Clinical epidemiologist Ziyad Al-Aly has access to a treasure trove many researchers can only dream of: millions of sets of electronic medical records from the U.S. Department of Veterans Affairs (VA), which provides health care to military veterans from the country.
With this data in hand, Al-Aly, who is based at the VA St. Louis Healthcare System in Missouri, and his colleagues looked at the long-term effects of COVID-19, cardiovascular disease1 to diabetes2. They also took on the challenge of studying long COVID – a condition in which people experience symptoms months after an acute SARS-CoV-2 infection appears to have resolved – and recently published results.3 which surprised some researchers. The team found that previous vaccination only reduced the risk of developing long COVID after infection by about 15%, which is significantly lower than some other estimates.4which suggested that vaccines cut the risk in half.
It’s the kind of whiplash result that people who follow long-running COVID research have grown accustomed to seeing, as data from various studies report discordant results. Differences in the definition of the syndrome, the types of data used to study it, and how that data is analyzed have left the public and policy makers struggling with disparate answers to fundamental questions. How common is long COVID? And how does vaccination or reinfection or the latest variant of SARS-CoV-2 affect the risk of developing the disease?
The answers to these questions can be used to craft COVID-19 policies, but the constant drip of jagged studies can also confuse. Al-Aly said. Having so much uncertainty doesn’t engender much confidence, adds Al-Aly: “The public doesn’t react very well to saying ‘between 15% and 50%’.”
Part of the problem is the definition of long COVID, which has been linked to more than 200 symptoms, ranging in severity from inconvenient to debilitating. The syndrome can last for months or years and has a tendency to reappear, sometimes months after an apparent cure.
So far, there is no agreement on how to define and diagnose long COVID. The World Health Organization’s tentative consensus, published in 2021, has not proven popular with patient advocates or researchers, and studies continue to use a range of criteria to define the disease. Estimates of its prevalence can vary from 5 to 50%.
A study of such a complex condition must be large enough to reflect the range of symptoms and the possible impact of characteristics such as age and severity of acute SARS-CoV-2 infection. This is where analyzes like Al-Aly’s offer a host of advantages: data from large healthcare networks can provide enormous sample sizes. Al-Aly’s study of the long COVID after “breakthrough” infection – the one following vaccination – included records from more than 13 million people. Although 90% of those people are men, that still leaves 1.3 million women in the analysis, Al-Aly notes, more than many other studies can muster.
These large numbers, along with the types of data available in some health records, allow researchers to perform complex statistical analyzes to carefully match the demographics of those infected with the coronavirus to an uninfected control group, says Theo Vos. , an epidemiologist at the Institute for Health Metrics and Evaluation at the University of Washington in Seattle, who has worked with a variety of data sources to study the long COVID.
But there are also disadvantages. “People confuse study size with study quality and validity,” says Walid Gellad, a physician who studies health policy at the University of Pittsburgh in Pennsylvania.
In particular, Gellad worries that studies that rely on electronic health records may be confounded by behavioral differences. For example, compared to someone who doesn’t seek medical care for acute COVID-19, someone who does might be more likely to report long-lasting COVID symptoms, he says.
Additionally, medical records and health insurance claims might not reflect a demographically diverse population, says computational epidemiologist Maimuna Majumder of Harvard Medical School in Boston, Massachusetts. This is especially likely in the United States, she says, where health insurance coverage varies widely. “The number of data points considered is often so large that we incorrectly assume that this data must be representative,” she says. “But that’s not necessarily the case.”
Majumder also wonders if studying the claims data might lead researchers to underestimate the number of people with long-COVID, because many people might not seek medical care for their condition.
Another issue is how symptoms are recorded in claims and electronic medical records. Doctors often record codes for multiple symptoms and conditions, but they rarely list a code for every symptom a patient experiences, Vos says, and the choice of codes for a given condition can vary from doctor to doctor. This could lead to differences in whether and how long COVID is reported. “Electronic health records definitely contain useful information,” says Gellad, who says the VA study was uniquely well-designed. “But to answer the question of how common something is, they might not be the best.”
Other methods also have their pitfalls. Some studies rely on self-reporting, such as the COVID Symptom Study app developed by King’s College London and data science company ZOE, also in London. Data from the app showed vaccination roughly halved the risk of getting long COVID 28 days or more after an acute infection4. But studies in which people voluntarily report their symptoms can be biased because people with symptoms are more likely to participate, Gellad says. And studies that rely on smartphone apps might not fully capture data from disadvantaged communities.
A particularly useful source of data has been the UK’s Office for National Statistics (ONS), says Nisreen Alwan, a public health researcher at the University of Southampton, UK. In May, the ONS reported that the variant of SARS-CoV-2 that people are infected with may affect their risk of developing long COVID. Among doubly vaccinated participants, those thought to have COVID-19 caused by the Omicron BA.1 variant were approximately 50% less likely to develop long COVID symptoms four to eight weeks after infection than participants whose the infections were probably caused by the Delta variant. This finding is consistent with the results of a June 18 article5 based on ZOE data.
In search of a common thread
Alwan, who has long had COVID and has advocated for collecting data on the condition, praises the design of the ONS study, which involved recruiting a group of people with particular attention to representing the UK population and then follow them to ask questions about their condition and symptoms of infection.
Other aspects of the study design, such as the use of a control group, can strongly affect the results, says Alwan. But taking into account disparate methods and definitions should not block research. “It’s not something new,” she said. “It’s something we had before COVID, for other conditions.”
For Al-Aly, the discrepancies between the results of the studies are neither surprising nor overwhelming. Epidemiologists often weave together evidence from multiple data sources and analytical methods, he says. While it’s difficult to precisely quantify the effect of vaccination on long-term COVID risk, for example, researchers can look for trends. “You are looking for the common thread,” says Al-Aly. “The common thread here is that vaccines are better than no vaccines.”
#long #COVID #Studies #Give #Answers