I have a hard time with reporting like this that doesn't quantify cost of biopsy on false-positives as well as cost of complications from examining false-positives.
Source: worked in clinical research for 2 years.
However, this is also one of the places where it's quite disappointing to see human squishiness in action. When we opt to not receive data because we know there'll likely be non-malignant anomalies, we are also depriving ourselves of the raw data that allows us to discern non-malignant from malignant, data which would help very much in deciding on the cases we do receive.
A stream of known false positives is worth its weight in gold in any detection system, it's just that this specific system happens to contain squishy feely human doctors that are susceptible to short-term human 'Please don't spend more money than necessary, we're already in the red for this quarter' kind of pressures from their superiors.
Had we enforced a full-body MRI for every patient, we would quickly amass enough imaging data to know which anomalies are malignant and which are not to a very high degree, which would counteract the temporary cost increase in chasing false positives. The fact that we don't want to eat the short term initial cost means some of those things that could be uncovered by the scan, things that could save a person from a malignant deviation, is left unseen, untreated, to its natural end.
In-depth investigations have massive opportunity costs for someone reimbursed on piece-work basis, takes up way too much time.
Better technology is unlikely to be cheaper, unless it is so new and novel that it can get its own unique billing codes, and if it becomes ubiquitous so it cheaper and more available than existing machines which are not yet depreciated.
Just typing that out depresses me.
There are different kinds of MRIs, but the actual direct costs for running an MRI without preparatory procedures is around $100 (preparatory dye injection may be another $50 cost). Marginal costs are lower still.
See also here:
https://www.npr.org/templates/story/story.php?storyId=120545...
https://www.pbs.org/wgbh/pages/frontline/sickaroundtheworld/...
<http://www.hiperdia.ro/servicii-medicale/>
It's a modern 1.5T MRI.
Determining whether something is (non) malignant isn't trivial. It often requires a more invasive test (e.g., biopsy, exploratory surgery). We try to avoid those procedures not just because we're cheap or afraid, but because they have a real risk of harming us 'squishy humans'.
For example, detecting prostate cancer a few months earlier is often not a huge win on its own. It is even less of a success if doing so requires that four people have allergic reactions to the anesthetic and two others acquire some kind of infection at the biopsy site.
It's certainly possible that more data would eventually let us avoid most invasive followups, that's nowhere near certain. Even if it were, the ethical calculus is still pretty tricky.
I think you're correct in the current context. However, this strikes me as a sign of some kind of paradigm shift waiting to happen, enabled by the reduced cost of data gathering, data processing, and improvements in AI.
There was a time when many things which are mass produced now were once handcrafted luxury items. Perhaps, as technology improves, it will one day be cost effective to take a whole body scan into account. I can imagine a day when an AI would digest the information from a full body scan, then autonomously plan robot keyhole surgery to take many of the biopsies.
It's certainly possible that more data would eventually let us avoid most invasive followups, that's nowhere near certain.
I think it's likely to happen in decades, if our civilization doesn't fall soon.
Perhaps, as technology improves, it
will one day be cost effective to
take a whole body scan into account.
Both options have risks and both options can kill you. Namely:
A) Doing nothing may allow a problem to fester until it either kills you or becomes detectable with standard treatments. However, you have no exposure whatsoever to side effects, infections, or medical errors for the things that are missed.
B) Doing whole-body MRI 'fishing expeditions' decreases the risk of missing something, but increases the likelihood of possible false positives. Each thing flagged by the scan now needs to be ruled out (otherwise, why bother?) and this usually requires invasive procedures with their own risks.
With our current imaging and understanding of biology, Option A may be less likely to kill people. Opinion doesn't even have to enter into this--you can literally run the numbers.
It's worth noting that we limit testing FOR EVERYTHING. When you get a blood test, you could be screened for everything from ADH to Zika. You aren't though, because you're almost certainly a perfectly healthy outlier on a few tests; that's just how statistics work (someone has to be 3 sigma above the mean). Instead, tests are picked and interpreted based on your symptoms and circumstances.
I tend to think it is the former but can anyone clarify?
The normative values aren't handed down from God, or some definite evidence of health. Instead, they were the typical range (mean ± 3SD, percentiles, etc) from a sample of healthy people, possibly matched for demographics (e.g., sex). Suppose we set each range so that it contains 99/100 healthy people. It would only take about 70 independent tests before you have a 50:50 chance of being outside that range on one of them--and that's assuming you're an exact match for the reference population.
