; Triage; Eye diseases/diagnosis

Purposes: To validate a teleophthalmology mobile system aimed at improving and providing eye urgency screenings in remote and poor area settings in Brazil. The system enables one or more ophthalmologists to remotely examine a patient's condition and submit a decision describing the gravity of the case. If necessary, the patient can be forwarded to a hospital for further consultation. Methods: A cellphone (Nexus One model, with a 5 megapixel camera) was used to collect data and pictures from 100 randomly selected patients at the Ophthalmology Emergency Room located at the General Hospital of the Federal University of São Paulo (UNIFESP). Data was then sent remotely to an online recording system to be reviewed by an ophthalmologist who provided feedback regarding the state of ocular urgency. Results were then compared to the gold standard diagnosis provided at the hospital. Results: The diagnosis of urgency was given by two ophthalmologists: one in the hospital (gold standard) and one remotely. When we compared both diagnoses we obtained results of 81.94% specificity, 92.85% sensitivity, and 85% accuracy, with a negative predictive value of 96.72%. This work also included a processing time analysis, resulting in an average time of 8.6 min per patient for remote consultations. Conclusions: This study is the first that has used only a cellphone for diagnosing the urgency of ocular cases. Based on our results, the system can provide a reliable distinction between urgent and non-urgent situations and can offer a viable alternative for the servicing of underprivileged areas. In screening techniques, the most important outcome is to identify urgent cases with a high level of sensitivity and predictive negative value. Thus, our results demonstrate that this tool is robust and we suggest that a major study aimed to verify its efficiency in resource-poor areas should be initiated.