Interrater reliability of videofluoroscopic swallow evaluation

The past two decades have brought an enormous widening of interest in and knowledge about swallowing disorders. The most frequently used technique for swallow evaluation is X-ray videofluoroscopy. Most interventions are based on this examination. Only a few studies assessing interobserver reliability of videofluoroscopy have been published. The aim of our study was to assess the interobserver reliability of videofluoroscopy for swallow evaluation. Fifty-one consecutive dysphagic patients referred for videofluoroscopy were entered into the study regardless of their underlying disorder. The first swallow (5 ml of a semisolid radio-opague contrast media) of each patient was assessed in the lateral projection by 9 independent, experienced observers from different international swallow centers. All studies were evaluated according to a standardized protocol sheet and the interobserver reliability was calculated. The interobserver reliabilities assessed as kappa coefficient for parameters of the oral and pharyngeal phase, for the temporal occurrence of penetration/aspiration, and for the location of bolus residue ranged from 0.01 to 0.56. High reliability with an intraclass coefficient of 0.80 was achieved only with the well defined penetration/aspiration score. Our study underlines the need for exact definitions of the parameters assessed by videofluoroscopy, in order to raise interobserver reliability. To date, only aspiration is evaluated with high reliability by videofluoroscopy, whereas the reliability of all other parameters of oropharyngeal swallow is poor.