Robert A. Harrington, MD: Hi. This is Bob Harrington from Stanford University here on theheart.org | Medscape Cardiology. I hope you listened to my end-of-the-year wrap-up with Mike Gibson, where we covered some of the important clinical research findings that emerged over the past year.
I’ve asked Mike to stick around for a brief conversation about some of the learnings we had on the public health front. As part of that, we’re going to talk about COVID-19, but we’ll largely focus on cardiovascular trials and we may move into some other areas.
Again, I’m joined by my good friend from Boston, Dr Mike Gibson, professor of medicine at Harvard Medical School, interventional cardiologist at Beth Israel Deaconess, and director of the Baim Institute and academic research organization. Mike, thanks for sticking around for part two.
C. Michael Gibson, MD: Thanks for having me again.
Harrington: Let’s get into public health. As a clinical trialist, I’ve applauded some of the design features of these massive public health studies that have tried to get at questions that some have said, “Boy, that’s a tough one to tackle.”
First, the Salt Substitute and Stroke Study, which is a cluster randomization study in rural villages in China designed to see if a salt substitute might produce better cardiovascular outcomes. It’s getting at that question of whether too much salt, particularly sodium, is bad for us as a population. Boy, what a neat design. It was challenging, but these investigators pulled it off.
Gibson: I think COVID-19 highlights the need to keep it simple. I think we need studies that are patient-rich and not really data-poor, but with less of an emphasis on large amounts of data per patient. We do need many patients to answer these important questions. You and I sweated through randomized trials where we hoped to have enough events and enough power to answer the primary endpoint. That’s not an issue here. These are important public health issues that have a lot of debate surrounding them, and it’s great to see the question answered definitively.
Harrington: Putting my advocacy hat on, having a foundation of high-quality evidence to push the policy discussions is so important. Salt is a policy issue because it plays into things like how do you regulate food, what goes into food, and what kind of recommendations do you give regarding schoolkids and the food that’s being served at lunch, et cetera. We’ve done these trials to look at therapeutics, but public health is actually a bigger field where we can have a bigger impact.
Gibson: I agree. When we do trials, it’s often a selected group of patients. The thing I love about these trials is that they’re broader, more real-world, with more generalizable results. Now, I will say that different populations genetically may have different salt handling. The same kinds of questions should probably be asked in different groups of patients in the United States and elsewhere with different genetic makeups. This is an important answer to the question in this one population, but it’d be good to see the answer in other populations.
Blood Pressure Control: No Doctor Required
Harrington: Good caveat, Mike. One of the issues with salt is blood pressure. During the pandemic, as you know, we’ve seen a loss of blood pressure control in the US, and we’ve actually lost a fair bit of blood pressure control as a population. Because of that, we’re going to need some new strategies.
There are two strategies in recent trials that I enjoyed hearing about. One was done outside the United States in rural villages in China, where the so-called village doctors — think of them as community health workers — were instrumental in a randomized trial of controlling blood pressure better than usual care. Boy, that has some relevance to our country with community health workers and a workforce not dependent on physicians.
Gibson: The thing I want to emphasize is the generalizability of these kinds of results. You’re not talking about some fancy academic medical center in a wealthy ZIP code. We’re talking about real-world outcomes from a simple, real-world intervention that sometimes is not applied in 100% of people with 100% compliance. This is a public health strategy that can be applied broadly.
Harrington: Mike, your healthcare system took that concept of a workforce other than physicians as a way to manage blood pressure and to use the digital world in part through messaging and devices at home. Kudos to your colleagues at Mass General Brigham. I think it was around in 10,000 subjects, and they were able to demonstrate better blood pressure control. Do you want to comment on that one?
Gibson: That’s the concept of healthcare navigators. I have to be honest, some physicians were uneasy with this: “You’re taking away my patient. I want to manage them.” I can understand that, but it’s hard to argue with the results. A guidelines-based algorithmic approach with a large amount of touch by someone who may not be a physician yielded really good outcomes.
The same thing is seen on the statin and low-density lipoprotein (LDL) cholesterol side of things — it’s not something that requires the level of care of a physician. Compliance, adherence, and targets are maybe something that other healthcare providers can manage.
Harrington: A number of years ago, our colleagues at Kaiser showed that a pharmacy-directed program with nursing clinical support can provide good blood pressure control across, in that case, a captive audience based on their payer situation. For a system that’s a little more heterogeneous, like Mass General Brigham, which would be more akin to our heterogeneous payer population here at Stanford, I was really impressed that bringing these teams of people together at scale is doable. If I remember correctly, this had 10,000 patients involved.
Gibson: Scale has taken on a whole new meaning in the virtual and digital world. We’re trying to do the Heartline trial of up to 150,000 patients. When you go digital and virtual with apps, these things are very scalable, particularly if you have guideline-based algorithms that require less human intervention. However, human touch is still very important for getting in touch with these patients and really showing them that you care about them. No one cares how much you know, unless they know how much you care. I don’t want to cut out the healthcare and replace it with an algorithm, but I think that the two can work quite well together.
Harrington: My friend Abraham Verghese, the well-known author who’s our vice chair of education at Stanford, and Nigam Shah, who’s one of our informaticians, we wrote an editorial viewpoint in JAMA a few years ago, where we said, “What this computer needs is a physician.” We were making the point that it’s this intersection of human intelligence and artificial intelligence that’s needed. Healthcare, at the end of the day, is a personal relationship. You want it based on the best available data, but you also want it in the context of that individual patient’s values, their preferences, and understanding of their social context. I cannot agree with you more.
