Hacker News Reader

Comment by Xcelerate | Hacker News Reader

The counterarguments presented in the article are silly. Of course we should have full body scans for preventative purposes. Could you imagine a utopian society where we don’t regularly have full body scans to detect problems early? Any new problems that arise as a result of this trend should be handled as exactly that — new problems.

Too many false positives leading to wasted time and costs? Raise the price for the supplemental scans and improve detection precision and recall. Unnecessary procedures that make a benign situation worse? That’s a sign to improve medical research in that area. Too much demand for MRI machines pulling away time from patients that need to use them? Build more MRI machines. This is one of those situations where the ideal end goal seems obvious, and I think we should just plow ahead and address new problems as they arise. In the long run, better preventative medicine will save time, costs, and minimize suffering for everyone. A few road bumps along the way shouldn’t dissuade us from even getting started, otherwise we never will.

View on HN

No medical intervention is without risk. Any drug or surgical intervention has the risk of negative side effects which must be weighed against the benefits of the intervention. Medical research won’t make surgeries much less risky than they are, even highly targeted laparoscopic procedures have risks. Cutting deep into you tissues is just inherently risky. Drugs too, even the safest drugs will have very negative outcomes for some number of people.

That you think this is a trivial problem shows that you’re largely ignorant of some 30 years of research and clinical debate.

What are the risks involved with getting an MRI?

I believe they're saying that this type of comprehensive early detection leads to subsequent interventions on issues that would never develop into anything serious where they left alone. by increasing the number of interventions happening you could actually be making the patient worse off

> early detection leads to subsequent interventions on issues that would never develop into anything serious where they left alone.

Then wouldn’t these be left alone ?

The more research and procedures done the more info gathered on the evidence

At least in the US, that's not how the healthcare system works. Some blame CYA. Some blame the fact that the system only makes money per procedure. The GP is saying in the US, patients often get procedures that are not worth the risk.

No, I'm saying that anywhere in the world you might get an unnecessary procedure. It is impossible to know in many cases if small tumors will become malignant, so judgement calls must be made. Finding more of them inevitably leads to more unnecessary procedures across a population. This is well established in European and American public health literature. This is why screening guidelines exist.

On an individual level, lets assume a polyp in the colon, there's no way to know which may become individually cancerous and which will remain small and benign. The genes are the same, the size and appearance may be the same. You can't know.

At a population level its well established that beyond a certain point you're injuring more people than you are helping by doing the colonoscopies at all. Where to draw that line is where the debate is playing out. The problem is well known, and I'm not aware of ANYONE seriously suggesting that it's possible to eliminate this problem.

If the MRI finds a dozen indicators in your body in a bunch of different locations that might be cancer, all with different likelihoods where would you draw the line on getting them biopsied? Would you be comfortable having heterogeneous masses in your body that have a minor chance of being cancer? Because the MRI is going to say "might be", not "is".

> Then wouldn’t these be left alone ?

The problem is that you don't know which ones would never develop into anything if left alone.

The reason they're left alone now is that you don't know they're there, because you never do any test that shows them being there.

But once you do a test that shows that something is there, it's a lot harder to just leave it alone. Something is there, and you don't know that it's never going to be a problem, so you end up not leaving it alone.

If you were going to leave the thing alone anyway regardless of what the first test showed, there's no point in doing the first test at all.

After getting another answer from them further below in this thread, I am under the impression that they are not entirely forthright about what it is they are claiming, or not claiming.

It is well established that more MRI lead to more procedures that are ultimately themselves risky and useless. This is extremely difficult to parse out on an individual level but obvious in the data at a population level.

Colonoscopies are a great example. As their use increases past a threshold more people are injured by additional surgery than you actually help. This is again at a population level.

Are you claiming that, at population level, the amount of false positives MRIs produce outweigh the benefits in terms of human health benefits?

Beyond a certain point all scans will produce more unnecessary procedures than beneficial ones. It's the nature of treating illnesses with uncertain presentations. Everyone has lots of odd bits and small benign tumors that are harmless. But these often look indistinguishable from the early stages of something serious. Even biopsy isn't without risk.

There's no free lunch here.

