« The paradox of screening | Main | Projecting iPad sales »


Feed You can follow this conversation by subscribing to the comment feed for this post.


Kaiser: Anyone testing positive in a screening test should go for a confirmatory test, as you suggested.

Well, no, I'm not suggesting that. In fact, quite the contrary. There is a segment of the population for whom it makes no sense to treat what is a slow-moving disease that will likely not cause them any problems before they die.


John: Sorry I wasn't being clear. What I meant was, even for a good screening test, with a targeted screening population, the false positive rate is likely to be higher than we like, and we should always seek confirmation.
I definitely agree with your point about the futility of learning about little things that if left uncorrected would in no way shorten our lives. That applies to the mammogram screening too.

William Ockham

[Don't take medical advice from me. I'm not a physician.]

No competent doctor is going to tell someone that they have cancer just because their PSA is "positive". Everybody here seems to be assuming that the benefit of the PSA is from a single event (getting the test once), but the real diagnostic value is being able to compare PSA results from last year with this years's. Lots of guys have "positive" PSA results (defined as outside the normal range), but not that many have results that are over 10 times as high as last year's results (that's when you should worry about prostate cancer).

Also, if you have an elevated PSA and a positive DRE (that's when the doctor sticks his finger up your butt and feels something that shouldn't be there), then you should get more testing.

With all that said, there's still an argument to be made that we're doing too many PSAs. It's just not as simple as you folks are making it.


William: I feel the need to reiterate that there are two separate issues here: (a) why broad-based screening is typically a bad idea (the original point of the post), and (b) why the PSA test in particular is a bad idea. Now on to your comment...

This is how one should think about your question of taking sequential tests. Assume Charlie does *not* have prostate cancer during the entire period in which he takes 5 PSA tests. What is the chance that Charlie will receive one or more positive findings out of 5 tests? Given the low level of accuracy of the PSA test, this probability is very high. But please note that any positive for Charlie is a false positive by our assumption.

Regarding the 10x normal range result: if the test shows a 10x normal range, it should cause alarm whether or not past tests have been negative. But, if you read Dr. Ablin's article, you will learn that PSA can be elevated for many reasons not related to having prostate cancer. Since he is the discoverer of PSA, I trust him at his words.

In fact, the healthy man who receives a positive PSA test can be compared to the clean athlete who receives a positive steroids test finding. I discuss the latter in Chapter 4 of the book. Imagine you are the athlete; try proving to the world you are clean. That is the predicament of the healthy guy who tested positive on PSA.

You're right this is a complicated issue and I'm glad we're having some serious discussion here.

The comments to this entry are closed.

Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR, Wired.

See my Youtube and Flickr.
Numbers Rule Your World:
Amazon - Barnes&Noble

Amazon - Barnes&Noble


  • only in Big Data

Next Events

Jan: 10 NYPL Data Science Careers Talk, New York, NY

Past Events

Aug: 15 NYPL Analytics Resume Review Workshop, New York, NY

Apr: 2 Data Visualization Seminar, Pasadena, CA

Mar: 30 ASA DataFest, New York, NY

See more here


R Fundamentals, Principal Analytics Prep

Numbersense: Statistical Reasoning in Practice, Principal Analytics Prep

Applied Analytics Frameworks & Methods, Columbia

The Art of Data Visualization, NYU

Signed copies at McNally-Jackson, NYC

Excerpts: Numbersense Ch. 1, 7, 8. NRYW

Junk Charts Blog

Link to junkcharts

Graphics design by Amanda Lee