May 16, 1:30 PM in the Grand Conference Room
  • Antipsychotics and Risk of Death NEJM 2005 Download:
  • Wayne Ray's editorials Download:

Wes' comments about this paper: The attached paper from this week's NEJM is an important one to us for many reasons. I think you should read it thoroughly.

Reason 1: The statistical methods are very similar to those that we have proposed for our observational cohort study. They use MVA and Propensity Scores and a similar approach to sensitivity analysis for estimating the magnitude of an unmeasured confounder that would be required to undermine their findings. Check it out and teach me from your thoughts. I am sure that this paper is going to generate TONS of letters, etc. Especially when you look at all of the imbalances in Table 1. Note, btw, that the mean age is 83 and the deaths in first 180 days were 3% points different.

Question 1: How confident are you that this is real given Table 1? There is no question that the RRs are large, hold up to dose effect, and methods don't alter the findings, and the senstivity analysis looked solid. Nevertheless, I bet that most people will look at that and say that they don't have confidence that any statistical techniques, no matter how mathy and savvy, could confidently tease out truth under such circumstances and confusion. On that point, I'd like to know your opinion.

Reason 2: This study, along with that in JAMA a few months ago on Atypical antipsychotics, are extremely timely considering that we plan to submit the MIND study (ziprasidone vs. haloperidol vs. placebo) as an RO1 on November 1, 2006. We are ~40 pts into our Phase II trial. I had not been worried about this observation of increased death with atypicals because of the "demented nursing home" paptient population and the duration of use of the agents. The same seems to be said for these NEJM pts, though there wasn't as much dementia. Interesting that the deaths happened typically early.

Question 2: Does this study alter your enthusiasm (up or down) for designing an RCT of typical vs. atypical vs. placebo?

Edit | Attach | Print version | History: r4 < r3 < r2 < r1 | Backlinks | View wiki text | Edit WikiText | More topic actions...
Topic revision: r2 - 17 May 2006, AyumiShintani
 

This site is powered by FoswikiCopyright © 2013-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback