Reliability and Recommended Settings for Pediatric Circumpapillary Retinal Nerve Fiber Layer Imaging Using Hand-Held Optical Coherence Tomography

Abstract

Purpose: To investigate feasibility and reliability of 3-dimensional full circumpapillary retinal nerve fiber layer (cpRNFL) analysis in children, with and without glaucoma, without the use of sedation and to recommend a protocol for hand-held optical coherence tomography use. Methods: A cohort of pediatric glaucoma patients and normal children were imaged with hand-held optical coherence tomography to assess the feasibility of obtaining full cpRNFL. Two consecutive scans were acquired in a smaller sample to investigate test-retest repeatability and interassessor reproducibility. The cpRNFL thickness was assessed in four quadrants, at several visual angles from the optic nerve center. Results: Scanning was attempted in both eyes of 90 children with pediatric glaucoma and 180 controls to investigate feasibility (mean age, 6.98 ± 4.42 years). Scanning was not possible in 68 eyes of glaucoma children mainly owing to nystagmus, unclear optical media, or high refractive errors. Where three-dimensional imaging was possible, success at obtaining full cpRNFL was 67% in children with glaucoma and 89% for controls. Seventeen children with pediatric glaucoma and 34 controls contributed to reliability analysis (mean age, 6.3 ± 3.63 years). For repeatability intraclass correlation coefficients across quadrants ranged from 0.63 to 0.82 at 4° and improved to 0.88 to 0.94 at 6°. Intraclass correlation coefficients for reproducibility were also highest at 6° (>0.97 across all quadrants). Conclusions: We demonstrate that acquisition and measurement of cpRNFL thickness values using 3-dimensional hand-held optical coherence tomography volumes in awake children is both feasible and reliable and is optimal at 6° from optic nerve center. Translational Relevance: Our recommended protocol provides guidance on how pediatric optic nerve pathologies are managed by clinicians.