Secret recordings captured physicians' concerns that more children seemed to fare poorly after heart surgery. Their hospital kept doing the operations. Ellen Gabler of The New York Times investigates.
Tasha and Thomas Jones sat beside their 2-year-old daughter as she lay in intensive care at North Carolina Children's Hospital. Skylar had just come out of heart surgery and should recover well, her parents were told. But that night, she flatlined. Doctors and nurses swarmed around her, performing chest compressions for nearly an hour before putting the girl on life support.
Five days later, in June 2016, the hospital's paediatric cardiologists gathered one floor below for what became a wrenching discussion. Patients with complex conditions had been dying at higher-than-expected rates in past years, some of the doctors suspected. Now, even children like Skylar, undergoing less risky surgeries, seemed to fare poorly.
The cardiologists pressed their division chief about what was happening at the hospital, part of the respected University of North Carolina medical center in Chapel Hill, while struggling to decide if they should continue to send patients to UNC for heart surgery.
"I ask myself, 'Would I have my children have surgery here?'" said Dr. Blair Robinson. "In the past I'd always felt like the answer was 'yes' for something simple. ... But now when I look myself in the mirror, and what's gone on the past month, I can't, I can't say that. And if I can't say it for my kids — and that should be our group discussion — if we can't all look ourselves in the mirror and think we're doing the right thing, then we need to change what we're doing."
That March, a newborn had died after muscles supporting a valve in his heart appeared to have been damaged during surgery. At least two patients undergoing low-risk surgeries had recently experienced complications. In May, a baby girl with a complex heart condition died two weeks after her operation. Two days later, Skylar went in for surgery.
In the doctors' meeting, the chief of paediatric cardiology, Dr. Timothy Hoffman, was blunt. "It's a nightmare right now," he said. "We are in crisis, and everyone is aware of that."
That comment and others — captured in secret audio recordings provided to The New York Times — offer a rare, unfiltered look inside a medical institution as physicians weighed their ethical obligations to patients while their bosses also worried about harming the surgical program.
In meetings in 2016 and 2017, all nine cardiologists expressed concerns about the program's performance. The head of the hospital and other leaders there were alarmed as well, according to the recordings. The cardiologists — who diagnose and treat heart conditions but don't perform surgeries — could not pinpoint what might be going wrong in an intertwined system involving surgeons, anaesthesiologists, intensive care doctors and support staff. But they discussed everything from inadequate resources to misgivings about the chief paediatric cardiac surgeon to whether the hospital was taking on patients it wasn't equipped to handle. Several doctors began referring more children elsewhere for surgery.
The heart specialists had been asking to review the institution's mortality statistics for cardiac surgery — information that most other hospitals make public — but said they had not been able to get it for several years. Last month, after repeated requests from The Times, UNC released limited data showing that for four years through June 2017, it had a higher death rate than nearly all of the 82 institutions nationwide that do publicly report.
UNC Health Care defends the surgery program, describing it as "very strong" today and citing its most recent data to support that. It denies any past problems affecting patient care.
"We determined," said Dr. Benny Joyner, chief of critical care at the children's hospital, that "there is nothing here that is systematic, or systemic that would lead us to be concerned about the performance of operations on children that are high-risk, low-risk, no-risk."
Other administrators, in a joint interview, said there was "a dysfunctional group" in 2016 that sowed mistrust, creating "team culture issues." Lisa Schiller, a spokeswoman, said in a statement, "They were handled appropriately, and today we have new team members." UNC cited leadership changes — most taking effect in 2017 or 2018, including the appointment of a new chief surgeon last year — to help improve the dynamics.
The turmoil at UNC underscores concerns about the quality and consistency of care provided by dozens of paediatric heart surgery programs across the country. Each year in the United States about 40,000 babies are born with heart defects; about 10,000 are likely to need surgery or other procedures before their first birthday.
