n engl j med 363;13 nejm.org september 23, 2010
The new england journal of medicine
Screening Mammography — A Long Run for a Short Slide?
H. Gilbert Welch, M.D., M.P.H.
No screening test has ever been more carefully
studied than screening mammography. In the
past 50 years, more than 600,000 women have
participated in 10 randomized trials, each involv-
ing approximately 10 years of follow-up. Given
this extraordinary research effort, it is ironic
that screening mammography continues to be
one of the most contentious issues within the
The juxtaposition of such a charged medical
debate in the face of such an exhaustive scien-
tific investigation is in itself instructive. For con-
text, one trial involving fewer than 150 men who
were followed for less than 2 years was suffi-
cient to convince physicians of the value of treat-
ing severe hypertension.1 That physicians are still
debating the relative merits of screening mam-
mography despite the wealth of data suggests
that the test is surely a close call, a delicate bal-
ance between modest benefit and modest harm.
In this issue of the Journal, Kalager et al.2
provide additional data that the benefit of mam-
mography is modest. Making use of the oppor-
tunity provided by the staggered implementation
of a national screening program in Norway, the
investigators were able to isolate the benefit of
the screening program from other factors that
may have changed over time, including increased
breast-cancer awareness and improvements in
treatment. They report that the benefit of the
Norwegian screening program was disappoint-
ingly small: a 10% reduction in breast-cancer
mortality among women between the ages of 50
and 69 years.
Moreover, this reduction in mortality reflected
the combined effect of the two interventions that
make up the Norwegian screening program:
screening mammography and multidisciplinary
teams instituted to better treat breast cancer.
Kalager et al. provide data that the latter may be
the more important of the two factors, since
women over the age of 70 years, who were ex-
posed to the program’s multidisciplinary teams
but were not invited to undergo mammography,
had an 8% reduction in breast-cancer mortality.
Thus, the relative reduction in mortality due to
screening mammography alone could be as low
Clinicians who follow the mammography de-
bate will reasonably wonder why the benefit es-
timated by Kalager et al. is so much smaller than
the reduction in mortality of 15 to 23% estimat-
ed by the U.S. Preventive Services Task Force.3
The easiest explanation would be that the Kal-
ager estimate is wrong. Although the task force
uses data from randomized trials, the Norwegian
data are observational — and as with all obser-
vational data, the primary threat to validity is
the comparability of the comparison groups.
But the staggered cohort design that was
used by Kalager et al. mitigates the concern that
the women in the four study groups are some-
how different, since many of the women in the
study actually contributed data to each group at
different points in their life. Contamination is a
more relevant concern. If the women in the non-
screening groups were exposed to opportunistic
mammography screening or began to benefit
from the multidisciplinary teams, which had to
be in place before the screening program was
initiated, then the background effect of time
may have been overestimated. This would have
led to an underestimation of the benefit of the
screening program. Furthermore, the follow-up
period may be too short to fully capture the ben-
efits of screening. The authors argue that these
effects are small.
So another explanation must be considered:
The New England Journal of Medicine
Downloaded from www.nejm.org at CAPITAL HLTH-HALIFAX INFIRMARY on November 16, 2010. For personal use only. No other uses without permission.
Copyright © 2010 Massachusetts Medical Society. All rights reserved.