Think about the some of the implications of that statement for a while.
It really is no wonder that the softer sciences have a reproducibility problem.
Even Further, I think it's also indicative of medicine being full of people that have a fatalistic, passive attitude towards their patients and medical care. Medicine and health is really hard, and so many health professionals essentially believe that the only thing they can do is wait until something horrible happens before doing anything. And as the case with this woman, when something bad does happen, they bin it in their per-organized mental filing cabinet and you're fucked if they're wrong. Much like police officers, I think many get so jaded by the job that they can't be bothered to give a shit when they really should. And I really hope technology will save us from this.
Iatrogenic disease (harm caused by interventions) is by some measures the leading cause of death in the US, outstripping heart disease and standing at more than ten times the rate of death by automobile.
Under those circumstances, not doing anything until something horrible happens is not a bad policy, and the people engaged in it are not expressing cynical indifference so much as a due awareness of the real risks involved in almost any intervention.
Which is, admittedly, kind of depressing, but for quite different reasons. Either way, technology and patient empowerment are likely to help.
What you have to remember is that for the vast majority of human history, medical interventions of all kinds were at least as likely to hurt the patient as they were to help them, because of our crude understanding of how the body works and how disease is communicated.
A textbook example is Oliver Wendell Holmes Sr.'s classic 1843 study of puerperal fever (http://en.wikipedia.org/wiki/Oliver_Wendell_Holmes,_Sr.#Medi...). It showed how doctors who were doing a completely routine intervention -- general examinations of the health of newborn babies -- were actually picking up that infection from babies who had it and then spreading it to others who did not, because they didn't sterilize their instruments and change their clothing between examinations.
This seems obvious to us today, but at the time it was quite shocking, because in thousands of years of medicine doctors had never sterilized their instruments or clothing when moving from one patient to another. It never occurred to them, because none of the prevailing theories of how diseases spread called for it. So untold numbers of newborns sickened and died because of well-meaning, completely uncontroversial interventions by their doctors.
Today doctors obviously do understand the role of germs in spreading disease better, and they have much improved medical technologies and practices available to them, which has helped reduce the risks of those types of interventions. But millennia of these types of unpleasant discoveries has given medicine as a profession a healthy regard for the possibility that what they think they know even today could be incomplete or flat-out wrong.
And yet study after study of hospital acquired infections show that simple hand washing between patient examinations is abysmally irregular (between 10 and 50%.)
http://en.m.wikipedia.org/wiki/Iatrogenesis#Incidence_and_im... has more links, which I didn't read.
To be fair, the majority of problems are probably addressable without thinking differently. So people get into habit/routine and that's their "job"; thinking otherwise is not a day-to-day operation. To your point, automating the redundant/repetitive problems should allow for creative problem solving where machines don't yet excel.
You'll be in for a treat.
A "treat".
Fun bonus anecdote:
One of my wife's principals was convinced the prayer jar she provided for the staff was effective because over half of the prayers the prior year had been answered. When she told me that story it damn near broke my brain.
Depending on the prior probability of the prayers that were answered (and those that weren't), 51% could be plenty to support the efficacy of the prayer jar.
A string of anecdotes of "I double-checked and it turned out the doctor was wrong" remains less than a trickle compared to the flood of patients doctors have to deal with, and a tiny fraction of the hypochondriacs they must deal with. Our medical system is broken enough without adding doubts on the competence of its medical professionals.
By all means, people should double-check what the doctors say, but realize the much worse consequences of sowing doubt.
"What happens if we do nothing? If we do watchful waiting?"
"What happens if I don't take these meds?"
"How likely is that bad event? Tell me in terms of numbers per 10,000 people rather than percentages".
If he can't justify his decisions, then he's probably using instinct which is unreliable in the case of rare diseases, or he doesn't know how to diagnose known diseases which means he's incompetent.
Some people are offended when you challenge them for reasons for their advice, but I think that's because they aren't confident in it themselves or feel superior. Doctors shouldn't be in either of those positions and should be willing to tell you how they came to their conclusions.
Gerd Gigerenzer has a book explainin it better than I do.
As the joke goes, "what do you call the guy that graduated last in his class at medical school? Doctor."
Doctor's aren't machines, they're more like mechanics, they see a symptom, use their knowledge to make an educated guess essentially, and then work from there. If their knowledge is soft it's likely an incorrect diagnoses could just be exacerbated by incorrect treatment.
Get another opinion, and then another, the next problem is deciding on siding with consensus or outlier.
Hence why we can automate away most of the problems with it.
It's really unfortunate that nearly 100% of the bins are disease or a disorder. A typical interaction with a doctor involves finding out which disease bin you should be in, and then putting you in there. Once you're there, they're not concerned with making you healthier, or looking at the variety of options that could improve your life. They're just concerned with doing the standard operating procedure for dealing with the disease. Instead of being a person, the patient becomes a checklist.
Many doctors don't even see you. They see the bin you're in, and the many potential bins you could potentially be in.
