Tu1454 Inter-Rater Agreement for Analysis of Pharyngeal High Resolution Manometry Parameters

Abstract

Introduction: Pharyngeal manometry using high resolution technology is a developing field that is being advocated for evaluation of dysphagic symptoms that are not explained by a conclusive or objective finding on routine evaluation such as radiography. Fundamental to the success of this approach, automated analysis techniques not withstanding, is the reliability and reproducibility of the measurements performed by the clinicians. Our aim was to determine the inter-rater reliability for measurements of deglutitive pharyngeal peristaltic pressure wave characteristics such as peak peristaltic amplitude, pharyngeal contractile integral (PCI), nadir upper esophageal sphincter (UES) deglutitive pressure and intrabolus pressure (IBP). Methods: We studied 11 healthy individuals (age 56±25 years, 6F) by high resolution manometry using a catheter with 36 recording sites spaced at 1 cm intervals that recorded from the entire pharyngeal contractile zone, UES and proximal esophagus. Each volunteer swallowed 0.5 ml of room temperature water 40 times with 20 seconds intervals between swallows. Three raters (one seasoned, two with recent training) each independently analyzed a total of 3080 pressure signatures, 440 pharyngeal contractile integrals and 440 intra-bolus pressures. Measured parameters included peak peristaltic wave pressures at specific manometric locations 2, 3, 4, 5, 6, 7, and 8 cm above the upper margin of the UES high pressure zone, as well as PCI, nadir UES deglutitive pressure and IBP. Inter-rater agreement was evaluated using Bland-Altman intra-class correlation coefficient (ICC) analysis. Results: As seen in the table, agreement among the three raters all for all manometric locations was highly significant (p<0.001) with sites p2 and p3 (2 and 3cm above the upper margin of the UES) having the lowest levels of agreement. Furthermore, metrics with a single value for each swallow like PCI and UES nadir pressure also showed highly significant agreement whereas IBP measurements did not. A graphical depiction of intra-rater agreement is shown in the figure for PCI as a representative parameter. Conclusion: Analysis of pharyngeal pressure parameters by individuals with proper training is reproducible as evidenced by high level of agreement between the results obtained independently by three different raters. Supported in part by EOPH grant and PPG. Supported in part by R01DK025731 and P01DK068051