The Philadelphia case soon took a turn, when the mother had another baby. At 7 weeks, the infant was diagnosed with a fractured right leg. Holick was interviewed again and said there was “a very high degree of medical certainty” that EDS could explain the injury. “I wouldn’t hesitate to have the children returned to the parents,” Holick said.

Seven months after Frasier complained to Holick’s dean, a pediatric geneticist at the Children’s Hospital at Albany Medical Center also raised concerns about him with B.U. Natasha Shur wrote in an e-mail to Holick’s department head, Alan Farwell, that Holick’s expert findings lacked scientific support. “I hope that you are interested in considering these issues, and the effect of BU and its involvement,” she wrote. Farwell replied that he was aware of Holick’s work and encouraged Shur to invite Holick to an upcoming conference to discuss the issues she had raised. Farwell declined to be interviewed.

Shur, who now works at Children’s National hospital, in Washington, D.C., is frequently called on by legal authorities to determine whether a metabolic bone disorder or other condition might explain injuries that look like they were caused by abuse. “In the majority of fracture cases, I don’t make a medical diagnosis of child abuse,” she said. “Something else has happened.” Nevertheless, she finds Holick’s assertions alarming. When she takes the medical histories of patients with EDS, she told me, they don’t mention frequent fractures or brittle bones. “My patients are the first ones to state, ‘I never fractured as a baby,’” she said. “They have had to deal with joint dislocations, joint pain, but as a whole have not had to deal with unexplained fractures.”

B.U., like most universities, guards the academic freedom of its professors by giving them wide latitude to express controversial views, particularly outside the classroom. B.U. administrators would have to show that Holick’s courtroom utterances lack scientific merit or violate accepted professional standards in order to discipline him, John Wilson, the author of the 2008 book, “Patriotic Correctness: Academic Freedom and its Enemies,” said.

But Holick’s critics argue that his statements fall outside the bounds of academic freedom, because they have profound repercussions for families and children. “Academic freedom is not just about saying what you want,” John Leventhal, the medical director of the Yale-New Haven Children’s Hospital Child Abuse Program, said. In an article for the Journal of the American Medical Association last year, Leventhal and his co-authors said that academic medical centers should take some responsibility for the testimony of physicians who promote “flawed theories” in child-abuse cases. “If medical faculty at these institutions testified under oath that smoking did not cause cancer or that HIV did not cause AIDS, would such testimony be tolerated?” he wrote.

While the medical school has largely supported Holick, under the banner of academic freedom, the affiliated Boston Medical Center — where he is the director of the Bone Health Clinic — has taken a dimmer view of his outside work. The hospital has the busiest trauma center in New England and its child-protection team reviews about 800 cases per year.

In 2014, Robert Sege, a pediatrician on the child-protection team, asked to meet with Holick. Sege wanted to know the basis of Holick’s assessment that Ehlers-Danlos or vitamin D deficiency could explain fractures in infants. Other than agreeing that Holick provided a detailed explanation of his thinking, the men have different recollections of the meeting. At the end of their conversation, Holick said, Sege “looks me in the eye and said: ‘Holick, it is my recommendation you cease and desist immediately. We are going to make your life miserable.’”

Sege said that this confrontation never happened, and that he left the meeting thinking that Holick was “very convincing.” Sege’s contemporaneous notes, which he shared with me, indicate that the conversation was “very cordial and collegial.” It wasn’t until later, after reviewing relevant literature in the field, that Sege determined that he could not validate the link between EDS and infant bone fractures. “I believe this is his firmly held belief,” Sege, who is now a professor at the Tufts University School of Medicine, said of Holick. “It is not backed by evidence.”

Still, Sege said, one comment of Holick’s unnerved him: Holick told him that he doesn’t recommend that parents in potential abuse cases test for osteogenesis imperfecta, the disease known to cause injuries that look as though they might be the result of abuse. Holick acknowledges advising against testing for the disorder. “It’s incredibly rare,” he said, and the test “will come back negative,” adding to the impression that “you are a child abuser.”

“That is a real red flag,” Sege said. “I am a doctor. I test for all things out there before I go to the thing that is my invention. I test for everything I can. It is alarming to tell someone not to get a test because it might impact your court case.”

Boxes and equipment inside Holick’s lab at Boston University. (M. Scott Brauer, special to ProPublica)

In May 2017, Boston Medical Center banned Holick from using “its facilities to treat or evaluate patients under the age of thirteen for any reason,” the hospital said in a statement for this article. It said that its “medical leadership” made the decision, but didn’t explain why. Holick said that complaints about his child-abuse work prompted the edict. The hospital did let Holick consult on one pediatric case that didn’t involve alleged abuse, he and the hospital said. However, he infuriated higher-ups when, at a mother’s request, he also examined twin 2-year-olds who had allegedly been abused. According to Holick, the medical school’s chairman of medicine, David Coleman, told him, “We would be thrilled if you retired and would leave.” He was warned that he would be fired if it happened again. Boston University declined to comment regarding Coleman and said that personnel matters are confidential.

