Two Years Later, We Still Don’t Understand Long Covid. Why?

Depending on the data you look at, between 10 and 40 percent of people who get Covid will still have symptoms months later. For some, those symptoms will be modest. A cough, some fatigue. For others, they’ll be life-altering: Debilitating brain fog. Exhaustion. Cardiovascular problems. Blood clotting. This is what we call long Covid. It’s one term for a vast range of experiences, symptoms, outcomes. It’s one term that may be hiding a vast range of maladies and causes. So what do we actually know about long Covid? What don’t we know? And why don’t we know more than we do? Dr. Lekshmi Santhosh is an assistant professor at UCSF Medical Center, and the founder and medical director of UCSF’s long Covid and post-ICU clinic. Her clinic opened in May 2020 and was one of the first to focus on treating long Covid patients specifically. We discuss the wildly broad range of symptoms that can qualify as long Covid; the confusing overlaps between Covid symptoms and other diseases; whether age, race, sex and pre-existing conditions affect a person’s chances of contracting long Covid; why it’s so difficult to answer a seemingly simple question like, “How many people have gotten long Covid?”; what to make of a recent study that seemingly undermines the biological existence of long Covid; how worried we should be about correlations between Covid and medical disasters like heart attacks, strokes and abnormal blood clotting; and more. Mentioned: “Post–COVID Conditions Among Adult COVID-19 Survivors Aged 18–64 and ≥65 Years — United States, March 2020–November 2021” by Lara Bull-Otterson, Sarah Baca1, Sharon Saydah, Tegan K. Boehmer, Stacey Adjei, Simone Gray and Aaron M. Harris “Long COVID after breakthrough SARS-CoV-2 infection” by Ziyad Al-Aly, Benjamin Bowe and Yan Xie “A Longitudinal Study of COVID-19 Sequelae and Immunity: Baseline Findings” by Michael C. Sneller, C. Jason Liang, Adriana R. Marques, et al. “Positive Epstein–Barr virus detection in coronavirus disease 2019 (COVID-19) patients” by Ting Chen, Jiayi Song, Hongli Liu, Hongmei Zheng and Changzheng Chen “Risk factors and disease profile of post-vaccination SARS-CoV-2 infection in UK users of the COVID Symptom Study app” by Michela Antonelli, Rose S. Penfold, Jordi Merino, Carole H. Sudre, Erika Molteni, Sarah Berry, et al. “Understanding and Improving Recovery From COVID-19” by Aluko A. Hope “Markers of Immune Activation and Inflammation in Individuals With Postacute Sequelae of Severe Acute Respiratory Syndrome Coronavirus 2 Infection” by Michael J. Peluso, Scott Lu, Alex F. Tang, Matthew S. Durstenfeld, et al. Book Recommendations: In Shock by Dr. Rana Awdish Every Deep-Drawn Breath by Wes Ely Mountains Beyond Mountains by Tracy Kidder We're hiring a researcher! You can apply here or by visiting nytimes.wd5.myworkdayjobs.com/News Thoughts? Guest suggestions? Email us at [email protected]. You can find transcripts (posted midday) and more episodes of “The Ezra Klein Show” at nytimes.com/ezra-klein-podcast, and you can find Ezra on Twitter @ezraklein. Book recommendations from all our guests are listed at https://www.nytimes.com/article/ezra-klein-show-book-recs. “The Ezra Klein Show” is produced by Annie Galvin, Jeff Geld and Rogé Karma; fact-checking by Haylee Millikan and Michelle Harris; original music by Isaac Jones; mixing by Jeff Geld; audience strategy by Shannon Busta. Our executive producer is Irene Noguchi. Special thanks to Kristin Lin, Kristina Samulewski, Dr. Ziyad Al-Aly and Lauren Nichols.

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:06Ezra Klein?

  I'm Ezra Klein. This is the Ezra Klein Show.

• 0:21Ezra Klein?

  Before we begin today, we're looking for a researcher to join the show. This is going to be someone who gets the show deeply, who is both broad and obsessive in their interests, and who can read and synthesize and communicate a lot of information fast. That is really the core of this. Location-wise, New York is preferred, but we're open to remote work, and we're going to be closing the position in two weeks. So now's the time to apply. You can find the listing at nyco.com/careers, and there will be a link in the description for this episode. But for today, depending on the data you look at, between 10% and 40% of people who get COVID will still have inexplicable symptoms months later.

• 1:03Ezra Klein?

  For some people, that's going to be modest, a cough, some fatigue. For others, it can be life-altering, debilitating brain fog, constant exhaustion, cardiovascular problems, blood clotting. This is what we call long COVID. It's this one term for this vast range of experiences, of symptoms, of outcomes. This one term that may be hiding a lot of different conditions, maybe even a lot of different causes. We've known long COVID is a thing for years now, but to try to dive into the literature here is to realize very quickly how little we actually know. Different studies give wildly different estimates of who gets it, how often, and why.

• 1:48Ezra Klein?

  Some studies make it seem unbelievably widespread, others cast doubt on the whole phenomenon. We're still trying to understand the cause or the causes, and of course, we're still trying to figure out how to treat it. Dr. Lexby Santos is an assistant professor of pulmonary and critical care medicine at UCSF Medical Center, and she's a founder and medical director of UCSF's long COVID and post ICU clinic. That clinic opened in May 2020 and was one of the first to focus on treating long COVID patients specifically. And so Dr. Santos helps me here wade through what we know, and frustratingly, but really importantly, what we don't know about long COVID, what it's doing to us, and what we can do about it. As always, my email, if you have feedback,

• 2:32Ezra Klein?

  guest suggestions, recommendations of things we should read or think about, is [email protected]. Dr. Lakshmi Santosh, thank you so much for being on the show.

• 2:46Speaker 01

  Thanks so much for having me, Ezra.

