Reliability of Publicly Available Digital Endoscopes in Otolaryngology: A Pilot Study

Abstract

To compare the utility, accuracy, and confidence of an inexpensive, commercially available endoscope (not specifically designed for medical use) with traditional methods in the diagnosis of otologic conditions. This is a prospective study. Following institutional review board approval, patients were recruited from a tertiary university-based otology/neurotology clinic. Complete history and physical were obtained by the resident, including both traditional handheld otoscopy and image captured from a commercially available digital endoscopic device (AnyKit Digital Otoscope with 4.5-inch screen). The patient was then presented to the attending without the endoscopic images and a putative diagnosis was made. The endoscopic images were then shown and the putative diagnosis was affirmed or rejected. The attending then examined the patient and determined the final diagnosis using the microscope. Data collected included resident year, resident and attending diagnosis before and after digital-otoscopic images, confidence in diagnosis (1–5 scale), and agreement between the initial putative diagnosis and the final diagnosis. Noninferiority testing was calculated using inter-rater agreement between digital-otoscopic and final diagnoses. Differences between resident and attending confidence were analyzed. A power analysis was performed and the sample size was calculated a priori. A total of 62 participants (114 ears examinations) were enrolled. Cohen’s kappa coefficient showed very high agreement between both resident and attending digital-otoscopic and final diagnosis (kappa = 0.868 and 0.882, respectively) suggesting noninferiority between the digital otoscope and the final diagnosis. There was no significant difference between attending confidence in diagnosis following resident presentation versus attending confidence in diagnosis after reviewing images (4.65 vs 4.61, P = 0.701). Average resident confidence in digital-otoscopic diagnosis remained above 4.2 throughout the study. Inexpensive and readily available digital endoscopes are not inferior to the traditional methods of resident-attending consultation and may provide some substantial benefits. Such devices have the potential to enhance both patient care and resident education when faculty are not immediately available (ie, inpatient and emergency room consults) and improve patient-initiated communications.