There's also error in the measurements themselves. Some of it is errors in the actual procedure or preparation: cross-contamination between samples, dust blows into the well, etc. However, many of the tests are stochastic too. Some measure binding between stuff in your blood (e.g., antibodies) and a "probe" that's designed to detect them. This usually works, but non-specific binding can cause false positives (the probe binds something that is similar to, but not quite the same as, the target). Other conditions can cause false negatives. The "front line" screenings are usually meant to be cheap, fast, and biased towards false positives.
You might be interested in a fairly common statistics question about screening. Suppose you had a test that is 99% reliable: 99% of sick people test positive; 99% of healthy people test negative. The disease itself is somewhat rare--only 1% of people have it. If you test positive, what are the odds you're actually sick? The answer[0] will explain why we don't test or scan people for tons of rare diseases, even with very reliable (and free) tests.
[0] ʎʇɟᴉɟ-ʎʇɟᴉɟ
Then the doctor should recommend not attempting that type of detection.
I find it quite hard to believe that, in an ideal world (!), less data would actually be better. The doctor should still have the ability (at least in theory) to decide what information is worth investigating further, and what should be ignored.
The issue is that this sort of information isn't free. The full body scan isn't free. Following up on it, which is the whole point, 'costs' increased risks of adverse outcomes (from biopsies), it 'costs' emotional anxiety, and it costs even more actual money.
The argument is that right now, these whole body scans aren't worth the cost. You're more likely to find a few things you should ignore (but will probably worry over) and a few things you should follow up on (but will be benign). You're less likely to catch something deadly and just in the nick of time.
Not necessarily, waiting can in many cases tell you whether something is malignant. If we had, say, a century of monthly full-body MRIs of a million persons, together with their history and cause of death, the technology to align those scans across time, and the technology to analyze such a data set, a retrospective cohort study (https://en.wikipedia.org/wiki/Retrospective_cohort_study) probably could uncover quite some interesting and useful information.
If, at every step, you pick the option with the highest expected QALYs (https://en.wikipedia.org/wiki/Quality-adjusted_life_year), and you manage to make doing the MRI very efficient, I think doing that experiment even might pass an ethics committee (at the cost of making the analysis harder)
(might because one could argue that the patient would be better of if the money spent doing those MRIs were spent on something else)
Unfortunately, doing such an experiment isn’t practical (¿yet?).
First, the cost would be insane. A century of monthly MRIs is 1200 scans per person, or 1.2B scans to finish your hypothetical dataset. We pay about $US 550/hr for scanner time, using a research scanner that's subsidized (i.e., we're just covering costs, not making a profit). The article doesn't say how long the scan is. You can burn as much scanner time as you want chasing resolution/quality, but an hour seems reasonable. That comes out to $660B, which is...a lot, and we haven't even paid staff yet!
Second, we have sorta tried this already. There's a massive neuroscience project to scan tons of brains called the Human Connectome Project. They have 1200 subjects, some scanned multiple times, behavioral measures, health outcomes, the works....
People have certainly found stuff in the data (I'm using some of it right now), but it hasn't lead to wild breakthroughs. There's a ongoing debate about whether this money would have been better spent on hypothesis-driven research instead.
“Because it’s too expensive,” seems like a great opportunity for some clever startup to figure out a way to make it less expensive. Computers used to take up entire rooms. Flying across the country used to be insanely expensive. Cell phones used to cost a ton of money per minute. Reducing the cost of scanning, or developing entirely new scanning tech isn’t science fiction, it’s the future.
https://www.pbs.org/wgbh/pages/frontline/sickaroundtheworld/...
The machines themselves are not magically cheaper in Japan; they're just being paid for through some other route. If the scanner were somehow free (gov't grant?), 98$ seems pretty reasonable for the labor.
They’re also “free” (at the point of service) in the UK, but the scanner and helium are not a generous gift from the fey folk, nor does the Queen volunteer as a tech.
These things cost money. Any study you do is going to have to cover expenses—-including their share of a temperamental, multimillion dollar machine (or find a way to dip into the same accounts that cover its clinical use).
The $550/hr you are paying may be because you are using a particularly fancy scanner, or because utilisation rate is low, or because the amount is loaded with overheads, or w/e.
That’s in the ballpark for a 1.5T, but those are fairly old. About $1M per Tesla used to be a decent rule of thumb, but it’s come down a little at the low end. Still, I would be amazed to see a 3T for anything below $2M.
