Utility of software analysis of esophageal manometry studies in patients with aperistalsis

Esophageal manometry is the gold standard for the diagnosis of esophageal aperistalsis. There is computer software that analyzes peristalsis on esophageal manometry, but this automated analysis has not been formally evaluated. Our primary aim was to evaluate the software analysis of esophageal aperistalsis by esophageal manometry in patients diagnosed with aperistalsis by an experienced clinician. Esophageal manometry studies from January 2006 to November 2007 were retrospectively reviewed for evidence of aperistalsis by an experienced clinician. All studies demonstrating aperistalsis were selected for further review. The automated analysis performed by our software program for each study was recorded. Agreement between the automated analysis and the clinician was measured by the proportion of agreement on the absence of peristalsis. Eighty-seven of the 962 esophageal manometry studies reviewed demonstrated aperistalsis. The automated analysis reported esophageal body peristalsis with wet swallows in 66 out of 87 patients (75.9%). In these patients, the software analyzed an average of 34.2% of the wet swallows as peristaltic. The agreement between the clinician's review and software analysis of aperistalsis was 24.1%. These data suggest there is poor agreement between the automated analysis of peristalsis and that of an experienced reviewer. Automated analysis cannot be relied upon in the diagnostic evaluation of esophageal aperistalsis as it overestimates the presence of peristalsis and may lead to incorrect diagnoses and management strategies.