Do Less, Heal More: The Case for Medical Conservatism (with John Mandrola)

John Mandrola: Well, thanks for having me. It’s a real honor to be on a podcast that hosted Milton Friedman and Thomas Sowell, so I’m quite excited.

John Mandrola: Well, when you go to the doctor, the doctor would probably assess you, determine what the examine is like, and then he or she will probably order an MRI [Magnetic Resonance Imaging] or some sort of X-ray, maybe after a period of time. And then, if it shows an anatomic abnormality, say, of the meniscus, or people get labral tears, other kinds of things, the doctor could suggest arthroscopic surgery.

And, what strikes me about these treatment plans is that a patient could get better because the illness just runs its course–so-called natural history. It could get better because of the surgery. Or it could get better from a combination of things, such as the placebo or placebo effect, which is an expectation that they should get better. In so many things in medicine that we treat, there’s multiple reasons why patients get better.

And, what struck me so much about the study that we’ll talk about is how they used really a proper placebo control, which we use in drug trials, but we don’t often use in surgery trials.

John Mandrola: Well, Russ, I think if someone’s sticking a tube inside of a joint, any time that we put things into the body, I would call it invasive. And of course, the definition of small surgery is surgery on somebody else. So, it’s not like open knee surgery, but it’s not nothing either.

John Mandrola: I read in the introduction to the paper that many, thousands. I think 700,000 arthroscopic partial meniscectomies have been performed annually in the United States. So, this is a big deal, 700,000 procedures.

John Mandrola: So, importantly, whenever you look at studies, you always have to think about who is in the study, and these were patients with chronic tears. These were not somebody who–not a complete, acute tear that was there immediately from an acute injury–but sort of a chronic knee pain, chronic things. And, what they did was they took 146 of these patients, and they all had arthroscopy–

John Mandrola: Arthroscopy: a tube looking into the knee. So, the surgeon looks into the knee and sees that there’s a meniscal problem. And then, there’s randomization to repairing it, shaving it, making it look clean. My impression of it is that it makes the knee joint look pretty. Or, pretending that they did: a so-called sham operation where they push around and ask for instruments. Basically, the patient doesn’t know whether they’re getting real meniscal surgery or a sham surgery.

John Mandrola: Yeah. Well, no, you’re under sedation, light sedation. And, there’s a block, a nerve block, and there’s sedation.

And, the way we do it in cardiology is patients wear headphones and they listen to music, and they’re sedated, so they don’t really know. And then, of course, later on whenever there’s sham controlled studies, there’s a sham index where people are asked whether they can guess their treatment assignment. Usually, if the blinding is pretty good, then they don’t know.

And so, then, the surgeon is off the case. All the follow-up is done by people who don’t know whether the patient has had true surgery or a sham surgery.

John Mandrola: No, correct. So, what’s so elegant about this study is that all patients had the arthroscope. So, they all had an incision in the knee, the tube placed into the knee; there’ll be a scar afterwards. It’s just one group had the actual repair through the arthroscope, and the other group had no repair–had a fake repair–and just left it, basically. And so, on the outside, the patient can’t tell.

John Mandrola: So, that’s the tension, isn’t it? Because, we just talked about 700,000 of these procedures are done annually. And then, what are the ethics of doing 700,000 procedures that really doesn’t have any better effect than sham surgery? Or, operating on 145 patients. And, we have many examples in cardiology where patients were getting far more invasive surgeries on their heart that were, once studied under placebo, were shown to be not beneficial.

And so, I’m not an ethicist, but I can understand that there’s tension. But, without doing this placebo controlled study, we don’t know whether some of these things are effective.

John Mandrola: Well, I guess you could–I understand your point, and I think it’s an important point. But, I would counter by saying that you could make an argument that the arthroscope is partially diagnostic. So, the surgeon is looking into the knee joint, might find something else, and the patient is having that diagnostic procedure.

In cardiology, we have a very famous study where this group in London looked at patients with single-vessel coronary disease–severe lesions of coronary that was causing angina. And, one group got a stent–got the thing fixed–and one group got a pressure wire and nothing was done. And, in the end, Russ, they showed that there was no difference in exercise time from fixing these things.

But, the way they got along with that study, called the ORBITA [Objective Randomised Blinded Investigation With Optimal Medical Therapy of Angioplasty in Stable Angina] study, was that they said, ‘Okay, you can have the stent procedure after a six-week period. You won’t know whether you had the stent or had the placebo stent. But, if you just wait six weeks, then we can do the stent.’

John Mandrola: And, patients in Britain have to wait more than six weeks anyways for their procedure, so they were able to learn so much. So, there’s that.

I think there’s some creative ways around the ethics of this. But ultimately, it’s just learning what works and what doesn’t work.