• 2:48Ezra Klein?

  If you were to list every symptom you believe is credibly associated with long COVID, what would that list include?

• 2:56Speaker 01

  I think that list matches a lot of the medical literature out there right now. COVID is not just a disease of the lungs, even though as a pulmonologist, a lung doctor, of course, I love the lungs, but it really can be a disease of the whole body acutely in the short term as well as chronically. So it's not surprising that some people have acute impacts in multiple organ systems and some people have chronic impacts in multiple organ systems. One thing that we know for sure is that some people will have different, what we call phenotypes of long COVID. And I think that's one point that I make all the time to patients and other providers is that there is no one long COVID experience.

• 3:38Speaker 01

  Some people clearly have a more cardiopulmonary phenotype with a lot of shortness of breath, wheezing, coughing, chest pain after COVID. Some people tend to have neurological symptoms, that cognitive impairment that's been colloquially called brain fog. And some people argue that using that term actually kind of diminishes the severity of what people are experiencing. So cognitive impacts, mental health impacts, headaches, difficulties with concentrating, neuropathies, different sensation of nerves, these have all been seen after COVID.

• 4:16Ezra Klein?

  So there is such a huge range of symptoms here. So somebody comes to your clinic, how do you make the decision? If they are feeling terrible two months after getting COVID, if what you think they have is long COVID or just possibly something else, another virus, they've developed a chronic ailment they would have developed anyway. Given how many ways this can present, how do you make the judgment as a clinician? If somebody fits in your clinic or they've just gotten sick sometime after they had COVID?

• 4:47Speaker 01

  One thing that's key is that you have to exclude other causes. What I tell people when I'm teaching clinicians all the time, and I tell patients too, is all that long hauls is not COVID. And that I personally and my colleagues have diagnosed metastatic lung cancer, metastatic prostate cancer, inflammatory bowel disease like ulcerative colitis, thyroid issues, postpartum depression, hypersensitivity pneumonitis, which is basically an allergic reaction in the lungs. I have personally diagnosed all of these, sarcoidosis in patients

• 5:26Speaker 01

  who their clinicians or themselves thought it was all, quote, "just long COVID." And so part of our visit is listening deeply, taking a really thorough history, physical exam and lab work to say, could anything else explain this? Because like you said, this is such a common and prevalent condition. And many people have also been out of healthcare for two to three years, have not seen their primary care doctor, have been afraid to seek medical care, or it's their first entry to medical care. And so we absolutely have to do our due diligence as clinicians to make sure that we're not missing other diseases and health conditions

• 6:09Speaker 01

  and just falsely attributing them to COVID. So that is a key thing that I teach clinicians and patients is all that long hauls is not COVID. Look broadly, keep your thinking cap on. And I tell patients and families too, if new symptoms develop in your body that are new for you, that are different than your long COVID symptoms, talk to your healthcare provider and don't assume, oh gosh, this must be a new weird long COVID thing. Please get checked out because it might be due to something totally unrelated.

• 6:37Ezra Klein?

  I want to go back to something you said a minute ago that's interesting to me about the phenotypes of long COVID. And I've heard you talk about this before. I've tried to move it in my head to this idea of clusters. So as I understand it, and you can tell me if this is wrong, you're saying that people present with different clusters of seemingly related symptoms. So one long COVID patient, you know, who has what gets called brain fog is probably more likely to have a cluster of mental symptoms. But a patient who's having, you know, a lot of heart palpitations, maybe is going to have a cluster of cardiovascular symptoms. That, that it seems that there's some way in which the different symptoms

• 7:17Ezra Klein?

  connect to each other, such that there's a higher correlation between some in one person than in others. Is that right? Is that what you're saying there?

• 7:27Speaker 01

  So researchers who have studied huge data sets of long COVID have noticed these patterns or interrelationships between these symptoms or clusters of these symptoms. Does that mean that somebody with brain fog can't also have heart palpitations? No. But even when they look across large data sets of patients, large patient cohorts, they see that some people tend to have more cardio pulmonary symptoms. Some people tend to have more neurological symptoms. Some people can tend to have more gastrointestinal symptoms. Why is that? I think it goes back to when we look at research and studying these vast

• 8:09Speaker 01

  illnesses like long COVID or like sepsis, right, which is infection in your body, or respiratory failure. This is something that we're seeing more and more over a variety of different diseases, that this is a very heterogeneous group of diseases. That's why one size will not necessarily fit all when it comes to diagnosis or treatment. That seems to be another phenotype that we're seeing in long COVID, that some people seem to be having an overlap between chronic fatigue syndrome, myeloidic encephalitis, and long COVID. Some people have this PICCs phenotype of post-intensive care syndrome. And so studying each of those patients and those groups of patients with a focus on what is the underlying

• 8:53Speaker 01

  biology, what is the underlying mechanism, can help us more effectively treat long COVID.

• 8:59Ezra Klein?

  So I think this brings up a somewhat deep question here, which is, we keep speaking of a singular condition, like the literal tense we're using is a singular, that has a plurality of symptoms, long COVID, but it has multiple clusters or phenotypes or symptoms or presentations. Do you believe long COVID to be one thing? Do you believe that there is one disease or condition happening here with a lot of presentations? Or do you believe we have this categorical term that is hiding a potential multiplicity of underlying conditions?

