Results
We enrolled 186 patients with suspected LPR who underwent 24-h MII-pH monitoring. All subjects were prescribed high-dose PPIs twice. However, 78 patients with suspected LPR were lost to follow-up after 2 months. Finally, 108 patients were included. Selection and grouping of patients with suspected LPR are summarized in the flowchart in Figure 1.
There were 28 patients (12 men and 16 women, mean age: 51.43 ± 12.62) with no reflux and 80 patients (31 men and 49 women, mean age: 55.24 ± 12.78) with LPR. There were no significant differences in age, sex, medical history (diabetes mellitus, hypertension), social history (alcohol, smoking, coffee), and pre-treatment RSI or RFS between patients with no reflux and those with LPR. However, the responder rates after treatment during 2 months were significantly higher in patients with LPR than in those with no reflux (57.50% vs. 28.57%, p = 0.0157; Table 1).
Patients with LPR were divided into non-responders (n = 19; 9 men and 10 women; mean age, 56.38 years) and responders (n = 61; 22 men and 39 women; mean age, 54.88 years). There were no significant differences in age, sex, medical history, social history, pre-treatment RSI or RFS, and medication periods between two groups (Table 2). Responders showed higher proximal MII parameters compared to non-responders (Table 3). All reflux time and longest reflux time in various proximal MII parameters were significantly higher in responders than in non-responders (p= 0.0040 and 0.0216, respectively; Figure 2A, B). However, there was no significant difference in the proximal all reflux episodes between two groups (p = 0.4781; Figure 2C). Also, there were no significant differences in distal MII parameters between two groups (Table 3).
The ROC curves used to determine the appropriated cut-off value of proximal all reflux time and proximal longest reflux time for predicting responders with LPR are depicted in Figure 3. The area under the ROC curves (AUCs) were 0.619 (95% confidence interval [CI], 0.488–749) and 0.624 (95% CI, 0.488–760) for proximal all reflux time and proximal longest reflux time, respectively (Figure 3). The cut-off values to predict responders with LPR were >0.000517% (sensitivity 47.5%, specificity 84.2%) for proximal all reflux time and >0.61 min (sensitivity 34.4%, specificity 94.7%) for proximal longest reflux time. The sensitivity plus specificity was higher for the cut-off value of proximal all reflux time than for that of proximal longest reflux time (1.317 vs. 1.291).