John Mandrola: That’s correct for any surgery, and it’s a really important point because what it would be for arthroscopic knee surgery is X. I’m not sure I could quote that; but it’s something. And, it’s the same with any cardiac procedure or any back surgery procedure. There’s going to be a finite risk. Infection, anesthesia complications, bleeding. And, if the surgery is no more effective than placebo, that’s really important to know.

John Mandrola: They found, interestingly, that when you look at the response curves of the pain scores–these were subjective scores–both groups improve exactly the same. There’s no difference. So, the placebo group improves, and the actual surgery group improves.

And so, the actual surgery was no better than a sham surgery.

Now, you don’t know–without a no-surgery group, you don’t know what the true placebo effect is. You just know that the surgery done 700,000 times per year is no better than a placebo surgery.

John Mandrola: It’s generally not done. I can’t explain exactly why. I can say that logistically, it would be difficult. It might be assumed that there would be a difference, but it would be the only way to, I think, truly quantify the placebo effect.

John Mandrola: Yes. There’s a very famous group in London–Imperial College of London, that did the ORBITA study–that really pushes back against using the term ‘sham’ because it’s really a placebo procedure. And, if the surgery doesn’t work, then the whole thing is the sham.

John Mandrola: Yes. It’s 12% versus 4%; they’re small numbers, but yeah. It’s not just the potential harm, and cost, and inconvenience of having surgery, it’s potential downstream problems.

The study that I wrote about in Sensible Medicine was a 10-year follow-up, and these are old studies. But, it’s truly remarkable that they’re done.

And I think, Russ, the other thing that, as a clinician–I’m a practicing clinician–the other thing that we learn: We have a research lens to these studies, and we learn what procedures work and how placebo-resistant they are. But, as a clinician, we also learn from these studies in using the placebo effect and avoiding the nocebo effect–the opposite of the placebo effect–in clinic. It helps us understand, I think, the component of the caregivers’ relationship in the caring program.

John Mandrola: So, the nocebo effect is directly opposite the placebo. The placebo effect is when a patient, a doctor expects there to be an improvement. A colored pill or a big capsule makes a bigger placebo effect than a white, small capsule, and surgery is certainly a bigger expectation signal than a pill.

But, the nocebo is when you expect harm, and the expectation of harm makes you feel ill.

There’s a very common medicine that we use for cholesterol called statin drugs, and statin drugs are a classic example of nocebo effect. Where the drug–there’s four or five different statin drugs–and they lower cholesterol, they reduce the chance of having a heart attack. And, in blinded trials, when a patient doesn’t know whether they’re on the statin or a placebo, there’s absolutely no difference in side effects. But, then, when we look at observational studies where patients know they’re taking the statin drug, there’s a tremendous amount of muscle pain and ill health.

And, in clinic, we see it all the time, ‘This statin makes me feel terrible.’ I’m a cyclist, and cyclists who take it just say that their legs aren’t as good on the statin because they know it could affect their muscles. And, that’s mostly all nocebo effect.

And, it was actually shown in one of the best trials ever, and if I could just tell you about it because it has–it’s called the SAMSON [Self-Assessment Method for Statin Side-effects Or Nocebo] Trial, and it’s such a great story. Again, the Imperial College of London group took 100 patients who they said could not take statins–these were statin-intolerant patients. And, very clever experiment: What they did is they randomized these patients on a month basis. In one month, the patient would take a statin. Another month, they would take a statin placebo, identical tablet. And, in one month, they would take no–

John Mandrola: Visually, visually.

John Mandrola: No, they’re not chemically identical. They just looked the same. So, one month, a statin. One month, a statin placebo. And, one month, no tablets at all. And then, they had an app, and every day they said how they felt on the app.

And, what they found, Russ, is–it’s just so amazing–is they found that the best months were the no tablets, and they felt identically bad whether they took a statin or a statin placebo. So, there was clear ill-effects from the statin, but it didn’t matter whether it was the statin chemical or a statin pill that looked like a statin. And so, they just elegantly showed the nocebo effects of statins.

John Mandrola: Yeah. I think what we can learn from it is that we–what’s so really beautiful about these studies as a clinician is that we learn from it as a clinician. And we can have counsel with the patient. And, we can say that we are sorry that they’re having pain–we understand that–but the pain will get better; and a surgery has been shown not to be any more effective than a placebo surgery. And, it will get better, and we’ll support you through it.

I once gave a lecture titled “Words Can Harm and Words Can Heal.” And, I think that what we can learn from these procedures, these sham surgeries, is that patients will get better via the natural history. They’ll get better from other measures. But they will get better. And, I think that we have to be partners with our patients and advisors. And, I know that sounds lofty, but it’s really true.

