Intro: From the JAMA Network, this is JAMA Internal Medicine Author Interviews, conversations with authors, exploring the latest clinical research, reviews, and opinions featured in JAMA Internal Medicine.
Dr. Mitch Katz: Hello, I'm Dr. Mitch Katz, and I'm the Deputy Editor of JAMA Internal Medicine. This month we're publishing a really interesting study on the association of self-reported COVID-19 and SARS-CoV-2 serology results and persistent physical symptoms among French adults. It gets very much to the question of long-haul COVID symptoms, an issue that has perplexed many of us. I have the pleasure today of speaking with two of the authors, Dr Joane Matta and Dr Cédric Lemogne, I welcome you to our podcast.
Drs Matta and Lemogne: [In unison] Thank you!
MK: Before we talk about your article, I think our listeners would be interested to hear, in general, what impact has COVID had on the French health system?
CL: Well, as you can imagine, a huge impact, as everywhere in the world, I guess. The pandemic forced the French system to reorganize itself urgently, especially to increase the number of places that are available in intensive care units. In France, this number was more than doubled at the peak of the epidemic. But this success was not achieved without collateral damage or something like this as it meant closing several medical and surgical departments for many months. So it reduced the activity of these departments. And it might have -- I think this has actually led to delays in the treatment of patients with disease considered less urgent at the time. Surgical intervention for cancer, for example. But also to difficulties in access to care for urgent pathologies. In order to face these difficulties, as I imagine it was the same in other places in the world, the development of telemedicine has been greatly increased. In psychiatry -- I am a psychiatrist, actually -- this has worked quite well for the follow-up of the already known patients. But it has not really solved the problem of access to care for patient consulting for the first time. And an important success at the end, in France, is the rate of vaccination coverage, which is now about 90% in the adult population, supported by the decision of the French government to make access to certain places or certain activities conditional on the presentation of a “health pass.” That is a certificate of recent vaccination or immunization or negative PCR test. So this vaccination coverage rate has probably allowed us to renew the effect of the last pandemic wave despite the appearance of more and more contagious and dangerous variants. Unfortunately, I have to say that in some French departments where there were not so good vaccination coverage [rates], actually, in these departments, people experienced very significant increase in mortality.
MK: Cédric, to work in a French hospital at the current time, must you be vaccinated?
MK: OK, so you cannot test out of that option.
MK: You have to actually have the vaccination. That's great. With that requirement, did you have many healthcare people leave the profession? Or did everyone just go and get vaccinated?
CL: A little. Only a little. In my department, for example, 2 people out of maybe 50 preferred to leave their occupation.
MK: Very sad to lose good people. Another aspect of this that I think people in the United States would wonder; here, the pandemic was disproportionately felt by low income people, African American and Latino populations especially. In France, you have a national health system, and I'm wondering if you saw disparities in who got sick and who died.
CL: Unfortunately, yes. We do not have any statistics regarding ethnic aspects because it's forbidden by law in France. But regarding low income people, yes, they were clearly more affected by the pandemic. We can see that, not on an individual basis but on the basis of the mean income of some locations, regions within the territory. So the less the income was -- the mean income was in some part of the territory, the greater was the impact of the pandemic.
MK: All right. Well, that's really interesting to hear how the things that are the same and the things that are different between our countries. Let's talk about your article itself. There's been a lot of interest in the issue of long-haul COVID symptoms, people who report symptoms following infection with COVID, like fatigue, difficulty breathing, digestive problems. You used an existing cohort of French adults to explore the association between self-reported COVID and having a serologic test that showed that there was a previous exposure to COVID. Tell us what you found.
JM: OK, so basically, as you said, we first hypothesized that the belief of having been infected with COVID would be associated with greater symptoms, more than infection would do so. So what we found, in brief, is that the belief of having had infection, the COVID-19, was much more associated with all persistent symptoms except for hearing impairments, and then actual infection was only associated with anosmia, or negatively also with the sleeping problems. So that is, in brief, what we have found.
MK: How do you interpret the findings? What do you think it means?
JM: Well, we actually think that in the studies where they studied the COVID-19 participants, we think that they have not checked about the beliefs of having been infected with COVID. So we think they overlooked this part, so we believe that it's a determinant that was found usually in studies where they associated long-haul with persistent physical symptoms. What we can see is that those physical persistent symptoms, they are actually more frequent in the general population that was thought when all the -- I mean the publications about long-haul were published or were discussed in the literature. Also these studies, they omitted the part of participants who did not have the COVID-19. So they only focused on participants who had COVID-19, so I think this study also adds this important part of comparing both participants who were actually infected versus not actually infected.
MK: One of the issues that comes up any time people report in a clinical setting common symptoms like fatigue or digestive problems, you always wonder, you know, is it the illness, or you know, is this just something that happens in the population? Is it related to feelings of depression? I know you controlled for depression, and so we know it isn't specifically depression. Is that correct?
JM: Yes, we have controlled for depressive symptoms. We have also controlled for self-rated health. And the result did not change with regard to beliefs being associated, I mean much stronger associated with the persistent physical symptoms.