> No medical intervention is without risk

no but an mri scan and blood tests are pretty damned close

30 years of research and clinical debate have largely come to the wrong conclusion because medical doctors are, by and large, mathematically illiterate; in some countries, like the usa, they are also strongly motivated by cya, which comes into direct conflict with the patient's well-being in cases like this

MRI scans and blood tests are, but what comes next? If they find something there's always a least a follow-up biopsy, and often times what's seen on the screen would have never impacted the person's life. Yes, there are outliers, but if you do enough scans you are guaranteed to find something wrong with you that you would never have known about otherwise.

if the mri scans and blood tests show strong enough evidence of a problem that will impact your life that the risk outweighs the risk of the follow-up biopsy, you should get the biopsy

if they don't you shouldn't

so either you're better off or you're no worse off

unless you have to be dealing with a doctor in cya mode

I have some familiarity with diagnostic imaging - a lot of the time with these scans there's enough to say "yes, there is something there you should definitely get checked out" but you can't make a full assessment without a biopsy. The MRI is going to find things you need to investigate further.

It'll also reveal a bunch of stuff you didn't need to know. Do you want to know about that minor slipped disc in your back that wasn't bothering you before? The nocebo effect is very real.

if investigating it further isn't worth the risk, you don't need to investigate further

if investigating it further is worth the risk, then it's a good thing you did the scan

this superstitious ideology that ignorance is good for you has always been popular; what's surprising is that some people still hold to it despite the industrial and scientific revolutions

knowing about the minor slipped disc in your back might be important information if it starts hurting; you'll know it isn't a metastatic bone cancer. you might also be well advised to avoid taking up olympic weightlifting

> if investigating it further isn't worth the risk, you don't need to investigate further

How do you make that call? What do you, as an eager but uneducated layman, use to decide that? You now know there's a heterogeneous cyst in your abdomen that the doctor says is super unlikely to be cancerous but it's not impossible that it could be. How do you evaluate your odds? Are you opting for the biopsy? If there's a dozen of such cysts in all corners of your body? Again, the MRI is a tool to help spot a problem, it's not going to give you the diagnosis.

usually i read the wikipedia articles about the disease and the common treatments, a couple dozen papers starting from google scholar via pubmed and sci-hub, maybe a relevant medical textbook via libgen. also talk to family members and friends about the issue. you can also find papers in medical journals and the mmwr on the prevalence of possible complications from different kinds of biopsies

this takes a little longer than undergoing the biopsy, but it's enormously more pleasant, and it lets you know when doctors are ignorant or lying, which is perhaps even more important information. it's orders of magnitude easier than undergoing chemo and radiation so i don't know why 'eager but uneducated' laypeople don't always do it. maybe they're too used to living in a world without sci-hub and libgen

it's also easier when it's not the first time

it's never impossible that any part of your body could be cancerous, so 'super unlikely to be cancerous but it's not impossible that it could be' is a terrible justification for taking action

> it's never impossible that any part of your body could be cancerous, so 'super unlikely to be cancerous but it's not impossible that it could be' is a terrible justification for taking action

If you knew that you or your spouse had something that could be an issue but was unlikely to be - and it's nearly guaranteed that somewhere in your body right now you have stuff that would come back as "possibly a problem" in imaging - you'd be comfortable not taking action? I doubt that's how most people would response.

yes, i have in fact done this

i'm not here to protect most people from the consequences of their own bad decisions

and i don't want to be subjected to a medical system that's oriented toward doing that to me

> You now know there's a heterogeneous cyst in your abdomen that the doctor says is super unlikely to be cancerous but it's not impossible that it could be.

That's up to the individual to decide - blanketing people in ignorance and leaving it up to chance should not be the only option.

But you're still making decisions out of ignorance and you're biased towards worst case outcomes because it's you. People are going to make the choice out of fear and that will lead to a ton of extra procedures and those people possibly having a worse quality of life as a consequence of all the elective procedures.

I would say you're biased because it's not you. You've turned everybody into a statistic to make your decisions based on the idea that it's more important to keep your percentages in line with expectations created in a broken health care environment where reactionary medicine is preferred over preventative medicine.

How are we supposed to ever make progress here if that's your mindset? This is the medical system that fears its own incompetence so much that it would prefer to NOT have data. How silly is that? What other industry operates this way?

