It begins with a mistake. A doctor botches an operation, misses something on an MRI, or fails to make a diagnosis. Medical lawsuits are nothing new. Doctors are human. They make mistakes, and injured patients expect to be compensated. But since 2001, the University of Michigan Health System has been taking a radically different approach to these inevitable problems: admitting its mistakes.

The goal of the “Michigan Model” is to limit errors by acknowledging and learning from them. Apologizing to patients and offering follow-up care and sometimes financial compensation also makes them less likely to sue. On the other hand, if a UMHS investigation concludes the medical staff did all that they could, the system will stand behind its employees.

That response, which medical researchers call “communication and resolution,” might seem obvious. But until recently the trend in the medical world was very different. The typical response to malpractice lawsuits was to deny that anything was done wrong and prepare a legal defense. Yet even if administrators believed their staff was blameless, they might pay a settlement if it was cheaper than going to court.

Tacked on the wall of Rick Boothman’s office is a quote from George Orwell: “In a time of universal deceit–telling the truth is a revolutionary act.” The UMHS’s chief risk officer developed the Michigan Model after concluding that the old “deny and defend” approach eroded the faith of doctors, patients, and hospital administrators in one another.

Doctors, Boothman says, often feel very guilty when a patient is harmed, even if they weren’t negligent. “Everything caregivers do is inherently risky. You can’t control that risk … when things go wrong, they feel betrayed,” he says. “Medicine made a Faustian bargain.”

When hospitals barricaded themselves behind lawyers, they also weakened one of their best defenses against lawsuits: the trust between doctor and patient. In 1997, University of Toronto physician Wendy Levinson conducted a study showing that patients sue more out of a sense of neglect than a conviction their doctor was negligent. And patients are more reluctant to sue doctors they like–even if the doctor admits to making a mistake.

Boothman says the main motive in developing the Michigan Model was to do what’s right, not to limit lawsuits. But malpractice litigation has plummeted since it was introduced. According to statistics provided by U-M, the rate of lawsuits has fallen from 2.13 per 100,000 patients to roughly 0.75 per 100,000 patients. The median time to reach a resolution has dropped from 1.36 to 0.95 years.

“We refuse to settle out of expediency. We have never had a case fall on us because we were honest,” Boothman says. “Once you admit you own it, you can improve it. You can’t fix anything if you don’t think you have a problem.”

Ann Arbor attorney Peter Davis has been handling malpractice lawsuits since 1966. He says he has seen many cases over the years where doctors altered records in an attempt to cover up errors. The Michigan Model, he says, gives doctors an incentive to be honest.

“We all fear being on the wrong end of a malpractice suit,” says Daniel Hayes, medical director of the breast oncology program at U-M’s Comprehensive Cancer Center. “At most centers, this becomes confrontational.” At other hospitals, he’d seen many colleagues get hung out to dry by administrators, so when he began working at the U-M roughly twelve years ago, it took him only about ten minutes to get on board with its policy of admitting mistakes.

Conventional wisdom teaches doctors that they should never make mistakes, and if they do, to deny them or risk destroying their careers. The Michigan Model, Hayes says, makes them more comfortable admitting when they could have provided better care. “If doctors recognize that when they make mistakes they are not going to get hammered, they are going to be allowed to do the right thing.”

Most patients, Hayes says, aren’t looking for a big payday; they simply want an acknowledgement of wrongdoing and modest compensation to cover their medical expenses. The health system saves money by paying reasonable compensation directly and eliminating the middlemen.

Norm Tucker is one of those middlemen–he’s a malpractice attorney with forty-two years’ experience. Like Davis, he credits the Michigan Model for encouraging honesty.

“The minute one side is trying to polish their story, so to speak, the whole thing doesn’t work,” Tucker says. “It’s not just about avoiding lawsuits; it’s about correcting a problem. It’s not just about the money; they don’t want it to happen again.” Most of his clients, Tucker says, would return to the U-M hospital after settling their cases.

