On a recent afternoon, a dark-haired newborn lay sound asleep beneath a blue crocheted blanket in a hospital crib in one of Pennsylvania's poorest cities.
Suddenly, the seemingly serene 8-pound boy was wailing shrilly, making a sound unlike typical infant crying. His balled fists quivered. Almost exactly three hours had passed since the 2-week-old had been given a bottle of formula and his medication — a small dose of morphine squirted into his mouth.
"Just enough to tickle the opioid receptors," said pediatrician Robert Noll, interim chair of pediatrics at Crozer-Chester Medical Center, as he listened to the baby.
About 2 percent of all babies born in Pennsylvania — 2,700 a year — are exposed to addictive drugs while in their mothers' wombs, according to the only comprehensive estimate, widely considered to be a low figure. Most of these newborns, like the boy at Crozer, spend weeks being treated for withdrawal, called neonatal abstinence syndrome, or NAS. They are known to be at higher risk of neglect or abuse when they go home with mothers battling addiction. Yet like most states, Pennsylvania has no up-to-date, complete count of these vulnerable, hapless victims of the opioid crisis.
Their ranks are clearly exploding, especially in areas hardest hit by the epidemic. An Inquirer and Daily News analysis of hospital billing records found the number of drug-exposed babies born in Philadelphia and four surrounding counties doubled in just eight years, from 513 in 2009 to 1,061 in 2016. More than half were diagnosed with NAS and spent an average of three weeks in the hospital. After hospital discharge, a quarter of them needed further medical care, usually from a home health worker, because withdrawal often comes with other complications.
When women use drugs during pregnancy, they are more likely to neglect prenatal care, and their babies are more likely to be born prematurely or underweight. While fetal heroin exposure has been a concern for decades, the explosion in opioid painkiller use is a fairly recent phenomenon that has renewed questions about long-term effects on babies' development, intellect, and academic achievement — effects that are often intertwined with poverty.
At Crozer, which serves the city of Chester, nearly 9 percent of newborns were exposed to drugs, the Inquirer analysis found.
Not coincidentally, Crozer has become a regional leader in developing rehabilitation programs to help addicted mothers and their babies.
The work is often heartrending.
"The name of the game is to keep families intact, but sometimes babies go home to desperate situations," said Noll, the pediatrician. "There is no easy answer."
Last month, in declaring the opioid crisis a disaster emergency, Gov. Wolf made NAS a condition that must be reported to the Department of Health, like whooping cough or measles.
This decision sounds straightforward and sensible.
In reality, it is complicated and full of pitfalls.
Counting drug-exposed newborns has long been a subject of debate among lawmakers, public health experts, and child welfare advocates. There is passionate disagreement about which babies to count, whom to share the information with, and for what purpose.
The governor's order covers only a subset of newborns whose mothers may be struggling: those diagnosed with withdrawal. Babies with suspected prenatal drug exposure, or even confirmed exposure that does not trigger withdrawal symptoms, are being excluded. In the Philadelphia area, that would be nearly 500 babies in 2016.
"Exposure alone does not qualify for reporting; the infant must show symptoms to be reported," says the health department instructions.
Another limitation on this surveillance: The governor's declaration lasts just three months, through April 10, with an option for a three-month extension. Even six months of data isn't much, public health experts say, given that the ambitious goal of the reporting is "to describe the burden of NAS in Pennsylvania, identify high incidence locations for targeted intervention and to reduce the statewide incidence of NAS."
State physician general Rachel Levine, the acting secretary of health, said the Wolf Administration is "working to make NAS a reportable condition to the department moving forward" after the declaration expires.
NAS was first formally described in the 1970s by Philadelphia pediatrician Loretta Finnegan. It remains a challenging condition to diagnose and report.
Although a woman's drug use during pregnancy can be confirmed by testing the baby's urine and first bowel movement, withdrawal is a cluster of symptoms with a range of severity — not an unmistakable disease. Most hospitals use a scoring system developed by Finnegan to assess NAS severity and the effectiveness of treatment. The hallmark symptoms are shrill, inconsolable crying and tremors. But the baby may also have muscle spasms, convulsions, vomiting, diarrhea, insomnia, trouble feeding, fever, nasal stuffiness, scratching, yawning, sweating.
Most NAS babies need decreasing doses of morphine or methadone, but some get sufficient relief with cuddling and swaddling. Those whose mothers took stimulants, such as cocaine or methamphetamines, don't display the classic signs of withdrawal from prescription opioids and heroin. With so much variation, medical professionals have to use their own judgment when applying internationally accepted diagnostic codes used for billing. Researchers who use these codes to track drug-exposed babies, meanwhile, may underestimate the true problem.
"One of the problems with the diagnosis of NAS is that there is no national or state guidance on when to code it," said Debra Bogen, a pediatrician and NAS expert at Children's Hospital of Pittsburgh of UPMC. "It's a messy term."
Levine, the state health official, acknowledged the difficulty.
"NAS is a clinical diagnosis. It involves lab testing," she said. "But some of the symptoms are nonspecific in terms of a jittery baby. It is not always obvious and can be challenging."