Can you imagine an engineer doing maintenance on a plane and saying "hey actually, we have no idea what we are doing so it's better if we don't do an inspection so we don't mess it up even more by trying to fix something that might not be a problem."

> What do you, as an eager but uneducated layman, use to decide that?

Is this a trick question? I use my doctor's expertise and advice, obviously. Just like I do for any other nontrivial medical question.

This has happened to me a number of times throughout my life. I have some problem, my doctor requests a scan (or x-ray or ultrasound or...), it turns up some anomalies, and my doctor has to evaluate them to decide if they are relevant to anything. For many of them the answer is, no, they're not. Now I know that I have a cyst here and some slightly misaligned vertebræ there and ... that's fine. They weren't causing the problem I had and they're probably not causing any other problems either.

The argument that we must not look at ourselves because if we do, we'll be compelled by some mysterious force to take dangerous interventions is silly and just seems like an excuse for medical incompetence or laziness.

> strong enough evidence

This is actually a bit subtle. For example I've heard more than one radiologist claim that if you look carefully, literally anyone's body scan will contain details that could justify work up.

The fact is, if you do this broadly in a population and you work up everything you see that would justify it, some small number of people will die or suffer serious problems from complications of follow up.

This is a fundamental problem in screening technologies (of any kind) with imperfect sensitivity. You can't get away from this, so you try and be clear that the net benefit is (much) higher.

I'm not even against the idea of baseline scans, but the idea that this part is simple is naive.

that's just because the radiologists' standard for 'could justify work up' are too low

Or it's because the problem is genuinely hard. See above ('imperfect sensitivity')... the same problem exists in e.g. blood tests.

I'm no radiologist but I've seen enough of this done that I'd lean that way. Especially if you consider realistic bounds on time and additional information.

yes, of course the problem is genuinely hard, but having less information about your body makes it harder, not easier

In the abstract more information is always better. In practical healthcare with real constraints, sometimes it makes things worse globally for reason above.

Think about say a diagnostic blood test with a poor sensitivity (but better than chance) on a very rare disease. As an oversimplification: for a bound take your population * workup rate * fatality rate due to workup and you have a baseline needless death due to the screening (NB there is basically always a fatality rate). The number you are comparing to is lives saved due to earlier detection.

Better than chance means it's strictly more information always. But as policy, you shouldn't do it unless you can be extremely confident that lives saves is much larger than lives lost due to the process.

This is a bit counter intuitive, but it is a fundamental issue in a lot of healthcare policy. The real calculations are more complex but in the end it is unavoidably actuarial in nature.

i explained this in https://news.ycombinator.com/item?id=37267690 except that i suggested using a better measure than crude fatality rate, such as daly

this is not a global optimization problem; it's a local optimization problem. one person is deciding whether or not to spend their own money on an mri, or a blood test or whatever, on themself. they are the ones in the best position to do that, and for very-low-risk tests like mris it is very rare that the optimal decision is for a rich person not to get them

another nuance is that you don't have to be extremely confident that the mri or biopsy or whatever will increase your expected daly or qaly or whatever; in the cases where you're not extremely confident, it's because it doesn't matter much either way, because you're weighing two extremely small risks against each other. if you make the wrong decision and lose (in expectation, risk-adjusted) a hundredth of a qaly to a blood test or an extremely unlikely cancer, it wasn't a terrible decision, just a small mistake

(even if that small mistake kills you in three months)

the cases where the decision matters, it's easy to get enough information to be extremely confident

Well I did say the actual calculations would be (and are) more complicated. And as I said, short of availability constraints, the MRI itself isn't an issue, it's the follow up.

Ignoring the fact that in the real world due to constraints this isn't really a local optimization, if we assume infinite availability you still have a problem that low sensitivity tests lead to excess unnecessary work up risks. Your assertion that it is always easy to get good information leading to correct choices here is just incorrect and probably naive.

The trade offs you are talking about are very well known and have been studied by professionals for decades; none of them will claim it easy - for good reasons.

Note I'm not appealing to authority here. Many people can educate themselves to the degree that in a narrow area of their own healthcare they are as or more knowledgable than the average professional. However, this is not at all easy, and will require a ton of work. The cost benefit for a lot of screening etc. just isn't there; for a particularly symptomatic disease it may well be.

