Video-Based Assessment of Tonsil Size in Children With Sleep-Disordered Breathing

To evaluate whether parent-captured oropharyngeal videos obtained using a handheld telemedicine device can assess tonsil size in children with sleep-disordered breathing compared with in-person clinical examination.

We conducted a prospective diagnostic accuracy study at a single tertiary pediatric referral center. A consecutive sample of 169 children scheduled for tonsillectomy, adenoidectomy, or adenotonsillectomy was enrolled. Parents obtained oropharyngeal videos using a handheld telemedicine device (TytoCare). Three blinded pediatric otolaryngologists independently graded tonsil size using the Brodsky scale. Interrater agreement, diagnostic performance relative to in-person examination, and video usability were assessed.

Among 169 children (median age 4.4 years, IQR 3.3-7.6), 53% underwent adenotonsillectomy and 46% adenoidectomy alone. Interrater agreement for tonsil size grading was strong (Kendall's W = 0.88, p < 0.01) and remained substantial when grades were dichotomized as 0-2 versus 3-4 (Fleiss κ = 0.65, 95% CI: 0.58-0.73). Using the in-person examination as the reference standard, sensitivity was 70%, specificity was 86%, positive predictive value was 84%, and negative predictive value was 73%. Twenty-one videos (12%) were unusable by all evaluators and 54 (32%) by at least one evaluator, with higher rates in younger children.

Parent-captured oropharyngeal videos may provide a feasible method for remote tonsil size assessment in selected children with sleep-disordered breathing. Strong interrater agreement and generally favorable parental attitudes support further evaluation of this approach, while additional validation is needed before routine reliance on video-based assessment for surgical decision-making.