Predictive Accuracy of Ambulatory 24-hour Impedance in Estimating Abnormal Esophageal Acid Exposure Time in Patients with Symptomatic Gastroesophageal Reflux

Abstract

Introduction: Twenty-four hour impedance-pH monitoring quantitatively describes gastroesophageal reflux disease (GERD) by detecting impedance reflux events and evaluating esophageal acid exposure time (EAET). However, clinical ambiguity arises when pH and impedance results differ in interpretation. Studies are often abnormal by acid exposure time, but normal by impedance measures of reflux. As the limit of normal total reflux events was determined in asymptomatic patients, we sought to analyze the predictive accuracy of 24-hour impedance in patients with symptomatic GERD. Methods: Seventy-two patients (mean age 55.5 ± 13.2 years, 75% female) with unmedicated, symptomatic GERD received 24-hour impedance-pH testing. The sensitivity, specificity, and area under the curve (AUC) were calculated for impedance reflux events using upright, recumbent, and total EAET values greater than 6.3%, 1.2%, and 4.2%, respectively, as gold standards. Values for number of reflux events were varied to determine the ideal combination of sensitivity, specificity, and AUC indices, which were then used to generate best cutoff values of impedance-measured reflux events. Results: Mean total EAET time was 10.5% ± 9.9%, and 46 (63.8%) patients had elevated total EAET > 4.2%. EAET in the upright and recumbent positions were elevated in 58.9% and 59.1% of patients, respectively. Patients experienced 46.4 ± 30.5 reflux events measured by impedance. Fifteen patients (20.8%) had an elevated number of reflux events. The current standard of total reflux events >73 had a sensitivity of 32.6% and a specificity of 100%. AUC analysis revealed a cutoff of >41 total reflux events had the best sensitivity and specificity, 69.6% and 80.7%, respectively (AUC 0.84, 95% CI [0.75-0.93]). AUC analysis confirmed that a cutoff of >41 was also optimal when compared to gold standards for upright and recumbent acid exposure. Conclusion: In patients with GERD, 62.8% had pathological levels of EAET, with a mean duration of total EAET twice that of normal. Only 20.8% had an elevated number of reflux events as previously determined by the upper 95th percentile of asymptomatic patients. The standard of >73 reflux events is weakly sensitive, but extremely specific for symptomatic GERD. AUC analysis indicates that a lower threshold of abnormal events, such as >41, optimizes sensitivity and specificity in symptomatic patients, and may be a more appropriate diagnostic measure for impedance-pH monitoring.