• 9:37Speaker 01

  Definitely the latter. I think this is a grab bag term or framework that we have now. And as the research accelerates, I'm hopeful that we'll get more precise, more nuanced categorization of, you know, you could imagine with different subtypes or phenotypes, even in the field of acute respiratory distress syndrome. So that's basically respiratory failure or lung failure after any infection. Could that be influenza or COVID? That whole grab bag phenotype called ARDS, acute respiratory distress syndrome, is something that's been studied for decades. And we're still trying to figure out how to differentiate

• 10:22Speaker 01

  people with ARDS and how those biological differences might lead to different treatments. For now, if you're in the hospital, on a ventilator, intubated with COVID or influenza, there are some treatments that we use, regardless of what you have, to treat this syndrome of acute respiratory distress syndrome. I think similarly, long COVID is a framework. We don't know the underlying biology behind many of these symptoms just yet. We're still in the early days. So what are we doing now? We're treating the manifesting symptoms. And until our understanding evolves, that is what we're going to have to do.

• 11:05Speaker 01

  And so I tell patients and other providers, there is no one long COVID experience. Therefore, there is no one magic bullet, magic cure as of now, especially because different people probably have different biological mechanisms at play. And so for now, you'll see on the NIH website for long COVID, on the WHO website and the CDC website, that they say a couple of different principles. One, believe your patients. Our testing may not capture the depths of what's going on right now. We're still actively figuring out what are the best tests to deeply look at the different organs. Two, treat the symptoms that people are having. We know how to treat headaches.

• 11:49Speaker 01

  We have some tools in our toolbox to treat chest pain and shortness of breath. We have some rehabilitation tools to treat cognitive impairment and fatigue. And so right now, what we're using is a symptom directed approach, which is definitely not sufficient, which is definitely frustrating to patients and advocates and clinicians and researchers. But until our biological understanding evolves, only then will we be able to truly deeply grok and understand these phenotypes and then have targeted therapies towards them.

• 12:23Ezra Klein?

  So at this point, given all the literature we have, these big data studies, these studies following individual people over periods of time, what do you believe the incidence to be of long COVID after getting COVID? What percent of people who get COVID, would you then expect to get long COVID, knowing nothing else?

• 12:45Speaker 01

  As you know, there are so many studies that have ranged in prevalence, ranging from 10% to 30-ish percent and everything in between. Some studies have higher numbers, some studies have lower numbers. Including meta-analyses, different studies, prospective, retrospective, big data studies, et cetera. And I was recently on a policy committee call with Dr. Paluso, Dr. Deeks and others, where that was a central question being debated by policymakers. What is the quote, "true incidence"? I would argue that these minor quibbles about

• 13:24Speaker 01

  the true incidence seem to fall apart in the face of massive multimillion infections, which is what we're experiencing right now. So 10% of multiple, multiple millions, or 30% of multiple, multiple millions is still a huge public health crisis that we're experiencing. And we're seeing just the tip of the iceberg in terms of people who are able to seek care, seeks subspecialty care. If anything, I worry about underdiagnosis and worsening disparities in populations who have difficulties accessing healthcare.

• 14:00Ezra Klein?

  You're, you're deep in this. You've looked at a lot of these studies. You see people coming in. You know what the flow has been. If it's gone way up or has not gone up that much after Omicron. Do you find 10% or 30% more intuitively credible to you?

• 14:15Speaker 01

  So the bottom line is, the long COVID risk is high. I think the number that the CDC released lately of quote, "as many as one in five," which kind of is right down the middle of that 10% and 30%, feels right experientially. Does that mean that every single one in five is going to look a particular way? No. I think, again, there's a huge diversity of experiences.

• 14:43Ezra Klein?

  Let me ask you about that one in five number, because I was struck by that CDC study. And I wanted to actually get at something technical about it, which is, I think there's what you might call technical long COVID, which gets you inclusion into the criteria for long COVID in one of these studies. And some of them, if you're experiencing symptoms four weeks, or six weeks, or eight weeks. And then I think there's this question of folk long COVID, maybe, which isn't to say it's not true, it's very true. But the super severe long COVID, the life destroying long COVID, the, the former marathoner who can't walk down the block, or the academic who has so much brain fog, they can barely do any work. When we say one in five there, what percentage of that is the kind of severe

• 15:28Ezra Klein?

  long COVID that we read about a lot in the press? And what percentage of that is people are still struggling with some symptomatic dimensions of an infection that seems to be clearing up? Like, how do you think about that distinction?

• 15:43Speaker 01

  I think the one in five distinction is broad. It's inclusive of both many patients in our clinic have asthma or COPD. And then after COVID infection, their asthma or COPD is not as easy to control. They flare up more than they used to. They cough more than they used to. They wheeze more than they used to. So for some people, they're having those respiratory symptoms. For some people, it's persistent chest pain. For some people, it's headaches and neuropathy. And for some people, it is, as you said, it's not folk at all. It's disabling, life threatening, totally disheartening and incapacitating chronic fatigue syndrome, long COVID overlap, right?

• 16:22Speaker 01

  Sometimes when I feel discouraged by science communicators, articles in the media, et cetera, trying to almost preemptively exculpate or absolve our whole society by saying, well, technically, that risk of the super severe, horrible disabling fatigue is so, so low. So it doesn't matter. We'll take that chance. I think that a lot of, quote, back of the envelope calculation articles seem to be written with a framing of preemptive, well, that is super rare, seems bad, but we don't have to worry about it. And I worry about that framing.

• 17:06Speaker 01

  Again, if anything, I worry about a lot of people not attributing some symptoms to COVID, whereas many people worry about the converse or people over attributing symptoms to COVID. I think that there is a balance, right? There is a balanced perspective that has to be taken. Of course, the classic, quote, rebuttal of some of these big data articles is, of course, if you saw your doctor the next week and said you were constipated, would that big data set analysis say the constipation is because of COVID? So that's a classic rebuttal to these big data type analyses.

• 17:47Speaker 01

  On the flip side of that, there are definitely symptoms that we don't necessarily understand, we can't necessarily link them to long COVID, and we wonder about that. I'll give you an example from the hospital, from the intensive care unit where I work and see this acutely. In the hospital, we've definitely seen where COVID has this abnormal balance of clotting and bleeding, where we've seen some rare diagnoses, rare disorders of anticoagulation, done a thorough workup and can't really come up with a clear cause. So it ends up being, as we call it, idiopathic, which means we don't know the real cause. And we wonder, could long COVID have caused this rare thing? We don't know, we can't prove it.