What if half the bins were about great physical health? And what if the doctors worked just as hard to get you into one of those, as they do "treating" the diseases you might have?
Ugghhh. It makes me sad thinking about how far we are from that world.
While I agree that technology and automation can help, in addition we should find ways to help professionals move beyond their routines.
why should a doctor care? they get paid weather you live or die. what happens to you basically has no bearing on their career. a doctor i know worked on a guy who was having chest pains but was 'diagnosed' with COPD. he was to be sent home, but that doctor refused to because of patient history: all 5 of his brothers died of a heart attack before 60. what would have happened to anyone there had he been sent home? nothing, just another heart attack death.
at least they got things right on brother #6 though.
in short, doctors have little incentive to pick up their game. the supply of doctors is kept short purposefully to ensure they make money++. Only way out of this mess is technology when people can finally take their health back into their own hands.
Even worse is when some of those guys open source it and give away the software and hardware schematics, just because they hate giving $30 co-pays and dealing with deductibles, just to get an antibiotic prescription for a viral infection.
Don't make professional problem solvers see you as a problem, or you might just get solved.
>the supply of doctors is kept short purposefully to ensure they make money++
No. It's costly to train them, and it's hard for an average person to afford.
I would disagree with your statement that medical people have a fatalistic passive attitude. In this case the decision is between 1) watching and waiting to see if the tumor grows and 2) acting immediately to take it out. The tumor in question is a meningioma, a relatively common tumor of the skull (meninges, to be precise) whose natural behavior is unpredictable (some are essentially benign, others are highly malignant). Sometimes high-risk features can be identified on MRI imaging (brain invasion, areas of tissue necrosis, rapid growth) and these will always be treated with intervention. I work at UCSF, one of the top neurosurgery programs in the world, and I can guarantee you the radiologists don't make the mistake on sizing tumors mentioned here (well maybe they do, but it is out of carelessness rather than a problem with there software. it is not hard to find anatomic landmarks to ensure you are measuring at the same level). I would argue that a small meningioma without worrisome features should definitely not be operated on because the risks of surgery are not negligible! These surgeries are no joke, they are extraordinarily complex and they are mucking around next to structures which can cause significant morbidity, neurologic deficits, and even death sometimes if they nick the wrong artery.
It is well accepted practice and backed by evidence (see the WHO:tumors of the CNS publication) that the best way to decrease overall morbidity is by watching and waiting with these types of meningiomas. This is because MOST meningiomas are low-risk/benign (they are often found incidentally during autopsies for other reasons) and the risk of adverse surgical outcomes would be high if you subjected each of these people to invasive procedures when the incidence of high grade meningiomas is low. It is a population based strategy and unfortunately on an individual level you can't predict the outcome, but that is unfortunately a problem with this data. Fortunately lots of new technologies are emerging that can tease out individual differences (the so-called personalized medicine technologies)
That being said the new techniques to map out anatomy to aid surgeons is very cool! And to be fair, it sounds like the patient in this article did not get great care and that some of the doctors were not following the standard of care (1 year between imaging is definitely too long).
> "well maybe they do, but it is out of carelessness rather than a problem with there software."
I think this statement precisely illustrates the point that the parent commentator has with much of the medical community. Think about what the "carelessness" of a radiologist that you're talking so flippantly about means for a family – the anguish, sleepless nights, and feeling of morbidity and helplessness.
I may be overreacting here, but I would be extremely worried if a radiologist can be careless and ruin a year or more of someone's life, possibly forcing them into opting for a highly risky surgery. If you think there is a chance that your radiologists are not measuring things properly, wouldn't it be beneficial to start a double-checking procedure of some kind whereby each MRI is checked by 2 radiologists and a doctor (or something of the kind)?
It's always funny when someone thinks they're the center of attention... yes in a perfect world I'd remember everything and care only about you but if I do that I won't be finished my job. After a while I've noticed even as a student (and certainly picked up that a few doctors) that if you try care on a more personal level about every person, you'll be knocked down more and it'll be harder to function. Hence why I think a middle ground or even lack of care is what you see.
It's proabably a waste of resources and time to have a 2-check method as protocol (doctors are already overworked). Often times it is present informally (radiologists/pathologists will often have their colleagus review problematic cases for agreement and neurologists and surgeons will frequently review the films)
What everyone should do is realize you can always get a second opinion and have another expert review the material.
However you can be assured before an actual surgery, the same films will be reviewed by many doctors. Meaning that a mistake falling through everyone's mind is less likely.
I do agree with your point that sometimes a mistake can have a large impact on people. It's one of the challenges of medicine I'm not sure anyone has a solution to. On the flip side there would be many who are properly told the correct analysis, but i guess sometimes you get unlucky and a human error is made.
A neurologist I know, she always checks the actual scans as opposed to relying on the report narrative alone. Her background is in computer science, before her MD.
The issue with new analysis tools is that most doctors won't stake their reputation by exclusion. If a doctor says the tumor isn't growing, and it is, they are liable. If a neurosurgeon removes a benign tumor, then its business as usual.
