Prescribing Yoked Prism in Patients with Autism Spectrum Disorder (ASD): A Study of Inter- and Intraexaminer Agreement Using the Kaplan Nonverbal Battery

Abstract

Background Double-blind studies have reported improvements in posture, spatial awareness, and behavior in Autism Spectrum Disorder (ASD) individuals wearing yoked prism lenses. The Kaplan Nonverbal Battery (KNB) is a performance-based test used to evaluate yoked prism lenses. To date, only Kaplan has reported KNB results. This study investigates inter-examiner and intra-examiner agreement obtained in scoring the KNB in nonverbal/ minimally verbal ASD subjects. Methods Ten nonverbal/minimally verbal ASD subjects between 9 and 17 years old enrolled and completed an eye examination. All subjects adapted to spectacles, wearing refractive correction for four weeks. Subjects completed two study visits that were videotaped. At each visit, subjects completed the KNB wearing habitual spectacle correction and then wearing yoked prism lenses of five-prism diopters base up or down. Two examiners scored the KNB live and two other examiners scored the KNB via videorecording. Examiners were masked to the base direction of the yoked prism lenses and to the other examiners’ responses. Agreement between and within examiners was computed using the weighted version of Cohen’s kappa. Results Nine subjects completed both study visits. Both inter-examiner and intra-examiner agreement were markedly better when examiners scored via video. When scored by live presentation, inter-examiner agreement by the weighted Cohen’s kappa coefficient was 0.674, 0.364, 0.726 and 0.357. When scored via video, inter-examiner agreement was 0.634, 0.620,0.702 and 0.678. Intra-examiner agreement for live presentation was 0.451 and 0.579, but increased for video scoring to 0.861 and 0.875. Conclusions Examiners assessing the KNB to evaluate yoked prism lenses for nonverbal or minimally verbal ASD children consistently showed good inter-examiner and intra-examiner agreement when evaluating by video, but not by live presentation. Clinicians may improve reliability by incorporating scoring of the KNB by video presentation.