Loading ...

Linking children’s birth diagnosis of congenital cytomegalovirus (cCMV) to their long-term health outcomes can help capture the healthcare burden of cCMV infection, according to research presented at the Pediatric Academic Societies (PAS) 2026 Meeting.

This informatics approach to investigating a large healthcare system’s medical records can help the system determine the actual cost of cCMV while also helping physicians know what later health and developmental problems they should watch for, said Ryan H. Rochat, MD, PhD, a pediatric infectious disease specialist at Texas Children’s Hospital and director of the biomedical informatics dual degree program at Baylor College of Medicine in Houston.

“Long-term outcomes in congenital CMV remain poorly described in a real-world setting,” Rochat said. Diagnostic confirmation of cCMV has to occur within the first 21 days of life, but the results of those tests are often siloed within the medical records. Children’s long-term health outcomes also live in those same medical records, he said.

Linking the two is the first step toward compiling data whose implications are far-reaching, both in estimating the societal and financial cost of cCMV and in helping pediatricians know what potential long-term sequelae to watch for in children years after they tested positive for cCMV.

“We know in congenital CMV that some kids will have late-onset hearing loss, but if you don’t know that, you don’t know to look for it,” Rochat said. Because delayed-onset hearing loss carries a risk for speech and language development disorders, the timing of identifying it is very important to access early interventions, he said.

Capturing the true burden of cCMV, meanwhile, can pave the way toward more funding for a vaccine to prevent cCMV and provide justification for universal screening programs like those already implemented in a handful of states.

It can also help create a diagnostic “fingerprint” from medical records that allows physicians to retroactively estimate the likelihood that a child had cCMV even if they were never tested for it at birth, thereby identifying children who warrant closer observation and screening.

It would be too difficult and expensive to conduct a longitudinal study that follows children for 18 years to identify all the health and developmental outcomes that occur following a cCMV infection, Rochat said. But it’s also not necessary to do that when the information already exists in the electronic health record (EHR) and can be linked.

The Study

His healthcare system includes four freestanding hospitals and over 70 pediatric clinics, with more than 6 million unique clinical encounters recorded in 2024 alone. So Rochat queried the EHR for children born between 2013 and 2024 using broad International Classification of Diseases, 10th Revision, codes to identify 324 children coded with cCMV.

Of those 324 children, 130 were already listed in a virologically confirmed registry of cCMV maintained by another researcher, Gail J. Demmler-Harrison, MD, at their institution. They included 88 children with symptomatic cCMV and 42 children with an asymptomatic diagnosis.

Symptomatic cases had at least one of the following: purpura/petechiae, jaundice, hepatosplenomegaly, microcephaly, unexplained neurologic abnormality, elevated liver enzymes, hyperbilirubinemia, hemolytic anemia, or thrombocytopenia. Asymptomatic cases were those with virologically confirmed cCMV infection but none of the symptomatic signs.

Of the other 194 children coded with cCMV in the EHR, 58 were virologically excluded because they had a negative CMV test at birth, and the other 136 had not been classified. Rochat then mapped long-term health outcomes of unclassified children and children with symptomatic and asymptomatic cCMV.

He created heat maps that showed the density of coding for 12 collections of possible problems: weight problems, vision problems, sleep apnea, otitis media, neurologic problems, musculoskeletal problems, hearing problems, hearing aids, gastrointestinal tube, developmental disability, central nervous system problems, and difficulties with activities of daily living.

The resulting data showed that the average patient with symptomatic cCMV had had 15 clinical encounters by age 3. Children with asymptomatic cCMV had fewer encounters, and the number of encounters for unclassified cases fell between the two.

To better understand what that middle group could involve, he examined the long-term health outcomes for children with symptomatic and asymptomatic cCMV. Doing so revealed that children with “asymptomatic” cCMV started developing several problems early on, and these persisted throughout childhood. While the proportion of billing events was not as great for those with symptomatic cCMV, these children were being managed for developmental disabilities, central nervous system problems, musculoskeletal problems, and neurologic problems.

Any integrated healthcare system can conduct a similar analysis, Rochat said, and doing so can provide several important insights. Top among these is understanding the true level of healthcare utilization in these populations. The heat maps Rochat created show higher levels of utilization than might have been expected in unclassified children and children with asymptomatic cCMV, “and this is the floor,” he said, pointing out that those clinical encounters represent a minimum amount of healthcare used.

“It doesn’t come as a surprise that someone with symptomatic congenital CMV [at birth] has long-term health issues, but it does say something to see that by 5 or 6 years of age, half of them will have diagnosable musculoskeletal problems or a third will have developmental disabilities,” Rochat said.

Implications for cCMV Research

Sallie Permar, MD, PhD, a pediatric infectious disease specialist, chair of pediatrics, and professor of pediatrics and immunology at Weill Cornell Medicine and NewYork-Presbyterian Komansky Children’s Hospital in New York City, has spent her career researching cCMV and developing a vaccine for it. She was not involved in this study but immediately recognized the potential value of it.

“It is asking a question that we’ve always wanted to ask: What are the long-term consequences of a congenital CMV infection?” Permar said. “We know some kids are fine with congenital CMV, but we don’t know how many are not fine and how many have subtle issues.”

This research reveals that it’s not just the first 2 or 3 years of life when problems can occur. Issues can arise through school age, Permar said. “So you have to be on the lookout for things for many years with a child,” she said. “This is such a common disease,” affecting an estimated 0.5% of US newborns, she said, “and the anticipatory guidance for kids diagnosed with congenital CMV who don’t have any signs at birth is kind of a great unknown for the whole pediatric community.”

Permar agreed with Rochat that interrogating a healthcare system’s EHR is a huge opportunity to look at the healthcare utilization of children diagnosed with a congenital condition.

“The piece I keep being more and more surprised by is how strong the associations are with neurodevelopmental conditions that are very common but could be either exacerbated or even caused by congenital CMV, including conditions like ADHD [attention-deficit/hyperactivity disorder] and autism,” Permar said.

If we have more insight into the association between cCMV and developmental issues, it can be easier to prescribe therapies in those first few years of life when they’re most effective, she said. “We could actually prevent some of the long-term outcomes just by offering the appropriate therapies that have been shown to work in other settings,” she said.

The implications of this kind of research at scale are tremendous, Permar suggested.

“We have got to figure out those who are at highest risk and make sure services can be delivered to them in a way that is family-friendly and time-friendly,” Permar said. 

The study was funded by Moderna. Rochat reported receiving past research funding from Moderna and Merck and funding from WebMD/Medscape for curriculum and CME development. Permar reported having served as a consultant to Moderna, Merck, Pfizer, GSK, Hookipa, and Dynavax Technologies on their CMV vaccine programs and having led sponsored programs with Merck and Moderna on CMV vaccines.

Tara Haelle is a science/health journalist based in Dallas.