The Problem With Silicon Valley Medicine

In this video, Rohin Francis, MBBS, reviews modern health trends and the dangers of unsupported medical claims.

The following is a transcript of this video; note that errors are possible.

Francis: There is so much that I could say about the collision of the worlds of Silicon Valley and the mindset that drives it, sometimes referred to as the “tech bros,” with the world of medicine and the strange bedfellows that they make.

I will explore many of the phenomena that arise when this happens in future videos, things like novelty bias, where you assume that something new must be better. While this normally holds true for computing and we’re all familiar with Moore’s law, it very commonly isn’t true in medicine, with many new and exciting therapies being quietly or occasionally loudly shelved years later for being useless or worse, harmful, or how Silicon Valley’s motto of “move fast and break things” can be catastrophic for medicine.

If you want to hear more about tech and medicine, then please do consider subscribing. But for this video, I want to focus on one aspect, the obsession with data.

The belief that if we just measure more and more we can unlock the secrets of the human body. We can use 100% of our brain, become immortal, and transform into supernatural beings comprised of pure energy.

If I say to you “medical grifter,” you probably think of people selling quack remedies like essential oils or some sort of herb. You might not immediately think of the enormous industry loosely centered around longevity. People, mostly men, focused on maximizing lifespan, optimizing performance and so on, which sounds fine.

Unlike the Herbalife or Mercola natural shtick, this is futuristic, sexy, cool-sounding stuff, sciencey-sounding stuff. If you’re not familiar with this world, to be honest they’re really not that different from other promoters of pseudoscience. But unlike magical thinking, the magical thinking employed by your common or garden alternative medicine practitioner, these guys feel like they are invoking science because they reference some study, apparently utterly ignorant to the fact that it’s complete garbage.

What particularly winds me up is that so many of them are medical doctors or scientists in relevant fields who either demonstrate what I have said many times, that you really don’t have to be that smart to be a doctor, or they are just shameless, money-grabbing grifters.

Now I see variations on this tweet a lot. This is just one example of the phenomenon. But the other day, tech entrepreneur Peter Diamandis posted, “Do an annual ‘Full-body Health Upload.’ Most of us have *no idea* what is really going on in our bodies. Every year, I do a full body MRI, an AI-driven CT, Genome, microbiome, and much more. A key to longevity is proper screening and preventative medicine.”

Well, first of all, Peter, the actual word is preventive. But I was stunned. I mean, this man claims to be interested in longevity, but only does these tests once a year. I mean, what an amateur. Every morning, I perform a full body MRI, colonoscopy, bladder flow cytometry, testicular volume analysis, lung function, liver percussion, and complete genomic sequencing in case I have been replaced by a clone in my sleep, who differs from me by just one base pair on chromosome 15.

My entire house has been converted into a PET detector. All my food is radiolabeled. I monitor my heart rate and blood glucose on a picosecond basis, anything less was too imprecise, and I have been in a continual state of CT scanning since 1996. It’s lucky I have already reproduced because I now require a 60-m radiation exclusion parameter and my sperm have been rendered as impotent as Boris Johnson’s influence on Russian foreign policy.

Now, I don’t want to focus too much on one individual because publicly talking about how many medical procedures you submit yourself to is a staple of many movers and shakers in the wellness and maximizing performance, life-extending space. But I think Dr. Diamandis is a useful case because he has been very into longevity in recent years and he is a really, really clever guy. He is a Harvard graduate medical doctor. He is also an engineer and, in a neat link to my recent video, he also founded Zero G, the parabolic flight company based in the U.S. He has had a whole career involved with space and medicine, he’s friends with Elon Musk, he founded the X PRIZE, he’s a best-selling author. On paper, this guy should be my hero. I mean, these are all things that I love. Well, maybe.

Why do I think that this idea of getting the most ridiculously detailed medical checkup is completely barmy? Actually, just before we move on, I only just learned this incredible alleged fact about Diamandis. But after he had held a $30,000 a pop party during lockdown, which was declared a super-spreader event as almost everyone tested positive for COVID, he tried to sell some of the guests inhaled amniotic fluid and ketamine lozenges as COVID therapy, which I mean fair play.

Anyway, the urge to measure everything is not limited to tech bros. It appears to be human nature. Ancient records of physicians show meticulous studies of what was administered to patients. Santorio Santori, a friend of Galileo’s in Padua, was a doctor and a genius — as I imagine, you’d have to be to be friends with Galileo — and he invented many measuring devices such as a wind gauge, a water meter, a device to measure the pulse, which is amazing, the precursor to the modern thermometer, and most famously he weighed himself and everything he ate and drank, as well as everything he excreted, using a massive set of scales that, you guessed it, he invented.

