Small-angle strabismus detection by telemedicine in an experimental model

To evaluate the ability of observers to accurately detect strabismus using an alternate cover testing approach via telemedicine and to assess the effect of various factors related to video conditions on the accuracy rate.

Videos were made by the authors in which different angles of strabismus, up to 12^Δ, were induced by prism and recorded using alternate cover testing. The videos were made under a variety of conditions that incorporated various head postures, lighting, fixation target position and motion, and viewing angle. The videos were shown to observers of varying levels of expertise. The accuracy of detecting the deviation was assessed and analyzed for statistical significance.

The overall rate of detection was significantly affected by the angle of deviation, with 12^Δ being correctly detected with 94% accuracy, 8^Δ with 72% accuracy, and 4^Δ with <50% accuracy. Esotropia and hypertropia were more accurately detected than exotropia for all angles tested. The level of training of the observer did not correlate with detection accuracy. However, accuracy was negatively affected by backlighting and movement of the fixation target during cover testing.

Our data demonstrate that detection of even relatively small angles of strabismus can be accomplished with a high level of accuracy using telemedicine and an alternate cover testing strategy. For optimal results, it is important to consider conditions related to the fixation target and lighting.