The best outcomes for patients with complex heart problems correlate with hospitals that perform a high volume of surgeries — several hundred a year — studies show. But a proliferation of the surgery programs has made it difficult for many institutions, including UNC, to reach those numbers: The North Carolina hospital does about 100 to 150 a year. Lower numbers can leave surgeons and staff at some hospitals with insufficient experience and resources to achieve better results, researchers have found.
"We can do better. And it's not that hard to do better," said Dr. Carl Backer, former president of the Congenital Heart Surgeons' Society, who practices at Lurie Children's Hospital of Chicago. "We don't have to build new hospitals. We don't have to build new ICUs. We just need to move patients to more appropriate centers."
At least five paediatric heart surgery programs across the country were suspended or shut down in the last decade after questions were raised about their performance; a Florida institution run by the prestigious Johns Hopkins medical system stopped operations after reporting by the Tampa Bay Times in 2018. At least a half-dozen hospitals have merged their programs with larger ones to achieve more consistent results. And more institutions are considering such partnerships.
There is no way to know if Skylar or other patients at the North Carolina hospital would have survived or done better had they been treated elsewhere. Surgeries to repair a child's heart can be risky, and some deaths and complications are inevitable. But doctors and advocates have pushed to make paediatric heart surgery outcomes more transparent, so institutions would feel pressure to improve and parents could better assess the care their child might receive.
About 75 per cent of the roughly 115 hospitals that perform paediatric heart surgery in the United States publicly share their mortality statistics on a website run by the Society of Thoracic Surgeons. UNC is not among them.
That information is risk-adjusted to help account for prematurity, some genetic abnormalities and other factors that could make a child less likely to survive, and to more fairly assess hospitals that take on the most compromised patients. The statistical method also helps evaluate if hospitals are losing patients who would not be expected to die. While there is some debate about whether the data fully captures the complexity of heart surgery, many experts say it is a strong indicator of a hospital's performance.
"You should think twice about going to a center that doesn't publicly report," Backer said. "People don't buy a car without knowing what the gas mileage is."
UNC Health Care contends that using the mortality data isn't an accurate way to assess its surgery program. The health system first told The Times it was "critically important" to use risk-adjusted data but then released only raw, unadjusted numbers. The hospital later said no current risk adjustment adequately accounted for the breadth and severity of its patients' medical issues. The Times is suing UNC for the risk-adjusted data.
This account is drawn from interviews with about two dozen current and former UNC physicians and nurses from multiple departments, as well as with patients' families and doctors from other institutions. Some people spoke only on the condition of anonymity, for fear of retribution from UNC. The Times also reviewed emails, state death data, medical records and the audio recordings, obtained from someone who felt the institution had not been responsive to repeated concerns about the surgery program.
The recordings caught hours of cardiology department meetings and other discussions that were corroborated by multiple participants. The Times gave UNC transcripts of relevant portions of the conversations and some audio clips but did not share the full recordings to protect the identity of the person who provided them.
UNC said it would not comment on some of The Times' findings without an opportunity to listen to the recordings. Dr. Michael Mill, the chief surgeon at the time; Hoffman; and the four other cardiologists who remain at the hospital declined for the same reason. Dr. Kevin Kelly, who led the children's hospital until retiring last year, did not respond to repeated requests for comment.
He met with the cardiologists in 2016, soon after they laid out their concerns to Hoffman. When discussing where to refer patients, Kelly advised: "Do what your conscience says."
But, he warned, performing fewer surgeries at UNC could hurt revenues and cost the cardiologists their jobs. "If it reduces the volume of things," he said, "I'll just — we'll just reduce the number of people that we have."
"We all took the oath"
Hoffman had been at the children's hospital just over a year when he was confronted with the cardiologists' mounting concerns. When he sent out an agenda for a routine department meeting in June 2016, one physician, Robinson, responded that the only topic should be the surgical results and said an emergency faculty retreat might be in order. Two others quickly agreed.
In their hour long session days later, the chief acknowledged things hadn't been going well. "We are nowhere near where we need to be," Hoffman said, according to a recording. Three children who had recently undergone low-risk procedures had experienced "issues," he said; two ended up on mechanical support, an unusual and potentially dangerous situation.