As for the $550/hr, it’s probably true that research scanners have lower utilization and higher costs to support all the weird stuff researchers want to do. An outpatient clinic specializing in knees can run much leaner. That said, that rate seems to be pretty standard across universities and I maintain that it's a very reasonable estimate of the cost. For example:
* Hopkins $668/hr (3T) or $538 (1.5T) during prime time; cheaper nights and weekend http://www.mri.jhu.edu/div_mri_res/ServCentPolicyFY17.pdf
* Yale: $539/hr (3T) https://medicine.yale.edu/mrrc/users/charges.aspx
* Harvard/MGH: $640/hr https://www.nmr.mgh.harvard.edu/core
* WUSTL: $710/hr (3T) https://www.mir.wustl.edu/Portals/0/Documents/Uploads/CCIR/F...
* McGill: $500/hr (3T) or $500-700/hr (7T) https://mcgill.ca/bic/files/bic/bic-rates-03052018.pdf
Most of these do not include F&A. It’s already coming out of the grants and external users pay a "Dean's Tax" on top of that to cover the missing overhead (which can often double the price).
(The NIH does have more money, but also funds trials, vaccines, and other stuff that we probably don't want to cancel for a century).
Out of curiosity, what determines the speed of the MRI scan? Are massive improvements theoretically possible?
Two ideas for time reduction I haven't seen discussed but perhaps might also help:
1. Don't scan at the same resolution across the whole body each time. Instead, focus on anomalous places that you wanted to monitor from the first high-res scan, or places that look anomalous in low res in the latest scan. Then dial up the resolution in those areas.
2. If better imaging algorithms existed that could account for very slight movements of the body (ala the heartbeat one above), perhaps prep time could be reduced by changing the physical layout of the scanner itself. The whole process of lying down and getting your head or whatever mounted in their plastic frame, then lifting the gurney and slowly sliding it into the machine.. it's all very slow.
Instead of laying down, what if you could just walk in and be upright and get scanned relatively quickly - basically a slightly longer but similar experience to a chest X-ray. There's upright MRIs right now but they're not very high res (0.7tesla instead of 1-3), very few exist, and I'm sure they still take some time to complete scans and have lots of error correction extra scans to correct for patient movement.
However, even with less time in scanner, there's a lot of fixed time cost of scheduling and patient prep (remove all metal things, wear this gown and these ear plugs, please don't freak out its really claustrophobic patient messaging), as well as still needing technicians.
I'm just a patient who's gotten some MRIs but it definitely feels like there's ways to cut the time down significantly.
[1] http://news.mit.edu/2011/better-mri-algorithm-1101
[2] https://phys.org/news/2017-10-technology-mri-scan.html
[3] http://news.mit.edu/2018/faster-analysis-of-medical-images-0...
Open, upright scanners do exist, but they're lousy. The goal of the (big) magnet is to produce an incredibly strong, uniform magnetic field. Due to physics, this is much, much easier to do with a torus-shaped magnet than a 'U'shaped one. Even so, there's one point ('isocenter') where the magnetic field is maximally flat and the quality will be best. The gurney moves to point the region of interest (e.g., your head) right to the isocenter. That's why the tech often uses a little light or laser to find your position, rather than just asking you to scooch. Motion is also, as you alluded, a huge problem.
I hate to be discouraging, but I am excited to see people actually thinking about MRI on HN!.
MRI is all about the protons. Under normal conditions, the protons in your body are all spinning ('precessing') around their own axis, but they're disorganized: each proton's axis is pointing in a different direction. They're like little wobbly tops drifting through space.
In an MRI scanner, the strong static magnetic field (B_0) forces the protons into alignment, so that their rotational axes are now lined up with the field's north-south axis. The field needs to be very strong for this to work, which is why MRI systems usually have expensive superconducting magnets.
Now that we've created a nice organized system, we're going to destroy it. A quick radio frequency burst energizes the protons and 'knocks them over' so they're no longer aligned with the field. Once the pulse ends, they 'relax' and realign themselves with the magnetic field, releasing some of that energy as they do so.
Sensitive detectors around the subjects' head detect those emissions and use it to determine how long it took for protons to realign themselves with the different components of the magnetic field. T1 is the time (or formally, the time constant) needed for relaxation parallel to the static field; T2 is the time needed for protons to relax to the transverse component. The T1 relaxation time for fluids is on the order of seconds, while fatty issue is more like 50-150ms. In the brain, grey matter has a relaxation time of 1.3 sec, but the fat-coated white matter relaxes much faster (~0.8), which makes T1 images very useful for examining brain anatomy.
Hopefully, this little crash course has revealed one of the bottlenecks in MRI: the actual signal being measured is slow.
Of course, I've massively oversimplified things and didn't explain at all how we localize these responses. The proton's precession frequency depends on the magnetic field, so by changing the static field slightly (across space), we can measure T1 at different locations, and sometimes even overlap measurements. You can't switch the field too fast though, or you'll start to induce currents in the subjects' nerves, which hurts. This is actually the principle behind a brain stimulation technique called transcranial magnetic stimulation.