What are you talking about? There’s a huge body of public health literature about the dangers of for example doing too many colonoscopies on people below a certain age.

This isn’t a controversial claim at all among people familiar with stats on outcomes across large populations. It’s tricky to get right and your glib dismissal comes across as arrogant and ignorant.

i'm familiar with the literature and the incorrect consensus in the field

The idea that you believe that the entire public health establishment on 3 continents is wrong about the risk of preventative scans is ridiculous. Surely you. have some evidence to support your heterodox position. It can't be that "doctors are innumerate," because I assure you that many aren't, and there are statisticians that work in public health.

How do you propose the medical field determines which small tumors will become life threatening in general? Can you beat the best research hospitals in this task? Even an average oncologist?

i already explained my reasoning above; it's obviously correct, and it doesn't depend on my knowledge about which small tumors will become life-threatening, so it sounds like you didn't understand it

you can calculate the cost and expected morbidity due to any particular candidate test (mri, blood panel, biopsy, ct, pet, whatever) and compare it against the reduced expected morbidity due to the probably-not-present disease, using whatever metric of morbidity you like best (ypll, daly, expected change in your malpractice insurance premiums, whatever), using whatever risk tolerance you like best

for mris and blood panels obviously this will favor doing the test in almost all cases, unless traveling to the hospital is a significant hardship or the money is enough to be important to you

for other tests it may not, and watchful waiting is often a better choice than doing further tests

with a little work, you can always beat the best research hospitals at this task because you care about different things than they do. this is not the same thing as beating them at guessing whether your prostate is going to kill you before something else does

I totally agree with you. This idea that “there is no problem if we don’t look for one” is absurd. We have plenty of statistical techniques to prevent unnecessary surgeries for every anomaly in an MRI scan; let’s utilize those instead of keeping medicine in the dark ages.

> No medical intervention is without risk

Risk will dramatically decrease with advances.

> That you think this is a trivial problem shows that you’re largely ignorant of some 30 years of research and clinical debate.

Biologists are still doing punch card-level stuff. If we started cloning monoclonal, antigen-free bodies in massive body farms and performing head transplants, that'd solve almost every non-cranial disease. I could see that becoming a preventative solution to aging, too. Renewed cardiovascular health, clean lungs, thymus, etc.

"This isn't a problem because of science fiction" isn't a terribly realistic answer to the concerns being raised.

> "This isn't a problem because of science fiction" isn't a terribly realistic answer to the concerns being raised.

This isn't a step function difficult thing to do [1-3]. Nobody is doing it because it's "icky".

[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5116034/

[2] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6511668/

[3] https://www.nature.com/articles/ncomms15112

But it is. Giant body farms of clones raised without the benefit of a womb that we then decapitate and reattach our heads to is a non-trivial enterprise.

Here we go: https://news.ycombinator.com/item?id=37308675

Again, for the people in the back: There's no such thing as minor surgery.

Every surgical procedure carries a real risk, which amortizes over the population of billions of people into thousands to millions of complications and deaths. That's the cost, and an intervention has to be good enough to justify that cost. Vaccines are good enough for that. Some types of cancer screening is good enough for that. Popping yourself in a full-body scan for a fishing expedition? Almost certainly not.

The result of this product is that a few more very rich people will die a bit sooner than they would have otherwise. Whether or not you think that's an upside is up to you and your personal politics.

> That you think this is a trivial problem shows that you’re largely ignorant of some 30 years of research and clinical debate.

Am I not allowed to criticize decades of research on astrology because I haven’t paid any attention to it?

If 30 years of research and clinical debate have led to this situation where you get 5 different opinions from 5 different doctors, then maybe it’s time to kick out the incumbents and replace them with people who adhere a little better to the scientific method. I would rather trust my medical care to a group of particle physicists who had one year to study medicine than the 30 year experts who can’t reproduce any of each others’ work and seem ignorant of any advances in statistics or inference over the same period of time.

> 5 different opinions from 5 different doctors

No, you get the same opinion from all five doctors. That the scan is unnecessary.

I think that long term, preventative scans do make sense. With that preface, I don't think we should just argue there are no problems, and it's all easily solvable.

"Too many false positives leading to wasted time and costs? Raise the costs for the supplemental scans and improve detection precision and recall."

