Post-Hoc Analysis Shows No Statistical Benefit of Adaptive Design in Stroke Prevention Trial
September 28, 2016
Leslie McClure, PhD, MS, professor and chair of the Department of Epidemiology and Biostatistics at Drexel University’s Dornsife School of Public Health, led a study looking at the impact of increasing the sample size in an adaptive design clinical trial.
Researchers analyzed the Secondary Prevention of Small Subcortical Strokes (SPS3) study, where an adaptive design was used and the study sample size was re-estimated midway through the trial. In the SPS3 trial, the original sample size of 2500 participants was increased to 3020 participants to account for slow rates of trial enrollment and newly published recurrent stroke rates that were well below those used in the original SPS3 study design.
Adaptive design, which by definition allows recalculation of needed sample size over the course of a trial, has become increasingly popular in public health and yet its benefits have rarely been investigated.
The evaluation, published in Clinical Trials, compared results from the SPS3 planned and adjusted sample sizes. Researchers found no difference in the conclusions regarding treatment efficacy; but in one instance, the results from the increased sample size suggested significantly different conclusions from the original samples. However, implementing the adjusted sample size presented many challenges and substantially increased the total cost of the study.
These results suggest that the benefits of adaptive design do not always outweigh practical issues, such as higher costs and additional resources needed to implement an adaptation.
A post hoc evaluation of a sample size re-estimation in the Secondary Prevention of Small Subcortical Strokes study