Symptom Index or Symptom Association Probability?: A Closer Look at Symptom Association in Suspected GERD Patients.

ABSTRACT

BACKGROUND: Symptom index (SI) and symptom association probability (SAP) are popular methods used to measure symptom association in patients with gastroesophageal reflux disease (GERD). AIM: To investigate whether these 2 methods yield similar results in analysis of both typical and atypical GERD symptoms. MATERIALS AND METHODS: Combined impedance-pH reflux studies of 1471 patients tested for possible GERD symptoms from January 2010 to May 2015 were reviewed. SI and SAP were analyzed for typical and atypical GERD symptoms including heartburn, regurgitation, indigestion, chest pain, cough, and throat clearing (TC). Patients who reported <3 symptom events during the 24-hour monitoring period were excluded. ON and OFF proton pump inhibitor (PPI) groups were reviewed. Kappa coefficient (κ) rather than simple percentage was used to measure the agreement rate. Simple percentage agreement is a less reliable method compared with κ. RESULTS: On PPI therapy, there was a good κ between SI and SAP for regurgitation (0.68) and indigestion (0.64), moderate for heartburn (0.48) and chest pain (0.51), and poor for cough (0.33) and TC (0.29). There was a lower κ OFF PPI therapy for heartburn (0.36), regurgitation (0.44), and indigestion (0.50). But there was no difference in κ for chest pain (0.61), cough (0.29), and TC (0.33). CONCLUSIONS: SI and SAP showed better agreement for patients with typical GERD symptoms and even better when tested ON PPI. A better symptom association method is needed for patients with atypical GERD symptoms.