[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: Wald interval and the WSJ

From   "Lachenbruch, Peter" <[email protected]>
To   <[email protected]>
Subject   st: RE: Wald interval and the WSJ
Date   Thu, 14 Aug 2008 09:03:51 -0700

I saw this too.  I responded to the ASA as follows:

There are several concerns here:  the p-value worship is only one.  I
think there are other potential problems and I'm a bit concerned that we
go off half-cocked.

1.  Non-inferiority studies are frequently used.  One issue is finding
an appropriate margin of non-inferiority.

2.  The historical control group is a problem for me.  When the response
is subjective 'how do you feel?'  An unblinded study is a big issue.
Another issue with historical controls is that they work with patients
from a different time and possibly place.  If health care for the
condition has changed, the control and treatment groups are no longer

3.  The Wald test is a bit of a concern, but far less than the first
two.  One could use a likelihood ratio test or a score test.

4.  I'm a little surprised that they didn't do a Bayesian analysis,
which is something that CDRH likes.

5.  I suspect most of us would regard two studies with p-values of 0.049
and 0.051 as showing about the same thing, so the worship of the p-value
is rather silly.  A confidence interval would be better.

There is an attachment showing the results of the various tests.  They
range from 0.0484 for the Wald test to 0.0547 for the exact double
binomial test.  The score test gave 0.0515.  They didn't publish a
likelihood ratio test

My point is that all the tests gave similar results.  We shouldn't
worship p=0.05.


Peter A. Lachenbruch
Department of Public Health
Oregon State University
Corvallis, OR 97330
Phone: 541-737-3832
FAX: 541-737-4001

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Scott
Sent: Thursday, August 14, 2008 5:18 AM
To: [email protected]
Subject: st: Wald interval and the WSJ

There is an interesting article in today's Wall Street Journal on the
use of the Wald interval in clinical trials.

"Boston Scientific Stent Study Flawed"  by Keith Winstein, August 14,
2008; Page B1.

"But Boston Scientific's claim was based on a flawed statistical
equation that favored the Liberte stent, a Journal analysis has found.
Using a number of other methods of calculation -- including 14
available in off-the-shelf software programs -- the Liberte study
would have been a failure by the common standards of statistical
significance in research.
Boston Scientific isn't the only company to use the equation, known as
a Wald interval, which has long been criticized by statisticians for
exaggerating the certainty of research results. "

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index