The next biggest problem is that there's a gold standard of treatment, and it's hard to get enough research to move the gold standard. In this case, the gold standard is essentially a diameter measurement, based on picking the visibly largest slice of a data set which happens to have an axis determined by the patient's somewhat random orientation in the CT scanner. Axial, Coronal, Sagittal. Those are not perfect, anatomical orientations - they're rough. Most radiologists won't take the time to create Multi-Planar Reformats along oblique or double-oblique angles. And even if they did, they'd still be eye-balling the measurement.
It's not very hard from a regulatory standpoint to create tools to allow a radiologist to click here, click there, and show a measurement of distance. Then the radiologist can look at a similar measurement from a different radiologist on the previous scan, and say whether it's bigger or smaller now. (But some PACS still don't handle anisotropic MPR correctly, which just blows my mind.)
It's vastly harder from a regulatory standpoint to say in a medical product for commercial use, "Hey, we automatically measured the volume of this tumor by automatically finding the boundary of the tissue and automatically ignoring the connected blood vessels and neighboring bone, and then we automatically found the largest cross-sectional diameter, and then we automatically compared the same measurements against the automatically non-linearly-registered longitudinal previous scan... diagnosis: the tumor is not growing, just changing shape slightly."
People think their CT scans are examined for hours by a team of doctors, like on "House." In reality, you're lucky if your CT scan gets 5 minutes of investigation, and that's by a radiologist who was trained in med school to read X-Rays, not CT. And your oncology head CT may be read by a radiologist who specializes in abdominal pediatric Ultrasound. And there's almost certainly no double-reading for Quality Assurance. And there's no rating on radiologists to tell you who has the most accurate diagnosis. And he's probably using 2D measuring tools on 3D data - because that's what they're familiar with, and that's what the reconstruction software manufacturers provide, and that's the medical gold standard. On a shared computer, where some other radiologist (or resident!) might have changed the display settings in a way he didn't notice. And he's probably been reading images for 14 hours straight.
It's worse than that: It would be sued out of existence by those people who rank badly (ie, exactly those bad practitioners that need to have their asses kicked.)
In a related example: A friend wrote a book about first-hand experience in dealing with a nursing home for her aging parent. There were many instances of problems, but one in particular was that a doc had prescribed the wrong drug.
The publisher loved the book, and made the publication offer contingent on a number of changes. One of those changes: You can't implicate a person or his institution because of the liability. It doesn't matter how good your proof is, the jury system is so broken that the publisher just can't accept the risk.
Now just imagine if there were a searchable database of outcomes tied to individual practitioners or institutions.
Medical Practitioners and Medical Institutions should be required by law to document outcomes, provide information about disputed outcomes, and update the information on a periodic basis.
And tampering with the information should carry heavy punishment, including jail time for the practitioner and officers of the institution.
Actions like the recent cover-ups in the VA hospitals about wait times leading to deaths should be nearly impossible to hide from the public eye.
If you are lucky, some eager, tech-savvy resident will use 3D Slicer or OsiriX on their personal laptop to coregister and segment images, and do longitudinal volumetric comparison. But those basic tools are not available on the platforms used to read most images.
Major vendors are trying to catch up, but from what I've seen, demand (and thus willingness to invest) is surprisingly limited.
You may not be able to get insurance to reimburse for it, but my impression is that there is no shortage of people ready to pay cash money, easily into 4 figures, for someone providing this service in a 'personalized medicine' context.
As far as packaging a tool-chain that actually makes these state-of-the-art measurements and applying it in this domain, it's not clear what the FDA approval / requirements would be.
Sadly, none of this is easily available in FDA approved packages from typical MRI/CT imaging vendors like GE or Siemens. Often need to pay extra for such features. You could theoretically walk to the surgery suites and borrow a stereotactic imaging station (basically an embedded Linux PC) and do it there, but radiologists typically aren't trained to do that.
FSL and ANTS aren't approved by FDA for clinical use so can't be officially documented in a patient's chart...
Most radiology training is based on visual recognition of abnormalities in anatomy - the "eyeball approach" has its value, which is why intelligence agencies employ human spotters and people are working on neuromorphic chips. But there isn't really a lot of quantitative/comp sci expertise among typical medical students (maybe more now, who knows?)
I had a similar experience when I was having a spinal curvature corrected. After a particular xray, several people -- a couple doctors, the xray guy, a med student or two, a mix of a couple people like this -- had measured my spine's then-current curvature and had come up with a huge range; to one, I had gotten worse... to another, I had some improvement... to the third, marginally the same with a smudge of improvement.
In the end I got the most improvement I was going to get given my specific situation and I'm happy for//with that. The point is, it was literally three measurements on the same xray from which three mutually exclusive conclusions//opinions were [literally?] drawn.
I'm therefore not surprised that there is similar with tumours.
Hm, people with 6-digit salaries and doctoral degrees whose main job is to figure out diseases perform worse than some random dude? That's the first thing that comes to my mind.