Abstract
BackgroundWe present a conceptual framework for simulations to determine the utility of biomarker enrichment to increase statistical power to detect a treatment effect in future Alzheimer's disease prevention trials. We include a limited set of simulation results to illustrate aspects of this framework. MethodsWe simulated data based on the Alzheimer's Disease Anti-Inflammatory Prevention Trial, and a range of sample sizes, biomarker positive predictive values, and treatment effects. We also investigated the consequences of assuming homogeneity of parameter estimates as a function of dementia outcome. ResultsUse of biomarkers to increase the sample fraction that would develop Alzheimer's disease in the absence of intervention from 0.5 to 0.8 would increase power from 0.35 to 0.69 with n = 200. Ignoring sample heterogeneity resulted in overestimation of power. ConclusionBiomarker enrichment can increase statistical power, but estimates of the expected increase are sensitive to a variety of assumptions outlined in the framework.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