Holick eventually won approval for a research project to study the genetic basis of Ehlers-Danlos, helping him to get around the ban. As part of that research, Holick now evaluates young children in suspected abuse cases in a building shared by Boston University and Boston Medical Center, just across the street from his office.

As concern about Holick’s work has mounted, he has faced increasing scrutiny in court. Lawyers with the Massachusetts Department of Children and Families recently spent several days attacking his credibility and trying to persuade a judge in a child-abuse case not to allow him to testify. Holick said that they brought in other experts to challenge his credentials. They subpoenaed his boss, Alan Farwell, to testify. Finally taking the stand himself, Holick endured a withering cross-examination, he said.

At times, Holick appears surprised and wounded by the criticisms aimed at him. He called some of the e-mails written to his superiors “a little disappointing” and “vitriolic.” He said that those who find fault with his views should publish commentary in medical journals, or do studies of their own. Academics, he said, should want to learn from one another.

Holick compares the attacks on his work to the Salem witch trials. Then, as now, Holick said, “clueless” doctors who were unable to make a proper medical diagnosis blamed scapegoats. “With no experience or knowledge about metabolic bone disease the present day witch hunters will dam [sic] any parent who innocently brings in their child for evaluation,” he wrote to the editor of a medical journal.

Holick said that he has noticed cars lurking at the end of his driveway. He’s tried to approach some of the drivers, but they sped away. “You know, maybe I’m a little paranoid these days,” he said.

This past January, Holick received a phone call so unsettling that he doesn’t even remember who was on the other end. The caller said that the child of a family Holick helped in an abuse case was in the hospital with a serious brain injury, and the father had been arrested.

“Look what you’ve done,” the caller said, according to Holick.

The child was one of Jenn and Robbie Ray’s twins. Officials in South Carolina had agreed to a plan to gradually reunite the family. They‘d had three weekends of unsupervised visits. Now the local newspaper featured a mug shot of Robbie Ray in orange prison garb.

Holick said that he called Jenn Ray to find out what happened, and that he’s confident the latest injuries weren’t caused by abuse, either, but were another complication of EDS.

The daughter of Jenn and Robbie Ray in a South Carolina hospital in January, 2018, after being treated for injuries including a broken right leg and a brain bleed.(Photo courtesy of Jenn Ray)

Holick told me that the girl, whom he described as about 2 years old, was in a swing and hit her head. Her father was with her, he said, “and I think that he brought her into the hospital and they gave a clean bill of health. And the next day the daughter had some significant neurological deficits and was bleeding in the head. And so, as a result, he was arrested for felony child abuse. But I believe that what likely happened is just like he was stating: that he was innocently swinging his child in the swing and bumped the head. Unfortunately, the pediatrician and or emergency doctor didn’t actually connect dots and didn’t realize that, if that infant or child has EDS, and if they’re having trauma, they should be more concerned about a brain bleed and not to have sent that child home.”

The explanation is, in many ways, typical of how Holick views these cases. He is quick to declare that a trivial accident has resulted in injuries that can be explained by his diagnosis, that the child has received incompetent medical care from less knowledgeable doctors and that the authorities have rushed to judgment.

Holick’s description of the incident, however, is largely inaccurate. The girl was 8 months old, not 2 years. Robbie wasn’t “innocently swinging his child in the swing.” By his account, he had strapped the baby into a motorized Fisher-Price swing, with the seat about a foot off the floor, while he sat on the couch and fed the other child. He didn’t drive the baby to the hospital; he called 911, and an ambulance took her. In a recording of that call, Robbie seems to be sobbing as he tries to awaken his unconscious daughter. “What’s wrong with her?” he asks the 911 operator. “What happened?”

There was no delay in treatment or missed diagnosis of a brain injury. When the infant arrived at the hospital, the trauma team met her and inserted a tube into her trachea to help her breathe. A CT scan showed that a hematoma on the right side of the brain was shifting the brain to the left. Nor could Holick’s comment that she “bumped her head” explain the damage to the rest of her body. Her right leg was broken near the thigh. A photograph from the hospital shows the infant lying on her back, with one eye swollen shut and bruises on her right eye and nose. There appears to be dried blood on her left hand.

Robbie and Jenn Ray in the backyard of their home in Columbia, South Carolina. They met in August 2016 and were married six months later. (Travis Dove for ProPublica)

Robbie’s explanation for his daughter’s injuries is that he forgot to snap in the right side of the strap, and she squirmed out, fell forward and hit her head on the metal base of the swing. Robbie says he didn’t see what happened, but heard a thud. When he looked over, the baby was face-down on the carpeted floor. He told state social workers, according to their report, that the baby wasn’t crying or moving and when he picked her up, her head was “just dangling back as he held her.” (Jenn told authorities that she was running an errand at Walmart when her daughter was injured, and showed them a receipt from the trip. As during the prior incident, hospital staff members found her demeanor troubling. She was noted to have a “flat” affect and to be “nonchalant and standoffish.”)

One of the emergency-room doctors who treated the child was skeptical about Robbie’s story. She wrote in her notes that the “mechanism of injury” — a short fall from an indoor swing — “does not fit clinical presentation on arrival.” In a repeat of the first time the girl was brought to the hospital, six months earlier, Susan Lamb, the child-abuse specialist, was called in to evaluate. She didn’t believe Robbie, either. The fall he described couldn’t have caused the girl’s injuries, she concluded.