The most comprehensive statewide count of drug-exposed newborns was done by the Pennsylvania Healthcare Cost Containment Council, an independent state-funded agency. The analysis, now out of date, used hospital billing records to find babies who were diagnosed with withdrawal, as well as those who were exposed to addictive drugs. (The council declined to update its analysis for this story.)
Between 2000 and 2015, the annual statewide number of babies exposed to drugs soared from 788, or 0.5 percent of all newborn hospital stays, to about 2,700, or 2 percent of newborn stays. The vast majority went through withdrawal. The care of these infants added 27,000 days to hospitalizations, at an extra cost of more than $20 million, the council analysis found.
While state health officials seek to understand how many babies are born with NAS and where they are located, their colleagues at the state Department of Human Services (DHS) have been caught up in a different debate about identifying babies exposed to drugs. At issue is how to keep infants safe without stigmatizing or punishing their mothers.
One side of the debate argues that hospitals should report all drug-exposed babies to local child protection agencies, while the other side says only babies whose mothers used illegal drugs should be reported.
Behind the distinction are good intentions. Many women work to control their misuse of legal or illegal drugs with counseling and prescribed opioids — namely, methadone or buprenorphine — that prevent wrenching withdrawal symptoms. Some are taking prescribed opioids for chronic pain under a doctor's care. What's more, women who are motivated to quit drugs altogether when they discover they're pregnant are typically prescribed methadone, because going cold turkey increases the risk of miscarriage.
But newborns exposed to legally prescribed opioids still can suffer withdrawal — and still can be at long-term risk of neglect or abuse, given the chronic relapsing nature of addiction.
The federal child abuse prevention law has been amended repeatedly to require that all newborns "affected by" prenatal drug exposure be reported, so child protection workers can evaluate whether the families need supervision and services. But this mandate comes with little funding, and no penalties. Many states, including Pennsylvania, are trying to figure out how to comply.
James C. Greenwood, who represented Pennsylvania in Congress for six terms, pushed for this comprehensive approach, stressing that the goal is to help families, not target struggling women.
"In crafting the federal law, I never envisioned that the 'referral' from a health care provider was the same as a child abuse report," said Greenwood, a former child protection caseworker.
Pennsylvania's child abuse prevention law, meanwhile, has gone back and forth. In 2015, the law was amended so that only babies exposed to illegal drugs had to be reported to child welfare agencies. Last year, state Rep. Katharine Watson (R., Bucks) introduced a bill to revert to the older language, which covered any baby with withdrawal symptoms.
"I am biased in this debate. I'm biased toward the child," she said. "Putting the onus all on the county children and youth services is wrong, but they [addicted mothers] should come under some kind of watchdog care."
No matter what the laws dictate, she believes fear of disgrace will shape what actually happens.
"I will tell you that in many parts of Pennsylvania, the deal is begging the doctor: 'Please don't report it. We're getting our daughter help.'" Watson said. "I think there are hospitals and doctors who don't put it down, and they think they're doing the best thing for the mom."
There are also hospitals that believe full disclosure is the best way to deal with conflicting directives. Abington Hospital-Jefferson Health reports all exposures, including marijuana. Denise Ellison, the nurse practitioner who has compiled the information for 14 years, relies on drug screening tests rather than billing records, which may accidentally omit an exposed baby.
"For me, the biggest challenge is how much of it there is now," Ellison said. "What's hard is the moms who are caught in that place where they don't want to confess and, lo and behold, the [drug] screen comes back positive. You really want to believe them. There are some really tough situations out there."
Four years ago, Cathleen Palm, a longtime child advocate, was reviewing summaries of fatal child abuse cases across Pennsylvania when she was struck by a recurring theme — a history of newborn drug withdrawal.
Her nonprofit group, the Center for Children's Justice, began asking the state Department of Human Services for data on NAS, starting with babies covered by Medicaid, the health insurance program for the poor.
"It was an insane process that took eight months to get data," Palm recalled from her home and office in Berks County. "They denied our right to know. Then they said, 'We don't have this data set, so we have to create it.' "
That much has changed.
For this story, DHS readily shared updated Medicaid data, showing NAS diagnoses climbed steadily from 1,080 in 2010, to 2,369 in 2016 — 3.5 percent of all Medicaid births.
But it is hard to reconcile this number with the far lower tally of newborns reported to county child protective agencies because of either withdrawal or drug exposure. In 2016, there were 972 such reports, according to DHS.
The state health department has an even more puzzling set of data that apparently isn't verified or used. Each year since 1996, an annual questionnaire completed by all hospitals as part of the licensing process has included "live births exposed to illegal drugs before birth."
The number has risen steadily for a decade and hit almost 3,900 in 2016, or about 3 percent of the 134,000 births that year. That's higher than the rate from the health care council's hospital billing data analysis, which was not limited to illegal drug exposures.
The 2016 questionnaire lists no data at all for this benchmark from Crozer — or from Jefferson University Hospital and Temple University Hospital, all of which delivered a large number of drug-exposed babies. Neither state nor hospital officials could explain the apparent oversight.
"What I have heard consistently from a number of people involved in this issue is that the commonwealth doesn't have a complete, timely database and that is hampering our ability to understand" the NAS crisis, said Joe Martin, executive director of the health care council.