Sample Size Verification for Clinical Trials

Abstract

In this article, we shall provide simple methods where non-statisticians can evaluate sample size calculations for most large simple trials, as an important part of the peer review process, whether a grant, an Institutional Review Board (IRB) review, an internal scientific review committee, or a journal referee. Through the methods of the paper, not only can readers determine if there is a major disparity, but they can readily determine the correct sample size. It will be of comfort to find in most cases that the sample size computation is correct, but the implications can be major for the minority where serious errors occur. We shall provide three real examples, one where the sample size need was seriously overestimated, one (HIP PRO-test of a device to prevent hip fractures) where the sample size need was dramatically underestimated, and one where the sample size was correct. The HIP PRO case is especially troubling as it went through an NIH study section and two peer reviewed journal reports without anyone catching this sample size error of a factor of more than five-fold.

Electroacupuncture Improves Cerebral Vasospasm and Functional Outcome of Patients With Aneurysmal Subarachnoid Hemorrhage

“…The incidence rate of delayed ischemic neurological deficit in the acupuncture group was 10%, and the control group was 38.9% ( Cho et al, 2015 ). To detect a significance level of 0.05 between the groups with 80% power at the assumption of a drop-out rate of 20% and a 1-tailed α = 0.05, 54 patients (27 participants in each group) were required for this study ( Shuster, 2014 ). Data were represented as mean ± standard deviation (SD) or number (%) and analyzed using SPSS version 23.0 (IBM Corp. Armonk, NY, United States) software.…”

Graphical displays for clarifying how allocation ratio affects total sample size for the two sample logrank test

“…Hsieh [3] compared the sample size formulas given by Freedman [4] (used in Nquery and STATA 10), Schoenfeld [5], Hsieh [6], and Shuster [7], and found that Freedman's formula gives the highest power for the logrank test if the sample size ratio of the two groups is equal to the reciprocal of the hazards ratio, whereas the other formulas give highest power when sample sizes in the two groups are equal. Hseih also observed that it was difficult to locate the optimal sample size ratio for the logrank test due to the flat portion of the power curves, but suggested that the the optimal sample size ratio may exist in the neighborhood from equal allocation of sample sizes to equal allocation of events.…”