A flawed study should not define a new standard of care.