Episode-level reflux characteristics: How experienced reviewers differentiate true reflux from artifact on...

Accurate reflux episode identification is crucial for pH-impedance interpretation. Individual reflux episode characteristics associated with inter-reviewer concordance are incompletely understood.

Ambulatory pH-impedance studies from 19 GERD patients (median age 52 years, 78.9% F) were analyzed by 5 reviewers. Metadata from pH-impedance studies were exported to a dedicated software tool designed to compare episode-by-episode identification between reviewers within a ±7.5 s window. Patient position, acidic vs. nonacidic episodes, acid clearance time (ACT), bolus clearance time (BCT), and proximal extent of reflux episodes were compared between episodes identified by all reviewers against those identified by automated analysis, and one to four reviewers, respectively.

Automated analysis identified 1644 episodes (median 78 episodes per patient, IQR 64-108), of which 84.9% were identified by ≥3 reviewers and 57.1% by all reviewers; 339 unique episodes were added by at least 1 reviewer. Characteristics defining 5 reviewer concordance included acid reflux episodes (88.9%), upright episodes (88.4%), high proximal extent (median 17 cm, IQR 15-17 cm), and longer acid clearance times (67.0 s, IQR 29.0-146.0 s) (P < 0.001 compared to 1-4 reviewer concordance for each). In contrast, 1 reviewer-identified episodes were 69.8% acidic, 76.9% upright, and limited to the distal esophagus. Using 5-reviewer concordance, designation of GERD evidence changed from automated analysis in 16%-19% of patients.

Acidic episodes with high proximal extent in the upright position and longer acid clearance times on pH-impedance studies have the highest concordance for identification by expert reviewers. Reflux episode identification may be influenced by reviewer opinion despite availability of established criteria.