The goal for that meeting, Hoffman said, was "to figure out where to go from here." He said he and his counterparts from other specialities had emergency conversations about options, including suspending surgeries and creating a task force to scrutinise the entire system — from critical care to surgery to anaesthesia. "This is a program issue. It's every aspect," Hoffman said. "We need to do something."
The physicians pointed to possible vulnerabilities. In 2015, UNC had lost two paediatric cardiac intensivists, who care for critically ill heart patients. The children's hospital also had closed its CIC unit, a specialized area for cardiac patients coming out of intensive care; nurses experienced with those patients left. And unlike some larger hospitals, UNC did not have a dedicated cardiac intensive care unit, or CICU.
Some of the doctors worried that a task force could take too long. "How do we get a change achieved within a reasonable amount of time when we all have this quandary as to what to do with our patients?" asked Dr. Sunita Ferns. She and the three other cardiologists who have since left UNC declined to comment for this article.
Referring patients to other hospitals had been "discussed at a high level," although nothing had been decided, Hoffman said. While noting that patient care was "paramount," he cautioned that referring patients elsewhere could have long-term implications. Administrators might stop investing in the program, he said, as those at some other hospitals were considering.
In a recent statement, UNC said that sending patients to other medical centers had been discussed but was determined not to "be in the best interest of UNC or its most vulnerable patients." Separately, the five remaining cardiologists, who include Hoffman, said in a statement that they had always "selectively referred particularly complex patients to outside institutions" but sent "the large majority of patients to our internal team at UNC, because we were confident that they would receive high-quality care."
The cardiologists acknowledged in the meeting that multiple factors went into successfully managing cardiac patients but expressed doubts about the chief heart surgeon at the time, Mill, who performed most of the operations. Just one other surgeon, who was more junior, was then on staff. They did not report directly to anyone inside the children's hospital but to the department of surgery within UNC's broader medical system, which includes five hospitals in a Chapel Hill complex and about a dozen other facilities around the state.
A few weeks earlier, Mill had not come in on a weekend to perform a transplant on a baby when a donor heart became available, the doctors said. The surgeon's reason wasn't clear, but the incident could have led to a shutdown of UNC's program if the United Network for Organ Sharing, which manages transplants throughout the country, had found out, Robinson said in the meeting.
The child's parents, who were not aware of what happened, soon transferred her to nearby Duke Children's Hospital, where she would eventually receive another heart a few months later. But the episode angered several UNC cardiologists.
"As a mother of three children, oh my God. ... It's inexcusable," said Dr. Jennifer Whitham. "As a physician, I mean, we all took the oath. We are supposed to do what's right for our patients. ... This is what you signed up for. And who is he to play God with some kid's life? I can't get past this. This is beyond horrifying."
Robinson, who has since left UNC, said in the meeting that he worried what it meant for other transplant patients if they could not count on the surgeon to be available. Hoffman told the group he did not think the surgeon had offered an explanation for not doing the transplant. But no matter the reason, he said, the episode suggested that UNC lacked the infrastructure for that type of care.
In a statement, UNC said that under state law it could not discuss personnel information but noted that no surgeon is on call 365 days a year.
Mill, now 65, has worked at the hospital for three decades. Parents of some patients described him as reassuring and skilful. Dr. Melina Kibbe, who leads UNC's surgical department, called him an "outstanding surgeon." He had never been subjected to disciplinary action by the North Carolina Medical Board and was sued in the state once for malpractice, but the case was voluntarily dismissed by the patient who filed it.
Colleagues said he was often affable but could bristle when questioned about a case or appear aloof in group discussions. Several physicians said he had not been an effective mentor or leader, a role a chief surgeon is typically expected to play.
Hoffman told the group that he had spoken to the surgeon, letting him know the cardiologists might discuss sending patients to other hospitals. "He's frustrated with the results as well," Hoffman said. "He's wondering what's going on, himself — with himself." But, the chief went on, he believed that Mill was also looking at things that could be improved, like a more consistent operating room team.
