What question did this study set out to answer?

To validate the Guided Oral-intake and Swallowing Test (GOST) against the Dysphagia Outcome and Severity Scale (DOSS) using instrumental assessments as a reference.

June 3, 2026Dysphagia0 citationsOpen Access

The Guided Oral-Intake and Swallowing Test (GOST): Validation of a Novel Bedside Composite Score for Dysphagia Against Instrumental Assessment

YTYohei TeramotoUniversity of Tsukuba KSKenji SuzukiUniversity of Tsukuba SMShin MatsumotoUniversity of Tsukuba

Key Points

To validate the Guided Oral-intake and Swallowing Test (GOST) against the Dysphagia Outcome and Severity Scale (DOSS) using instrumental assessments as a reference.
Retrospective analysis of 67 inpatients who underwent GOST and DOSS assessments via videofluoroscopy or videoendoscopy.
Correlation, agreement, and categorical concordance between GDOSS and DOSS were examined using Spearman's ρ, MAE, RMSE, and weighted kappa.
Ordinal logistic regression was used to evaluate GDOSS as a predictive tool for DOSS levels.
GDOSS demonstrated a strong correlation with DOSS (ρ=0.716, p<0.001).
Good agreement was shown in various metrics (MAE=0.881; RMSE=1.203; weighted κ=0.749).
GDOSS was identified as a robust independent predictor of DOSS levels (OR=3.79, p<0.001).

Abstract

Videofluoroscopy (VF) and videoendoscopy (VE) remain the gold standards for evaluating dysphagia severity. The Dysphagia Outcome and Severity Scale (DOSS), based on these tests, is widely used as a reference. However, such assessments are not always feasible in clinical practice, highlighting the need for a practical alternative. In response, we have developed the Guided Oral-intake and Swallowing Test (GOST).To validate the Gauge of DOSS (GDOSS), a key indicator derived from GOST, by comparing it with instrumental assessments (VF/VE) using DOSS as the reference.We retrospectively analyzed 67 inpatients who underwent both GDOSS assessment and DOSS scoring via VF or VE. The median DOSS level from three independent raters was used as the reference. Correlation (Spearman's ρ), agreement (MAE, RMSE), and categorical concordance (weighted kappa) between GDOSS and DOSS were examined. Predictive validity was evaluated using ordinal logistic regression. GDOSS demonstrated a strong correlation with DOSS (ρ=0.716, p<0.001).Furthermore, it showed good agreement across various metrics (MAE= 0.881; RMSE= 1.203; weighted κ=0.749). GDOSS also emerged as a robust independentpredictor of DOSS levels (OR = 3.79, p<0.001), even after adjusting for potential confounders.Rather than replicating exact DOSS levels, GDOSS effectively estimates a clinically meaningful severity range, reflecting both functional capacity and safety. Its integrative and probabilistic nature supports GDOSS as a valid and practical tool, especially where instrumental assessments are limited. Future studies should explore its longitudinal responsiveness and role in clinical decision-making.

Perguntar à IA

Bookmark

View Full Paper