• 18:30Speaker 01

  We wonder about COVID, quote, "unmasking." That's a concept we talk about a lot. Perhaps you had a little bit of asthma or COPD before, and now COVID has unmasked that, or accentuated that, or exacerbated that, or brought that out. We're seeing now with studies by Dr. Michelle Mange and others about the cognitive effects of COVID, even a mild respiratory infection. It's disturbing, we don't want to hear it, but seeing how she clearly has done compelling biological research comparing brain tissue in post-influenza studies and comparing that to post-COVID brain tissue, those are compelling, well-done research studies that show a long-term

• 19:13Speaker 01

  impact even after a mild respiratory infection on the brain. And so the one in five burden of disease feels right to me because of the huge heterogeneity in symptoms that I see after COVID. And I think, again, it's this huge spectrum from mild worsening of current symptoms to truly life-altering disabling COVID, and everything in between. And I think when we, as science communicators, journalists, scientists, policymakers,

• 19:45Speaker 01

  Speak about it in these black and white terms of either the binary of you died of COVID or you survived COVID, that's a false binary. Similarly, I think it's a false binary to say you can get COVID and have zero symptoms afterwards, or you will get long COVID and you'll be in horribly debilitated shape for the rest of your life, too. So I think that's a false binary as well.

• 20:11Ezra Klein?

  So I want to go back to something you said a minute ago about the way in which this conversation gets used to absolve society of prospectively worrying about, you know, what can happen, right? Maybe if long COVID, severe long COVID is really rare, eh, you know, a lot of rare things happen, they're very bad, car accidents are bad, people don't need to worry about it. And I see this a little bit differently. One frustration I have personally with the COVID debate, for lack of a better term for it, is that we keep speaking about it in almost the same way we were two years ago, as if COVID is something that is coming, as opposed to something that has come. I've seen estimates that as many as 70% of Americans have had COVID.

• 20:54Speaker 01

  Right, once, have had it once.

• 20:56Ezra Klein?

  Have had it once. And people get it, you know, more, more often. I already know people who have had it twice. But nevertheless, a world in which you have 20% of 70% of the country in a long COVID scenario is a really different world, potentially, than the one people realize they live in. And I've frankly been struggling with trying to understand if I even think that's the world we're in, because basically, everybody I know has gotten COVID. And I can name a couple really tough cases, but nothing that looks to me like 20% of people still exhibiting or telling me about noticeable symptoms. But on the other hand, as you say, they may not attribute to them, they might not tell

• 21:37Ezra Klein?

  me. But either way, I don't really think this is a conversation that is still as prospective as people seem to think it is. It seems to me to be much more conversation about where is the health of the country, really? Because if the health of the country is 30% of people, you know, or 20% of people, rather, are going to be in a, like a post-COVID scenario that is not better, but is actually still sick, that's a profound public health crisis. And at the other hand, if it's not really that, then, you know, that's also worth knowing.

• 22:11Speaker 01

  You're exactly right that it is worrisome, the impact that this is going to have in our public health. I will also point out that, that you and I, right, are people, people who have certain privileges. The people that you and I know who've had COVID may be having a really different experience than many individuals in this country who don't have access to regular primary care, don't have access to early Paxlovid, haven't had three shots necessarily, haven't been plagued by misinformation about vaccines or about hospitals or about ventilation.

• 22:52Speaker 01

  And so I do hear you that many people that you and I might know, people who have some privileges in our society, might be getting off, thankfully, okay.

• 23:07Ezra Klein?

  Does long COVID track those kinds of demographic questions? Because one thing I've been, I'd heard about it earlier was that it didn't. But I've also heard people say, in fact, that as we've gotten more data, it does. So would you say that from what you've seen, long COVID, sort of like COVID itself, correlates to lower income, space in society, social status, access to the care system, et cetera?

• 23:32Speaker 01

  So I think that the social stratification, systemic racism, classism, and the disproportionate impact of COVID on people of color, particularly Black people, Latinx people, has been seen at every facet of COVID, right? And so absolutely, social stratification, systemic racism, impact acute COVID, exposure to COVID, as well as long-term outcomes of COVID. I think that these larger data sets, like you said, have the opportunity where we can look and say, does this cut across races, classes, sexes, et cetera?

• 24:14Speaker 01

  And so we need better research in communities of color to make sure that they are not underrepresented in clinical trials, in trials of pathogenesis, in all the ongoing research related to long COVID. And we need to, research community needs to actually overcome those barriers to access and make sure that communities of color are represented in research. I will say that in our clinic, because our clinic's kind of catchment area or auto referrals come from people who are hospitalized at our hospital, we see, and we've looked at our metrics to see, we see people across the spectrum of race, ethnicity, age, sex, and

• 24:57Speaker 01

  insurance status. And I will tell you that long COVID does not respect boundaries of age, sex, race, ethnicity, or socioeconomic status. And I've had in my clinic, healthcare workers, physicians, lawyers, executives, who are suffering from intense symptoms, as well as workers in the field, workers in airports, workers in restaurants, experiencing nearly identical symptoms. And that is really profound and humbling. I think my point is that in order to truly answer that question, I worry that communities of color who've been underrepresented in many research studies are similarly underrepresented in studies

• 25:40Speaker 01

  that are answering this fundamental prevalence question, right? I worry about that huge impact and that our communities of color or disadvantaged communities may not have access to long COVID diagnosis, research, or treatment.

• 25:54Ezra Klein?

  Are there any other demographic characteristics or preexisting conditions that seem to predict, in your experience, how often people get long COVID, or how severe the cases that they get are?