Several cardiologists expressed a lack of confidence in the surgeon, without specifying why. Dr. Pamela Ro, who is Hoffman's wife, and Dr. John Cotton said the hospital needed a new chief surgeon. Two others said they were not comfortable sending patients to Mill.
Some doctors also complained that Mill was not providing mortality data intended to help hospitals assess their performance, improve results and identify trends beyond just one surgery or death. The Society of Thoracic Surgeons organises and audits the data collection and prepares extensive reports for each institution.
It also rates hospitals. UNC has one star, the lowest rating. The two other hospitals in the state with comprehensive paediatric heart programs are Levine Children's Hospital, in Charlotte, with two stars, the middle rating; and Duke, in Durham, which will be upgraded to two stars this summer, according to a report provided by the institution. Each of those hospitals performs a larger proportion of high-risk surgeries than UNC, according to a data analysis, and two to three times as many heart surgeries overall.
UNC told The Times that during part of 2014 and 2015 it lacked a database coordinator — a position it said was difficult to fill — and did not submit its data for analysis then. The cardiologists acknowledged this during the meeting but said Mill should nevertheless have shared basic statistics that might help them advise parents.
Administrators also told The Times that cardiologists could have tracked patient outcomes by attending morbidity and mortality conferences, where cases are discussed in detail. But the information the doctors sought — broader data to identify trends over time — was not available in those meetings, several said.
Hoffman told his team in the session that each department, including critical care, anaesthesia, surgery and cardiology, would do a "deep dive" to look at its processes. He later added that things would be starkly different in two years, noting that UNC might not even have the surgical program anymore. "This is the fork in the road," he concluded.
But, he said about his patients, "I'm not sending them out of this hospital right now."
"He's not going to make it"
Skylar Jones was a few months old when a cardiologist diagnosed a hole between the upper chambers of her heart, one of the most common and treatable heart defects. Surgery to correct it could wait until she was older, doctors said. She grew into a chatty toddler who bossed around her two older brothers.
By the spring of 2016, Skylar's mother noticed the 2-year-old was sweating more than usual. Tasha Jones could feel her daughter's heart racing when she held her close. The right side of Skylar's heart was enlarged, a UNC cardiologist said. She would need surgery soon, her parents were told, but her prospects were good.
Skylar had no other major health problems and was a "normal, active child," according to medical records. (Thomas and Tasha Jones and other parents shared their children's medical records with The Times.)
The operation in June took about seven hours, twice as long as planned. Mill said the surgery had been more complicated than expected, Tasha Jones recalled, but Skylar should do well.
In fact, her diagnosis before surgery had been inaccurate, according to her medical records. Instead of the more common condition, a vessel going into the right side of her heart was not fully enclosed, allowing blood to leak into the wrong chamber. The condition, known as an unroofed coronary sinus, required a different repair by Mill but one still considered relatively low-risk, with a mortality rate of about 2%.
But even given the procedure's low risk, and that a child of Skylar's age and overall health would typically be expected to fare well, there were complications. Records show that during surgery, Skylar's heart stopped beating properly; Mill put in temporary wires to send electrical signals to the heart.
That night in the intensive care unit, a critical care doctor was called to the toddler's bedside: Her blood pressure was low. She struggled to breathe shortly before she went into cardiac arrest. She was put on life support, with a machine doing the work of her heart and lungs. That puts patients at risk for a cascade of other problems including bleeding, kidney damage and stroke.
After five days, Skylar improved enough to come off life support. But her condition fluctuated over the next seven weeks. According to the medical records, her vocal cords, or the nerve controlling them, had been damaged. That can happen during surgery or when a person is put on life support, making it possible for food or liquid to slip into the windpipe. Eventually, Skylar inhaled something into her lungs and ended up back on life support. Several times, she had bleeding in her lungs.