Boothman gives several examples of how U-M’s program has led to changes in hospital policy. For instance, two patients with pre-existing sleep apnea experienced difficulties while on narcotic painkillers, which inhibit respiration. The administration then opted to install pulse oximeters throughout the hospital to catch problems sooner. The hospital also improved the way surgeons pass instruments in the OR, changed the way it cleans bronchoscopes, and addressed a disparity in the way ER docs and cardiologists were interpreting EKGs.

“Health care is not like building a Toyota,” Boothman says. “If you whack a mole here, you can cause unintended consequences somewhere else. So, everything is so interconnected that if you don’t do it publicly and you don’t do it in a collaborative and integrated way, you don’t make a lasting fix.”

On a gray afternoon inside Ruby Tuesday, Vickie Patterson grabs a tortilla chip, dips it into a spinach-artichoke dip, and shovels it into her mouth. After a few seconds, her face twists into a grimace.

“I don’t feel hot until it touches the side of my teeth,” she says.

Patterson doesn’t believe the Michigan Model has patients’ best interests at heart. During a 2002 jaw surgery, her surgeon accidentally bored a screw into her eye socket, leaving her with substantial nerve and hard palate damage.

What followed over the next couple of years would change Patterson’s view of U-M’s hospital. Patterson says that members of Boothman’s staff refused to allow her to speak with the dean of dentistry and verbally abused her over the phone. When the hospital contacted her in writing, she said it did so as a wolf in sheep’s clothing.

“They never identified themselves as lawyers,” she says. “This has got to stop. They need to inform patients that they are lawyers … I was never told I needed an attorney.”

Boothman says that some of the people who make the initial contact are lawyers, but many are not. “We tell people, ‘You are entitled to a lawyer, and we will not treat you any differently.'” Whoever makes the contact, he says first priority is to inquire what the hospital can do to ease the patient’s trauma.

“Somebody has a bad outcome, a bad result, something bad happens to them, we want to engage those people immediately but along a certain descending algorithm,” Boothman says. “So, the very first concern isn’t [legal] claims. The very first concern is how do we grease the skids so that patients’ care, whatever new care needs they have, can be improved–we can cut through the red tape and get them the very best care they need right now.”

Another critic, retired U-M pathologist Douglas Smith, believes the real purpose of the Michigan Model is to reduce the cost of lawsuits. If the hospital resolves a problem early, he points out, it does not have to report it as a malpractice suit to the National Practitioner Data Bank, which influences hospitals’ rankings and accreditation.

Smith cites an example in an area on which he focused: graft-versus-host disease–a complication of organ transplants where healthy tissue rejects donor tissue. A transplant patient died of what at first appeared to be a colon infection but also showed signs of graft-versus host-disease. Smith says Boothman denied his request for a DNA test to check for the condition, citing a Michigan law that stipulates that a hospital cannot be sued for wrongful death if there is a greater than 50 percent chance the patient would have died regardless of treatment. Research showed a patient with graft-versus-host disease had only about a 15 percent chance of survival.

Smith knew that statistic well–he contributed to the research that developed it. However, he said Boothman failed to give credence to new methodology that put a patient’s survival rate closer to 75 percent. He says he told Boothman that if the hospital was going to increase its number of transplants it should examine how to catch graft-versus-host disease earlier. The suggestion was ignored, he says.

“They are not thinking of this program: how do we fix these problems so we prevent malpractice? They just don’t want to get sued for malpractice,” Smith says.

Boothman responds that interpreting the test Smith wanted would have required before-and-after data for both the donor and recipient, which didn’t exist. Even if it had, he says, making sense of the results is tricky. And he is vehement that the hospital’s investigation was exhaustive. Looking back, he says, it appears that the patient had graft-versus-host and a negative reaction to antibiotics. “There was not a way in the world this patient was going to survive this,” he says.

Boothman says that there’s a positive correlation between fostering good relations between caregivers and patients and the quality of patient care. He says many hospitals, including the U-M, refer to the National Practitioner Data Bank only for a few situations, and even then rarely. Chalking up Smith’s criticism to a disgruntled former employee, he says the case he describes did not lead to any changes in the U-M’s procedures for identifying graft-versus-host disease. Perhaps the transplant team and other health care providers could have communicated better, he says, but the issue of how to handle graft-versus-host was not a problem on a fundamental level.

