Number Needed to Treat in Multiple Sclerosis Clinical Trials

Macaulay Okwuokenye, Annie Zhang, Amy Pace, Karl E. Peace

Clinicians are expected to select a therapy based on their appraisal of evidence on benefit-to-risk profiles of therapies. In the management of relapsing-remitting multiple sclerosis (RRMS), evidence is typically expressed in terms of risk (proportion) of event, risk reduction, relative and hazard rate reduction, or relative reduction in the mean number of magnetic resonance imaging lesions. Interpreting treatment effect using these measures from a RRMS clinical trial is fairly reliable; however, this might not be the case when treatment effect is expressed in terms of the number needed to treat (NNT). The objective of this review is to discuss the utility of NNT in RRMS trials. This article presents an overview of the methodological definition and characteristics of NNT as well as the relative merit of NNT use in RRMS controlled clinical trials, where endpoints are typically time-to-event and frequency of recurrent events. The authors caution against using NNT in multiple sclerosis, a clinically heterogeneous disease that can change course and severity unpredictably. The authors also caution against the use of NNT to interpret results in comparative trials where the absolute risk difference is not statistically significant, computing NNT using the time-to-event endpoint at intermediate time points, computing NNT using the annualized relapse rate, and comparing NNT across trials.