“At this time with no plausible accident history, [the child] clearly has suffered abusive head trauma and physical abuse,” Lamb wrote.

Three days later, Robbie was at home and heard a knock on the door. He peered out a window and saw several police officers in tactical gear. Robbie was shackled at his wrists and ankles and led away. He was charged with abuse to inflict great bodily injury upon a child, a felony that carries a sentence of as long as 20 years in prison. He pleaded innocent and was released on bail.

Before Robbie faces a criminal trial, the child-abuse allegations will be weighed in family court. If the judge finds that there has been abuse, the D.S.S. has said that it will move to revoke Jenn and Robbie’s parental rights, which would likely lead to putting the children up for adoption.

Police and the D.S.S. haven’t said what they think happened. When D.S.S. investigators came to the house, they wanted to know about the family’s trip to see Holick, according to a recording Jenn made of the interview. They asked for dates, receipts and the names of everyone who traveled to Boston.

I met with Robbie and Jenn at their home on a steamy day in May. The only sign that something was amiss was the Fractured Families window decal on the back of the Ford Explorer parked in the driveway. Toys were neatly stacked in the living room. The twins’ bedroom remained just as it was in January, when their daughter was hurt. Photographs of the children adorned the refrigerator. Medical and court records were spread across the kitchen table, near where the mechanical swing was set up before police seized it.

Jenn and Robbie talked to me without a lawyer present, taking occasional cigarette breaks on the back patio. The only other people in the home were Jenn’s parents, who stopped by toward the end of the interview, and who are draining their retirement funds to pay for the legal fight to return the twins to Jenn and Robbie. The twins, Jenn told me, are in foster care about 90 miles away.

Joking that she has earned a “Google medical degree,” Jenn offers detailed alternative explanations for her daughter’s injuries and for the twins’ prior fractures. She insists that mistakes by doctors aggravated her daughter’s condition. The only time she became emotional was when talking about the stress that the abuse charges have put on her marriage.

Robbie and Jenn believe that Lamb, the abuse specialist, took an immediate dislike to them after the Memorial Day incident and saw the swing injury as an opportunity to separate them from their children permanently. Jenn called the doctor a “liar” who is “evil,” “wicked” and a “narcissist.” A spokesperson for South Carolina’s Department of Social Services declined to comment. Lamb said that medical-privacy laws prohibited her from discussing the case. A spokesperson for Palmetto Health Children’s Hospital said that child-abuse experts there, including Lamb, are consulted roughly 50 times a year on suspicious injuries. Since 2007, they have found that abuse is medically indicated in about half of the cases they’ve seen.

Ray family photos on the wall of their home in Columbia, South Carolina.(Travis Dove for ProPublica)

The couple contends that their daughter suffers from a metabolic bone disorder — though not necessarily EDS — that worsened the damage from the fall. Jenn thinks the baby may have brittle-bone disease. An orthopedist who evaluated the child concluded otherwise, but Jenn wants a genetic test performed. Both the orthopedist and a new guardian ad litem representing the interests of the infant attribute the injuries to abuse, according to the guardian’s report. Robbie said that he will fight the charges. “If anything, I saved her life,” by calling 911, he said. “I would not hurt my daughter. And that’s why I keep trying to tell people, I know I didn’t do it. I know I didn’t do anything wrong.”

Those weren’t the only criminal charges that Robbie emphatically denied. I asked him about a 2015 misdemeanor conviction for public intoxication that had turned up when the police did a background check. A truck carrying Robbie and a friend, both drunk, had been found on its side in a ditch at 2 a.m., according to police. Robbie was adamant that he knew nothing about it and insisted that it must be a case of stolen identity.

But an intoxicated Robbie is clearly identifiable in a video taken by a police-cruiser camera. “I got really drunk,” Robbie told one officer. “I can’t handle my alcohol.” The video shows him being handcuffed.

When I confronted Robbie about the video, he first reiterated that the incident was “news to me.” He eventually acknowledged that he was at the scene and offered a series of new explanations — for example, that a third person had brought him there after the accident — that are contradicted by the recordings. Nothing has surfaced to indicate alcohol was a factor in either of the alleged abuse incidents.

Holick will be watching the Ray case from Boston. Mistakes happen in medicine, but he realizes that one high-profile misstep on his part could greatly damage his credibility.

“I’m very, very concerned this could upend this whole activity, if even one of the cases of the hundreds turns out to be a child-abuse case,” he said. But he acknowledged the possibility that the case might not turn out as he would like. “You know, these types of things are going to happen,” he said.

In another conversation, he reflected on the impact of his crusade to reunite families divided by child-abuse allegations. “To date,” he said, “almost all of the kids I helped return to parents are happy and well.”

Senior reporter David Armstrong is an adjunct professor of journalism at Boston University.

This story is not subject to our Creative Commons license.

Correction, Sept. 26, 2018: This story originally said that Michael Holick has tenure. B.U.’s medical school does not grant tenure.