Claudia Megaro, a registered nurse now living in New Jersey, underwent surgery at U-M in 2006. While she did not pursue damages, her story, she says, also illustrates problems in the Michigan Model. Following an infection, a surgeon grafted a piece of bone from one of her legs to her jaw. Shortly afterward, Megaro began having intense swelling in her face. Though blood and pus were seeping from her drainage tube, nurses never contacted her surgeon. Instead, Megaro was discharged.

To make matters worse, after an ordeal that spanned roughly three years and five surgeries, Megaro says she discovered another error. During her third surgery, to remove some hardware that had worked its way loose, the surgeon noted that he could find only nine of the sixteen screws that earlier records showed had been installed but dismissed the disparity to a clerical mistake. It was only after her dentist noticed several bright spots on an X-ray that Megaro learned hardware was still embedded in her face.

Megaro says she did not pursue damages because all she ever wanted was an acknowledgment of wrongdoing. As someone with a medical background, she didn’t want to contribute to a culture of litigation she believes is responsible for the defensive behavior that made U-M reluctant to admit its mistake in the first place.

Once hospital officials turned her case over to risk management, Megaro says the communication turned sour. Risk officers sent Megaro a plethora of emails and letters expressing concern, but she didn’t feel they were sincere. It’s as if the hospital expected her to be grateful, she says, simply for using the words “sorry” and “apology” frequently.

“It’s obvious now that [Boothman] is trying to cover it up,” Megaro said. “They think they can browbeat people … they try to intimidate you. It’s a bullying thing … If it’s being done to me, how many other people is it being done to?”

A 2011 article in the Chapman Law Review contends that patients are better off going to court. Apology programs, the author writes, “allow for the manipulation of injured patients as a means to persuade them not to pursue money damages … not only is a person less likely to pursue litigation following a doctor’s apology, but even if a patient does still pursue money damages, the patient is more likely to adopt a more pliant negotiating position.” The article cites U-M’s own research that found 71 percent of patients who participated in the apology program accepted settlements worth less than what they would have sought in court.

A book titled Family Centered Patient Advocacy: Understanding Patient Safety sits on Boothman’s desk. Despite his best efforts to make the Michigan Model seamless, he concedes that it is far from perfect. For instance, the system lacks a mechanism to determine the number of patients harmed by hospital-acquired infections, which cause far more injuries and deaths than medical errors. Still, U-M hospital staff closely monitor any potential outbreaks, visiting rooms and swabbing furniture, including the curtains between beds, and keeping an eye on air quality. But he acknowledges that the easiest and most cost-effective way to control infections–more frequent hand-washing–is still not being done often enough.

It’s also hard to get a handle on how many mistakes are being made, because the system relies heavily on self-reporting. According to medical literature, only between 8 and 11 percent of medical mistakes nationally find their way to an incident report. At U-M, the number of incidents reported skyrocketed from 2,400 in 2006 to more than 23,000 in 2013.

While Boothman cautions that the numbers are “squishy,” he believes they show that caregivers are becoming more willing to report problems. “We are clearly making progress, because we are not seeing repeat mistakes, and we are seeing improvements,” he says. And he says the Michigan Model has led to more decisive action when dangerous caregivers are identified.

“All I know is there are multiple strategies. It’s a push-and-pull kind of thing,” Boothman says. “It’s a changing landscape all the time.”

Another issue that is hard to address, Boothman says, is what constitutes medical malpractice. He offers as an example a heart surgery performed on a 420-pound man during which surgeons lacerated his vena cava.

“Is it a medical error, or are you just pushing the limits of what a human being can do?” he asks, a poster of Lou Gehrig with the caption “The Luckiest Man on the Face of the Earth” hanging behind him. “If you are supposed to get antibiotics every four hours, and a dose comes every four hours and fifteen minutes, is that a medical error? …

“You can do everything right and still have a problem,” he points out. That said, “We don’t want to defend care we shouldn’t be defending,” he says. For example, as surgeons age, some see their technical skills decline. Since they still have value as teachers, he says, the U-M may shift their roles to reduce the risk of mistakes.

