Hacker News Reader

Comment by redis_mlc | Hacker News Reader

Although you're trying to appear logical, you're not. There is no black or white in health care.

For example, most men who get prostate cancer die of something else first. The treatment often destroys quality of life, so is worse than the disease.

Also, you can't compare women's diseases to male diseases. Women have all kinds of body changes that men don't experience.

And doctors generally don't signup for heroic measures after retirement. They know when your time is up, prolonging life in an ICU is not living.

You're not going to live forever.

View on HN

> For example, most men who get prostate cancer die of something else first

While that is true, it is also true that men who get prostate cancer die several years earlier than men who don't get prostate cancer, and treatment do increase life expectancy significant.

Two clear truths, but the implications are not as clear. A common theory is that the combination of older age and cancer results in a much lower margin for error, turning non-deadly "something else" into deadly "something else".

It becomes a question of definition if "something else" should be seen as independent of the cancer or not.

In fact as well as appearance I am logical. Screens are just that, they screen for things, and by their nature some slip through - one hopes later screens will catch the problem as it gets larger

You're talking about false negatives, but the problem with cancer screening is the false positive: telling someone they have cancer when they don't.

https://www.harding-center.mpg.de/en/fact-boxes/early-detect...

One of the reasons cancer screening is pushed so hard is because people use the wrong metrics. People look at 5 year survival rates, or cancer detection rates, to judge success.

Imagine 1000 men with aggressive prostate cancer who don't have screening. After 5 years about 440 of those men will die. That's a survival rate of 44%.

Now imagine those exact same men but pooled with 2000 men who do not have prostate cancer (because this is what screening does). After 5 years those men will still be dead. But the survival rate is now 81%.

This is the bias caused by over-diagnosis.

Yes, the flaw is inherent in the accuracy or lack of it of the test method. Things like needle biopsy of the prostate can miss cancerous cells, unless you take 100 samples. An ideal tests would detect cancer markers in the blood for prostate and other cancers. I understand there is progress in that area.

> Things like needle biopsy of the prostate can miss cancerous cells, unless you take 100 samples.

biopsies can cause a stable tumor to metastasize [0]. Sometimes medical screenings are useful, but when you’re looking for problems and don’t actually have symptoms, practitioners have an incentive to find something to treat you for. Ideally you’ll be fired as a patient, but I’ve heard of doctors who do procedures and operate every time they can possibly justify it (“he owes money to everyone.”)

[0] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4015162/

Yes, I have heard stories from a friend who had needle biopsies, he likened it to being machine gunner by a Japanese Zero in WW2. I am over 80, so I have chosen the finger for years, and when I became enlarged I took the prostate shrinking pills, since I was a widower, and I could probably only have sex with desire with a hooker, as I would not be attracted to an older woman for the same reason no young woman would be attracted to me, yes the shrink meds also made me not in the least bit horny, I do not miss it at all.

In a study here in Sweden they did a test with 20000 men where half got PSA test every second year since 1998 and other half acted as control and did not. The group that got the PSA test had 30% less deaths by prostate cancer.

That seems like a pretty decent way to judge success.

https://gupea.ub.gu.se/handle/2077/57825 page 92: Our data confirm that an organized screening programme reduces mortality by ~30%, a level that is similar to evidence from the ERSPC and the PLCO trial (showing a 25–32% lower risk) (175).

Reply on news.ycombinator.com