Babies born with genetic diseases — especially rare diseases — may begin to show symptoms very early. But when they’re admitted to the hospital for help, getting the right diagnosis so doctors know how to appropriately treat them is a major challenge.

In many of these cases, clinical teams will run a battery of tests to understand what’s going on. All too often, though, results are inconclusive. This painful process is often called “the diagnostic odyssey” and, with babies suffering from rare diseases, it can take years to get useful answers. During that time, families and doctors must try to manage the symptoms without knowing what’s really going on.

As DNA sequencing tools make their way into hospitals, researchers have been making the case that analyzing a baby’s entire genome may be the best way to provide answers in these challenging situations. Now, clinicians and scientists have published results of a new study demonstrating that genome sequencing generates conclusive results more often for babies with genetic diseases.

This effort, known as the GEMINI study, was conducted in six children’s hospitals around the U.S. For 400 infants, researchers compared results from sequencing each baby’s entire genome to results from a more traditional test that looks for a set of known genetic culprits commonly associated with neonatal diseases. The major difference: the traditional panel tests restrict the search to known causes, while genome sequencing offers the ability to look at all genes and spot potential problems that may never have been seen before.

While genome sequencing generally took longer to return results — about six days versus four days, on average, for traditional testing — it was able to provide a diagnosis in almost twice as many cases. Nearly half of all babies got a definitive answer from genome sequencing, while just over a quarter of the babies got conclusive results with the panel test. Perhaps most importantly, the information generated from genome sequencing allowed physicians to adjust to more appropriate clinical interventions for 19% of infants in the study.

“More than half of the babies in our study had a genetic disorder that would have remained undetected at most hospitals across the country if not for genome sequencing technologies,” said neonatologist Jonathan Davis at Tufts Medical Center, one of the leaders of this project, in a statement. “Successfully diagnosing an infant’s genetic disorder as early as possible helps ensure they receive the best medical care.”

Even though genome sequencing results typically took a couple of days longer than traditional testing, that difference vanishes compared to the diagnostic odyssey faced by so many families. In one case, for instance, researchers were able to diagnose a baby with Kabuki syndrome less than two weeks after birth. The average time to get a diagnosis for this particular disease is around five years.

One of the reasons genome sequencing performed so well was its ability to spot new genetic errors that could be linked to disease: even in this relatively small study, researchers identified 134 genetic diagnoses that had never been seen before.

However, genome sequencing is not without its limitations, and this study clarified that too. With the deluge of data generated from this technology, trained specialists have to sort through to identify genetic variants that may cause health problems, and from that list pull out the single variant most likely to lead to the baby’s symptoms. While AI-powered tools are already helping to make a difference in this process, genome interpretation remains as much art as science. In this study, specialists at different laboratories disagreed about the cause of a genetic disorder in a stunning 40 percent of cases.

“It’s clear there [is] a lack of consensus among geneticists on the causes of a specific patient’s medical disorder,” said pediatrician Jill Maron at Women & Infants Hospital of Rhode Island, another study leader, in a statement.

Another potential problem comes not from the genome sequencing process itself but from the realities of clinical care: no test, however helpful, can go mainstream in medicine without reimbursement from insurance companies. Many insurance companies have already balked at covering panel tests, claiming that they cast too wide a net in the search for clinical answers. Genome sequencing covers the entire ocean, and costs more than traditional panel tests. Insurance companies will have to be confronted with undeniable evidence before they are likely to reimburse for genome sequencing in most cases.

That’s one of the reasons the GEMINI project is so valuable: it’s one of the broadest studies of newborn genome sequencing to date. Its results could help sway some insurance companies to consider reimbursing for this approach, at least for the babies in greatest need.

“Genome sequencing can be costly, but in this targeted, at-risk population, it proves to be highly informative,” Maron added. “We are supportive of ongoing efforts to see these tests covered by insurance.”