Editor's note: Find the latest COVID-19 news and guidance in Medscape's Coronavirus Resource Center.
Long COVID worries me. The vastness of its symptoms, the uncertainty of its causes, and the immense attention it gets provide a setup for medical iatrogenesis.
A middle-aged patient reports unusual fatigue months after SARS-CoV-2 infection. A clinician orders electrocardiography (ECG) to investigate post-COVID cardiac disease. The ECG shows nonspecific changes, which leads to a nuclear stress test. The coronary angiography done to evaluate the abnormal scan shows no coronary artery disease, but a catheter-induced coronary dissection requires a stent.
This case is fictional but is similar to one recently published in JAMA Internal Medicine.
Four factors drive most cases of medical iatrogenesis: good intentions, uncertainty, the incentive to act rather than observe, and, most important, fear.
Learning to Fear
The intense attention paid to long-haul COVID creates a social learning phenomenon whereby people learn to expect (and fear) symptoms long after infection with SARS-CoV-2.
In the digital age, one hardly needs a citation to document social learning.
But here is one: in this review of drug trials for major depression over two decades, the authors found a marked increase in response to placebo over time. Placebos are inert and shouldn't become more effective over time. One explanation is that as treatment for depression became common, people learned to expect positive results from therapy.
A Google search for "long-haul COVID" returned 83 million results. Numerous influential media outlets have featured compelling anecdotes of people experiencing a wide array of mysterious symptoms months after their initial infection.
Long-haul COVID also has the attention of academics. Consider that a review of post–acute COVID-19 syndrome in Nature Medicine submitted in October 2020, just 7 months into the pandemic, included 226 citations. And given that the National Institutes of Health will spend $1.15 billion to study what is now called postacute sequelae of SARS-CoV-2 infection (PASC), expect even more attention from grant-supported scientists.
Clinicians also read the news. We, too, are aware of the numerous patient advocacy groups pushing for recognition and the reputational danger of being perceived as insufficiently concerned about physical suffering from long COVID.
For instance, we know from the influential Atlantic writer Ed Yong that any suggestion that stress or anxiety plays a role in the symptoms of long COVID is akin to medical gaslighting.
There are two basic ways that a clinician can demonstrate a caring signal. One is to sit and actively listen and express empathy. That is hard. An easier way to is to order tests.
Social norms hold that physical symptoms deserve testing: "My doctor must really care because she ordered a battery of tests, including an MRI." But tests can lead to more tests, and the cascades can result in psychological, physical, or financial harm.
Long COVID and Uncertainty
It's hardly provocative to describe the science of long COVID as uncertain.
Start with the symptoms. Some of the most common—fatigue, weakness, insomnia, dyspnea, and brain fog—are difficult to quantify. Fatigue compared to what? As for cognition, the past 12 months hasn't been an easy time for staying on task. And I can almost count on two hands the number of people I have seen in clinic who report sleeping well.
Then there is selection bias in many of the initial reports. A research letter (with nearly 900,000 page views and 132 citations) enrolled patients from a COVID-19 postacute outpatient service. Another survey enrolled patients from the Mount Sinai Hospital's COVID-19 Precision Recovery Program. And, in a Bloomberg news story titled "COVID Everlasting," the study in question enrolled patients by tweeting: "If you consider yourself a Long-hauler, please let us know which symptoms you have experienced."
Reports like these sample groups of people more likely to have symptoms. There have been 126 million cases of COVID-19 worldwide; why not do a random sample of people at different intervals after their initial infection? This way we could learn the true prevalence of long COVID symptoms.
Perhaps the most perplexing aspect of long COVID are reports of symptoms without evidence of SARS-CoV-2 infection. A recent preprint reported 205 different types of symptoms in the months after initial infection. Approximately 1700 of the patients surveyed had a negative result on a reverse transcriptase polymerase chain reaction test and 600 tested positive. Of the 205 symptoms, only loss of smell and taste were significantly more frequent in those who had tested positive than in those who had tested negative. Other groups have replicated these findings.
Testing for SARS-CoV-2 is imperfect; tests were not available early on and some symptomatic patients without positive COVID test results were likely false negative. But it also seems likely that a finite number of people reporting long COVID symptoms were never infected with the virus. That is a puzzle.
When faced with patients who have symptoms, clinicians in many health systems, especially in the United States, are incentivized to order tests. A cardiologist who orders echocardiography, nuclear perfusion scanning, or cardiac monitoring in a patient with long COVID symptoms not only appears caring but also earns productivity.
Although it would be overly cynical to tie the rise in post-COVID care clinics solely to financial motives, it is also not wrong to note the highly competitive nature of profit-driven healthcare in many U.S. markets. From the perspective of a hospital administrator, long COVID is an opportunity to increase market share and make money.
The Medical Conservative's Approach
My best answer to the problem of long COVID is empiricism.
First is to remember that most medical interventions don't work or have modest effect sizes; magic bullets, such as antibiotics, insulin, and HIV medicines, are the exception, not the rule. Thus, not only should we be pessimistic that long COVID symptoms will have an easy fix, we should also be alert to the danger of iatrogenic harm.
Postviral syndromes are real. The pandemic has caused a lot of infection; there will be a lot of postviral syndromes. In electrophysiology, we have long seen patients with autonomic disorders, such as inappropriate sinus tachycardia (IST), that mysteriously began after an infection.
The history of IST is instructive. Its discovery coincided with the heyday of catheter ablation for supraventricular tachycardia (SVT). Powered by the success of SVT ablation, many doctors favored ablation of the sinus node for IST. This turned out poorly because of both the lack of efficacy and serious complications. A medically conservative approach would have saved many patients with postviral IST from iatrogenic harm.
Long COVID is also well suited for a medically conservative approach. Diagnostic pathways and therapeutic interventions should be studied in trials that include proper control arms.
People with any symptom are, by definition, suffering. Caring, empathy, and the strength to let time and nature do the bulk of the curing seem to me the best default. Make this the control arm; then prove that intervention is better.
That way we are more likely to fulfill our promise to first, do no harm.
John Mandrola practices cardiac electrophysiology in Louisville, Kentucky, and is a writer and podcaster for Medscape. He espouses a conservative approach to medical practice. He participates in clinical research and writes often about the state of medical evidence.
Follow John Mandrola on Twitter
Follow theheart.org | Medscape Cardiology on Twitter
Medscape © 2021 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: John M. Mandrola. Why Long COVID Worries Me - Medscape - Apr 05, 2021.