RECOVER EHR study provides additional information about pediatric Long COVID symptoms

Researchers have learned a lot about the variety of symptoms associated with Long COVID. However, most Long COVID research to date has studied adults, meaning that less is known about the Long COVID symptoms that children and adolescents (individuals aged 21 or younger) experience. 

Previous RECOVER research from the initiative’s pediatric observational study documented specific symptoms in children and adolescents between the ages of 6 and 17, and recent research detailed Long COVID symptoms in children between the ages of 0 and 5. 

A new RECOVER study published in PLOS Digital Health adds to this collection of knowledge. In this study, researchers analyzed electronic health records (EHRs) to better understand how Long COVID impacts children and adolescents, and how the condition appears differently among specific groups of people aged 21 and younger.

What did researchers do?

Researchers analyzed EHR data from pediatric patients who received care at 38 medical institutions across the US. These data are made available to RECOVER through a research network called PEDSnet.

To identify children and adolescents younger than 21 who likely had Long COVID, researchers used a computer program available within PEDSnet called a machine learning algorithm. The machine learning algorithm reads symptoms and other data within the EHR to recognize patterns that indicate a person likely has Long COVID. Although there is a formal diagnosis code within the EHR that clinicians use when they recognize Long COVID, not all children or adolescents with Long COVID receive an official diagnosis. Therefore, the machine learning approach is important because it enables researchers to identify and include a larger group of children and adolescents who are impacted by Long COVID. 

In this study, the algorithm identified a study group that included 17,525 people younger than 21 who were likely to have Long COVID based on symptoms and other patterns in their EHRs. Nearly a third (30.4%) of the patients in the study were between the ages of 16 and 20, and over half (54.5%) were female.

In addition to using machine learning to identify the study group, researchers used a similar approach to define different ways that children and adolescents may experience Long COVID. Using a process called model training, they taught a machine learning algorithm to recognize patterns and identify common types of symptoms.

What did researchers find?

In total, the machine learning algorithm identified 6 types of symptoms that children and adolescents with Long COVID commonly experience. In the individuals who were included in the study, cardiorespiratory symptoms (having to do with the heart or breathing) were most common, with over half (53.8%) of patients experiencing these symptoms.

Researchers also identified 5 other common types of symptoms (listed in order from most frequently experienced to least frequently experienced), including:

• Musculoskeletal pain, or pain in the body’s muscles, bones, joints, ligaments, and tendons.
• Neuropsychiatric conditions, or conditions having to do with the brain (such as anxiety disorder, which was the most common diagnosis in this category).
• Gastrointestinal symptoms, or symptoms that affect the stomach, intestines, or other parts of the gastrointestinal system. These include abdominal pain, nausea and vomiting, and diarrhea.
• Headache.
• Fatigue, or extreme tiredness.

Within each group, researchers performed analyses to better understand which groups of children and adolescents may be more likely to experience certain symptoms. The study found that:

• Study participants who had severe cases of COVID-19 were likely to experience heart and breathing symptoms upon developing Long COVID.
• Although fatigue was the least common symptom group, study participants who were categorized into other symptom groups sometimes also experienced fatigue. This finding suggests that children and adolescents often experience fatigue along with other Long COVID symptoms.
• Most study participants younger than 4 experienced heart and breathing symptoms.
• Hispanic and non-White study participants were disproportionately affected by heart and breathing symptoms. This result confirms results from a previous RECOVER study that found Hispanic children and adolescents younger than 21 were more likely to have breathing problems associated with Long COVID.

Why is this research important?

RECOVER provides researchers with access to millions of EHRs, meaning that they can conduct large studies that include many different people who are impacted by Long COVID. A large study enables researchers to analyze how specific groups are affected, such as different age groups within childhood and adolescence, as well as different racial and ethnic backgrounds.

Symptoms of Long COVID vary widely, and the way that different people experience Long COVID is affected by many factors. In addition, this research shows that children and adolescents often experience different Long COVID symptoms than adults do. Better understanding of these differences can inform future pediatric Long COVID research, including the development of personalized treatments. Findings like the ones from this study can also support clinicians in providing individualized care for pediatric patients to help them address and manage their symptoms.

Check out the full publication to learn more about Long COVID symptoms that children and adolescents commonly experience.