Comparison of methods for analyzing gastric isotope emptying

The gastric emptying of 99mTc-sulfur colloid, ingested in a nutrient liquid test meal, was followed by gamma camera for normals, diabetics, and diabetics receiving intravenous metoclopramide. Gastric emptying patterns of 99mTc by these groups present as normal, slow, and rapid on simple graphic inspection. Half times of indicator emptying were computed from inspection and from least-squares linear regressions of log(base e) residue versus time for the total postpeak curve and for the remainder of the 15-min postpeak curve. In addition, the percent of 99mTc residue leaving at, and the area under, the residue curve until 6, 12, 24, and 60 min postpeak and the indicator mean transit time (MTT) were computed. Standard half-time determinations revealed no significant differences among the three groups, despite obvious visual differences among them. In contrast, the mean transit time of the rapid group was significantly less than that of the slow (P less than 0.01) and normal (P less than 0.05) groups. However, no statistical difference was noted between the slow and the normal emptiers (P greater than 0.1) using the MTT measure. The percent of particles leaving the stomach and the area under the residue curves demonstrated significant differences among the three groups, reflecting the obvious visual perceptions gained from simple curve inspection. Because the percent of indicator particles leaving the stomach region at a given time reflects a single time, it was concluded that the residue area represents the most reliable, objective, and quantifiable parameter for testing of significant differences.