COVID-19 and Randomization
Harrington: Mike, let’s make the pivot to COVID-19. Last year, you and I talked often about COVID-19. We were coming to the end of the first 9 or 10 months of dealing with the pandemic. Now we’ve been through another year. Let’s discuss the clinical trials perspective. I don’t want to get into the vaccine myocarditis stuff. We can do that another time. Manesh Patel and I have had good conversations about that. Let’s focus on the trials, infrastructure, and apparatus. Our friends in the UK led by Martin Landray at Oxford — boy, did they do a public service for the globe.
Gibson: Yes, they did. We know what the questions are. We have so much experience in designing good case report forms. The case report forms don’t need to be rebuilt all the time. We need to do what they and other countries like Sweden have done. We need to build an infrastructure/backbone to not only help with this pandemic but also to get ready for the next one, so that when something like this happens, we can answer public health questions definitively, safely, and quickly in large numbers of patients.
It’s not a matter of can we do it; it’s a matter of do we have the political will to stop acting like individual clinical researchers and come together as a community.
Harrington: Well said, Mike. You and I were both involved in the Cardiovascular Clinical Trialists (CVCT) meeting, and I was in one of the sessions on COVID-19. I’m very involved with the National Heart, Lung, and Blood Institute (NHLBI) clinical trials in the COVID-19 space. I made the comment that in the US, we have this individual approach to clinical trials, and that’s why we see so many small trials that are hopeless in terms of being able to yield a real answer. It’s the individual “I” who gets rewarded or promoted. You and I got promoted because of our individual accomplishments, but what we need if we’re going to affect the public health is more of a group ethos, which certainly our friends at Oxford have tapped into.
Gibson: This isn’t a scientific question. It’s more of a geopolitical question. All of us have to look into our souls and ask, “Am I willing to give up some of my own personal autonomy, credit, and control to participate in something that’s bigger than myself to help the world?” I hope the answer will be yes.
Harrington: We hadn’t prepped for this one, so I’ll put you on the spot. If you were to give the NHLBI — or maybe the National Institutes of Health (NIH) more broadly — one piece of advice about their clinical trials, what would it be?
Gibson: Build a reusable backbone and stop reduplicating or reinventing the wheel over and over again. We can all do our own trials off that backbone if everyone wants their little individual trial. We’ve got a backbone: same endpoint, same definition, same everything. We need a good backbone.
Harrington: You mean a so-called platform to be able to do these things. It goes beyond the networks, which are very narrowly focused on certain diseases. What you’re really talking about is a national research infrastructure.
Gibson: In the world of subatomic particle physics, they come together at CERN. Everyone gets credit, and everyone’s going to become famous as being part of it.
Harrington: A group of physicists once told me that for your entire career, you can be the best at a piece of what those experiments need. You’re always fifth sixth, or seventh author, but you’re never the lead or the senior author. In their community, though, they know what that means.
Gibson: That’s what’s important — that your own community knows what it means.
Harrington: That’s a good piece of advice for a national infrastructure to be able to do the trials that need to get done in a timely way. When I hear Martin Landray say that 2 weeks into it, they were already beginning to think about randomizing patients, that’s extraordinary. It took the US much longer. We’ve done some good things, but nowhere near that national focus on contributing to the public good.
Gibson: You don’t have to be a wealthy nation like the UK or the US. There was a cluster randomized trial of hundreds of thousands of patients looking at masks vs no masks that was conducted in Bangladesh. If Bangladesh can do it, why can’t we do it? I think we know why: We are our own worst enemies.
Harrington: Yes. Renato Lopes from Duke, who’s Brazilian and who runs the Brazilian Clinical Research Institute, was talking about how that country came together to do some COVID-19 trials. Kudos to them. They had one of the cleverest trials of the of the pandemic — the angiotensin-converting enzyme (ACE) inhibitor withdrawal trial, which was designed to try to answer whether ACE inhibitors were dangerous in patients with heart disease in the COVID-19 era — and they did it quickly.
Gibson: They showed that ACE inhibitor withdrawal and all the worries about it really didn’t have much impact.
Harrington: Many of us were poring over the observational datasets, whereas they actually said, “Let’s do a randomized trial to get at the definitive answer.”
Gibson: Another thing we should learn from this is that observational data are no substitute for randomized trial data.
Harrington: Gregg Fonarow has a classic piece in JAMA Cardiology about there being no substitute for randomization. Observational data are good for generating hypotheses and learning about things like practice of care, patterns of care, et cetera, but when you want to show causal inference, you need to randomize.
Gibson: You may have seen on Twitter that I lined up the wastewater in Massachusetts and it lined up perfectly with COVID-19 cases and hospitalization, but you wouldn’t want to conclude that wastewater is causing the pandemic. You have to get causal inference right.
Harrington: Well, Mike, this has been a terrific discussion as we wrap up our 2021 overview. The first segment focused on the cardiology clinical trials. The second segment has dealt with some of the public health issues, including our response to COVID-19.
Mike is my good friend from Boston Harvard Medical School, Beth Israel Deaconess, and the Baim institute. Thanks, Mike, for joining us here on Medscape Cardiology.
Gibson: Bob, it’s always one of the highlights of my year to do this with you. Thanks for having me.
Harrington: Thanks for being here.
Bob Harrington, MD, is chair of medicine at Stanford University and current president of the American Heart Association. (The opinions expressed here are his and not those of the American Heart Association.) He cares deeply about the generation of evidence to guide clinical practice. He’s also an over-the-top Boston Red Sox fan.
Follow Bob Harrington on Twitter
Follow theheart.org | Medscape Cardiology on Twitter