Over a decade ago, The Atlantic ran a piece about a computer scientist named Larry Smarr, who was at the vanguard of a movement called the “Quantified Self,” which I’m sure you’ve subsequently heard of. He was measuring everything: weight of food in, constituents, and how many steps he took. He collected all of his feces and urine, and had it analyzed. I remember reading how he said that stool had a capacity of 100 terabytes of data per gram. I guess that’s what you’d call a semi-solid state drive.

Fast forward 10 years, and we are all counting our steps, our oxygen saturations, our heart rates, and diligently entering our food intake into apps. What tech guys think today often society adopts tomorrow. I’m not saying that gathering data is bad, if the urge to do it appears to be very human. But it’s what we do with that data that is the problem, how to interpret it. Yes, I know what some of you are saying that, how can we ever figure out how to interpret it unless we collect it? I agree, but that should be as actual research, not individuals making decisions based on half-cooked data and huge assumptions.

Let’s go through some examples so you can understand the kind of thing I mean. I’ll start with why I would advise someone who is feeling perfectly well against having that full body scan. Diamandis says that he has a yearly CT scan, which is completely nuts because each one gives you a fair whack of radiation. One CT, a few in your lifetime, okay, not such a big deal. But every year? I guess we have just figured out why he checks his genome so often.

But, okay, let’s say full body MRI scan, free of ionizing radiation. Some of you might recall that I actually did my doctoral research in MRI. I have reported many thousands of images. The main reason it’s not a good idea to get a screening MRI scan is incidentalomas. There are other reasons as well, but this is the key one.

Now I go into this in more detail here, but let me break it down super simply. The human body is complicated. It’s got squishy bits, hard bits, walls between those bits, cavities, protuberances, and holes. It’s not color-coded like in textbooks. If you just think about how different we look from the outside, why assume it’s going to be any different on the inside.

Humans are wonderful and varied and messy. Sometimes you see something slightly weird: a spot, a blob, a smudge. If you come to a hospital with abdominal pain, I do a scan on you, I’m paying particular attention to your abdomen. Well, I’m not. I mean, a radiologist is. I have no interest in anything underneath the diaphragm.

If I see a blob in your abdomen and you’re having abdominal pain, well, I’ll feel confident in saying something is wrong. But if you’re having a screening scan and you have no symptoms, and I see some blob in your abdomen, or a cyst on your kidney, or a mass on your uterus, a shadow on your liver, plus more blobs and smudges in other body parts, I have no idea if some of them, none of them, all of them, are significant.

What do I do? Is that blob a benign polyp that’s of no consequence, or is it early cancer? I tell you I can’t be sure and I suggest repeating the scan in 3 months.

Now, you have a potentially serious diagnosis hanging over your head. You have to declare this on insurance forms. You spend 3 months worrying. You then shell out for another scan and it kind of looks about the same, maybe slightly bigger. It’s hard to know. Scanning isn’t some perfect science. There is a bit of ambiguity.

I tell you again, “I can’t be sure. Let’s do a biopsy just to make sure.” Now you’re having an invasive procedure, with some risks, to remove a sample of tissue from a mass that has caused you no symptoms at any point. It comes back benign. Thank goodness, but what a stressful few months.

Okay, Rohin, you say, that’s one scenario. I have got common sense. I know how to interpret these findings. What about the scan where it does pick up early cancer and it’s removed and a life is saved? Well, in spite of multiple studies looking at this, that has never been shown to outweigh the risks. You can pick out some anecdote of someone whose life was saved by a screening test, but that ignores all of those whose lives were made worse.

Screening scans do not improve mortality. They just increase health anxiety. They increase the costs to the individual and often to the health system as well, if that has to pick up the additional aftercare, and they over-medicalize normal life, thinking these risks won’t apply to you, and only the benefits, is something called exceptionalism bias.

Talking of exceptionalism, I saw a news report, from America of course, where they cheer on any money-making scheme at the expense of someone else’s health, where a fit retired athlete had a screening heart scan, which showed some coronary artery disease and he went on to have open chest heart surgery, a coronary artery bypass, in his 40s, having never once had any chest pain. There was a photo of him grinning with his doctor, who he says saved his life, but unfortunately not a single study in cardiology shows that operating on asymptomatic heart disease like that saves a life. All it does is expose him to the risks of surgery.