• 26:06Speaker 01

  This is a great question that's hard to interpret because dominant strains have largely correlated with the time of the year or the surge that we've been in. So, like I mentioned before, we've seen a lot of patients with severe symptoms from that very first wave, as well as from the Delta wave. And then with the Omicron wave, the first Omicron wave, rather, in January, We noticed that some of the symptoms were different, right? Some people were not having as much anosmia, loss of smell and taste. Now with this latest surge, it seems to be kind of coming back a little bit. And so it's interesting that even in the acute surges, we've noticed that some symptoms have been more dominant than others with different strains.

• 26:50Speaker 01

  So it is totally probable that with that, some long COVID symptoms might be different than others, just as they are in the acute setting. So it makes total biological sense. And again, you have to factor in the impact of the vaccines, that the breakthrough infections, the breakthrough long COVID, how might that look a little bit different than the initial unvaccinated Delta period or that first spring of 2020 wave? And so I think we're still teasing all of that out. Often, when you get a COVID test, as you know, the average person, even in a research study, they're not doing that molecular level identification of what strain it is. So most people, we say, I assume this is Omicron BA1

• 27:34Speaker 01

  because of the time of the year that you got it, but we don't know for sure. So that's still being teased out. It's still a little bit earlier to tell, but it's absolutely plausible that there could be differences because we've seen differences in the acute presentation as well.

• 28:05Ezra Klein?

  I've heard it argued that there was quite a bit of protection from vaccinations and that a lot of the severe cases of long COVID we've seen came fairly early in the first wave before vaccination really took hold. Is that something you've seen or has the flow to your clinic, both in terms of numbers and severity, simply tracked case numbers pretty steadily?

• 28:25Speaker 01

  So there was a big study that came out last month that looked at breakthrough infections and long COVID. And that's that study that found only a 15% protective effect in this American cohort. There was another study, a UK study, that looked to be more like a 40 to 50% reduction or protective effect. Two take homes, I would say, is that, yes, the vaccines do seem to attenuate the number and severity of long COVID infections. And absolutely, people are still getting long COVID, including severe long COVID, in the post-vaccine era.

• 29:06Speaker 01

  I will say that our clinic numbers, because they're auto referrals from our hospital, for one, in addition to referrals from primary care and community referrals, our clinic has seen a lot of patients who have very severe symptoms from the earliest way, from New York, March 2020. And we still have a waitlist of several months, as people are still coming in with post-vaccination, post-breakthrough infection, long COVID as well. So I think it's a bit of both. I think that, yes, the vaccines are helping somewhat. Again, that makes you wonder, and scientists are studying this, what is the interaction between viral load and your immune system? What is that interaction?

• 29:50Speaker 01

  How does that manifest with long COVID symptoms? And so that is being actively studied. So we hypothesize, many people hypothesize, and this is being studied, that perhaps the higher viral load from that time in March 2020, when people were unmasked, unvaccinated, do those patients have higher burden of long COVID symptoms ever since then? And hopefully, in a society where more people are masked, more people are vaccinated,

• 30:17Speaker 01

  We're seeing a little bit less, hopefully, long COVID burden, but still people are definitely getting it, including severe cases. So it's not either or, it really is both.

• 30:28Ezra Klein?

  Why do you think so much of the research here ends up so all over the place? And to give an example, we've talked about this CDC study that came out in May that found that one in five adult COVID patients may develop long COVID. It found that COVID-19 survivors have twice the risk for developing pulmonary embolism or respiratory conditions, really, really frightening study. And then also in May comes out this NIH study finding, quote, "no evidence of persistent viral infection, autoimmunity, or abnormal immune activation in patients reporting long COVID, finds an association between long COVID and underlying anxiety." So is this study that seems like it's downplaying it a bit, that isn't able to find as much of a reservoir here? And I've seen this a bunch of times where, you know, I turn on the news in the morning

• 31:13Ezra Klein?

  and it's like one day it's this terrible study about long COVID and the next day saying, well, you know, it's good, there's clearly something there, but it may not be as prevalent, some of it might be psychosomatic. Why is the research ending up so all over the place? How do you, how do you read that?

• 31:28Speaker 01

  The NIH study that you're referring to is the study by Dr. Sneller and colleagues, published in Annals of Internal Medicine. And that paper, when many people read it, the accompanying editorial to that paper was written by Dr. Aluko Hope, who is a pulmonologist, an intensivist, and a long COVID clinic director as well. And he wrote in that editorial that these results were not surprising to many people who have been treating patients with long COVID. And what he means by that is that so many of us have taken care of patients where the repertoire of lab studies or imaging studies that we currently have

• 32:09Speaker 01

  Is not capturing the vast array of symptoms or the biology underlying this. So when you come up with a quote negative study, like the Snell and paper, there's so many different questions to ask, you know, and I think Dr. Aluka Hope raises these in his editorial. This is a relatively small sample. It's about a little bit less than 200 patients. The control group was not age matched or sex matched. That has to be considered. You wonder again about the bias in the population who participated in the study. Was this population a little bit of a healthier population than many of the most severe symptoms

• 32:53Speaker 01

  experienced by patients with long COVID? So to be in the study, the control patients and the patients with long COVID had to undergo these battery of serial tests, blood tests, as well as organ function tests, which again, we do a lot in our clinics. And you wonder again, if the people who have enough energy to do all those tests and participate in this trial, if that's truly representative of that most severe spectrum of population. So you wonder about that. You wonder about the age and sex match controls. And then you also wonder about, are there more sophisticated tests, more sophisticated imaging techniques, different autoimmune studies or cytokine profile studies that were not really

• 33:37Speaker 01

  capturing these patients' symptoms? In contrast, you have that CDC study that you mentioned, the one in five. You have the Journal of Clinical Investigations paper by Dr. Paluso, Dr. Durst and Phil and colleagues that I mentioned about the autoimmune profile of patients with the predominantly cardiovascular impact. So is this so hard to navigate as scientists, clinicians, policymakers, journalists, science communicators? Absolutely. This is a difficult web of studies to navigate. And again, I think it goes back to that fundamental heterogeneity of the patient population.