There is still tons of room for improvement. Stronger fields lower the relaxation time, so the scans are faster (and the relaxation time estimates are better). Improvements in the RF coils help a lot too: the signal being measured is very faint and there's a lot of self-cancellation. On the software side of things, a lot of effort has already gone into designing clever pulse sequences—and the sophisticated signal processing needed to interpret their results.
Things will obviously continue to get better; I was just looking at some data from ten years ago and it looks awful compared to more recent scans.
That said, the "just use machine learning!!!" tone in some of the comments is kinda frustrating. Most of the people in this field aren't dummies--if it were as easy as downloading PyTorch, someone would have done it already. It turns out that the biology and physics are both stupendously complicated (and fascinating too).
> Had we enforced a full-body MRI for every patient, we would quickly amass enough imaging data to know which anomalies are malignant and which are not to a very high degree
How, pray tell, are we actually determining which are malignant and which are not?? I’ll give you a hint: the charitable explanation is you are completely ignorant, the less charitable one is that you consider Mengele and Ishii modern medicine heroes. And that is even if the determination by imaging alone is even possible.
Once again I have to point out on HN... not everyone in the medical world is an idiot incapable of analytical thinking, it is just that a lot of us care about ethics. Our patients, even in research, are more than just numbers.
When people on this forum speculate about things like this, I assume they are thinking of some future ML application. Of course you're right though, that is not doable today. The point of the OP is that if we collected more data perhaps it would be possible.
> the less charitable one is that you consider Mengele and Ishii modern medicine heroes.
The ad hominem attacks against the OP are completely unnecessary.
> I am actually an MD.
The start of your post feels like "appeal to authority."
Sounds like a declaration of fact to me
Should I have started my post "I am actually a computer scientist." No, right? That would be an appeal to authority that doesn't prove I know anything about what I'm talking about. Now, if the person had started with "I am a radiologist who has looked at empirical data about full body scans versus more selective scans." that would be a more useful and valid statement.
Also the vast majority of practicing physicians are expected to have detailed expertise to understand the indications, application, and interpretation of imaging within their speciality. Most imaging orders are issued by medical or surgical staff, not radiologists.
Looking through your post history you often jump to "appeal to authority" combined with name calling and it's a pattern that reeks of arrogance to me.
For example one of your posts starts out along the lines of, I'm not a 747 pilot, but I am a pilot... in a post about something specific to the 747 and you say the other guy was "talking out their ass." Again, not super relevant appeal to authority and some mean-spirited name calling.
> Not a 747 pilot (not sure why that matters here)... but a pilot. The flight positions are equal. The plane is designed to be flown the same from either flight position and this is routinely done. The guy saying otherwise is talking out their ass.
https://news.ycombinator.com/item?id=19675391
Another of your posts both claims you were trained as engineer and manages to call someone "full of crap" in the first two sentences.
> I personally think the gp is full of crap. I’m trained as an engineer and worked in industry a few years.
https://news.ycombinator.com/item?id=19674031
You just seem to love to tell everyone you're super special and smarter than them because of your accomplishments as a means of arguing and then put them down. It gives you instant validity to some of course, but I think it's just lazy. Your arguments should stand on their own merit. Not request to authority or anecdotes from your kind-of-related experience.
In the case of these detections, the ground truth can be very subjective and is certainly not a simple explanation. Is the ground truth read by a radiologist? A radiologist with fellowship training? Multiple radiologists? Biopsied? Followed for 5 years to see if the patient actually died of their tumor?
The complexity is a lot more than the underlying algorithms and computer science (though they are important!).
A biopsy? Whichever method you would normally use to find this out?
You could even ignore all the information and just train your algorithm on whether the patient died in the next 5 years (after the scan/s).
Obviously you need data before you can determine anything automatically. But I don’t think everyone would be averse to this even if you do tell them it might be more likely to kill them due to excessive testing.
Had we enforced a full-body MRI for every patient, we would quickly amass enough imaging data to know which anomalies are malignant and which are not to a very high degree
Wouldn't it be more useful as a baseline? Scan once and store it, then scan every few years and compare. If any of the "anomalies" has grown or new ones have appeared between scans, then they might be worth a second look?
In the case of blood chemistry that is caused by fluctuations due to time of day, time of month, season, what you recently ate and so one.
In the case of MRI the problem is that internal organs shift to the point that it is impossible to automatically compare. And a experienced doctor comparing the scan millimeter for millimeter takes hours and would be very expensive. Even just tracking the volume of a known tumor in the brain (that shifts very little) to track the effect of radiation is tedious, time consuming and error prone.
