To PSA test or not to PSA test: That is the discussion

Steven J. Atlas, MD, MPH

Though it seems Americans don’t agree on much, screening for cancer is an exception. Who wouldn’t support preventing or identifying cancer at an early, more treatable stage, when the alternative is pain, toxic therapies, and a shortened life? That may be why people get confused when news headlines don’t reinforce a “just do it” message. A recent example of the disconnect between public perception and medical evidence is screening for prostate cancer using the prostate-specific antigen (PSA) test. The United States Preventive Services Task Force (USPSTF), a non-governmental expert panel that produces guidelines for primary care providers, proposed new recommendations saying doctors should only order the PSA test for older men after discussing its pros and cons and eliciting preferences for screening.

Screening for prostate cancer with the PSA test: The backstory

To understand the new draft recommendation requires a brief history of this test. Introduced in the 1980s as a way to follow patients already diagnosed with prostate cancer, it began to be used to screen for new cancers. Given that the PSA is an easy blood test to perform, it was quickly adopted — without waiting for evidence that it actually worked. For many years, the USPSTF said there wasn’t enough information to recommend for or against the PSA test.

That changed in 2012 when the USPSTF released a controversial recommendation against screening. It was based in part on a large US study showing no decrease in prostate cancer deaths among men screened using the PSA test. The recommendation also reflected concern about the test causing a surge in prostate cancer diagnoses, many of which were small, low-risk cancers being treated with surgery or radiation — treatments with common side effects.

I was uncomfortable with this “don’t screen” recommendation and am happy about the proposed change. Here’s why: while the US screening trial was negative, another large study in European men showed a small decrease in prostate cancer deaths after more than 10 years of follow-up. Moreover, specialists had devised new strategies to avoid overtreating low-risk cancers.

Having a conversation about screening with the PSA test

I discuss the pros and cons of the PSA test with my patients and ask about their personal preferences for screening. I tell them that while screening can reduce death due to prostate cancer by 20%, the “bang for the buck” is small. It takes screening of over 1,000 men to prevent one death. I also highlight that the benefit of screening is years off, but the risks of treatment — impotence, incontinence, and bowel problems — occur right away.

I also emphasize the PSA test isn’t very accurate. There can be anxiety due to false positive results, meaning further testing shows no cancer. I mention the potential for diagnosing a low-risk cancer where the treatment may be worse than the disease, and that following them closely without treatment may be preferable. How much a man wants to know something like this can differ — some view it as useful information, others see it as an endless source of worry.

Finally, I share my own perspective. As a medical student in the late 1980s learning about the PSA test, my grandfather was dying from prostate cancer. He was an otherwise healthy man who had many good years left, and I wondered if the PSA test could have helped him. Because of this family history, I have decided to have the PSA test. But I’m also unsure what I would do if I didn’t have that history. The small potential for benefit must be weighed versus the risk of false positives or of finding a low-risk tumor that may never cause harm. I can see how two men without risk factors for prostate cancer could make different decisions.

So, I think the USPSTF got it right. This is a decision best made by a well-informed patient in collaboration with his doctor. The challenge in implementing this is practical: the short time I have with each patient. I could save time by simply ordering the test without a discussion. But in my role as health advisor, I need to be able to not only say when I think we should or should not do something, but also when there is a choice. And when there isn’t one right decision for everyone, my patient is the best person to make the choice that’s right for him. I can state my personal preference, but need to highlight why that may not be the right answer for him.


The US Preventive Services Task Force 2017 Draft Recommendation Statement on Screening for Prostate Cancer: An Invitation to Review and Comment. JAMA, published online April 11, 2017.

Screening for Prostate Cancer: U.S. Preventive Services Task Force Recommendation Statement. Annals of Internal Medicine, August 2008.

Understanding Task Force Recommendations, Screening for Prostate Cancer. The U.S. Preventive Services Task Force, May 2012.



  1. andrew goldstein

    How PSA cancer screening with its inadequate specificity but good sensitivity can be wrong for some men escapes me. Isn’t the question of the health care provider’s skills an important part of getting the maximum value from serial PSA data while minimizing unnecessary, invasive treatments? Men get only one shot at curative treatment for PCa that requires treatment. There are many more tests now available that help men make the right treatment decisions. But first, men need clear and up to date guidance from their doctor.

    • Steven Atlas

      False positive are only one problem with the PSA test. Probably more important is that the PSA test can identify cancers that will never cause harm. Yes, not all prostate cancer is deadly. It is estimated that 30-40% may be so slow growing that they may never end up causing trouble. However, almost all men diagnosed are treated – and those treatments can cause harm – impotence, incontinence, bowel problems. So the evidence does not support what you say – despite your intuitive argument. For more detail, take a look at the USPSTF document.

  2. Michael Baca-Atlas

    Nice overview. It’s important to note differences in mortality by race. North Carolina has significantly higher mortality in AA patients. May be multifactorial but I make a point to mention this for my AA patients or who have extensive fam hx.

  3. Bjorn Tore Hammer Engelstad

    In Norway a PSA test are offered to all men over 50 y . old , as a part of the yearly health check at your GP , free of charge . My GP has informed me that the test has it weakness, but I prefer a PSA test instead of the old practice where the doctor asked you to bend over and you got a finger in your anus for a prostate check . My two Medical Doctors have both told me , that most men over 70 y. old will get prostate cancer , but they will live with , and die of other causes when you reach 70 years .For the records . Prostate cancer are still the number one killer , out of all types of cancer , for men in Norway .

Post a Comment:

This blog aims to provide reliable information as well as healthy dialog about the topics covered. We reserve the right to delete comments for any reason, particularly those that do not relate directly to the contents of this post, are commercial in nature, contain objectionable or inappropriate material, or otherwise violate our Privacy Policy. Promotional URLs will be removed from comments. Comments on this blog do not represent the views of our editors or Harvard University, and have not been checked for accuracy. All comments submitted to this site become the non-exclusive property of Harvard University.