Another asymptomatic case ended up with a patient suffering a complication from an angiogram and invasive tests of the heart, and they ended up needing a heart transplant. He started with a normal heart. This was a tragic waste that he ended up having a test that he didn’t need and he lost a perfectly healthy heart.

If you’re in any doubt that these are more about money than medicine, then here is a company advertising their $2,500 full body MRI that you can sign up for, which can detect a bunch of conditions where early diagnosis really achieves nothing, and also pre-kidney stones.

Now, full disclosure, I have no idea what pre-kidney stones are. Actually, I mean, this is genius. If you don’t have kidney stones, you’re healthy, and we can’t make any money from healthy people. We hate them. So no, no my friend. You have pre-kidney stones, very serious. In fact, this is a new pandemic. Everyone without kidney stones actually has pre-kidney stones and you need annual scans. Dr. Francis’ renal revelation, just $200 a month, to keep your kidneys clean. I’m afraid we don’t accept pre-dollars.

Next example. I have made a couple of videos about the Apple Watch. I love my Apple Watch. But for the vast, vast majority of people who buy it, maybe 98%, the ability to detect atrial fibrillation and irregular heart rhythm will be useless. Not just useless, but potentially could lead to unnecessary involvement with the medical profession, tests, and maybe even treatment.

Because AF is so uncommon in young and middle-aged people — in fact, it’s rare AF — false positives will dwarf true positives, even if the detection accuracy was one of the highest on the market, which it isn’t. It’s extremely low and it’s designed to be that way. It’s designed to be more sensitive than specific, i.e., the engineers feel that it’s better to catch more cases, even if that means misidentifying normal rhythms for abnormal ones, rather than missing an important diagnosis. They’ll pick up a good amount of the real problems out there, but also loads of bycatch flapping around on the deck needlessly.

There are going to be way more false positives than true positives. If you get an alert from your watch, it’s almost certain to be incorrect. Unless you’re elderly, as the only population with potentially useful data are the oldest. For under 65’s, the likelihood of a positive AF detection episode on an Apple Watch actually being AF are less than 5%.

But let’s assume the Apple Watch is 100% accurate at detecting atrial fibrillation, and it tells you that you had one 20-second episode in a month. What do we actually do with that information? All the literature in medicine up till now has been about AF recorded on a heart tracing or maybe over a day or two, but we have never monitored people constantly like we can do with wearables.

What is the cutoff between where normal ends and abnormal begins? Does 20 seconds of atrial fibrillation have any implications for your health? Should you do anything else? What about 5 seconds? What about 2 hours? We don’t know the answers to these questions, and these are really genuinely interesting questions. I asked this one on Twitter and I got a dozen different answers from 100 different cardiologists.

Again, I’m not suggesting that asking these questions or gathering data is wrong. I’m just saying that we have still got a lot of knowledge left to gain. This leads me on to my next example about the diabetes.

Type 2 diabetes is abnormally high blood glucose. We measure long-term blood glucose levels because obviously a one-off test at a specific point will miss the diagnosis a lot of the time because blood sugar goes up and down all day long, depending on what you eat, when, stress, sleep, illness, medication, and lots of different factors.

Recently, continuous glucose monitors have become easily available. They are brilliant devices and can wirelessly link up with your phone all day long and record accurate, near-continuous body sugar levels. They are life-changing for diabetics, an example of tech and medicine coming together beautifully. But the longevity, high-performance guys have got their hands on them and started using them, even though they don’t have diabetes.

Peter Attia, another very, very smart guy whose opinion I value on many things, he is a doctor, but he also, I think it would be fair to say, has a tendency to promote unevidenced medicine and of particular interest — which I didn’t know before researching this — he was a paid consultant for a company that makes continuous glucose monitors, or CGMs. He has talked about them quite a bit.

Now, people can watch their blood sugar respond to things they eat, and just kind of goes here on the arm and connects to your phone. Awful refined sugar like a can of Coke or a donut sends it up. But complex carbohydrates can also cause an elevation, like potatoes, brown rice, whole grain bread, and fruit, whereas pure fat and protein to a much lesser extent, which has led a lot of these people to conclude that this is objective evidence that carbs are bad for you and that you should stick to a purely keto or carnivore-style diet. Which I understand that seems very logical, but that’s based on the assumption that a physiological variation in your blood glucose, as all of our blood glucoses do vary, that’s assuming that a normal variation is actually pathological.