You can't improve precision/signal on an inherently weak and noisy signal. There are many things showing up on MRIs that currently require a biopsy. You need alternative DX routes to disambiguate, MRI alone can't do that. As for raising the costs - who pays? Medical costs are already completely bonkers.

"That’s a sign to improve medical research in that area. "

We know we need to improve search. We don't need to cause negative outcomes to reinforce that signal.

"Too much demand for MRI machines pulling away time from patients that need to use them? Build more MRI machines. "

Again, who pays?

"we should just plow ahead and address new problems as they arise."

There is the problem of opportunity cost. Is that the best possible way to spend sparse resources? The medical community currently mostly thinks that it isn't.

Targeting wealthy people first does solve the "who pays" problem. It might make the "too few researchers" issue worse. It will cause a lot of unnecessary biopsies (because often that's the only alternative dx route) - but you could certainly lean into informed consent there. You do run the risk of poisoning MRIs as a diagnostic tool, because media will report any inconvenience to sufficiently wealthy people as a major problem.

All of this might be solvable, but "let's just press ahead and see" is a very bad approach here.

If the "things showing up on MRIs that currently require a biopsy" don't actually pass cost-benefit analysis on their own, that 'requirement' can & should be adjusted down. Or, new diagnostics with less cost/risk developed. That today's "MRI alone can't do it" isn't a hint what novel techniques might be able to do – once demand created.

A sufficiently-flexible & high-res set of downstream decisions can always choose to ignore data that isn't really helpful. Enough scans & experience might discover the new patterns.

Our ossified medical system, with those "already completely bonkers" costs, has a hard time adjusting, & sticks with habits & tech that's stale. Thus things the legacy system isn't ready to handle efficiently – novel diagnostics with floods of new data – get blamed.

But the only way out is through. Letting those with discretionary funds try speculative things is a way to improve faster than waiting for a burnt-out & bureaucratically-deformed "medical community" to rustle up the will/resources.

'If the "things showing up on MRIs that currently require a biopsy" don't actually pass cost-benefit analysis on their own,'

So if we can't tell if it's malignant or not, you're saying we should start cancer treatment just in case? It's not a question of "tech that's stale" or "an ossified system" - it's a question of not being able to tell all malignant and benign cells apart until you look at them under a microscope.

The last thing that space needs is a bunch of uninformed tech folks going "I don't know, let the market solve it, tech is great". I really recommend getting at least a basic education on the subject.

If it looks bad enough that you'd risk starting cancer treatment right away (!) with no further confirmation, it's definitionally something they'll have been glad to detect early.

But for most things, the advice would be: watch & maybe run some other tests, if cheap & low-risk, if and only if the preponderance of the data implies it's a net-gain.

If the MRIs become cheap enough, high-resolution enough, and the ML/stats-assisted classification becomes good enough – possible with more imaging/progression data – the most common thing might be: reimage in a month or two, compare results. Do more if and only if that calcs out for maximum welfare.

These are ultimately probabilistic equations, not things you can hand-wave off based on old anecdotes and mere hunches about what's net worthwhile, and what's not. You have to run the tests, run the numbers, and reach a data-driven answer.

On the other hand, people who don't trust the numbers, or their own ability to dispassionately follow the numbers if more data arrives, should absolutely be free to remain in blissful non-knowing.

Many people can – and many more with proper coachin should be able to – use new info only if it improves their welfare. They don't have to follow automated scripts from a less-informed era.

> There are many things showing up on MRIs that currently require a biopsy

they don't require a biopsy; you're still free to not get the biopsy after the mri

sometimes getting the biopsy is actually worse for you, on average, than the small chance that you'll have a very rare disease. probably in those cases you shouldn't get the biopsy, which means the mri didn't actually help you, but it didn't hurt you much either; you're US$650 poorer and worried, that's all

in other cases getting the biopsy is better for you, in which case the mri did help you

> Is that the best possible way to spend [scarce] resources?

fortunately, we have private property, so we already have a system for allocating scarce resources: their owner decides how to spend them. this turns out to work much better than the alternatives that have been tried, which routinely result in famines and genocides. so we don't need to come to a global community-wide consensus on that question; instead we can consider each case as it comes up

> they don't require a biopsy; you're still free to not get the biopsy after the mri

Have you ever talked through a scan with an oncologist? Or are you just reasoning in the abstract because you feel you can first-principle it?

