Figure 3: Diagrams displaying symptom-specific algorithms for (left to right) ear pain, runny ear and hearing loss. Demographic data, as well as past medical history, drug history and allergies were also collected by the questionnaire.
Next, the same nurse administering the questionnaire captured brief otoscopic videos of each ear using the Cupris TYM device (a platform we have previously shown to be valid for expert tele-diagnosis5) attached to an Apple iPhone 5S mobile phone (Apple Inc., Cupertino). Videos were encrypted and stored. Because nurses were naïve to the device, they were given a short tutorial (author [removed for blind peer review]) before the study commenced, and trialled the device on ten consenting individuals.
Finally, each participant underwent assessment by one of two resident ENT specialists. They took a targeted clinical history (with open responses) in the Khmer language, performed oto-endoscopy, and recorded their diagnosis. In cases where wax precluded a full oto-endoscopic view, data were removed from final analyses. Specialist assessment was blind to previous assessment, and conducted in a separate area.