Hacker News Reader

Comment by CogitoCogito | Hacker News Reader

> While this may be the case for some tests, there are probably a great number of tests that could be scaled to be done yearly on the whole population.

Isn't this already the case? Tests that are considered worth doing regularly even without extra symptoms are done regularly. E.g. mammograms and colonoscopies/stool tests.

(I'm ignoring your focus on "yearly" since really you care about "regularly" given that yearly is totally arbitrary and timing that makes sense is clearly dependent on the tests in question.)

View on HN

No. Not all screenings/tests are done, even when it's super low cost and non-invasive. For example, one factor that is used to determine whether to screen for certain types of cancer is "Possible harms from follow-up procedures". [1] Another factor that is taken into account it the chances of a false positive. [1] Both of these factors are irrational, since more information is strictly optimal given a rational agent. They should not even merit consideration in choosing whether to undertake a screening (this information should only be used after a screening is done, to determine follow-up action).

In my opinion, the only justifiable factors are probably along the lines of:

- invasiveness (e.g. if it's non-invasive or minimal cost)

- benefit to the patient if detected

- cost relative to other screenings/actions that can be done for the patient

Everything else seems strictly suboptimal.

[1]: https://en.wikipedia.org/wiki/Cancer_screening#Risks

> For example, one factor that is used to determine whether to screen for certain types of cancer is "Possible harms from follow-up procedures". [1] Another factor that is taken into account it the chances of a false positive. [1] Both of these factors are irrational, since more information is strictly optimal given a rational agent.

Your entire argument here seems to require patients to be rational agents. They aren't.

> Your entire argument here seems to require patients to be rational agents.

Yes, my argument does rely on patients being rational agents.

> They aren't.

That may be true, but it's certainly paternalistic (in the formal, definitional sense) to act as if they are not rational agents and withhold information/reduce autonomy. This is a case of pure paternalism (again, in the formal philosophical sense).

Different ethical systems, of course, make different judgements on whether this behavior is moral. It's also up for debate whether this is a desirable feature of the medical system.

> That may be true, but it's certainly paternalistic (in the formal, definitional sense) to act as if they are not rational agents and withhold information. This is a case of pure paternalism (again, in the formal philosophical sense).

Call it whatever you want. Idealizing patients as rational agents instead of considering how they are in reality results in worse outcomes. If your goal is to actually help people, you should base your arguments on how things are in reality instead of some idealized dream world.

> If your goal is to actually help people, you should base your arguments on how things are in reality instead of some idealized dream world.

That's a fair position (you seem to be a utilitarian), and probably quite defensible. But, one could argue that patient choice is an important feature for our medical systems to have. I certainly want to be able to refuse medications that my doctor recommends (e.g. opioids) or seek alternate advice/second opinions. My cost-benefit analysis equation is probably not the same as my doctor.

To be clear, I'm not talking about an "idealized dream world", as you put it -- I'm talking about patient autonomy in the real world, even if it means allowing patients to make what seems like a suboptimal decision.

> I certainly want to be able to refuse medications that my doctor recommends (e.g. opioids) or seek alternate advice/second opinions.

You can do this.

> I'm talking about patient autonomy in the real world, even if it means allowing patients to make what seems like a suboptimal decision.

Patients are allowed to have these tests done against the recommendations of doctors (of course the costs might not be covered by insurance).

So I'm not really sure what you're arguing anymore...

I made that point because I argue that patient autonomy is important. Further, I argue that in order to protect patient autonomy, information should not be withheld from the patient. Accordingly, the patient should be given everything necessary to enable them to make optimal choices as a rational agent. Not providing low-cost, minimally-invasive screen tests to the patient, under the premise that the patient is an irrational agent, reduces their autonomy insofar as it limits their decision-making capacity. I argue that there is no compelling justification to do this for low-cost, minimally-invasive screenings. Furthermore, assuming that the patient is rational and finding ways to advance the rational patient's well-being protects patient autonomy. Therefore, the only justifiable factors that should be considered when screening patients that preserve patient autonomy is what I stated here: https://news.ycombinator.com/item?id=31450997

Does the position seem clear now? I apologize if this line of reasoning was not clear earlier.

> Not providing low-cost, minimally-invasive screen tests to the patient, under the premise that the patient is an irrational agent, reduces their autonomy for no justifiable reason.

The reality that patients are irrational agents _is_ a justifiable reason.

> Therefore, the only justifiable factors that should be considered when screening patients that preserve patient autonomy...

Your position is clear. I disagree. I think given real constraints (economic, legal, psychology of patients/doctors/etc.) focusing on overall quality of outcomes makes most sense. You can disagree if you want, but don't pretend it's not justifiable.

edit: Besides all the tests are available to you. You can pay them if you'd like. Given they are low-cost (at least those you're focusing on here) it really isn't a burden on you to get the tests done yourself.

Ok, let's agree to disagree, then.

I'll note that I never "pretended that it's not justifiable". I did explicitly say earlier:

> That's a fair position (you seem to be a utilitarian), and probably quite defensible.

I'm disengaging, for now, since it seems like we understand each other's position, and we had a productive, lively discussion.

Reply on news.ycombinator.com