By late July, Skylar seemed to have turned a corner. But one day, while watching "VeggieTales," her favorite TV show, blood began dripping from her mouth, her mother recalled. Nurses said she was all right, but hours later it happened again. Her eyes rolled back in her head. Doctors rushed her into surgery to find the source of bleeding. After more than 30 minutes of CPR, she died.
Skylar's autopsy noted multiple postoperative complications and said she had been chronically treated with blood thinners that interfered with her body's ability to properly clot. She died with severe bleeding in her digestive tract. The autopsy report also noted that her heart might have been damaged after being deprived of oxygen during rounds of CPR, including the nearly hour long effort the night of her heart surgery.
At home in Holly Springs, North Carolina, Skylar's bedroom door remains closed. Tasha Jones has gone in only once in the nearly three years since her daughter died. One of Skylar's brothers, Nathan, who is 11, taped a sign on the door to keep his friends out. The closet inside is still full of small pink clothes.
Skylar's parents, of course, were not aware back then that doctors were concerned about the hospital's cardiac surgery program. "We could have maybe went elsewhere," Thomas Jones said in a recent interview. "But all that, I feel, was taken away."
"I just wish somebody would have said something," Tasha Jones said.
Like other health institutions, UNC said, it is prohibited by federal law from discussing individual patients. The hospital said details about patients' cases in this article were inaccurate or incomplete but declined to discuss any specifics, even if the parents signed a waiver allowing it.
Another parent, a registered nurse named Ana Crow, remembers Mill telling her he had treated cases like her newborn's before, typically with good results. Her 5-pound baby, Adam, was rushed to UNC Children's Hospital from a Raleigh hospital in March 2016, soon after his birth. He had transposition of the great arteries, a condition where the two main arteries carrying blood out of the heart are reversed.
In the operating room, it became clear that Adam's condition was slightly more complicated. In addition to swapping the arteries, Mill needed to detach and reattach smaller blood vessels, which was more challenging given their unusual placement, according to his medical records. Still, the procedures are considered relatively low-risk, with a mortality rate of about 2%.
After the surgery, Mill sat on the edge of Crow's bed, she recalled. "As soon as he said, 'He's not going to make it,' I just tuned out," Crow said. "You go through denial."
Muscles supporting a valve in the baby's heart were damaged, according to Adam's medical records and autopsy report, apparently when Mill tried to remove a band of tissue in the heart. Without a functioning valve, the child couldn't survive.
Adam was brought out of the operating room on life support. He was 1 week old when he was disconnected from the machine. As his mother held him for the first time, he took his last breath.
"I've heard you loud and clear"
When Kelly, then the head of the children's hospital, met with the cardiologists in late June 2016 to hear their concerns, he said he had also spoken with people in other departments about the problems.
"They are not worried about you as cardiologists; they are worried about the surgical end of this and it's a very broad worry right now," he said. "You are the front-line filter."
Kelly listened as one physician after another said that things seemed to be going badly for too many young patients. "The cardiologists that are talking to you in this room today are saying that we are not on track," said Dr. James Loehr, who is no longer at UNC. "And I would argue that we have not been on track for some time."
In discussing options, Kelly, an allergist and immunologist who had become interim head of the hospital just a few months earlier, acknowledged the moral dilemma for the doctors, whom he described as a "very, very upstanding group of cardiologists." But in the two-hour session, he also clearly felt the weight of the situation and the possible consequences for his institution.
"I've been worried about this," he said. "The implication of somebody like me declaring 'paediatric cardiac surgery moratorium has occurred' will spread like wildfire from North Carolina, and this will never be the same for five or 10 years."
He mentioned he had seen the hospital's data for several years through 2013, which showed a mortality rate of 6 per cent. (The national average was 3.5 per cent at the time.) The Times obtained UNC's information from 2010 to 2013, showing 38 deaths out of 623 surgeries. The data is not risk-adjusted and has never been publicly disclosed by the hospital.
Kelly added that the hospital was working on more recent data and said that while he believed the statistics "were OK" for a while, "we have just hit a really rocky piece of road," according to a recording.