I didn't even know that is a thing, I just sorta assumed they stayed put.
Seems like a good area to apply machine learning if we had enough data?
There are whole studies done to try and quantify this, and it requires sending "human phantoms"- test subjects- around to different scanners, which requires you to physically take one subject around to several scanners operated by different people and stick them in there for a while. It's really involved.
In the same way astronomers can make new discoveries by digitizing photographic plates from a century ago.
Scanners cost a lot: ~4M for a 3T (the current 'standard') and much more for higher field strengths. We just got a 7T and, with all the trimmings (scanner, console, coils, the shielded room to hold it all, training, etc), it's in the $10-15M range.
As a result, time on these things is already in pretty high demand, even at facilities that are used only for research. The projects could be reprioritized, of course, but most of the on-going research is much less speculative than "someday, this can be training data for something."
It's not even clear to me that old MRI data will be particularly worthwhile. Photographic plates were, until recently, about as good as digital sensors. New MRIs are much, much better than old ones, and the tech will only improve more. Longitudinal data is useful in both fields, but for astronomy, there's no way to rewind time, so old plates are the only possible source for events from the past. We get new old people, however, every day!
Where is that? Might be worth the trip in some cases.
I know Yale also has a 7T, as does the NIH, Stanford, and MGH (I think--it's somewhere in Boston). The University of Minnesota has THREE which is crazy.
There are hospitals that have switched brands of scanners over protocol times.
I got up at 4am to run scans in grad school because the even the research scanners were booked solid, never mind the clinical ones. Scanners are really expensive: the machine costs a lot, it needs fairly particular conditions (shielded room, cryogens), and trained operators and analysts.
There's maybe a market for fleecing the extremely wealthy into undergoing MRI fishing expeditions, but you'd have to charge a fortune if you're going to buy scanners to do it.
$ many of these MRIs require contrast material.
(2) Malpractice for a missed cancer in most states has an enormous lookback period and an automatic “doctor loses.” It doesn’t matter if the doc made a statistically appropriate call. So docs follow-up on anything but the most unambiguously benign lesions.
but factually incorrect call is apparently OK, because they are treating "median person", and not the actual patient.
This has to do with money - and nothing else.
And of course we make statistical calls! When you’re deciding -prospectively- whether a course of testing or treatment is in a patient’s best interest, you have to look at the stats on likely benefits and harms of various courses of action. I’d love to know what my patients have before hand, so I could avoid dealing with probabilities, but I haven’t been blessed with that particular power. I’d love for the data to exist to tell me how to interpret a result in my precise patient rather than a larger population she belongs to, but “solitary pulmonary nodules in 33-year old men who smoked twice in college and live half a mile from a freeway with two episodes of bronchitis in their teens” is a study that hasn’t been conducted yet. One day perhaps it will be, and if so, I’ll be grateful for it. But right now that data doesn’t exist, so we use what we have, and use our judgment to tailor it to the patient in front of us - imperfectly.
I recommend reading “overdiagnosed” by Gilbert Welch.
> make statistical calls!
Really? Your fuzzy feels are not statistics.
Do you even at least try to know how large is the type 2 error in your guesswork? Is it better than flipping a coin?
Mostly it is just how the medical industry works. A previous co-worker of mine had a wife that ran a small family practice. He was always griping about the money they spent on insurance for the place ($20k a year or quarterly, can't remember). I asked why it was so much, and case in point, he told of a story just in the prior year of a woman that came in for a knot on her breast. She was informed to come back in a month as sometimes hormones can cause them to come and go. Well they never saw her again but a year later they got a malpractice lawsuit from her since she now had breast cancer...
The same idea applies to most tests in medicine. That is why tests are mostly done conditionally.
Even if you (or anyone, for that matter) might be comfortable with that, I think you'd find that a lot of patients wouldn't be.
Why can't a doctor see an anomaly and decide not to chase it down based on the same reasoning that they use to not scan for it?
I think Doctors (I am one) definitely err on the side of overtreating in those uncertain situations. Just like politicians, it feels more caring to act than to not act.
If you MRI 1000 people and leave the vague results in the hands of a bunch of physicians, they will tend to be interventional. This will lead to a spate of overtreatment.
Ways around this would be to consent patients for watchful waiting in the case of uncertain lesions. Even so, telling people they have weird looking lumps that you're not going to treat will cause serious anxiety.
There are many ways to deal with incidentalomas. If full body scans were more widespread, maybe doctors would have confidence to deal with them properly. For example, watching them over time.