Also, seriously, you want to turn this into a communism debate? fwiw, capitalism has already decided. The vast majority of resources goes to anything but full-body scans, because actual experts made the call what's more effective. Repeatedly. On many cases.

But you are right, you are of course free to make bad decisions with your money, and there will always be somebody happy to separate fools from their money. Capitalism does work.

> Have you ever talked through a scan with an oncologist?

yes, earlier this year

also, fuck you

people like you are why i always regret succumbing to the temptation to participate here

I was thinking of this when a friend passed away from pancreatic cancer a couple of weeks ago. One of the reasons pancreatic cancer is so deadly is that it is deeply internal and symptoms of it are not present. This is true with all of the deadliest cancers. We already recommend colonoscopy to older adults which is a fairly invasive procedure to screen for colon cancer with no other history. If we could scan people with low or no ionizing radiation that would be a great tool.

> We already recommend colonoscopy

You're behind the times. Now it's Cologuard. Not invasive although it is gross.

It's the 16% false positive rate, which they helpfully do not provide a dramatic anecdote for, that should give you pause.

A full body scan is a bit like p-hacking. If you have a hypothesis of what might be wrong and perform a specific test to see if some value is within normal range, there’s relatively little chance of a false positive.

If you test everything at the same time, some values are bound to be outside of the normal range, just because you’re measuring so many things at once. Doing any kind of procedure to further investigate or treat the symptoms has a risk associated with it.

In the end, doing a full body scan, with current medical technology (both in terms of the test equipment as well as our means to treat anything that comes up as a result of the test) is a risky thing to do. It sounds nice in theory, but unfortunately we do not live in the utopia you speak of.

It isn't just wasted time and costs, it's a risk beyond that.

If you scan any middle aged adult you will find abnormalities in their scan. In almost EVERY case those abnormalities will be benign, but once you've found them depending on risk factors a biopsy may be required. Almost all of those biopsies will also be negative, the tiny calcification in your lung or breast isn't a tumor. The swollen lymph node isn't a sign of lymphoma. A spot on a scan is FAR more likely to be down to habitus and scanning issues than some underlying medical issue.

But if you scan everyone, biopsy everyone, you're going to end up killing some of them too in addition to the waste of limited resources.

Wouldn’t that be largely solved by doing two scans a few months (or so?) apart?

You’d see if any of those potential issues grew, or disappeared.

Why not simply do nothing?

If you get the scan and do nothing, or you choose not to scan and do nothing, and it turns out to be something fatal, the outcome is the same: you die.

The difference, is if you do the scan, you might have an epiphany about your own mortality, and live your life truly savoring every moment, instead of being blindsided and quickly dying someday.

If you want preventive care. Exercise one more hour a week that what you are already doing. This actually reduces all-cause mortality. Anything else is snake oil. Including full body scans.

Well, yes, but all of that would need to be optimized for expected patient quality-of-life. If instead it'd be optimized to reduce risks as much as possible, we'd ultimately be spending all of our lives in medical facilities.

In non-utopian reality it'd be even worse, as the system would be optimized for shareholder value instead. Shareholders profit by scaring patients into believing they need to buy more medical interventions.

Maybe show a little sympathy? It only looks easy because it's not our problem to solve.

> Raise the price

Although raising a price does lower demand, a different price for a test doesn't make decision-making any easier. Predicting future costs and benefits (based on test results, not the test cost) is the hard part.

> a sign to improve medical research in that area

Decisions about whether to do a procedure assume current technology, not imagined future technology.

For me the biggest problem is the anxiety it produces. Both false positives as well as the fact itself of waiting for a result.

At least in the US consumers really don't have much experience / have a lot of trouble managing medical costs.... even getting the cost.

I think raising the price would just mean raised prices and effectively where we are now.

In the spirit of parent post (and for a whole lot of reasons by itself) then change the system.

I think the spirit of the parent post put more thought into it than "change the system".

are you suggesting controlling the price as a centrally-planned public healthy policy, or that costs drive medical prices? in the US free market the price goes to administrative overhead and profit margins

Reply on news.ycombinator.com