CL: The only change we've got is that when controlling for depression, back pain and joint pain were no longer connected to self-reported COVID, or maybe it was mediated or confounded by depression. But for the other symptoms, the results remain the same.
MK: So it puts all of us as clinicians in a bit of a conundrum. So I take care of many patients in my primary care practice who have COVID, had COVID, or as you say, believed they had COVID because in the real world many people, especially early on in the United States, we didn't have sufficient testing. So we have many people in the height of the pandemic in New York City where I practice, we couldn't test everybody. We didn't have the ability. And we were urging people who were not sick enough to need a hospital, we urged them to stay home. So in my clinical practice I had many people who, based on their antibody levels, I know that they were exposed to COVID. I have people who believed that they had COVID because they had symptoms consistent with COVID, but they never went for testing. For some of them, they report no further symptoms. But for some of them, they clearly report the kinds of symptoms that you looked at; digestive problems, shortness of breath, fatigue. And obviously with, you know, their experience, is I didn't used to have these symptoms and now I have them. Right? So it's a different sort of view on the same thing that in a cohort where you can ask yourself, OK, are these symptoms more common with self-reported infection or with proven antibody levels? What I have is the patient in front of me. The patient in front of me who says I had COVID and now I have these problems. What is the best way for a doctor to approach such patients?
CL: Well, I think you well described the clinical situation where many of us as physicians are facing. I think what we can learn from the results of this study is that in many cases the newly persistent symptoms -- because we only investigated symptoms that appear after the pandemic first wave but persisted at the time we conducted the study. We know that these symptoms are not really connected with the serology results. That means that they may be connected to other cause than exposure to the coronavirus. So this, in fact, is leading us, when we face the particular patients, maybe with not any tests regarding the coronavirus, to say OK, you have these symptoms. This is for certain because you experience them, so you have these symptoms. From the data that we have, we will first consider that these symptoms may be a signal for many diseases. It might be linked to the COVID. It might be linked to the COVID but not in a specific way. You might have had another virus, and maybe you could have experienced the same symptoms. It might also reveal another disease. It could be any disease because, as we know, the symptoms are not really specific. Almost -- I think today almost all the symptoms that people could have, have been related by at least one patient to a potential exposure to the coronavirus. So I think the point is that you have to cover all the hypotheses, the possible hypotheses. It could be another disease. This disease could be serious, could be more serious than COVID-19 itself.
It might deserve some diagnosis procedures. This disease could be, as well, a depression, as you said. It could be a non-psychiatric disease. But it could be also a psychiatric disease. After all, fatigue, difficulties in concentrating, pain symptoms are symptoms of depression. And also, we have to consider the fact that the COVID-19 pandemic and the concern about long-haul COVID may spread the idea in the population that there is a huge risk of developing persistent symptoms due to the COVID. And so they may attribute their symptoms to the COVID whereas there is another cause to find. And they may also be more -- tend to be more focused on their bodily sensations and so they be more prone to experience these symptoms. And I'm very surprised, not as a physician, but as a researcher, you know when you read some state-of-the-art literature review about possible mechanisms of long COVID, sometimes you can read 10 pages without the hypothesis of nocebo effect. I think the level of evidence of nocebo effect is quite high in our field. But from a scientific point of view, I think nocebo effect could explain a part of the symptoms in some patients or the symptoms of some patients in the population of those suffering from these symptoms. It's difficult to introduce this hypothesis to the patient. Sometimes they will feel not recognized. They feel as if they will suffer from imaginary symptoms. So it's very difficult to have to explain what you think nocebo effect is. But I think this possibility has to be considered seriously.
MK: Well, I think there's a way that we can go forward based on your advice and the research. I mean the key issues that you're highlighting is, first, as clinicians, we have to consider all possibilities.
MK: So if somebody, you know, is coming in with these symptoms, right, maybe a young woman. Perhaps she's anemic because of heavy menstrual periods. Perhaps the middle-aged person, their thyroid is not working appropriately. Right? So we have to first make sure that they don't have another treatable illness. Second, I think a key to always the care is the, as you say, the person's experience is their experience. And we don't take away people's experience. We don't, we can't, if somebody says their head hurts, it's not my role to say no, your head doesn't hurt. That would be a ridiculous statement.
MK: If the patient's head hurts, the patient's head hurts. What we ascribe the fact that their head hurts to, we're not always able to do. And one of the ways I might meld it is if somebody says that they have terrible persistent symptoms from COVID and they don't have a positive antibody, one thing that I would think of is, well, COVID has had such a profound impact on all of our lives. It has created all kinds of anxieties, all kinds of stress, hypervigilance about people near us. Is that person wearing a mask? Is my mask tight enough? Am I wearing the right mask? Should I get a vaccine? Which vaccine should I get? I can't see my family. Or should I see my family? Is it too dangerous? Is it OK to see my family outside? It's created this new world where many of us are more detached from the people we work with because we're doing virtual visits. We're more detached from our families because we're not able to connect with them for fear of exposing ourselves or exposing them. Huge seismic effects within the economy, losses of jobs, losses of people we love due to COVID, that if you came out of such seismic things with a set of symptoms, in a sense it would be due to COVID, but perhaps not the actual infection. Does that make sense to you as a psychiatrist?