In my mind this is the key... a way to frequently and conveniently do a full body scan so that anomalies can be detected via the delta. Scanning and detection is a problem that likely could be tackled via machine learning.
I'm hoping it's not too long before a method is developed where a low res MRI like scan is feasible at home (e.g. using radio or ultrasound).
Done on a regular basis the system could detect anomalies like the appearance or enlargement of a growth inside the body.
Small point aside: an ultrasound scan -- even if it was every feasible at home -- would not be an MRI. MRI means you use the spin echo and relaxation of nucleii (mostly protons for medical MRI). It is not a generic word for "please give me a 3d image". Words have meanings, especially when they are technical jargon, and it is important to not dilute that if we want to be able to communicate efficiently and effectively.
Also: technology that is more mature than software tends to improve at a more measured pace, as all but the hardest breakthroughs have been obtained already. It is therefore unlikely that you will be able to replace a 10 million dollar machine by things in your home any time soon.
Still not something you’d have in your home, but not nearly 10M dollars. Maybe you can have one in your community medical center?
In other words, ISTM the problem isn’t with a full-body scan per se. The problem happens if every little thing found gets treated or tested invasively.
So, you don’t mammogram every person with a breast. You mammogram people with advanced age, genetic predisposition, etc. so as to select for a population with a reasonable baseline frequency. Various professional bodies and the USPTF put a lot of effort into determining what the optimal screening population is.
Clinical decision rules are an extremely active part of medical research. Trying to figure out a reasonable predictive model to say “definitely follow up on this” or “definitely don’t” is a huge field in medicine. For instance, in patients ≥35 years old with incidental pulmonary nodules on CT imaging, we have the Fleischner Guidelines for when and which follow-up is appropriate. But it required a number of prospective external validation studies in order to properly validate it, and it’s for one super specific population: under-35 incidental pulmonary nodule on non-cancer-screening CT. So, you can imagine how much work there is to be done for the other 10,000 kinds of incidentalomas.
I’m sitting bedside wading through pull requests for software installation scripts. How is it the opinions of folks on my team for a bullshit process get more scrutiny than a potentially life-altering assessment that requires expertise, a keen eye and sharp mind every time?
In this case we had a scan in Nov. that was marked clear. Scan in Feb that had an unsized spot noted as ‘possibly metastatic’ with a note that ‘in retrospect this focus was present in previous exam’. Oncologist explains lack of measurement as being too small to precisely ascertain. Then two weeks ago we have a scan report indicating (among other new disease) a 4cm mass that was ‘previously 2cm.’
People are guaranteed to make mistakes, and my issue is with the process...not the personnel.
I am not a doctor, but I cannot fathom why a doctor would ever advocate against more data. Even if that doctor chooses to ignore it, at least let the patient utilize it as he sees fit. As someone who has regularly had to argue with doctors to get a test done, it is incredibly frustrating.
“Data” implies something neutral. Nothing about lab tests is neutral. Without contextual information and studies on how to -interpret- a finding, it’s just potentially terrifying noise. And terror usually results in action. Poorly informed action is often harmful. So... we are trying to prevent harm to our patients.
We are earnestly working on all fronts to build the studies to better interpret these findings. But we just aren’t there yet.
If the added data is noise, it can easily be ignored. If it's not noise, then you're lucky to have it. And figuring out whether or not it is noise is important, but you can't do that without the data.
But again, I am not a doctor, and how much bandwidth they have and where their priorities lie is not something I intimately understand. However something that has become blindingly obvious to me is that most doctors do not have a firm grasp of statistics, and will advocate for new drugs when the actual test results are borderline statistically insignificant and easily explained away by confounding factors (most obviously the placebo effect). Sadly not all trials are double-blind, something else I do not understand.
Possibly what I’ve failed to communicate is this:
MRI, CXR, etc are not images of the body. It’s not like getting a photograph of a liver and saying, at least we know this is or isn’t going on in the liver. They’re indirect measures of certain attributes of the body, such as tissue density, which we use - coupled with their medical information, and the mechanisms of likely diseases - to infer what’s happening. That’s why reading radiology is a medical specialty, and not something anyone with an anatomy background can do. (There’s a radiologist currently browsing the thread - he’s very welcome to correct me if I’m wrong about what radiology “is”.)
Because of this, every such finding has to be interpreted in a context, and studies tell us how.
Completely out of context findings aren’t a big problem if they’re completely unambiguous: hey, that bone is in two pieces and it should be in one.
