Biology in the News Explained

The USPSTF deals with data, not hyperbole

The U.S. Preventative Services Task Force agrees with Bioblog that it is finally time to admit that there are risks to cancer screening, in addition to benefits. It turns out that the benefits are marginal at best for women under 50 being screened for breast cancer.

Not too surprisingly (in fact, rather laughably, given their blatant conflict of interest), the American College of Radiology has gone, shall we say, hysterical on the issue, declaring that if the USPSTF’s recommendations are followed, “countless American women may die needlessly from breast cancer each year.” This hyperbolic statement is completely belied by the data that show clearly that screening affects mortality minimally at best.

In fact, the “classic” breast cancer screening study, known as the “HIP” (Shapiro, 1997) found a modest drop in mortality associated with mammography; in absolute terms, the difference in death rate was 5 deaths over 10 years over 1000 people, versus 7 deaths – hardly an example of “countless” women dying needlessly.

In addition, this and several other studies were reviewed (Olsen and Gøtzsche, 2001) with the conclusion that it had major methodological flaws (explained in more detail in Welch, 2004). The first was that deaths whose cause was unclear were more likely to be called breast cancer deaths in the non-screening group than in the screened group. This effect could account for half the observed difference in death rate.

Second, more women were excluded from the mammography group for prior history of breast cancer (once the trial started) than from the non-screened group, because women in the screened group were examined more closely than those in the control group, who did not actively participate in the trial, but rather went about their business. Because women in the screening group were far more likely to be identified as previous breast cancer patients, there were probably hundreds more previous cancer patients in the control group.

Both of these methodological problems clearly skew the results towards a conclusion that screening prevents deaths.

And here’s another serious problem with screening resulting in overtreatment that you have not read about in any of the articles describing this controversy – surgery itself can cause the spread of cancer (Retsky et al., 2003):

As a result of screening, cancers are found at an earlier stage than would be found without screening, which is favorable, but then surgical intervention to remove the primary tumor accelerates metastatic growth, offsetting the early detection advantage.

The fact remains that the panel changed its recommendations with regard to screening (since recommending it much more favorably in 2002) because the data simply aren’t there to show an obviously positive effect, and more and more negative effects of overtreatment are coming to light. It would be most beneficial to be able to separate screening from treatment in our minds; but the fact is that our lawsuit-happy society combined with the “cancer=death” attitude promoted by the very people who push screening on us in the first place makes it nearly impossible for any doctor/patient combination to adopt a “watchful waiting” approach when cancer is found.

Another fact is that radiologists rake in a pile of money for cancer screening. Obviously they are scared to death that their cash cow might be slaughtered, so they are resorting to extremely un-Hippocratic scare tactics to make sure it is not. The USPSTF’s report lists the breast-cancer screening recommendations of various national and international groups, and their new suggestion is far from radical; it fits right in a large range. The assertion that “countless” cancer deaths will be caused by their analysis is outrageous and irresponsible and ignores reams of data that say otherwise.

All the health advice given by all kinds of official-sounding groups can be dizzying. But a good rule of thumb is to ignore recommendations from groups that have a financial stake in what they are recommending. That’s pretty obvious when it comes to pharmaceutical companies, but organizations representing doctors are influenced by money too. It’s another example of the irrational distortions created by the “fee for service” system, which needs to be eliminated in favor of salaries for doctors, in order for health care reform to succeed.


Olsen, O., and Gøtzsche, P.C. 2001. Cochrane review on screening for breast cancer with mammography. The Lancet 358(9290):1340-1342.

Retsky M, Demicheli, R., and Hrushesky, W. 2003. Breast cancer screening: controversies and future directions. Current Opinion in Obstetrics and Gynecology 15(1):1-8.

Shapiro S., 1997. Periodic screening for breast cancer: the HIP Randomized Controlled Trial Health Insurance Plan. J Natl Cancer Inst Monogr 22:27-30.

Welch, H.G., 2004. Should I Be Tested for Cancer?: Maybe Not and Here’s Why
University of California Press, Berkeley, CA.


2 Responses to “The USPSTF deals with data, not hyperbole”

  1. Mark Faasse says:

    Nice post – and kudos for calling out the ACR.

  2. barbara jackson says:

    The truth is always so refreshing! Why doesn’t everybody tell it?

Leave a Reply

Your email address will not be published. Required fields are marked *

× six = 42

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>