Romeo Ranoco / Reuters

For many parents, the specter of sudden infant death is one of the dominant anxieties in the chaotic early weeks with a newborn. It is a fear that drives 3 a.m. internet searches, and prompts incessant checking to make sure baby is still breathing.

Part of what makes Sudden Infant Death Syndrome, or SIDS, so terrifying, is that it kills newborns who seem otherwise healthy. It does so without warning. And no one really knows how common it is.

Each year, about 3,500 infants in the United States die from Sudden Unexpected Infant Death, or SUID, a broader category that includes SIDS, according to the Centers for Disease Control and Prevention. Most of those deaths remain unexplained. About one-quarter of them are eventually attributed to accidental suffocation or strangulation in sleeping environments.

But scientists have reason to believe that many of the deaths that remain classified as “unexplained” may not be SIDS after all. This isn’t the first time researchers have refined their understanding of this frightening syndrome. In the 1960s, more than 10,000 infant deaths per year were attributed to SIDS in the United States. In the 1970s, researchers theorized that some cases of SIDS, then called “crib death,” were actually caused by botulism poisoning. Other theories have focused on possible breathing disorders, brain abnormalities, and damage to the inner ear. Since the 1990s, caregivers have been instructed to place babies on their backs to sleep. The “Back to Sleep” public health campaign is widely considered one of the most successful such efforts in modern history. It’s credited with cutting SIDS deaths in half, an astonishing decline, in just two decades.

Yet SIDS still remains the leading cause of death for babies between 1 month and 1 year old in the United States. And researchers are still scrambling to figure out why.

Seven years ago, public health officials at the CDC launched a case registry to log SUID deaths. The idea was to track deaths at the state level, monitor trends, and to gather data that could improve death-scene investigations and might ultimately help prevent future deaths. Since 2009, the registry has expanded from five states to 18.

Despite that growth, it is still a fairly “junior surveillance system,” says Sharyn Parks Brown, an epidemiologist at the CDC who manages the database.

“We are just getting to the point where we have the capacity to conduct a lot of analyses,” she told me. “In the next decade, we may be able to start to do genetic-type analyses of cases to see if there’s something underlying [linked to SUID deaths].”

In the meantime, though, there are many challenges. For one thing—and this is perhaps the biggest thing—there are stark differences in how various states, and even separate jurisdictions within the same state, handle death-scene investigations and classifications. “That’s one of the inherent challenges,” Brown said. “There have been a lot of differences in how ‘SUID’ is defined. There is not a standard state-by-state definition.”

There are also cultural factors at play. Some infant deaths that are eventually explained—a baby who suffocates because of loose bedding in her crib, for example, or when an adult rolls onto her in a shared bed—are sometimes classified as SIDS deaths anyway, out of sensitivity to traumatized parents grieving the death of their newborn.

“We definitely know that coroners and medical examiners have tendencies to certify deaths differently—and particularly when we are working with the coroner population, because they’re elected officials and oftentimes they’re working in smaller communities where they may be on first-name basis with these families,” Brown said. “It’s difficult for them in some situations to work on a case, and say to the family that something the family did or did not do led directly to the death of their child.”  

“A really important part of a SUID-scene investigation is doing a doll re-enactment,” she added. But it’s difficult to ask a grieving family to participate in such an exercise.

For these and other reasons, many death-scene investigations are not as thorough as they ought to be. Some jurisdictions may simply lack the resources to consistently handle such investigations properly. And although the CDC offers guidance on best practices for death-scene investigations, it has no authority to mandate how such examinations are carried out. Further complicating an already murky situation, is that sudden unexpected infant death does not have any clear biological markers.

“Unlike most causes of death, you can’t do an autopsy and say, ‘Yes, the baby died of SIDS,’” Brown said. “Even suffocation, unfortunately, does not have solid biological markers. There’s huge variation in what takes place during autopsy and death-scene investigation. And because determination of SIDS and SUID are so dependent on those investigations, we’re kind of in a catch-22 here.”

So autopsies alone are not enough—but in most sudden and unexpected infant deaths, there are no witnesses. To begin to unpack the variances in data that flows into the CDC registry, scientists designed an algorithm to help standardize information, and tested it by applying it to hundreds of cases from the registry in 2011. The way it works is by running each case through a series of questions—things like: Any unsafe sleep factors? and Factors for suffocation present?—then using the answers to determine a classification for the death. In a 2014 paper about the algorithm, researchers found that 88 percent of the cases remained unexplained, but that most of them—73 percent—also occurred in an unsafe sleep environment.

In a separate study, published in the journal Pediatrics in April, scientists took a closer look at death scene investigations and autopsy practices among 770 infant deaths across seven states in 2013. Though the vast majority of cases involved an investigation of some kind, comprehensive information about the sleep environment where the baby was found was produced in only 85 percent of cases. In most of the states examined, for instance, investigators noted the status of the baby’s airway when he or she was found less than 80 percent of the time—including just 35 percent of the time in one of the states, which wasn’t identified.

“Right now, things that are getting labeled as SIDS are kind of a mixed bag,” Brown said. “We suspect that the SIDS category is over-inflated because there’s so much variation in what’s getting put into that category. Having a very clear decision matrix to classify [deaths] will hopefully help.”

In other countries, scientists and public health officials are considering SIDS and SUID registries of their own. Ireland already has a national SIDS database. New Zealand is also considered a research leader in the field. And there are similar efforts in the United States to collect data about unexplained deaths in older children.

Jonathan Kaltman is a pediatric cardiologist at the National Heart, Lung, and Blood Institute, where he helps run the National Institute of Health’s Sudden Death in the Young Case Registry. Kaltman and his colleagues are coordinating an effort to track cases of sudden death among young people—scenarios in which a child believed to be healthy drops dead while playing sports, or dies unexpectedly when otherwise engaged in routine activities. As with the SUID Case Registry, data standardization is a concern. But there are other difficulties, too. For instance, when a child dies suddenly but an autopsy doesn’t reveal abnormalities of the heart or other tissue, families often want genetic testing on the deceased. “But insurance doesn’t pay for testing done on patients after they’ve died,” Kaltman told me. Even if a family can pay the “couple thousand bucks” to do the testing, he says, “you need a medical examiner or coroner that had the foresight to keep a vial of blood.”

Despite these obstacles, both registries are aiming to expand. “We would love to see the SUID registry eventually be a national registry,” says Brown. Many researchers agree that gathering data on the broader category of unexplained infant deaths—and not just SIDS—is critical if such an effort is to produce life-saving medical findings.

“A national registry is an excellent idea,” said Peter Blair, an epidemiologist who focuses on SIDS at the University of Bristol in England. “The key advantage of [including both] is that the epidemiology and pathology of sudden but explained deaths can be compared against the sudden unexplained deaths.”

For now, the CDC registry for Sudden Unexpected Infant Death in the United States has reinforced the importance of existing guidelines for sleep safety. In 2011, a pediatrician at the Children's National Medical Center in Washington, D.C., told NPR that, despite intense fear of SIDS among parents, it had been years since they’d seen a “true SIDs case,” in the District; all the babies who had died suddenly and unexpectedly in recent years had been in unsafe sleeping environments. In other words, their deaths could have been prevented.

“Our key takeaway in general is risk reduction,” Brown says. “Making sure babies are always in a safe sleep environment, on a safe surface, on their back.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.