How many swallows are required to establish reliable esophageal peristaltic parameters in normal subjects? An on-line computer analysis.

Esophageal manometric parameters are often examined using from four to 10 or more swallows to obtain a manometric profile without addressing the question of the reproducibility of these data. We asked the question of "how much is enough?" by statistically comparing five sets of 10 wet swallows (5 ml) each separated by 20 min in eight volunteers using constant infusion manometry. Peristaltic waves were evaluated for amplitude, duration, and velocity using an on-line computer program. Statistical analyses included analysis of variance and the Newman-Keuls test for uniformity among sets of swallows. The sets of 10 swallows showed no evidence of a training effect over time. Less than 5% of all comparisons (Newman-Keuls) showed significant differences. Using 5% variability outside of 1 SD as a cutoff (95% confidence interval) amplitude of normal wet swallows could be adequately defined by more than or equal to five swallows and velocity by more than or equal to seven swallows. Duration shows the greatest overall variability and required more than or equal to eight swallows for accurate assessment. These data suggest that at least five to eight wet swallows should be performed to describe "typical" manometry parameters for normal subjects.