endpoint hierarchy
Statistical strategy that orders multiple endpoints to control Type I error while preserving study power.
Full Definition
An endpoint hierarchy is a pre-specified statistical testing sequence that arranges primary and secondary endpoints in order of importance to control overall Type I error rate across multiple comparisons. This approach allows researchers to test multiple hypotheses while maintaining statistical validity without overly conservative corrections. In psychiatric clinical trials, endpoint hierarchies typically place the most clinically relevant outcome first (such as symptom severity), followed by secondary measures like functional improvement or quality of life. Testing proceeds sequentially, and if a higher-order endpoint fails to achieve significance, lower-order endpoints cannot be tested for statistical significance.
Usage
Usage note: Hyphenate when used as compound modifier: 'endpoint-hierarchy approach'.
In Context
- "The endpoint hierarchy placed MADRS total score first, followed by CGI-I and functional outcomes." — Statistical analysis plan
- "Due to the endpoint hierarchy, secondary efficacy analyses could only be interpreted descriptively." — Clinical study report