Sequential Global Statistical Tests: A Novel Method for Evaluating Primary, Secondary and Exploratory Variables in a Hierarchy with an Accumulated Evidence Approach

The high heterogeneity of AD results in low correlations between endpoints, making it more prone to misinterpretation of overall treatment effects than diseases with higher correlations. Expecting significance on all outcomes is likely at least partly responsible for the high failure rate in AD. Using an sGST addresses this concern, by statistically ensuring no double counting of results with highly correlated outcomes, and no underestimation of the evidence provided by nearly independent outcomes.