Predictive Efficacy of CAPE-V, pVHI, and CVA in Detecting Vocal Fold Pathology in Children.

Abstract

To evaluate the diagnostic accuracy of the Consensus Auditory-Perceptual Evaluation of Voice (CAPE-V), Pediatric Voice Handicap Index (pVHI), and computer-assisted voice analysis (CVA) against flexible fiberoptic laryngoscopy in a pediatric cohort. A retrospective cohort of 116 children (4-18 years) underwent same-day CAPE-V, pVHI, CVA (Visipitch), and flexible fiberoptic laryngoscopy at a tertiary pediatric voice clinic. Encounters were classified as dysphonia evaluation versus vocal fold surveillance, and age was grouped as preschool (≤ 5 years) versus school-aged (> 5 years). Diagnostic performance was assessed with receiver operating characteristic (ROC) curves, area under the curve (AUC), and optimal cutpoints by the Youden index. Multivariable logistic regression-adjusted for age category and encounter type-modeled odds of abnormal endoscopic findings. CAPE-V demonstrated the highest individual accuracy (AUC 0.827; p = 0.0034). Combined use of CAPE-V, pVHI, and CVA yielded superior discrimination (AUC 0.846; sensitivity 0.886; specificity 0.771; positive predictive value 0.886; negative predictive value 0.771). Dysphonia encounters had lower odds of abnormal findings compared to vocal fold surveillance (OR 0.09; 95% CI 0.02-0.36). School-aged children had lower odds of pathology than preschoolers (OR 0.18; 95% CI 0.03-0.09). Peak sensitivity occurred at Age 7; the probability of abnormal findings declined with increasing age. CAPE-V is a robust single-test predictor of pediatric vocal fold pathology, and integration with pVHI and CVA enhances diagnostic accuracy. A tiered, age-tailored screening algorithm beginning with CAPE-V may optimize referral pathways and improve early detection of vocal fold pathology while preserving laryngoscopy as the diagnostic standard. Level 3 (retrospective cohort study).