|Saturday, February 22|
|PS3 Poster Session III & Continental Breakfast||
Sat, Feb 22, 7:30 AM - 9:00 AM
Test of Equivalence for Repeated Measurements in a Clinical Study for Hepatocellular Carcinoma (HCC) Patients (302839)*Yiyi Chen, Oregon Health and Science University
Zunqiu Chen, Oregon Health and Science University
Keywords: equivalence test, repeated measurements
We share a real-world consulting and data analysis experience to a pilot study to assess a novel tumor motion suppression system for patients with hepatocellular carcinoma undergoing stereotactic body radiation therapy (SBRT). The primary objective of the pilot study is to determine whether the gross target volume (GTV) remains stable for10 respiratory phases when using the novel system. The clinical hypothesis is that GTV will remain stable under the new system. The investigators have conducted pairwise comparisons through 45 paired t-tests and have adjusted p-values using Bonferroni correction for multiple testing. The major problem with their approach is that fail to reject the null hypothesis does not imply that the GTV remains stable. In addition, Bonferroni correction (or any other alternative adjustment method) makes it harder to reject the null hypothesis, which in turn favors the clinical hypothesis. This poster will show how to conduct test of equivalence for repeated measurement (an appropriate method for this type of study) and discuss how to ad-hoc assess the power for such studies.