What about the finding that isn’t, though? This is equivalent to not having any information on a test’s false positive / false negative rate, only now it’s open-ended to “every condition that could look like that thing” because the defining characteristic of an incidental finding is that it’s -incidental-. It’s not related to any symptoms. So what do I do with “every disease or non-disease process that could potentially look like a spot on the lung, without any accompanying symptoms of that disease”?
What I believe is the responsible answer is: “if I think the pre-test probability isn’t borderline zero, AND the post-test probability would change my course of treatment or diagnosis, order the test. If the pre-test probability is so low that any positive test result would be very likely to be a false positive and thus force me to act in a manner harmful to the patient, don’t order it - it shouldn’t be allowed to change the course of treatment. If the pre-test probability is already so high that any negative result is likely a false negative, don’t order it - it shouldn’t be allowed to change the course of treatment. Only order tests whose results should impact the course of diagnostics or treatment.” What do I do with findings that haven’t been studied in a given context, so I have no clue what their impact on the post-test probability of a diagnosis is? I don’t know. But “test just in case” isn’t the zero-risk option. There aren’t any zero consequence options.
Not every test is an RCT because grant funding agencies don’t provide the budget, plus or minus, many RCTs we’d like to do are unethical (if you have good reason to believe one course of therapy is superior to another, you don’t have the clinical uncertainty to ethically allow randomizing people into an inferior therapy), plus or minus many sub-populations are just too uncommon to build an RCT on without a gigantic budget that facilitates long collection periods across multiple institutions.
I'm reminded of a recent visit to a doctor who refused to say anything quantitative about a test. Sure, your average client knows little decision theory, but this doctor either didn't either or pretended not to follow. To make a good decision you need both probabilities and utilities; the doctor presumably best knows the former while the client best knows and cares about the latter, so they need to be able to communicate this way.
So, should we now do a work-up for every disease known to man with an inflammatory component? If you have one, if, every test - with their attendant complications and costs - for every disease you don’t have will have net negative effect on your health. You baseline have no symptoms or clinical suspicion for any of these (again, tautology or incidental finding), so the baseline probability is you actually have no disease at all, and this is a spurious value. So, what’s the advantage of this incidental finding?
This isn't that hard. You look at it again in 6 months. And you ask "Did it go up? Did it go down? Is this just your baseline?" And you keep your eyes open a little more strongly for issues that might correlate.
And, you know what, I, the patient, am FAR more invested in keeping tabs on whether things are going right or wrong in my own body than any doctor.
As someone who has had to figure out his health on his own after multiple doctors have failed or many never even attempted to help, I cannot agree with this more.
That feels like cheating though--this information isn't free.
My argument is that the expected value of this information, including acquisition costs, can very easily be negative, since you can't just "ignore" being dead. These costs are extremely large in my snarky autopsy suggestion, but dominate for whole-body MRIs + followups too. The catch is that the scan doesn't have huge benefits and the costs of the followups, while non-trivial, aren't exactly open-heart surgery either, so both are in the neighborhood of zero, except for the monetary expense, which will be crazy high.
(To say nothing of the extra $1000 surprise bill...)
Anyway, what is raising my hackles doesn't seem rational in the same way. It seems to have the flavor of "when we look and see certain signs, our system too often takes stupid harmful costly actions on that information. Therefore, don't look!" It'd be reassuring to continue "This is a workaround to an admittedly irrational system while we have the following people researching better ways to improve." But that's not the sort of thing I've read.
It becomes more complicated when you consider doctors have limited time and resources and cannot spend too much time with any single patient, at least the way the US healthcare system is designed. I believe this is a large contributing factor for why they behave the way they do - it's how the system is designed. And it's why I have all but given up on doctors even trying to help me with my bizarre symptoms, where I have been accused of lying multiple times.
The lesson a lot of people who are not familiar with public health is that you need to sorta take the long view for these things since we don't make rational decisions.
To bring back your analogy, it's then sorta like requiring an all hands meeting every time your IDE catches a typo. It's still monitoring a something that MAY cause a problem in the future, but to a good enough degree the costs associated with monitering may not be relevant to your actual end goals.
Of course it is often impossible to know which cancer will or won't kill you, but it should be recognized that there is a risk to scanning oneself without a proper indication, for example, getting a prostatectomy for indolent prostate cancer and risking impotence.
In the same way, skin doctors take photos of body moles to see if any changes occur between visits.
I think internal doctors should start viewing this as inevitable.
Then up-front develop "watchful waiting" plans that over the course of time and several scans chase down any suspicious changes.
Ignorance is bliss until you miss that easily removed tumor... etc and you are stage 3 going for chemo, or a brain anurism...
More data is always better. Period.