Kelly: "And what you just described to me is, surgery — cardiac surgery — here is not in a good situation, right? And you know it's bad and I don't have to give you the data."
Loehr: "We don't need the data. We know it."
Kelly: "And I've heard you loud and clear. I don't know how long it's going to take to do this. A lot of these things that get vetted with large groups of people, the final decisions of what we're going to do happen behind closed doors in offices — you know, I think of scotch and cigar smoke and things like that. You know, some of these decisions about what we're going to do: Is it a junior surgeon? Is it a senior surgeon? Is it a relationship with another local university? Is it a relationship with a further university? Is it close down the program?"
Kelly noted that having two surgeons is a minimum standard. But, he said, he would never recommend that "another junior surgeon walk into this institution."
Several had come and gone over the years, working under Mill. "Each time, it has resulted in a negative outcome for the junior surgeon," Kelly said. "We've killed four." He and the cardiologists agreed that the junior surgeon at the time seemed likely to leave soon. In fact, she did leave later that summer, and for nearly two years afterward Mill was the hospital's only cardiac surgeon.
The best option, Kelly said, was to combine UNC's surgery program with Duke's. For years, physicians at both children's hospitals talked informally about joining forces, but nothing came of it. They were "basically destroying each other's capacity to be great," Kelly said, by running competing programs less than 15 miles apart. But even combining the programs was not an instant fix: It would take at least a year and a half, he said.
Kelly went around the room, asking if there were any patients the cardiologists would refer to Mill for surgery. One doctor said no. A second said she had a hard time recommending UNC even to parents of low-risk patients. Another, Cotton, suggested he might send some of the simplest cases, adding that he always encouraged parents to get a second opinion at top-ranked hospitals in Philadelphia, Boston or Michigan if the family could afford it.
Not everyone had a chance to answer Kelly, but five of the nine physicians expressed hesitation about sending certain types of cases to their own hospital. Dr. Scott Buck noted that there had been "some great recent outcomes." But, he said of referring patients, "I do feel increasingly morally, ethically uneasy about this."
Kelly said he would take the doctors' concerns to leaders above him — deans in UNC's medical school — and the incoming chairwoman of the surgery department.
He told the cardiologists to let ethics guide their decisions about referring patients. But he also noted that a drop in internal referrals would hurt revenues and could lead to job cuts. "'Cause that's all I can do," Kelly said. "All I get to do is manage money."
Soon after, an anonymous letter was sent to the UNC chancellor and hospital leaders. The letter, a copy of which was shown to The Times, complained about Kelly's "threat" to eliminate doctors' jobs if they stopped referring patients internally and warned of "unsafe and unethical practices" in paediatric cardiology and cardiac surgery.
In a statement, UNC Health Care said that the institution had conducted a "thorough internal investigation" of the concerns described in the letter and that "criticism of the program was found to be unsubstantiated."
The inquiry involved looking at specific cases and interviewing "key informants," said Dr. Thomas Ivester, UNC's chief medical officer. Someone also sent a complaint with similar allegations to the state medical board, which found them baseless, he said. The board's findings are confidential unless public action is taken.
UNC would not say whether a task force with outside experts, which was discussed in the 2016 meetings, was ever assembled, or what became of the "deep dives" by individual departments Hoffman had mentioned.
Several physicians said they had not been briefed on any findings from internal inquiries. UNC said in a statement that "not each and every physician would have been involved and aware of reviews" of other personnel.
"A pilot who does this every day"
At a conference last fall, Backer, the Chicago heart surgeon, urged fellow surgeons to consider "rational regionalisation," or joining forces in an effort to reduce mortalities nationwide for congenital heart defects, potentially saving hundreds of lives.
Reaching adequate case volumes to keep up skills is a challenge because so many hospitals are competing for patients — surgical programs are an important driver of revenue. The Orlando, Florida; Phoenix; and San Antonio metropolitan areas, for example, each have three hospitals doing paediatric heart surgeries. Cleveland has two about a mile apart. A study last year by Backer and other physicians found that 66% of hospitals doing the surgeries were within 40km of another one.
