Tip:
Highlight text to annotate it
X
The National Lung Screening Trial was designed
with the purpose of determining whether screening
with low-dose helical CT was more effective
in reducing deaths from lung cancer or resulting
in fewer deaths from lung cancer than screening
with standard single-view chest X-ray.
And this study began in the fall of 2002.
I, as a chief statistician, one of the two chief statisticians
for the trial, have been involved in it
since the inception, going back to the days of design in 2000.
The chief result of this trial and the one
for which the study was designed and powered,
statistically powered, as it were,
is that there is a 20 percent reduction
in lung cancer mortality in the helical CT arm relative
to the chest X-ray arm.
So we use lung cancer mortality as the end point.
Now, since these are healthy people,
most of these individuals never experience the end point.
Only about 2 percent of them do.
So that means, in order
to get enough individuals having the end point,
we need a big study.
We're also looking for rather modest reduction
in lung cancer mortality, about 20 percent.
So this combination of a rather rare event
with a rather modest reduction in mortality means
that we need a big study.
And, in fact, it was in the vicinity of 50,000 participants.
There are two consequences of this fact.
Number one is that it's a very expensive study.
It's in the vicinity of a quarter of a billion dollars.
Number two, it's an extremely complex study to coordinate.
It requires multiple centers.
And to ensure or to enhance the representativeness
of the results, we conducted it throughout the country.
So we needed a -- we needed the expertise of designing
and actually conducting,
managing very large-scale clinical trials with very clear,
closely monitored communication.
And I think for this purpose,
the National Cancer Institute was absolutely essential.