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

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 (Delta 244 [419.7, 69.52; p < 0.05]) and penetration (Delta 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.