John Mandrola: Any doctor will tell you, Russ, 100% that it’s much harder to just do the darn surgery than it is to explain why we shouldn’t do it. [Maybe Mandrola meant the opposite–that it’s harder to explain why you shouldn’t do the surgery than to just do it?–Econlib Ed.]

John Mandrola: You get reimbursed enough, I think. And, you get the pleasure of helping somebody with your words, which is, I think, more elegant than with a knife or a catheter, honestly.

John Mandrola: No. That’s where the doctor comes in. You can tell the patient–now, I’ve had a surgeon tell me this. I had a surgeon–I had a labral tear of my hip, and it was terrible. I couldn’t run, I couldn’t even cycle. And, he says, ‘John, you’ve got a labral tear. It’s going to get better; it’s going to take some time.’ And, I said, ‘Well, what about labral surgery?’ He goes, ‘I wouldn’t do it: you’re just going to get better on your own. It doesn’t really help.’ And, that was enough for me, and I really appreciated that advice.

But, that advice wouldn’t have occurred if we didn’t have brave investigators doing these kinds of studies.

John Mandrola: Yeah. Thank you. Yes, of course. My labral tear got better on its own.

John Mandrola: Absolutely. I don’t want listeners to think that I think that we should just tell patients they don’t need surgery. There are clearly things that get better.

But, for instance, for instance, Russ, my example, and I have colleagues who’ve had it, is: we’ve developed atrial fibrillation, an irregular rhythm of the heart. And millions of people get it. And, I got it. I couldn’t believe it: I’m a heart rhythm doctor with atrial fibrillation. And, the most frightful thing for me of having this arrhythmia disorder was having a procedure. Scared the crap out of me, even though I do the procedure every day.

John Mandrola: So, I just said, ‘I’m just going to see what happens.’ I just slow rolled it. And it got better. And, when I give lectures–

John Mandrola: No. I just dealt with it and took some medicines for a while, and it got better.

I have this beautiful example of this cyclist who came to me, he was having all this atrial fibrillation and he has this monitor. He’s an engineer. So he put all the AFib [atrial fibrillation] episodes in red boxes and he had this calendar full of red. And, I said, ‘That’s terrible.’

And then–he was also conservative and didn’t want to do anything. So, then I say that, in July, the AFib’s all better, all green squares. It’s basically gone. And, I say if he had had a procedure in February, we would have called that procedure successful–

John Mandrola: even though their condition gets better. Labral tears are like this, atrial fibrillation is like this. Many things are like this.

So, as a clinician, I think it’s elegant, and beautiful, and fun to harness all of the ways that patients can get better. And, rather than intervening early and often.

John Mandrola: You and I are 100% similar because I’m much more afraid of intervention than I am of any condition.

But, we should set out that conditions differ. And there are some things that are very reasonable to watch under careful surveillance of an intelligent physician who is an expert, and he or she partners with you; and it’s very okay to watch certain things. There are other things that are more scary, and you need a wise counsel to say, ‘Russ, a lot of things we can wait and we can slow roll, but this is something that we shouldn’t, and it’s hazardous and we should intervene.’

But, in a vast majority of these chronic conditions–leg pain, numbness and tingling in the foot or something or in a hand, even atrial fibrillation–you can watch and be conservative, really. Voltaire famously said, ‘The best physician humors the patient until nature heals the patient.’ And so, I call it the Voltaire Approach. It’s underused, but it’s really effective and elegant.

John Mandrola: Well, yes. I see where you’re going with this question. There are a lot of reversals, more than there should be, mainly because of our hubris and our ability to be bamboozled by observational non-random studies. There’s a lot of those. There’s too many, in my opinion. And there’s a way around that, I think, which gets to the medical conservatism.

But, there are many, many more things that work in medicine, and it’s never been a better time to be a doctor or a patient. We have many drugs. We’ve transformed, for instance, congestive heart failure care. Patients with congestive heart failure used to die of their condition in a year or two, and now it’s not uncommon to take care of people with congestive heart failure 20 years. Cancer has been transformed: many cancers are chronic diseases. And so, I would not want to be pessimistic or nihilistic about medicine now, it’s wonderful.

But, what we learn from medical reversals, I think, Russ, is to have the humility to understand that many things don’t work; and before things get accepted, they should be evaluated in proper trials.

So, I’ve been a big advocate of rather than accepting some of these procedures, that we should have proper trials initially to show that there’s an effect. Now, people would push back a little bit on that and they would say, ‘John, you have to be careful with that because first generation devices, first generation surgeries are often not as effective; and procedures iterate, and you have to let the procedure iterate so that things get better.’

But, again, I really, really believe that we should be very careful about observational non-random things that we think make people better because we observe it, and we should have more of a culture of randomization.