• 34:16Speaker 01

  We do a disservice as science communicators, journalists, clinicians, talking to media, when we say, "This study clearly shows that this is psychosomatic." I think that is absolutely a flawed conclusion. And the authors would similarly argue to say, "That's not what we meant by the study. That was not our conclusion either." Long COVID is absolutely real. It is not psychosomatic. And I always tell our patients in our clinic that it is understandable why people who would have symptoms of anxiety, depression, PTSD, after going through these horrible health experiences that they've experiencing, and having people doubt their symptoms on top of that.

• 34:57Speaker 01

  What we tell people when they come to our clinic is, again, rejecting that false dualism of mind versus body. But we say, "Let's treat everything you have. Let's treat your shortness of breath, your dyspnea. Let's treat your depression. Let's get you into a structured rehab program for your, for your physical function issues, for your weakness. And let's treat everything holistically." And so I do think that that study has often been misinterpreted as that shorthand of, "Oh, you know, they found a high level of anxiety, therefore it must be just anxiety." And Dr.

• 35:39Speaker 01

  Luca Hope's editorial, and hopefully these issues that I've mentioned, add some nuance to that simplistic interpretation, which I think is flawed, and helps us see that this is one dataset. And can we learn from this dataset? Yes. Is this dataset applicable or generalizable to all patients with long COVID? Not necessarily. And so when we paint all of these patient populations with one brush stroke and try to apply a small study to that hugely heterogeneous population, that's, I think, where we get into trouble as science communicators.

• 36:13Ezra Klein?

  How much better have we gotten on understanding the biology around this? Because I've heard different theories for what long COVID is, or maybe there are different long COVIDs too, but you hear a kind of extended immune response theory. You hear a theory that maybe the virus is persisting in some people for longer than we think. I've heard theories about it activating latent Epstein-Barr virus that people have. What to you are the dominant or most probable suspects for what is causing long COVID in people?

• 36:49Speaker 01

  You outlined a bunch of hypotheses that we have lots of tantalizing leads about. And so they're more than theories in that we actually, there's multiple studies supporting each of these different hypotheses. The trouble is, A, different patients may have different mechanisms at play. And B, how do we treat these mechanisms, which will only all be understood as we better understand the biology? So there is evidence of viral persistence in some studies. It's been demonstrated in the gut. It's been demonstrated in other organ tissues. That's one hypothesis that's actively being tested where there's a lot of evidence towards that. You would hypothesize that for that group of patients who have that mechanism,

• 37:32Speaker 01

  might there be a role for antivirals like Paxlovid? You know, there have been small case series, including one out of our institution that showed that some patients' long COVID symptoms got better with Paxlovid, adding credence to that theory. But again, one size doesn't fit all, and we know that that's not true for all patients. I've had patients with long COVID who've gotten reinfected. I've had some who had benefit from Paxlovid and others who didn't, again, showing the heterogeneity of this condition. The second theory that you talked about was Haywire autoimmune response. There are so many different studies from the basic science literature to patient level

• 38:11Speaker 01

  studies, looking at the abnormal inflammatory response, abnormal cytokines, basically, again, that interaction between the host, the human, their immune system and the virus, looking at different subunits or receptor binding domains of the virus and seeing how that dysregulation may be impacting long COVID symptoms. So again, lots of studies in this area, lots more than we knew from a year ago. Is it at the place where right now we have several therapeutics in the pipeline? Unfortunately not. So still an active area of research. The other area of research is we get some information from autopsy studies, what happens when people pass away of COVID.

• 38:56Speaker 01

  But again, acknowledging that there's selection bias in that those people were so severely ill that they died of complications of COVID. But we have a lot of information from autopsy studies to see how does COVID, the virus, attack the different organ systems in the body? So we have a lot more information than we did a year ago. What's frustrating to patients and clinicians alike is that that biological information that we have hasn't been translated to therapeutics yet. But we're getting there. We're making that progress inch by inch at a historically breakneck pace of research and of course, not yet fast enough to deal with these huge numbers of people who are suffering from these conditions.

• 39:37Ezra Klein?

  One of the ambiguities in the term long COVID has always been how long? How long are we talking about? And obviously, COVID is still a relatively new disease, or at least the novel coronavirus is still relatively new, but we have now a couple of years of experience. And so I wanted to ask if you see any clarity there that we didn't have before. When somebody comes into your clinic, what is the likelihood you're seeing them and they have not improved very much six months later and a year later?

• 40:10Speaker 01

  One of these things that you mentioned is that when did that definition of how long is long even come up? And the first whole year of the pandemic, 2020 to 2021, you have all these studies done with different definitions. In mid 2021, there came a good WHO consensus definition that shows that people looking at it, trying to put some guardrails on that to help us conduct research when we're all speaking the same language, when we all have a shared mental model. And they said, symptoms for at least two months and symptoms that have not been attributed to any other cause. And so I like that WHO definition because at least it gets us all on the same page of that symptoms for at least two months without any other cause identified.

• 40:55Speaker 01

  But that definition came late, about one and a half years into the pandemic. And so when I'm meeting people now, if they're super early, if they're still in that, I had COVID two weeks ago, I'm still coughing, setting some expectations and some reassurance that you're still in the active inflammatory period of this illness. And let's treat your symptoms, let's treat that cough, let's watch you closely, and let's see you back and see how you're doing. So that's something that we emphasize. And then similarly, if we see somebody who's been having symptoms for one to two years and they're coming to us for a second or third opinion, and they say, how long is this going to last?

