Missing Clinical Trial Data: "Mixed Model" Approach Gains Ground At FDA
Executive Summary
Clinical trial sponsors should adopt a mixed statistical model to cope with missing data rather than relying on the "last observation carried forward" model that was formerly perceived as regulators' favorite