Inter-rater reliability and validity of automated impedance manometry analysis and fluoroscopy in dysphagic patients after head and neck cancer radiotherapy

Abstract

Background: Pharyngeal automated impedance manometry (AIM) analysis is a novel non-radiological method to analyze swallowing function based on impedance-pressure recordings. In dysphagic head and neck cancer patients, we evaluated the reliability and validity of the AIM-derived swallow risk index (SRI) and a novel measure of postswallow residue (iZn/Z) by comparing it against videofluoroscopy as the gold standard. Methods: Three blinded experts classified 88 videofluoroscopic swallows from 16 patients for aspiration and degree of postswallow residue. Pressure-impedance recordings of the patient and age-matched control swallows were analyzed using AIM by three observers who derived the SRI and iZn/Z. Intra-class correlation coefficients (ICC) were calculated for videofluoroscopic and AIM measures. Patient pressure/impedance measurements were compared with videofluoroscopy scores and control subjects to determine validity for detecting clinically relevant swallowing dysfunction. Key Results: Agreement among observers assessing presence of penetration and aspiration was modest (ICC 0.57) for videofluoroscopy and good (ICC 0.71, 0.82) for AIM-derived SRI and iZn/Z. When compared with age-matched controls, the SRI was higher in patients with aspiration (mean diff. 28.6, 95% CI [55.85, 1.355], p < 0.05). The iZn/Z had moderate positive correlation with bolus residue on fluoroscopy (BRS score) (rs(86) = 0.4120, p < 0.0001) and was increased in both patients with aspiration ({increment}244 [419.7, 69.52; p < 0.05]) and penetration ({increment}240 [394.3, 85.77]; p < 0.05) compared to controls. Conclusions & Inferences: AIM-based measures of swallowing function have better inter-rater reliability than comparable fluoroscopically derived measures. These measures are easily determined and objective markers of clinically relevant features of disordered swallowing following radiotherapy.