Can machine learning improve randomized clinical trial analysis?

Despite over 24 approved anti-seizure medications, seizure freedom eludes 1 in 3 patients [1]. Meanwhile, new drug development costs have accelerated into billions of dollars. With these factors at play, it is time to re-assess basic assumptions about evaluating randomized clinical trial outcomes. It was recently shown that using “median percentage change” (MPC) [2] increased trial efficiency and lowered cost [3]. Is there a better metric, one that can reduce trial size while maintaining quality, reproducibility, and surveillance for adverse events?
Source: Seizure: European Journal of Epilepsy - Category: Neurology Authors: Tags: Short communication Source Type: research