Thoughtful analysis of the USPSTF and prostate cancer screening

I asked one of our medical editor/story reviewers, Richard M. Hoffman, MD, MPH, to write about the news stories predicting what the US Preventive Services Task Force will recommend on prostate cancer screening.

Rich Hoffman_67x100.jpgHoffman is a general internist, is a Professor of Medicine at the University of New Mexico School of Medicine and a staff physician at the Albuquerque VA Medical Center. He also serves as Interim Director for Cancer Prevention at the University of New Mexico Cancer Center. His areas of research interest are prostate and colorectal cancer screening and prostate cancer treatment outcomes, with expertise in clinical epidemiology, health services research, and meta-analysis. He is a medical editor for prostate cancer topics for the Foundation for Informed Medical Decision Making and works with the Foundation to develop shared decision making tools for prostate cancer screening and treatment of localized prostate cancer. The Foundation is also the sole financial supporter of this website.
Once again, the United States Preventive Services Task Force seems headed for a storm of controversy. The New York Times reported today that the Task Force will issue draft recommendations next week advising against prostate cancer screening. The Task Force concluded that screening should be discouraged because it has no net benefit or the harms outweigh the benefits.

Not even two years ago, the Task Force’s recommendation against routinely performing mammography in average-risk women before age 50 outraged professional organizations, advocacy groups, celebrities, politicians, and many other vocal critics. Unfortunately, the somewhat awkward message was distorted to imply that these women should never undergo mammography. In fact, the Task Force was highlighting that mammography for these women involved important trade-offs between potential benefits and harms–and that women needed to make informed, personalized decisions. This is an appropriate message.

However, the new recommendation on prostate cancer seems troubling. In 2008, the Task Force gave prostate cancer screening an “I” rating because the available evidence was considered insufficient to recommend for or against screening–largely because there were no valid data from randomized controlled trials of screening. The Task Force suggested that men be informed about the risks and benefits of screening, though strongly advised against screening men age 75 and older.

In 2009, long-awaited results from the major randomized controlled trials were published. The European Randomized Study of Screening for Prostate Cancer (ERSPC) found that screening reduced prostate cancer mortality by 20%. In contrast, the American Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) found no benefit with screening. However, evaluating PSA in America was challenging because screening was so pervasive that investigators had difficulty enrolling men who had never been screened–or preventing screening in the control group. The validity of the negative PLCO results is uncertain. The ERSPC study was more credible, but showed only a small absolute survival benefit. This benefit must be balanced against the harms of overdiagnosis–finding cancers that would never cause clinical problems during a man’s lifetime–and the resulting overtreatment, which can lead to urinary, bowel, and sexual dysfunction.

Unfortunately, we cannot confidently identify the “overdiagnosed” cancers found with screening, so most men will undergo surgery or radiation therapy. A new strategy of active surveillance might mitigate the harms of overdiagnosis. With this option, men are closely monitored with PSA tests, digital rectal examinations, and biopsies–and will be offered surgery or radiation only if the cancer shows signs of being aggressive.

Ideally, men should be making an informed decision that best reflects their values for the potential downstream consequences of screening. This is indeed the conclusion (based on similar clinical data) that the Task Force reached regarding mammography for women before age 50. The Task Force classified this as a grade “C” recommendation–which is what I expected for prostate cancer screening.

Publisher’s note: Anyone who is not familiar with the USPSTF’s grading system should read about it. The difference between a C and D recommendation is significant.

Another very thoughtful commentary has been posted by Dr. Len Lichtenfeld of the American Cancer Society. Excerpt:

“We have invested over 20 years of belief that PSA testing works. Catch it early, treat it early, and get it out. Save a life. That’s the mantra many of us-including me, as a practicing physician-believed. And here we are all of these years later, and we don’t know for sure. That is not an acceptable situation. Plain and simple, we have not done our homework to prove our point. And the chickens are coming home to roost.

Unfortunately, those “chickens” are men like me who dutifully get our blood tested every year. We have been poked and probed, we have been operated on by doctors and robots, we have been radiated with fancy machines, we have spent literally billions of dollars. And what do we have? A mess of false hope?”


A decision aid on prostate cancer screening, produced by the Foundation for Informed Medical Decision Making, is now available online for a limited time. One interesting aspect is that it features two doctors – one who chooses to be screened using the PSA blood test, and one who declines.


You might also like


Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.

Juliet Mavromatis

October 10, 2011 at 5:59 pm

Thanks for this. I spent the afternoon reading about using psa for prostate cancer screening and trying to sort through the truth about the utility of the test. I would like to hear more about what Dr. Hoffman finds troubling about the USPSTF new recommendation. PSA screening is the bastard child of evidence based medicine fanatacists. Not that I have a problem with them, but the trend toward reporting and emphasizing non-efficacy seems concerning to me. My reading of the literature suggests that we should study alternative screening intervals. I’m still not clear how I will advise my male patients, but will continue to read to seek a better answer.


October 10, 2011 at 10:32 pm

seems that it would behoove us to wait till we have good evidence before recommending screening tests. this backtracking business leads people to think that care is being rationed.
but truly, we should also think about the costs of tests and treatments and in our current health care system, we cannot afford basic medical care so please do not keep adding more tests and treatments.