For instance, I’ve been blessed to go to Denmark many times. And, when you walk through a hospital in Denmark, they’ll take you through, and probably a third or more of patients in a Danish hospital is in a randomized trial of one sort or another. And so, there’s patients getting care, but they’re randomized to one thing or another–maybe a different temperature in cooling in the ICU [intensive care unit], or maybe a different saline solution, or this or that. And they’re just constantly studying things. And so, they have a culture of randomization.

And then, we would be better off, I think, as a field if we did more of that. Even now, Russ, because we’ve done so many good things that we’re on the plateau phase of medicine where it’s harder to make big improvements. So, I think it’s even more important to study things before we accept them.

John Mandrola: Bravery is to randomize patients to sham surgery. I think it’s brave.

John Mandrola: No, that’s a really good point. And, I think that I’m not an orthopedic surgeon, but I know that many of the issues with translation of these trials to the clinic is that these were 146 patients or 148 patients–so these were very highly selected patients. And that’s a problem with clinical trials. So, a clinical trial, they highly select their patients; and that’s okay because you want to know for that group of patients, does the surgery work? does the medicine work? But then, patients we see in clinic come in many different varieties; and you’ll often see 1,000 patients screened, 140 enrolled. And, what about those other 800 to 900 patients?

So, I think a surgeon would say for this highly select group, it’s not beneficial. But there are many, many different kinds of meniscal injuries or knee injuries that are different from this study. And, I think that we have to be careful translating highly selective studies to a clinic where there’s all different manner of patients.

John Mandrola: It is true, it is true. And, I think an example in my world of cardiology is we have known for probably 15 to 20 years that doing a stent or fixing a coronary blockage doesn’t reduce heart attacks, doesn’t make someone live longer; and yet it’s very difficult to walk away from those things. And, it’s not just financial: it’s also your career is in fixing things.

John Mandrola: Vinay Prasad talks about this: it’s like the double-whammy. You get paid and you get good feelings about fixing these things.

But again, what strikes me about proper placebo-controlled trials is we learn humility. We learn that things get better, and we learn that not everything that we do works as well as we think it does. I just find it beautiful and fascinating.

John Mandrola: I couldn’t agree more, and I’ve heard you talk to Vinay about screening. And, one of the things that I don’t allow my doctor to do is a PSA [prostate-specific antigen]. Because I don’t even want to know. Because if I know the PSA, then I’m going to worry about it. Then it might lead to a biopsy, and then that might lead to whatever. It’s like if you give a mouse a cookie; and it’s best for me to not even start the cascade. And, it’s the same with colonoscopies; since there’s a study that shows it doesn’t change your longevity, I don’t do it; and I don’t have any regrets about it. And it is true, when you find out these things. So, I’m okay with it. I’m okay with not looking.

John Mandrola: Another example, since we’re talking about orthopedics, is I see tons of patients who the surgeon has told them they have ‘bone-on-bone.’ Bone-on-bone is one of the most harmful phrases–

John Mandrola: in medicine. And I’ll ask the patient, I’ll say, ‘Does your knee hurt?’ ‘No.’ Why are you having knee surgery? ‘I have bone-on-bone.’

And so, the notion that there’s bone-on-bone just gets into people’s heads and they say–I said, ‘You don’t really need to worry about it.’ They’re, like, ‘You’re not an orthopedic surgeon.’ I’m, like, ‘Well, if my knee didn’t hurt, I wouldn’t be having surgery.’

John Mandrola: We are just as guilty in cardiology. We have the ‘widow maker.’ So, if you have a widow-maker, you’re going to want to have a stent in it–

John Mandrola: even though studies show it doesn’t make a difference. When a patient comes in and they have a stress test that’s positive, we say they ‘failed’ the stress test. And so, we have all of these words and syntax that create a milieu where patients are more willing to have intervention.

John Mandrola: There are–I could show you empirical studies that show the exact same thing, and I try not to do this when we give local anesthetic. We say, ‘This isn’t going to hurt hardly at all. This is going to be fine.’ And, rather than–I used to say, when I was young, ‘This is going to feel like a bee sting,’ which is a terrible thing to say–

John Mandrola: because bees hurt like hell.

John Mandrola: Yeah, yeah.

John Mandrola: Well, number one, nothing has changed, and I think that as medicine has plateaued and as the push to do more and more, it’s made me even more medically conservative. Number one.

Number two, Russ, I have a small podcast in cardiology called This Week in Cardiology Podcast. It comes out every Friday. And, I hear from young people all over the world: when I go to a meeting, young people come up to me and they say, ‘I can’t say anything about this to my professors, but I just want you to know that you’ve influenced me.’ And, Andrew, and Adam, and Vinay, and I will get notes from young people who say that influences. So, I think it’s kind of a quiet, below-the-surface effect on young people, and it gives me great pleasure to hear that. [More to come, 52:28]