Cracking the long-haul COVID code: How a basic symptom-tracking app may predict monthslong cases

Ray Bradbury once said, "mysteries abound where most we seek for answers." Case in point: the COVID-19 pandemic.

Despite more than a year's worth of observation and intense study, there’s a lot we still don’t know about COVID-19. One of the most enduring mysteries is “long COVID syndrome,” where symptoms stick around for weeks or even months longer than the virus’s typical life span—and often in younger, healthier patients who had particularly mild cases of the disease.

Studies estimate as many as 30% of COVID-19 patients will suffer prolonged effects, and nearly half of long-haulers report health issues so severe they’ve been unable to work full-time.

One research team may have solved at least a small piece of the puzzle. They built on a recent analysis of self-reported symptoms in the COVID Symptom Study smartphone app—released in March 2020 by King’s College London, St. Thomas Hospitals and the startup ZOE Global Limited—to develop a simple algorithm that might identify those most likely to contract long COVID.

COVID long-haulers experience sustained fatigue, insomnia, anxiety, shortness of breath and “brain fog,” among other common symptoms.  


The model was built using data from more than 4,000 people who had begun using the app before they developed any symptoms, then received a positive COVID-19 diagnosis and continued tracking their health with the tool.

Using only age, gender and the number of initial COVID-19 symptoms, the algorithm was able to accurately predict almost 70% of long COVID cases, according to a study published in Nature. It was equally effective at avoiding false positives.

The researchers found that women, older people and those with at least five early symptoms—regardless of what they were or in what order they arrived—are most likely to experience long-term symptoms. The only underlying condition so far that they’ve been able to link to long COVID is asthma.

RELATED: Treating COVID-19 long-haulers with Amgen's sleeper heart drug Corlanor

Of course, as with almost everything pertaining to the ongoing pandemic, long COVID has its fair share of naysayers.

In a recent op-ed for Stat, for example, Adam Gaffney, M.D., a pulmonary and critical care physician at Massachusetts’ Cambridge Health Alliance, asserted that while the symptoms reported by long-haulers are “debilitating and real” and likely tangentially connected to the isolation and distress caused by the pandemic, they shouldn’t automatically be classified as a direct result of a COVID-19 diagnosis.

“We should expect a surge in both mental anguish and physical suffering that, while connected to the once-in-a-century pandemic, will not always be directly connected to SARS-COV-2 itself,” Gaffney wrote.

Still, plenty of other scientists and doctors argue long COVID is in fact a combination of psychological effects caused by the circumstances of the pandemic and both neurobiological and biomedical effects caused by the actual virus. That stance is backed up by increasing anecdotal evidence that COVID long-haulers are seeing their symptoms finally go away only after receiving the vaccine.

Long COVID syndrome is now under rigorous study by a plethora of research groups across the globe. The National Institutes of Health, for one, launched a four-year initiative in February for just that purpose, backed by $1.15 billion in congressional funding.

The COVID Symptom Study group, meanwhile, continues to urge new users to add their symptoms to the app in hopes that they'll continue to shed light on the many unsolved mysteries of COVID-19 and its long-lasting effects.

“It’s important we use the knowledge we have gained from the first wave in the pandemic to reduce the long-term impact of the second. Thanks to the diligent logging of our contributors so far, this research could already pave the way for preventative and treatment strategies to reduce the long-term effects,” study author Claire Steves, Ph.D., a senior clinical lecturer at King’s College London, said upon the release of the study’s preliminary results. “Using the app daily can help affected people and their doctors better categorize and judge their risks of developing longer more severe disease.”