DOC News October 1, 2007
Volume 4 Number 10 p. 5
© 2007 American Diabetes Association
Reanalyzing the Meta-analysis
Homer L. Staley, MD
Lawrenceburg Family Practice Associates Lawrenceburg, Tenn.
Regarding the latest article from Editor-in-Chief Dr. Irl B. Hirsch on
thiazolidinediones (TZDs), "TZDs: Where Do We Go From Here?"
(DOC News, August 2007, page 3), it appears to me that there is some
protection afforded by the TZD rosiglitazone (Avandia, GlaxoSmith-Kline).
However, never having taken a course in statistics, my skills in interpreting
statistical information are somewhat limited. Nevertheless, it appears that
[in the study Dr. Hirsch describes that was published in the New England
Journal of Medicine] if there were 86 deaths in a group of 14,371
patients, this would correlate to one death in 167 patients in the Avandia
group.1 When
compared with 72 deaths in the control group of 11,634 patients, this is
equivalent to one death in 161.6
patients.1
It appears to me that if I had my choice of which group of patients to be
in, if I were told one patient in each group was going to die, I think I would
prefer to be in the group that had 167 patients in it, which is listed as the
rosiglitazone group of 86 events in 14,000 patients.
If there is something here that I am missing, I would appreciate it if you
could explain my dilemma, but it appears that rosiglitazone does offer some
protection from myocardial infarction.
References
1. Nissen SE, Wolski K: Effect of rosiglitazone on the risk of
myocardial infarction and death from cardiovascular causes. N Engl
J Med 356:2457–2471, 2007.[Abstract/Free Full Text]
CiteULike 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