CL: Of course. And I will add that the persistent symptoms may also be due to the acute symptoms that you may have experienced due to the coronavirus or to another disease. But in this sense, the persistence of the symptoms because of hypervigilance or because of the health behaviors or the new health behaviors that are caused by the concerns or beliefs about the symptoms may have contribute to perpetrate the symptoms. So in this sense, the cause of the persistent symptoms is the COVID-19 as a physical disease, not only as you mentioned, and I agree, not only as a consequence of all the issues raised by the pandemic, but as a consequence of the very physical issue experienced at the time of the COVID. For example, if you spent one week controlling your breath, the respiratory frequency, to decide whether or not calling the emergency department, obviously you trained your brain during seven days to be extremely -- to become an expert to detect breathlessness. And even if after the COVID, a few weeks after your respiratory system has recovered, something, probably in your brain due to conditioned response and to memory system devoted to that kind of learning, something is remaining. And it was caused by the COVID but not by the virus.
MK: Right. Let's, for our listeners, cover a couple of other important points. First, we know that there are people who truly have long-term COVID symptoms. Some people that we've seen in the hospital, who required actually months and months of rehab due to intense deconditioning, being on vents for months and months, and clearly that's a different issue than what we're talking about here.
CL: Not sure Mitch, actually the deconditioning is sometimes influenced by beliefs. If you -- I went through, you know, a forum of discussion through the internet, groups of patients. For example, the belief that the symptoms, when they are increasing after an effort could be the signal of the fact that you are worsening your condition. This belief leads to more and more deconditioning.
MK: I just want to make sure that our listeners understand that there are people who have dramatic cases of COVID who have, you know, persistent and definable on echo or cardiac echo or radiologic tests that show ongoing damage.
MK: Kidney damage, patients who are still on dialysis following COVID. Right? I just want to make sure those patients are also seen and heard. I agree with you. Deconditioning is a phenomena that can occur at all stages of illness. Let me ask one other place where our readers, listeners may wonder. What if the antibody test is wrong? No test is 100%. There are a variety of antibody tests. How do we know that the people who, in your study, who reported that they had COVID but had negative antibody tests, maybe in those people the antibody test just wasn't sensitive enough.
CL: Sure. It was a critical question raised by our results. And the main reasons that we add to that is that we carefully searched for a statistical interaction between self-reported COVID, belief, yes or no, and the results of the serologic tests, yes or no. And the effect size of the odds ratio, that said the odds ratio of the belief was exactly the same, whatever the results of the serological tests. So, if our results would have been explained to some extent by particular subsets of participants, let's say, for those with negative serology, but having actually been infected by SARS-CoV-2, we should have expected a significant statistical interaction and we found none. Out of more than 20 symptoms, there was no significant interaction. But clearly this hypothesis should be present in your mind, and I'm sure that further studies will go, exactly, further, that was investigating this point.
MK: Is there anything else our readers or listeners should know about your study or about long-haul COVID symptoms that we haven't covered?
CL: Maybe the experience of the management plan that we set up in Paris in my hospital?
MK: Yes. Tell us -- that's a great -- tell us for patients who report long COVID symptoms, how do you treat them?
CL: Briefly speaking, we set up a multi-professional pluri-disciplinary consultation with three consultations. First, by an internist or a specialist in infectious disease, a long consultation, about one hour. Then a consultation with a psychiatrist with two part; first, diagnosing, searching for common mental disorder, especially anxious disorder or depressive disorder, and the second part devoted to the investigation of cognitive or behavioral mechanisms that could have contributed to perpetrate the symptoms. And third, a consultation with a physician specialized in physical activity to quantify the possible deconditioning and propose rehab if it's effective. And before this consultation, we send some documentation, some recommendation to the GP in charge of the patient. And sometimes it helps a bit and sometimes it helps a lot, even before the pluri-disciplinary consultation. We only started this program three months ago, so we are far from, I think, quantitative data to draw some conclusion about the relevance of this management. But we're quite confident that it helps many, many of these patients.
MK: Well, and that's what matters the most. And again, as clinicians, our focus is always on trying to relieve pain and discomfort for our patients who are experiencing it. And I think you've clearly set up a very comprehensive assessment process for your patients, and also through the possibility of exercise and rehab we can, you know, help people to do better. And I think that's also -- has always been, for me, a key touchstone in the care of people with symptoms is that I can't always change someone's symptoms, but maybe I can help them improve their function. And if people can function better, even though they may still have symptoms, over time they will be better off. Well, I want to thank both of you, Dr Joane Matta and Dr Cédric Lemogne, for talking with me about your article and about how COVID affected the healthcare system in France. This was really interesting for me, and I know it will be interesting for our listeners. Thank you both very much.
MK: For more of our podcasts, visit us at jamanetworkaudio.com. You can listen and follow wherever you get your podcasts.