Author Topic: Medical Tests and Probability  (Read 370 times)

0 Members and 1 Guest are viewing this topic.

Online The Latinist

  • Cyber Greasemonkey
  • Technical Administrator
  • Reef Tank Owner
  • *****
  • Posts: 9264
Re: Medical Tests and Probability
« Reply #15 on: May 16, 2017, 07:59:40 PM »
I would like to add that the in incidence of breast cancer peaks in women 75-79 at 450 per 100,000.  The chance that a positive screening mammogram represents a real cancer therefore never rises above about 4.5%
« Last Edit: May 16, 2017, 08:02:25 PM by The Latinist »
I would like to propose...that...it is undesirable to believe in a proposition when there is no ground whatever for supposing it true. — Bertrand Russell

Offline Soldier of FORTRAN

  • Stopped Going Outside
  • *******
  • Posts: 5545
  • Cache rules everything around me.
Re: Medical Tests and Probability
« Reply #16 on: May 16, 2017, 08:05:16 PM »
But will you?  There's plenty of room for doubt about that.

Personally, yes. 

'100% chance of irritating test' vs '10% chance of horrible death'
Every soup ladled to the hungry, every blanket draped over the cold signifies, in the final sense, a theft from my gigantic paycheck.

Offline Friendly Angel

  • Frequent Poster
  • ******
  • Posts: 2858
  • Post count reset to zero in both forum apocalypses
Re: Medical Tests and Probability
« Reply #17 on: May 16, 2017, 08:15:56 PM »
This is why, with my doctor's agreement, I've declined to have a PSA test done at 40.  Most studies seem to suggest that PSA screening of asymptomatic people has very little chance of significantly affecting outcomes and a high risk of needless treatments that can negatively affect quality of life.

Aye, there's the rub.

Friend of mine had a high PSA, they scheduled him for a biopsy, he worried for a couple weeks, had a very unpleasant procedure, and after waiting a while got the all-clear.  That is the emotional cost of these tests.

I had a PSA test and prostate exam at 50.  At 55 I had a new doctor and he said if I'm not having symptoms I didn't need them again.

But another friend (much older) had his prostate removed and gets on his soapbox about exams all the time.
Amend and resubmit.

Online The Latinist

  • Cyber Greasemonkey
  • Technical Administrator
  • Reef Tank Owner
  • *****
  • Posts: 9264
Re: Medical Tests and Probability
« Reply #18 on: May 16, 2017, 08:21:58 PM »
But the problem is, Soldier of Fortran, that there is not a lot of evidence that such screenings actually do extend lifespans significantly.  Doubtless for some they do, but for others you may be treating cancers which would never have presented clinical symptoms or for which the outcome would have been the same regardless.  In some cases there are questions about whether the early detection of certain cancers lengthens life or merely the duration of cancer diagnosis.  If detecting my cancer 5 years earlier means that I will live 7.5 years after diagnosis instead of 2, what have I really gained?  Perhaps six months of life and five years of cancer treatment that I could otherwise have avoided.  The calculation is just not as simple as you suggest; you appear to be ignoring a host of very relevant factors.
I would like to propose...that...it is undesirable to believe in a proposition when there is no ground whatever for supposing it true. — Bertrand Russell

Offline 2397

  • Seasoned Contributor
  • ****
  • Posts: 831
Re: Medical Tests and Probability
« Reply #19 on: May 16, 2017, 08:39:48 PM »
This recommendation of not having general screening before 50 might be relevant reading.

https://www.uspreventiveservicestaskforce.org/Page/Document/RecommendationStatementFinal/breast-cancer-screening1

Offline gmalivuk

  • Well Established
  • *****
  • Posts: 1340
    • http://gmalivuk.livejournal.com
Re: Medical Tests and Probability
« Reply #20 on: May 17, 2017, 04:59:11 PM »
when a doctor prescribes a test, she thinks you are no longer part of the general population
As others have said, this is generally false for things like breast cancer, where screenings are recommended based on age alone absent any other risk factors, and age alone isn't enough of a risk factor to change the numbers significantly from the story problem in the OP.

The Latinist already cited the incidence statistics, and here's a source. Even between 75 and 79 only about 0.45% of women are diagnosed with breast cancer per year, and all other 5-year age ranges have lower incidence than that.
The world is so exquisite with so much love and moral depth, that there is no reason to deceive ourselves with pretty stories for which there's little good evidence. Far better...is to look death in the eye and to be grateful every day for the brief but magnificent opportunity that life provides.

Online superdave

  • Stopped Going Outside
  • *******
  • Posts: 4848
  • My name is not dave.
Re: Medical Tests and Probability
« Reply #21 on: May 17, 2017, 06:13:10 PM »
when a doctor prescribes a test, she thinks you are no longer part of the general population
As others have said, this is generally false for things like breast cancer, where screenings are recommended based on age alone absent any other risk factors, and age alone isn't enough of a risk factor to change the numbers significantly from the story problem in the OP.

The Latinist already cited the incidence statistics, and here's a source. Even between 75 and 79 only about 0.45% of women are diagnosed with breast cancer per year, and all other 5-year age ranges have lower incidence than that.

I hate to play the semantic game, but  I was referring to a diagnostic exam and not a screening exam. 

Offline gmalivuk

  • Well Established
  • *****
  • Posts: 1340
    • http://gmalivuk.livejournal.com
Re: Medical Tests and Probability
« Reply #22 on: May 17, 2017, 06:36:52 PM »
You included simply being 55 as a criterion indicating a test. That sounds an awful lot like a screening that isn't based on any particular risk factors other than age.
The world is so exquisite with so much love and moral depth, that there is no reason to deceive ourselves with pretty stories for which there's little good evidence. Far better...is to look death in the eye and to be grateful every day for the brief but magnificent opportunity that life provides.

Offline HanEyeAm

  • Not Enough Spare Time
  • **
  • Posts: 221
Re: Medical Tests and Probability
« Reply #23 on: May 17, 2017, 11:46:04 PM »
A quick note here that I hope folks find useful.

It is easy to confuse "prevalence" vs. "incidence."

Prevalence: the number of individuals in the population (or proportion thereof) that has the condition at a particular time. This is often called "point prevalence" because it a count (or proportion) at a particular time. "Period prevalence" is just how it sounds, the number (or proportion) having the condition at some point during a particular period of time.

Incidence: the number of new cases in a population (or proportion thereof) appearing within a particular period of time.

Prevalence is usually quite a bit higher than incidence (unless it kills you quickly). For example, the proportion of American males with prostate cancer right now (prevalence) is likely higher than the proportion of the population who are new cases appearing in 2017. To illustrate, if I did some quick math right, there are ~400k new diagnoses each year (incidence) and ~3 million living with prostate cancer (prevalence) (https://seer.cancer.gov/statfacts/html/prost.html).

To calculate the diagnostic accuracy of a screening test, prevalence, not incidence is used.
« Last Edit: May 18, 2017, 12:06:56 AM by HanEyeAm »

Offline lonely moa

  • A rather tough old bird.
  • Stopped Going Outside
  • *******
  • Posts: 4115
Re: Medical Tests and Probability
« Reply #24 on: Today at 02:05:57 AM »
The NZ breast cancer screening pamphlet declares that 7 unscreened women out of 10, 000 will die of breast cancer in NZ this year.  Next line, 5 screened women out of 10,000 will die of breast cancer this year.  They don't talk about the bad outcomes of false positives.

"The home of the brave and the land of the free; the less you know, the better off you'll be"

Warren Zevon