Reliability of Observations from Modified Barium Swallow Studies.
This project addressed reliability of judgments made from modified barium swallow (MBS) studies. The specific intent was to minimize potential difficulties in the interpretation of the findings by limiting the focus of the judges and the types of judgments to be made. Ten speech-language pathologists (SLPs) viewed 10 single swallow videos, 1 from each of 10 patients with a common medical diagnosis. Each SLP made seven observations per video on two separate occasions. The judgments, solely from the pharyngeal phase, addressed structural movement and bolus presence. All judgments were based on 0-3 scales. Calculations included “Exact” agreement using all four scale values, and “Clinical” agreement using dichotomous scale divisions. Intra- and interjudge data were reported as percents and Kappa correlations, and subsequently organized into qualitative categories (good, fair, poor). As expected, Exact agreement was less impressive than Clinical agreement. For the latter, within-judge agreement appeared relatively strong except for one structure judgment. Most between-judge Clinical values (and the associated Kappas) were stronger for observations of bolus presence than structural movement. However, within bolus judgments, agreement regarding presence of aspiration appeared weak. Clinical ramifications of these findings are noted but should be considered within the acknowledged limitations of this study. **********