• 41:40Speaker 01

  Is this going to be my life forever? Right? Those questions are so hard to answer. We don't have those crystal balls. And I do firmly believe that, as you've done in our conversation today, that if you talk openly about what's uncertain, what's not, and you combine that frank discussion of uncertainty with what can we do now to treat your symptoms, that helps patients and families navigate this quagmire of uncertainty. People end up leaving the clinic visit saying, I have more hope than when I started here. At least I know what's to expect and what we can deal with.

• 42:16Ezra Klein?

  That's always, I think, a really powerful feeling to have some hope when you leave the doctors. For most patients you see, a year later, have they improved?

• 42:24Speaker 01

  Yes. The large majority of our patients that we've seen, of our 400-ish patients that we've seen, by the one-year mark, they have improved. Again, reflecting the vast differences in the phenotypes. I've seen people who were in the ICU on death store, who were in the ICU for months, on a tracheostomy on a mechanical ventilator for a couple of months, come out on home oxygen and then slowly but surely recover their lung tissue. We've also seen people in that same situation, ICU for months, where now they're at this horrible plateau, where they're on a couple of liters of oxygen. They're not quite sick enough to get a lung transplant, but they're not healthy enough,

• 43:07Speaker 01

  not even anywhere close to where they used to be. I've seen people make a full recovery, complete. They're back to running, athletics, their jobs. And also people who have the debilitating, life-altering chronic fatigue syndrome, long COVID overlap. I would say, again, the majority of people that we're seeing have improved over time. And I'm hopeful that as this pace of scientific research escalates and accelerates, and as we get more dedicated research into these symptom clusters and phenotypes, I hope that we find new tools to offer the group of patients who has been, their lives have been utterly ruined and altered.

• 43:50Speaker 01

  And so people are working on this at a fast pace around the clock. They're passionately working on this. My colleagues, Dr. Paluso and Dr. Deeks, Dr. Kiko Iwasaki at Yale. There are so many investigators around the country who are working to bring hope and alleviate suffering and understand the biology behind this. So majority of people are getting better and we have to keep searching for answers for the people who haven't gotten better yet.

• 44:36Ezra Klein?

  We've talked a lot so far about symptomatic long COVID. But there's this other category of study, which you gestured towards earlier, that unnerves me quite a lot, which maybe we'll call non-symptomatic long COVID, although some of it can become violently symptomatic all at once. But so there are these studies or a class of studies suggesting neurological damage, even in mild cases, even when people don't realize it's happened. I've seen studies suggesting routine IQ loss in people who've had COVID. On the other side, there are these studies suggesting that a lot of very, very dangerous medical events might be becoming likelier in the aftermath of long COVID.

• 45:17Ezra Klein?

  Dr. Ziad Alali and a bunch of colleagues did studies that found higher long-term rates of heart attack, of stroke, of diabetes, of blood clots after COVID. You might not have any symptoms and then you have a heart attack. And as you said in an earlier answer, maybe we don't connect that to COVID, but maybe there is a connection in the way there's a connection between stronger hurricanes and climate change. How do you think about this dimension of long-term effects of COVID? How do you read this research, which many of us would very much like to shut away in a dark room and never think about ever again?

• 45:53Speaker 01

  Yes, you are right that this research is unnerving and important. I think another mechanism that we didn't talk about is endothelial dysfunction. Basically, are the cells lining the blood vessels, becoming abnormal from long COVID? This is a hugely worrisome field of research about the increased risk that Dr. Alali's study found of cardiac events up to one year after COVID infection, including mild infection. And we wonder if that might be related to endothelial cell damage or vascular damage. Very, very highly plausible because again, even in the acute setting with COVID in the ICU, COVID in the hospital, we've known for a long time that these folks tend

• 46:38Speaker 01

  to have blood clots at higher rates. So this tendency of people with COVID to both clot abnormally and bleed abnormally has been seen in the acute setting and in the chronic setting for a while. And it's very worrisome, especially when at baseline in America and in the world, our cardiovascular health is not great. And so having this additional cardiac risk factor of prior COVID infection is something that seriously needs to be studied. Again, with parallels to, just like I've talked about parallels with research in other areas of science, there's parallels to research in HIV, of course, where we realized with

• 47:19Speaker 01

  study over time that HIV similarly seemed to impact atherosclerosis plaque in the heart through similarly endothelial dysfunction. So that's a hypothesis that's very plausible. We've seen it with other viruses. That's what we're worried about is that, like you said, you could have symptomatic long COVID, it could be mild or severe. And it's also worrisome that even if you don't have symptoms right now, perhaps there's something brewing in your vasculature that can be a risk factor for future heart attacks and strokes that needs to be actively managed. And so this area of research is new, it's concerning, and there are a lot of people working on this.

• 48:04Ezra Klein?

  So let me try, because it's such a depressing thing to think about, let me try my one optimistic take on this on you. And you can tell me if you think it's credible or if you think I'm getting something obviously wrong. One thing I try to do when I see this data is test it against what we're seeing in aggregate data. Because I think the implicit prediction being made here is that population level health is getting much worse, right? A world where lots of people are having mortality level events because of the damage COVID has done is a world where we should be seeing a lot of excess death. And in general, looking at the CDC dashboards, you see a lot of excess death when COVID counts are very high.

• 48:48Ezra Klein?

  And then as they get lower and more recently, the excess death data goes away. So, you know, looking at the more recent dashboards, you know, through March, through April, through May, which is when we have data for, there isn't really an excess death count nationally. Things sort of look like what we'd expect them to look like. Should that give me any comfort that even if there are these connections or correlations we're drawing, that they're not so large that they seem to be affecting population level health at the population level data sets?

