Changes in symptom reflux association using dynamic pH thresholds during ambulatory pH monitoring: an observational...

Symptom reflux association (SRA) assesses symptoms associated with reflux events defined by pH <4.0, but limited symptoms associate with reflux events. We evaluated the impact of alternate pH thresholds on SRA in a large ambulatory pH database. Acid exposure time (AET), reflux events, and associated symptoms (within 2 minutes following a reflux event) were extracted from ambulatory pH studies performed off antireflux therapy (722 patients, 49.1 ± 0.5 years, 66.8% F) over a 7-year period. Symptom association probability (SAP) and symptom index (SI) were calculated at pH 3.5, 4.0, 4.5, and 5. Receiver operating characteristics (ROC) were generated using SRA at any pH as gold standard; areas under the curve (AUCs) were determined. Discordant cases were reanalyzed to determine changes in SRA and predictors of change using multivariate regression. At pH 4.0, 41% had a positive SAP, and 34% had a positive SI. While there was sustained gain in SI positivity from acidic to more weakly acidic pH thresholds, SAP positivity was highest at pH 4.5. On ROC analysis, performance characteristics were best at pH 4.0 (AUC 0.97) for SAP, and at pH 4.5 and 5.0 (AUC 0.92-0.94) for SI. On multivariate logistic regression adjusting for age, gender, and change in AET and reflux events, only number of associated symptoms predicted change in SRA (P < 0.0001). Changing pH thresholds for reflux events augments SRA by increasing reflux events associated with existing symptoms, while symptom recording remains the principal determinant of SRA.