Correlation of Reflux Finding Scores on Laryngoscopy to Clinical Outcomes

Abstract

Objectives:

Evaluate the reflux finding score as a clinical tool to diagnose laryngopharyngeal reflux disease by testing its reliability and reproducibility.

Methods:

Two scorers independently and blindly reviewed 330 videostrobes of patients that were performed from 2005-2009. Scorers gave each videostrobe a reflux finding score. A chart review was then performed to determine the diagnosis given at the time of the videostrobe, smoking status, whether the patient was treated with anti-reflux medication, and any abnormalities noted on recent esophagoscopy. The reflux finding scores of the 2 scorers were statistically compared for agreement with each other and to the findings of the chart review.

Results:

A total of 168 out of 330 patients met final criteria for inclusion in the study. The interrater agreement between the 2 scorers had a weighted kappa of 0.30. The c-statistic scores for correlation of the scorers’ reflux finding scores (RFSs) to the chart diagnosis were 0.554 and 0.609 for scorer 1 and scorer 2, respectively.

Conclusions:

The reflux finding score has a fair inter-relater reliability in our study. Blinded rater use of RFS in our study did not correlate well with clinical diagnosis of laryngopharyngeal reflux disease. No statistically significant differences in RFS validity were found based on gender, age, or treatment status.