Accuracy and reproducibility of 12-h esophageal pH monitoring.

Seventy-five consecutive 24-h intraesophageal pH recordings performed according to the methods of Jolley et al. were evaluated. Total scores and the mean duration of sleep reflux (ZMD) were calculated for the first and the last 12 h, respectively, and compared to each other, as well as to the full 24-h recording. The accuracy of the first and last 12 h in predicting whether the 24-h study was abnormal was 85% and 87%, respectively. The accuracy in predicting whether the ZMD for the 24-h study was abnormal was 73% and 83%, respectively, with an overall 39% false-negative rate. The 12-h study reproducibility was 72% for the total score and 56% for the ZMD. Since the original standards were determined from 18- to 24-h recordings, the test was restandardized based on 12-h recordings from patients with normal 24-h studies. Restandardization did not improve the accuracy or reproducibility. The high false-negative rate for the ZMD and the poor reproducibility cast serious doubt on the utility of 12-h pH recordings, especially in patients with respiratory symptoms.