• 49:20Speaker 01

  I think it's too early to tell, unfortunately. Dr. Jeremy Faust, who's an ER doc and health policy researcher, published a paper on excess mortality during the Omicron surge in January as well. And so the current surge that we're in, I don't think that we'll see that impact right away. And so kind of like the analogy that you made earlier with climate change, just because the weather's okay right now, I don't think that should reassure us necessarily about the pace of where the future is going in a scenario, of course, of unmitigated spread as well.

• 50:01Speaker 01

  My optimistic take on it is this, we have amazing teams of researchers and scientists looking at this from so many different angles and hypotheses and animal models and immune interactions and longitudinal cohort studies from so many different lenses, that I'm hopeful that we are going to get answers and actionable trials sooner than later. I still worry that in a scenario where millions are getting infected with unmitigated spread, I still worry about the population level impact. And even though it's not raining today, or it's not overly hot today, that doesn't falsely reassure me about where the temperature is going to be or where the excess mortality

• 50:46Speaker 01

  is going to be six months from now or one year from now, especially given that January Omicron data where the public health messaging was, this is mild. But when we looked at the excess death, it was actually higher.

• 50:59Ezra Klein?

  Let me ask you then about another, I don't want to call this one optimistic, but potentially something that could come out of this that is positive, which is something you hear from the chronic disease community, particularly people with a lot of these more mysterious conditions like chronic fatigue syndrome, or MS, or a lot of things that seem autoimmune oriented, is that there have long been these conditions people get, they are potentially, though not obviously, or certainly related to viruses. It's been very hard to get doctors to take them seriously. They've taken them more seriously in recent years, but it took a long time. It's been hard to get enough research done on them. And then COVID hit, and because it was such a huge wave then of long COVID,

• 51:44Ezra Klein?

  all of a sudden, there was a lot of attention. And that that attention is also filtering backwards, that there's just more going into studying a lot of these conditions that might be long-term consequences of other diseases. And so some of the work going into long COVID, and also just some of the cultural change towards taking patient reports very seriously, is going to also, you know, help a lot. The much larger percentage of people with a lot of chronic ailment than we, than we always realize get help that the medical system, the research system, has not always given them. Do you think this is leading to some kind of cultural change or some kind of research change that will actually reverberate beyond COVID in a, in a useful way?

• 52:28Speaker 01

  I agree with you completely that I hope we are listening now more than ever, as we should have been all along, to our patient advocates in the chronic disease community, in the disability advocacy community, and the rare diseases community. And I hope, as you said, that we are getting lessons learned that spill over, biological insights that spill over from long COVID to other conditions, the link towards MS and virus was a huge scientific discovery as well.

• 53:08Speaker 01

  I think you said the global, hopefully more, understanding that we have to believe patients, we have to take it seriously. We cannot label these symptoms as, quote, "all in your head." We cannot label them as psychosomatic or not real. And having that humility in the medical and research community that perhaps this is something that we just don't understand yet. And we've seen that time and time again, right? If you look at the old reports of women being labeled with hysteria, and you can draw that line to where we are today and say, now we have this opportunity to listen to our patients, to study the biological underpinnings of these illnesses, and to,

• 53:53Speaker 01

  hopefully, change our society for the better to being a more inclusive space in general, to help all people, right? And so I think that's another key concept that these communities have taught us, is that when you have protections for all people, everyone benefits. And I hope that we can also use these insights to advocate for better protections for all, including, you know, paid sick leave, which we're one of the few countries in the world that doesn't have paid family leave, paid sick leave, et cetera. Thinking about, can we take this opportunity to clean our indoor air better? Can we take this opportunity to build up our safety net systems to support our most

• 54:37Speaker 01

  vulnerable patients that will ultimately positively improve the health of the whole country?

• 54:41Ezra Klein?

  I think that's a good place to end. So always our final question, what are three books you'd recommend to the audience?

• 54:46Speaker 01

  One is called In Shock by Dr. Rana Awdish. It really talks about Dr. Awdish's journey on both sides of the drape, being both a patient and a physician, and what she has learned from that experience and her lessons to all of us. She's a beautiful writer. She is an astute thinker and- Her words are more poetry than prose. It's an amazing book. You have to check it out. The next one in a similar vein is Every Deep Drawn Breath by Dr. Wes Ely. Dr. Ely is a leading researcher in delirium, in the complications of critical illness and post-intensive care syndrome, and now has been a huge advocate for people with long COVID.

• 55:30Speaker 01

  And in this book, he talks about his whole personal journey and the journey of research in the field. And I recommend this book to patients and families of patients who were in the IC or people who have long COVID. There's even a chapter on resources for patients and families at the end. It's a beautifully written book and another wonderful clinician, researcher, advocate and writer. And the last book, just because to honor his memory, is Dr. Paul Farmer, Mountains Beyond Mountains. That book is just really inspired a whole generation of doctors to serve and to always think about the global health impacts of what we're doing and to not forget about,

• 56:13Speaker 01

  in all this work that we're doing, to serve, to work towards justice, to try to get to global vaccine equity for COVID and for other conditions, and to honor Dr. Farmer's memory that we can all strive to be more like Dr. Farmer.

• 56:28Ezra Klein?

  I love that book, too. I don't think there's anything you can read that is more inspiring than that one. Dr. Lexi Santos, thank you very much.

• 56:35Speaker 01

  Thanks so much, Ezra. Have a great day.

• 56:49Ezra Klein?

  The Ezra Klein Show is produced by Annie Galvin, Jeff Geld and Rojai Karma. Fact checking by Michelle Harris and Haley Milliken. Original music by Isaac Jones. Mixing and engineering by Jeff Geld. Audience strategy by Shannon Busta. Our executive producer is Irene Noguchi. Special thanks to Kristin Lynn, Christina Samulski, Dr. Ziad Al Ali and Lauren Nichols.