Friday, February 3, 2017

Tom Price is a physician who doesn't understand cancer screening

If confirmed by the Senate, Dr. Tom Price will become the first medical doctor to lead the U.S. Department of Health and Human Services in 24 years. One might think that having completed medical school and practiced orthopedic surgery before entering politics might give him some extra insight into what works and what doesn't in medicine. But judging by a letter to then-HHS Secretary Kathleen Sebelius that he signed in 2011 objecting to the U.S. Preventive Services Task Force's draft recommendations on prostate cancer screening, Dr. Price either failed to learn anything in evidence-based medicine class or forgot everything he learned.

Price and colleagues wrote: "Since the [prostate-specific antigen] PSA test came into widespread use for cancer detection in the mid-1990s, the rate of deaths due to cancer has fallen by 40 percent." This statement reflects an association, not causation, and there is a serious problem with positing the latter based on the natural history of PSA-detected prostate cancers. In the European Randomized Study of Screening for Prostate Cancer (ERSPC), the only trial to conclude that PSA screening reduced deaths from prostate cancer, it took 9 years to observe any difference in prostate cancer deaths between the screening and control groups. But not only was the prostate cancer death rate falling in the U.S. long before any possible screening effect could have occurred, it was also falling in other countries (such as the United Kingdom) that were not employing PSA as a screening test.

The letter goes on to state that the percentage of potentially "curable" prostate cancers rose from 35 to nearly 90 percent after doctors started routinely screening with PSA, so it's obvious that the test works. Dr. Price, have you heard of lead-time and overdiagnosis bias? Even implementing a completely ineffective screening test, such as chest x-rays for lung cancer, will artificially increase survival rates and inflate the number of cancers amenable to surgery because a large percentage of these would never have been detected at all, if not for the test. (On this erroneous conclusion, Dr. Price is unfortunately in good company: a 2012 survey of primary care physicians found that two-thirds would recommend a cancer screening test supported by irrelevant changes in 5-year survival rates.)

Finally, the letter accused the USPSTF of having "cherry-picked" information rather than reviewing the totality of the evidence on PSA-based, which is laughable since it came in the same paragraph that Price and the other legislators highlighted the Goteburg, Sweden randomized trial (a subset of ERSPC) as "the best designed and controlled study." Talk about cherry-picking! Goteburg also happens to be the country with the most impressive-appearing benefit of screening; most of the other countries involved in ERSPC found no statistically significant mortality benefit, as did the U.S. in our own randomized trial.

These cancer epidemiology concepts I've mentioned aren't difficult to master; I teach them all the time to undergraduates and first-year medical students. Since Dr. Price clearly needs a refresher, I refer him to a previous instructional post I originally wrote for urologists. Patients can understand these concepts too, thanks to this excellent video decision aid that persuaded men to make more evidence-based decisions about PSA screening in a recent study published in Annals of Family Medicine.