Several countries have consolidated programs to help ensure better, more consistent care. In Sweden, mortality rates dropped to 1.9 per cent from 9.5 per cent after surgeries in the early 1990s were consolidated to two hospitals from four. In the United Kingdom, the mortality rate fell from 4.3 per cent to 1.9 per cent — despite an increase in complex cases — after a series of babies' deaths led to an overhaul in the late 1990s. Hospitals there must have multiple surgeons who each perform at least 125 operations annually.
There are no such requirements in the United States, although some hospitals have formed partnerships to better serve patients. Children's Hospital of The King's Daughter in Norfolk, Virginia, joined in 2017 with the University of Virginia Children's Hospital, nearly three hours away in Charlottesville. Dr. James Gangemi, the chief of congenital heart surgery at UVA, collaborates with surgeons at both hospitals.
The goal is to provide care for patients as close to their home as possible, while minimising risk. Children with the most complicated conditions have surgery in Charlottesville, where the hospital has a transplant team and cardiologists who specialise in paediatric heart failure.
"When you get on an airplane, you want a pilot who does this every day," Gangemi said. "By having a dedicated team, you reduce variability."
Glimpsing the data
In February 2017, for the first time in several years, the UNC cardiologists got a glimpse of their hospital's heart surgery mortality data. They were ushered into a conference room, along with other staff and Mill, who presented some of the information, according to a recording.
Doctors were told that the rate had improved in recent years, but the program still had one star. The physicians were not given copies or summaries of the statistics and were cautioned that the information was considered confidential by the Society of Thoracic Surgeons. In fact, surgeons at other hospitals often share such data with cardiologists from competing institutions.
While UNC said in a statement that it was "potentially reckless" to use the data to drive decision-making about where to refer patients, doctors across the country said it was simply one factor, among several, that should be considered.
In October 2017, three babies with complex conditions died after undergoing heart surgery at UNC. In a morbidity and mortality conference the next month, one cardiologist suggested that UNC temporarily stop handling some complex cases, according to a person who was in the room. Kibbe, the surgery department chairwoman, said in a recent interview that the hospital had never restricted surgeries.
In December, another child died after undergoing surgery a few months earlier for a complex condition.
The four deaths were confirmed by The Times but are not among those disclosed by UNC. It has declined to publicly release mortality data from July 2017 through June 2018, saying that because the hospital had only one surgeon during most of that period, releasing the data would violate "peer review" protections.
Other information released by UNC shows that the hospital's cardiac surgery mortality rate from July 2013 through June 2017 was 4.7 per cent, higher than those of most of the 82 hospitals that publicly report similar information. UNC says that the difference between its rate and other hospitals' is not statistically significant but would not provide information supporting that claim. The hospital said the numbers of specific procedures are too low for the statistics to be a meaningful evaluation of a single institution.
UNC also said it is the state's "safety-net hospital," serving "a high volume of Medicaid and uninsured patients with complex medical and socio-economic factors that complicate their conditions." It contends that no risk adjustment can account for those factors. But other hospitals in North Carolina and across the country also treat a high percentage of Medicaid patients and publicly report their data.
The surgeons' society frequently works to improve its statistical models. "It's better than anything else that is out there, for sure, but it's not perfect," said Dr. John Mayer, a surgeon at Boston Children's Hospital, who is leading an effort now.
UNC has said it will publicly release its mortality data through the Society of Thoracic Surgeons once it has had at least two paediatric surgeons for several years. Last June, Dr. Mahesh Sharma, previously of Children's Hospital of Pittsburgh, replaced Mill as the chief surgeon. The hospital has said that was part of a succession plan as Mill nears retirement.
The hospital in April released raw data for the nine months Sharma had been at UNC, showing two deaths out of 82 surgeries — a mortality rate of 2.4 per cent. That information does not meet the Society of Thoracic Surgeons' standards for public reporting, which require, among other things, four consecutive years of data to more reliably assess how a program performs over time.