There is no evidence that these gradual, short-lived rises in blood glucose are harmful. In fact, they are totally normal. Indeed, you’re only getting half the story by concentrating on CGMs. If there was a continuous triglyceride monitor, which is a kind of fat molecule, you’d see that a high-fat meal causes elevations in this after eating. It’s almost as if you should have a varied diet.

Do you know what else causes blood glucose to rise in a similar manner to the same kind of level and with the same return to baseline? Exercise. Blood glucose can really shoot up, as can your blood pressure, but is anyone going to suggest that exercise gives people diabetes or hypertension? Of course not.

Patients now say that they have stopped eating fruit because they saw their blood glucose shoot up after eating an apple or an orange. That’s really sad and kind of crazy. Please don’t believe all the claptrap about fruits being bad for you. I certainly wouldn’t advise a fruitarian, all-fruit diet, nor would I advise drinking juice, but an apple a day may not keep the doctor away if you’re accompanying it with crystal meth and a Whopper. But the little transient blood glucose elevation from the apple is not bad for you. Think about it like a duck’s legs underwater. The up and down perturbations are normal, but the overall function is ideal.

Maybe you’re not convinced. “A high sugar is a high sugar,” you say. “I don’t want my blood glucose going up.” Well, you’ve heard of the concept of anti-fragility, Nassim Nicholas Taleb’s term about systems that are designed to be tested. Some things are fragile, breakable. Other things are tough and robust. If you knock them around, they don’t break, but neither improve, when they are stressed.

However, biology is somewhat different. Exercise stresses you, but makes you stronger. The immune system develops by being exposed to pathogens. If you replace glucose with a blood test called a CRP, an inflammation marker, you will see CRP go up with pathogens. But if you employ the same rationale that a raised level is bad, and you should try to avoid any raised CRP, and you meticulously avoid anything that might provoke an immune response, what happens? Well, you end up with the rate of severe allergy shooting upwards, as it has done in recent decades.

Obviously, the immune system and the endocrine system are very different. This is not a perfect analogy, but we all understand that moderate exercise, moderate immune stresses, and moderate temperature variation are better than attempting to rigorously limit challenges to the body. Why assume that we can’t handle moderate challenges to the endocrine and digestive systems?

My point is that we are gathering more data about this than we have ever done before. We’re drawing conclusions and making assumptions, which might well be wrong. The bottom line is we are not computers. We are biological and susceptible to chaotic and stochastic processes.

A bit of abnormality is normal. The more intensely we look at ourselves, the more we learn that human biology has an enormous signal-to-noise ratio. One day I have got no doubt AI and machine learning will be able to see through all of that, but right now it is way too early to be using this raw unvalidated data to make big decisions about your health.

By all means, buy these gizmos. I never want to tell anybody what to do, but do it for curiosity and fascination at watching the human body doing its thing, not to inform you and instruct you what to eat, in an attempt to let you grow as old as Yoda.

Just believing that you can throw AI at something and expect an improvement is automation bias. Please, many of you are adept at pointing out financial conflicts of interest in medicine. You will accuse doctors of being on the payroll of Big Pharma for prescribing a drug and yet you hand over wads of cash to quack doctors peddling unproven therapies. Or you’ll shell out for expensive tests, even when you’re feeling absolutely well.

You are right to point out the problems in modern medicine. I often do the same thing, particularly the financial conflict, but please see these full body scans and longevity or wellness influencer plans for what they are, money-making schemes. I don’t want to see you. It’s not me, it’s you. I don’t want you seeing other doctors behind my back either. If you are healthy, then enjoy your life. Eat your apple a day. Keep away from the medical profession.

Every time I hear about some new app or screening program, I think of my 80+, 90+-year-old patients that I know and look after, who are still going out dancing, hiking, sailing, golfing, and they are mentally just as sharp as ever. They didn’t do any of this nonsense, but they all emphasize the importance of staying fit, staying active, eating healthy food, and enjoying life.

And the same messages emerge from studies of the oldest populations on the planet. Do those things and let’s keep you out of my clinic room and off my operating table.

Rohin Francis, MBBS, is an interventional cardiologist, internal medicine doctor, and university researcher who makes science videos and bad jokes. Offbeat topics you won’t find elsewhere, enriched with a government-mandated dose of humor. Trained in Cambridge; now PhD-ing in London.