25 Years Later Pediatric infectious disease specialist Asuncion Mejias knows all too well the deadly unpredictability of respiratory syncytial virus (RSV), an infection that hospitalizes 80,000 children under the age of 5 in the United States each year.
“This is a very rapidly changing disease,” says Mejias, who is at St. Jude Children’s Research Hospital in Memphis, Tenn. “I always tell my colleagues that for every two kids who come in, one will be in the ICU within three hours, and the other might be able to go home the next day. It’s completely unpredictable.”
RSV infections are very common, with almost all children infected at least once before age 2. Most children experience cold-like symptoms such as coughing and sneezing, but some develop severe lung disease. RSV is responsible for more than 100,000 infant deaths worldwide each year, nearly half of them in infants under 6 months of age.
The problem is, aside from a few known risk factors, like premature birth or pre-existing lung disease, it’s hard to tell which children are most affected. “80 percent of children who are hospitalized with RSV appear completely healthy,” Mejias says. “They were born full term and have no risk factors for severe disease.”
That’s why various research groups around the world are working to train machine learning algorithms and develop statistical models that can show which children are most susceptible to RSV. Based on vast databases of electronic medical records, these tools aim to identify groups of risk factors that can help predict which children are more likely to become infected and be hospitalized. Healthcare providers can use this information to prioritize vaccines and other preventive measures for the children most at risk.
Earlier this year, respiratory epidemiologist Tina Hertert and her colleagues at Vanderbilt University developed a tool to train a statistical model using data from more than 400,000 infants in the Tennessee Medicaid program to identify 19 risk factors for RSV infection. “This allows us to calculate each infant’s individual risk at birth,” Hertert says.
Some of the variables used in the tool aren’t surprising. For example, prenatal smoking is known to impair fetal lung development, making babies more susceptible to viral pneumonia; meanwhile, low-birth-weight babies already lack the strength to breathe normally. But often, a combination of risk factors makes a child less susceptible to infection, says Hartert. “If we only evaluate individual factors, we’re going to miss a lot of at-risk infants,” she says.
In 2023, U.S. regulators approved a vaccine called Abrisbo, designed to be given to mothers between 32 and 36 weeks of pregnancy with the goal of ensuring babies are born with protective antibodies against RSV. They also approved a drug called Bayfortus, a lab-made protein called a monoclonal antibody that can be given in a single injection to provide protection before the winter RSV season.