Until recently, it was an open question whether the Michigan Model could work outside U-M. In January, the medical journal Health Affairs published a pair of studies that examined communication and resolution programs around the country, including a group of five New York City hospitals directly “inspired” by Michigan’s example. Lead author Michelle Mello, professor of law and public health at Harvard, had previously co-authored a 2007 article that argued that full disclosure ran the risk of increasing rather than decreasing litigation.

Mello says she has “become more of a believer in the model over time.” While the studies did not have the long view U-M has after twelve years, the results were positive overall. The hospitals didn’t copy the Michigan Model exactly, instead tailoring the idea to their comfort level. Some adopted a more passive approach, simply asking patients what they expected of the hospital and shying away from talks regarding concrete figures of compensation. All faced challenges similar to Michigan’s, including mitigating patients’ financial expectations and convincing doctors to go against a hardwired instinct not to admit mistakes.

While Mello believes the Michigan Model can work elsewhere, she also credits a “Boothman factor” behind its success here. Not all risk officers will be as adept at anticipating snags and selling physicians and administrators on the value of full disclosure, she says. “I don’t think there is anything distinctive about U-M’s setting that gives them a monopoly” on communication and resolution programs, Mello says. “It’s an approach that can work in many other settings, but whether they will be able to marshal the other issues is another story.”

That question may soon be answered: as part of an initiative announced by President Obama in 2009, the federal Agency for Healthcare Research and Quality is promoting a variety of efforts to reduce medical liability costs, including grants for communication and resolution programs around the country. And Mello’s article in Health Affairs noted an encouraging pattern among the hospitals that have followed Michigan’s lead in adopting communication and resolution programs: all encouraged other hospitals to give them a try.

This article has been edited since it was published in the May 2014 Ann Arbor Observer. The description of Vickie Patterson’s jaw surgery has been corrected (see Calls & letters item, below).

The following Calls & letters item was published in the June 2014 Ann Arbor Observer:

Vickie Patterson’s surgery

Last month’s article on the U-M Health System’s “communication and resolution” program for patients who suffer medical injuries (“The Honesty Policy”) mistakenly described the procedure that left Vickie Patterson with bone and nerve damage.

“I had Obstructive Sleep Apnea Jaw Surgery (OSA),” Patterson emailed. “Screws were bored into my eye socket and nasal bone. My nerve and facial injuries are multiple, and permanent due to dishonest and incompetent doctors, Helman and Ward. I am a former University of Michigan employee and patient that was terminated with a Non Union of the Maxilla, and left without insurance to get the surgery corrected.”

The following Calls & letters item was published in the July 2014 Ann Arbor Observer:

The Michigan Model’s history

Former U-M Department of Surgery development officer Alan Cotzin called to say that our feature on the Michigan Model (“The Honesty Policy,” May) should have credited another key player in reshaping the U-M’s response to medical errors. While Rick Boothman has led the program since 2001, the groundwork was laid by Ed Goldman.

Goldman headed the health system’s legal office from 1987 to 2009. When he started, Goldman recalls, the university’s malpractice insurer told administrators to “never admit a mistake—in law that’s called an ‘admission against interest.'” But if a patient sued, caregivers would eventually have to tell the truth under oath in court—so Goldman remembers thinking, “It would be better if we said [to the patient], ‘This happened, this is why, and this is what we have done to make sure it doesn’t happen to others. What can we do to make it right for you?'”

Two changes made it possible to implement the honesty policy. When the insurance company quoted a $4.2 million premium–on a $4 million ­policy–the health system switched to self-insurance, taking the overly cautious company out of the picture. And in 1993 state malpractice reform required injured patients to give six months’ notice of their intent to sue.

“Well, you can ignore that, or you can take advantage of it,” Goldman says. At Michigan, administrators used the time to review the patient’s care—and, if they concluded it had fallen short, to try to reach a settlement. Goldman hired Boothman from an outside law firm to work on managing pending cases.

“I think Rick internalized these ideas,” Goldman says, “and worked hard to implement them.

“What I care about is not my fifteen minutes of fame, but that the policy I started gets used—so if there is a mistake, the university sets it right.”