Problem is that we lack some kind of protocols of "calculated ignorance" that they could follow, like be allowed to ignore large classes of semi-normal findings that would show up often and lead people on wild goose chases, and be safe from malpractices and stuff if they follow this "calculated ignorance". This is because in the end medicine is not about doing "the best thing at all cost", but at doing the best thing that can be done at population scales, under limits of financial cost and "comfort cost" (there are limits after most of us would prefer (high risk of) death instead of further investigations and treatments and these limits vary (wildly) by individual).
There's two things that should be drilled down into the heads of all domain experts: (1) more data is always better, and (2) it's ok to ignore as much of the data as you want if it doesn't make sense cost-wise to investigate it further. But it's always better to fish in a larger lake!
Several years ago I worked on a study looking at the spread of bacterial lung infections in long term hospital patient. In order to get the data, we would have to intubate patients that did not need it. In this case, the people who approve these studies rightfully ruled that getting that data would be too painful for the patients to allow.
Biopsies caused by benign findings from a full body scan are not without pain, emotional distress, risk or cost to the patient. Period. It must be weighed against the benefit that the patient would recieve (all in the aggregate, or course).
That’s asinine, obviously he doesn’t mean more data is always better even if the test is invasive or dangerous.
> Biopsies caused by benign findings from a full body scan are not without pain, emotional distress, risk or cost to the patient. Period. It must be weighed against the benefit that the patient would recieve (all in the aggregate, or course).
Both need to be judged on time and cost, but only the biopsy needs to be weighed against risk, not the scan. If people see a scan with anomalies and want dangerous procedures even if the risk is not demonstrated that is a cultural, education, and legal problem. Both for doctors wanting to avoid malpractice ordering unjustified tests, and for patients “wanting it out”.
But if the scan can demonstrate some marginal elevated risk, not the need for immediate intervention but for increased screening, that data could potentially save many lives, and it should be used, not hidden away because of the possibility of overreaction. Researchers need to find how best to use the data and patients and doctors need to be educated in our current best understanding, not pandered to.
You're setting an arbitrary boundary because you don't want to consider the downsides. If you set the same boundary on the benefits of the MRI, then you wouldn't be able to include treatment of cancer as part of the upside to doing an MRI scan.
Doctors are far wiser about these things than HN commenters that have never dealt with the complex tradeoffs of dealing not only with populations of people, but also with dealing with the psychology of humans beyond people that prioritize rationality over everything else.
Doctors are responsible for treating all of humanity, not just the subset that behave the way we want them to.
With him classifying all MDs as idiots, no it is not asinine.
I’ve known few academic physicians that would turn down data if it can be ethically obtained.
The difference between the medical profession and, say Google, is that we as a general rule do not consider the populace one big lab to experiment on as we please. The consequences are more than a little different.
It’s not like things were always this way in medicine either.
And then, that medicine needs to become much more sophisticated in dealing with risk and uncertainty, given the absurd complexity of the problem domain. Doctors are not idiots but they are no more than human while trying to do an impossibly difficult job, and are far too cocksure given that context. They should be asking for as much help as they can get, which is not the reality of the situation. See something as simple as the resistance to checklists in surgery for example, despite the evidence.
I see your point about ethical breaches which have taken place. I would tend to think of those as abuses of power rather than coming from an ideological position, but that may not be correct. You’re certainly right about the dangers of a beta test mentality.
My point was not that either side was correct but rather that a simple statement like "more data is always better" is not as simple as they seem to think it is. Those of us who work in this field deal with balancing the risks as best we can so that we can detect the marginal elevated risk of something in a way so that the benefit outweighs the costs (financial, physical and emotional).
The cost of full body imaging is not just in the scan itself but in the procedures that follow it that require anasthesia (non-zero chance of killing the patient), cause pain and may cause secondary infection. In fact, just being in the hospital or doctors office to get the procedure has a non-zero risk to due to the concentration of sick people there.
I wonder what the patients who died because they had especially fast-spreading infections whose pace was not detected quickly enough would think of these altruists' wise decision to weigh the median level of "emotional distress" more highly than their lives.
If you order an unnecessary test and it shows something abnormal, even though in your clinical judgement it is not likely to be worrisome, you now feel a strong impulse to spend thousands more dollars to completely rule out it’s not some 1 in a million rare cancer lest you get sued. So some systemic issues at play here, not that “doctors are idiots.”
In data science, sure more data is always better but you’re not worried that more data is going to lead you to a traumatizing malpractice lawsuit.
Insurance reimbursement is exactly the same for your appointment, regardless if the patient lives or dies, or even gets correct diagnosis, nevermind correct tests.
/s
Do tumors from various cancers typically share any discernable features?