Abstract

BACKGROUND:

In 2008, American consensus recommendations for performing gastric emptying (GE) scintigraphy were published. It was recommended that data are acquired only at 0, 1, 2, and 4 h and that the results are expressed as percentages of meal retention. Until now, it was established that the GE time-activity curves should have many points (every 10, 15, or 20 min) to reflect the GE process accurately and to be optimally adjusted by a mathematical model. In this study, we have evaluated the curve fitting using only a few points as proposed by the consensus protocol.

MATERIALS AND METHODS:

GE scintigraphy tests of 224 patients were retrospectively analyzed. Two curve fittings were done for each patient, either using data acquired every 20 min or using data acquired every hour. A comparison of these two methods was made based on the values of the computed GE parameters.

RESULTS:

We observed strong correlations between the two methods (r=0.81-0.99, P<0.05). Using the Bland-Altman analysis, more than 95% of the differences were included in the mean difference 95% confidence interval. The mean differences were weak with a relatively small SD and Cohen's k coefficients ranging from 0.84 to 0.93, indicating an excellent agreement between the two methods.

CONCLUSION:

Our results showed the feasibility and accuracy of curve fitting using only a few points. The curve fitting is easy to perform and allows the computation of reliable and reproducible parameters that reflect the whole GE process.