Erratum / apology
I am indebted to Gavin Giovannoni for pointing out a major error in
the calculations I made for my rapid response dated 17/02/06.
For some reason, I have read the annualised reduction in relapse
rates from the AFFIRM and SENTINEL trials as a %rate per patient year,
rather than absolute rate per patient.
Consequently I have underestimated the treatment benefit by a factor
of 100, and the NNT should indeed be around 2, rather than 200. Which
makes Natalizumab highly effective, and the cost issue wholly different at
around $47000 per relapse prevented.
Once other cost-utilities (quality of life, prevention of progression
to disability, costs of disability) are taken into account I imagine
Natalizumab may well be approaching the £30K per QUALY NICE uses as it's
benchmark for approval.
I will try to remember to refrain from commenting on NNTs in future
without a chance to peruse the raw data first (somewhat chastened, I am
typing this with both March 2nd NEJM articles (1,2) at my elbow. There are
a lot of potential NNTs, for multiple secondary outcomes, and all rather
low - in the order of 2-3 at worst).
Once more, my apologies.
(1) Polman CH, O’Connor PW, Havrdova E, et al. A randomized, placebo-
controlled trial of natalizumab for relapsing multiple sclerosis. N Engl J
(2) Rudick R, Stuart WH, Calabresi PA, et al. Natalizumab plus
interferon beta-1a for relapsing multiple sclerosis. N Engl J Med
Competing interests: No competing interests