Purpose: A standardized 4-hour gastric-emptying protocol has been recommended by a consensus panel. However, a study length of 4 hours may be inconvenient. The purpose of this investigation was to determine if the percent emptying at 2 hours could predict abnormal or normal results at 4 hours.
Methods: Our data consist of 2 distinct patient groups, containing 174 and 158 patients, respectively. All patients were given a standardized meal according to the Tougas et al protocol. Percent gastric emptying was measured at 1, 2, 3, and 4 hours (or until gastric emptying was equal to or more than 90%). The sensitivity, specificity, and the likelihood ratios of the 2-hour cutoffs were determined using the 4-hour data as the standard. We have estimated the accuracy and percent of studies stopped at 2 hours, resulting from various "definitely normal" and "definitely abnormal" threshold combinations.
Results: Using gastric emptying of less than 35% at 2 hours for diagnosis of delayed gastric emptying, and emptying at 2 hours greater than 55% for a diagnosis of normal gastric emptying yielded accuracies of 0.966 and 0.949 in the 2 datasets, respectively, compared with the Tougas standard values. The observed 2-hour gastric-emptying values would have allowed early termination in 79% and 78% of the studies, respectively.
Conclusion: Adopting criteria for the termination of gastric-emptying study at 2 hours in selected patients yields significant reductions in the average length of a gastric-emptying study, while maintaining study accuracy.