Modeling the course and outcomes of multiple sclerosis is statistical twaddle--Yes.

Our ability to modify the outcomes of multiple sclerosis (MS) is reliant on our ability to measure disability and change in disability reliably, reproducibly, precisely, and ideally at all stages of MS.1 We cannot determine whether interventions, pharmacological or otherwise, alter the natural history of MS if we cannot model the natural history accurately at the individual and group level. These apparently simple metrics, that is, whether the person with MS getting worse, staying the same or improving, have proven elusive, and to date, this problem has not been solved. What MS desperately requires is such a metric of severity and change. Since at least the 1930s, various MS diagnostic criteria, MS clinical course definitions and clinical and para-clinical disability measures have been proposed. The fact that none of these measures has survived the test of time without modification, or significant qualification, attests to the complexity of the disorder.