Abstract
Purpose: :
To determine the accuracy and sensitivity of a single-field nonmydriatic digital fundus image interpreted by an ophthalmologist and performed within a primary care setting.
Methods: :
Fundus photography using a digital non-mydriatic camera (Topcon NW6s) was performed on both eyes of 1175 consecutive patients as part of an executive health program. The fundus images of all patients included a 45 degree field of the posterior pole capturing the optic nerve and macular area. Diagnostic findings were recorded and appropriate recommendations for follow-up were made. Patients were then contacted to see if appropriate follow-up was successfully completed and chart reviews were performed to determine biomicroscopic findings.
Results: :
Photographs were adequate in both eyes in 1117 patients (95.1%). Exam findings were normal in both eyes in 944 84.5% of patients. Abnormal findings were noted in either eye in 173 (15.5%) patients. The most common abnormal findings were macular degeneration (57/173, 32.9%), optic nerve cupping (45/173, 26.0%), hypertensive retinopathy (15/173, 8.7%), and choroidal nevi (10/173, 5.8%). In all patients with abnormal findings, routine follow-up ophthalmologic examination with an eye care specialist was indicated, and none of the patients required urgent attention. Sensitivity was found to be 87% and stratification was performed based on the initial diagnosis. Those with false positive results were from confounding diagnoses rather than true false positives.
Conclusions: :
Single field non-mydriatic fundus photography is accurate and sensitive for screening retinal disease in a primary care setting.
Keywords: imaging/image analysis: clinical • retina