UNC says its recent results show that its surgical program is performing well, thanks in part to years of quality-improvement efforts including extensive reviews of each death; specialized training for nurses; and protocols to decrease infections, standardise preoperative care and manage post-surgical patients with irregular heart rhythms.
By 2018, four of the nine paediatric cardiologists had left the hospital. A new cardiologist was hired last year, and UNC says three more will start this summer. Although UNC maintains that its surgery program had adequate resources in 2016, it is now taking steps the cardiologists discussed back then, like hiring more cardiac intensive care doctors and opening a cardiac intensive care unit.
"The program today is not the program from 2016, and the program in 2016 was not the program from 2012," said Ivester, the chief medical officer. "We are continually critically appraising and evaluating our performance and opportunities to improve."
UNC has not joined forces with Duke. Last summer, Duke formed a partnership with East Carolina Heart Institute in Greenville, about 100 miles away, handling all paediatric heart surgeries for the hospital.
In recent weeks, as The Times story neared publication, medical personnel reached out to the families of some patients who had died or had complications, offering to discuss their cases. UNC began searching for a paediatric heart surgeon who could fill in at the hospital on a temporary basis, according to recruitment emails. Mill is taking a leave for family medical reasons, according to UNC.
"We trusted them"
In May 2018, Jeremiah Moore came to the North Carolina hospital for what was expected to be his final surgery. The child's condition, hypoplastic left heart syndrome, requires a series of three surgeries to reconstruct the heart and blood vessels.
He had the first two operations at UNC — including the riskiest, which has a mortality rate of 15 per cent — but given the complexity of his condition, his parents, Quanasia Dean and Sean Moore, said they considered taking him to a Philadelphia hospital that specialises in the third procedure.
They had gotten the idea a year earlier, when Jeremiah's cardiologist had said she would send referrals for second opinions to the Charlotte medical center and Children's Hospital of Philadelphia. Part of Jeremiah's aorta had started to narrow — a possible complication from his second operation — putting strain on the heart. While the third surgery could likely wait a year or two, the aorta problem might need to be addressed sooner.
By the beginning of 2018, Jeremiah saw a different cardiologist at UNC, Dr. Elman Frantz. After Frantz discussed Jeremiah's case with other physicians at the hospital, a plan was made for the almost 2-year-old to have his third surgery there that spring, according to his medical records.
When the hospital called Jeremiah's mother to schedule the operation, she was surprised, she said, because of the earlier recommendation about the Philadelphia hospital. It reports a mortality rate of less than 1% for the surgery, known as the Fontan procedure; the national rate is 1.1 per cent.
When Dean called Frantz to ask about taking Jeremiah to Philadelphia, he assured her the surgery could be done at UNC, she said in an interview.
Nearly two years earlier, Frantz had said in the meeting with Kelly that the cardiologists had come to feel that UNC wasn't "where we need to be for complex cases," according to a recording. "We all sort of have our own take on the results and it's not favorable, and we've sort of lost confidence in the program," he said.
Jeremiah was on Medicaid, which can make it harder to go out of state, but his mother said Frantz never mentioned that as a reason to keep him at UNC.
It is unclear whether the cardiologist considered Jeremiah's case to be complex, but Mill, the surgeon, noted that the little boy was a "high-risk candidate" last May, shortly before his operation, according to his medical records.
The surgery took twice as long as expected, his mother said, but after about 11 hours, he was wheeled into the paediatric intensive care unit.
Over the next two weeks, he struggled. He developed an irregular heart rhythm. A test showed his heart was not functioning well. A cough doctors had noted before the operation progressed into a cold, a dangerous development for a child with Jeremiah's condition. His organs began to shut down. Early one morning, his heart stopped and he died.
"They said Jeremiah was strong," said Inez Holmes, the boy's grandmother. "We trusted them."
Written by: Ellen Gabler
Photographs by: Travis Dove
© 2019 THE NEW YORK TIMES