Richard Hoffman

October 11, 2011 at 2:53 pm

I appreciate your comments. I’ve now been able to read the draft recommendation statement and can address some specific issues.
First, I found their interpretation of the evidence to be troubling. I completely agree that screening carries substantial risks for overdiagnosis, overtreatment, and false positive tests. However, we do have data from a large randomized trial (ERSPC) that shows a survival benefit, albeit small. The negative randomized screening trials had important methodological limitations that undermined their validity. Citing meta-analyses that include these flawed studies to argue the lack of benefit is unconvincing. I also found it disingenuous to note that screening does not reduce overall mortality (a point highlighted by the Task Force Chair in the initial news report)–that’s also true for colorectal cancer screening (USPSTF Grade A) and breast cancer screening (Grade B for 50+).
Another issue is that the Task Force’s grading seems inconsistent. The USPSTF estimated that over 1900 women ages 39-49 need to be invited to undergo mammography to prevent one breast cancer death. They cited harms of mammography, including false positive tests, unnecessary additional procedures, and overdiagnosis. With similar harms, ERSPC showed that over 1400 men needed to be invited for PSA testing to prevent one prostate cancer death. Given these trade-offs, I think patients need to engage in informed decision-making about both prostate and breast cancer screening and make personalized decisions that reflect their values–which implies a Grade C recommendation.
Finally, the logistics of the “don’t ask, don’t tell (unless they ask)” approach are unclear–and could place unnecessary burden (including medicolegal) on primary care providers.

Bill Morgan

October 14, 2011 at 9:09 am

Hi Gary
I think eliminating PSA testing (combined with a DRE) for healthy men between the ages of 40 and 70 or so is absolutely ludicrous..I have been following the articles concerning the recommendations of the USPSTF that healthy men no longer be given the PSA and really don’t get it. I got my first PSA (and DRE) when I began working at an oil refinery in 1989 when I was 45 as part of the initial physical to establish baseline data and had my last one in 2001 when we were acquired and I was laid off. At that time my PSA was 1.9 and the DRE was negative. I resumed annual physicals in 2005 and had one each year through 2009 including PSA testing and DRE’s. My PSA scores slowly rose from 2.4 to 2.8 to 3.2 to 3.6 and suddenly to 4.3 with DREs. In 2009 (when I was 63) my primary Physician felt a lump so off to urology. The urologist felt the same lump , scheduled me for a biopsy in March 2009, which was not pretty, but had results back in mid March showing Gleason scores of 4,5,and 6 in 19 of 20 core samples. I then went to a urologist highly skilled in the use of the Da Vinci Robotic surgical machine who removed my prostate in May 2009.( 1 night stay in the hospital.) Pathology results by then were 8’s and 9’s everywhere, but because of the prompt surgery I have had no recurrence..My PSA’s have been less than .01 every six months. I had no chemotherapy and no radiation because of the combination of PSA testing, DRE’s in conjunction and prompt surgery.
So what is it with this task force???Certainly older men or unhealthy men should not go through these procedures, but I believe they stop PSA testing at age 70 or 75 anyway. There are no symptoms until the cancer has metastasized, and self testing is impossible. (Think about it)
PS I am a Vietnam War Veteran and tromped around in the areas that had been defoliated with Agent Orange, so probably I was higher risk..but a blanket statement that healthy men forgo these tests doesn’t make sense Do they want us to wait til we’re sick??

Tim Bartik

October 14, 2011 at 4:26 pm

Dr. Hoffman’s comment is by far the most thoughtful comment I have seen on this issue, and the most concrete. A C recommendation makes much sense than a D recommendation.
I note that as I have already indicated in a comment on another blog entry at this blog, in the case of prostate cancer, the reason why we can’t find an overall mortality effect is simple: statistically, there’s no reason to expect a 0.07% reduction in mortality after 9 years (the European study’s estimate) to be statistically detectable when the overall mortality rate is 10% in a sample of 80,000 each in the screening and control arms. The standard errors will be much too big. You would need a sample size of over 1 million. In contrast, given that the prostate cancer-specific mortality rate is about 0.36% in the European study in the control arm, it’s easier to detect that relatively large prostate cancer specific mortality reduction.
I don’t know if that is also the problem with detecting the effects of colorectal cancer screening or breast cancer screening on overall mortality, but it might be. There is some basic statistical analysis here that is being ignored.

Juliet Mavromatis

October 16, 2011 at 6:42 pm

Thanks for your reply Dr. Hoffman. It sounds like you and I are on the same page with respect to PSA testing and the evidence. The problem for clinicians is that you know that our patients are going to ask–“what do you think I should do Dr. XYZ ? ” I agree that it is useful to explain to patients what the risk of the test could be, but the bottom line is that you and I will be left with advising our patients. Do we advise for or against, in the case of PSA ? For men between ages 50 and 65 my personal inclination, based on the evidence, might still be to advise for. This is an example of an instance where medical evidence could potentially lag behind clinical judgement. There is little we can do to prevent cancer death and the numbers needed to treat remain high. I would like to see trials of alternative intervals for screening.

Howard Towt

October 17, 2011 at 12:33 pm

Here is a link to my thoughts on this issue…
Like Mr. Morgan (see comment above), I see merit in the test and don’t admire the “kill the messenger mind set.”
If we’ve got a good diagnostic technique to indicate the prostate is abnormal in some fashion, why not use it? If we later determine there is no high-grade cancer, why rail against the test? Treat the BPH,low Gleason score, etc. and continue with life.
Submitting to any invasive medical procedure has risks (note the form you sign before surgery), but to single out a blood test as the demon? That seems inappropriate, and clearly motivated by something other than science.