Abstract
The Dynamic Imaging Grade of Swallowing Toxicity (DIGEST) rating method was developed in head and neck cancer populations to describe severity of aspiration and residue. The purpose of this study was to assess criterion validity of DIGEST in a post-stroke cohort. In this retrospective analysis, two raters (using version two criteria) performed DIGEST(V2) rating on recordings of modified barium swallow studies (MBSS) from 88 post-stroke patients that were extracted from a larger de-identified database. Modified Barium Swallow Study Impairment Profile (MBSImP) scores and Functional Oral Intake Scale (FOIS) scores were used to determine criterion validity. Inter-rater and intra-rater reliability for overall DIGEST(V2) grade were substantial (ƙ = 0.69 and 0.73, respectively), however inter-rater reliability for efficiency were only moderately reliable (ƙ = 0.52). Reliability for MBSImP scoring was excellent for Pharyngeal Total (PT) scores (ICC = 0.81-0.93). Overall DIGEST(V2) grades were significantly associated with PT scores in the expected direction (τ = 0.51, p < 0.0001), and there was no association between Oral Total (OT) and DIGEST grade (τ = -0.01, p = 0.889). Pairwise comparisons using PT scores indicated significant differentiation between DIGEST(V2) grades 0 from all other grades (p < 0.0001), with overlap in intermediate grades (p = 0.102-0.711). Functional Oral Intake Scale (FOIS) scores were significantly associated with DIGEST(V2) grade in the anticipated direction (τ = -0.43, p < 0.0001). Expected psychometrics and acceptable reliability for DIGEST(V2) grading were shown in this post-stroke cohort. A larger dataset would clarify mid-grade differentiation and potential influence of oral phase impairments in this sub-population.