Taking down Dr. Death

Dallas County prosecutors did something possibly unprecedented:  They tried a local doctor for injuring his patients in the operating room.

Mary Efurd walked into Dallas Medical Center full of confidence in her smart, charismatic neurosurgeon, who was supposed to fix her lower back pain. After all, he had promised he could make her better.
    But the surgery went horribly wrong. Her fellowship-trained surgeon, Dr. Christopher Duntsch, put spinal fusion hardware into her muscle and soft tissue instead of the bone. He amputated a nerve root, leaving her with a permanent drop foot, a gait abnormality due to nerve damage. He twisted a screw into another nerve, which caused her to scream in pain for days when she awoke. He left hardware so loose in her back that it moved when touched. The list goes on.
    Every person in the operating room that day told the defendant he was doing the surgery wrong, but he refused to listen. When the operation was over, the medical device representative left his tools in the operating room because he knew someone would have to go back in and perform a revision surgery. As word spread throughout the hospital, the staff began to panic. Another surgeon was brought in to take care of Mary, lawyers were called, and staff was interviewed. But Dr. Duntsch simply said the surgery had gone well—and he wanted to operate on Mary again on Monday.
    “He had to have known how to do it right, and then did the opposite.” That is how another doctor, Robert Henderson, characterized Duntsch’s treatment of Mary Efurd. What Dr. Henderson saw was so unthinkable that he thought an imposter had done it. He even sent Duntsch’s picture to the University of Tennessee, where Duntsch had trained, to find out if it was the same man. Tennessee confirmed. A real doctor had done this.
        Mary now spends her days in a wheelchair.
        If this had been a one-time occurrence, a single horrific surgery, the Dallas County Criminal District Attorney’s Office would probably never have heard of Dr. Christopher Duntsch. Unfortunately, his bad outcomes on patients began piling up over a relatively short period of time. When doctors are lining up to testify against another doctor, you know something is wrong. Normally physicians will not go on the record to say bad things about another doctor. There is typically hemming and hedging because they understand that medicine is not an exact science, that patients react differently to treatments, and that even good doctors have bad outcomes. They are taught to fear lawyers and malpractice lawsuits. So when doctors, lawyers, and patients came banging on the district attorney’s door, insisting that what Dr. Duntsch was doing was criminal, we paid attention.
    Our investigation started with six patients and eventually expanded to include every surgery the defendant ever conducted in Dallas and Collin Counties. He went through four hospitals in just under two years, operating on 38 patients. Thirty-three of those patients were injured. Some were minor, strange complications or misdiagnoses, but most were life-altering mutilations. Two patients died.
    Clearly, something had to be done.

Time constraints
By the time our office started looking at the case seriously, we had two months before the statute of limitations ran on Mary Efurd’s case. We had to move fast. Lead prosecutor Michelle Shughart worked with four of the plaintiffs’ attorneys who had represented patients in the civil suits against Duntsch to learn about the cases as quickly as possible. Kay VanWey was the attorney for 10 of the injured patients and was the most helpful in getting our investigation underway. The initial meeting with her was a flood of information and made little sense to us at the time because there was so much to digest. She gave us access to her files so we could filter through them to find what we needed. Michelle spent days going through binders and boxes of information to learn about the patients’ injuries and the multitude of medical professionals they had seen.
    The next step was to understand how bad Duntsch’s behavior was from a medical standpoint; we had to determine whether his actions could constitute criminal behavior. Was this just professional rivalry between doctors? Were plaintiffs’ attorneys trying to use the criminal system to aid them in getting money? Or was it possible that a surgeon could indeed be held criminally responsible for things he did to a patient during surgery? There was no way to know without understanding the medicine, so we had to talk to the doctors. Drs. Robert Henderson, Martin Lazar, and Randall Kirby were our starting point. Dr. Henderson was one of the first to ring the alarms about Duntsch after he was brought in to fix multiple patients (the first being Mary Efurd). Dr. Lazar is a brilliant neurosurgeon who analyzed many of the defendant’s patients as an expert for the plaintiffs’ attorneys. He testified the defendant had “no conscience” and that it was “inconceivable” Duntsch didn’t know what he was doing to the patients. Dr. Kirby was a colorful vascular surgeon who had operated with Duntsch once and made it his mission to stop him from operating again. They spent hours with us three wide-eyed prosecutors as we barely followed all the medical terminology. We began to piece together what happened to each patient, and a bigger picture formed. The defendant was a narcissistic—possibly sociopathic—person with a license to legally butcher people. He had no empathy for his patients and accepted no responsibility for the trail of destruction he had left behind.
    It had all the trademarks of criminal behavior, but what, if any, criminal offense could he be charged with? We did a great deal of research and found that doctors are rarely prosecuted for bad outcomes on procedures for which they had consent to perform. The few cases we found dealt with actions by doctors that resulted in patient injuries or deaths where the doctor’s intent to assault or kill the patient was clear. Although we had two deaths, we knew proving intent to kill in those two cases would be difficult with our facts, and although Duntsch’s behavior clearly indicated a pattern, there was no course of conduct offense in the Penal Code (like continuous sexual abuse, engaging in organized crime, etc.) that applied in this situation. We believed we had enough evidence to prove recklessness, so we settled on the idea of charging him with aggravated assault in the most egregious cases that occurred in Dallas County. Then it occurred to us that Mary Efurd was over 65 years old at the time of her surgery—bingo! We could charge him with injury to an elderly individual. That charge accomplished two things: It offered us multiple mental states, and it gave us a larger punishment range to work with (third-degree up to first-degree felony, depending on which mental state the jury found).
    After two months, the statute of limitations was banging on our door.  The investigation into all of Duntsch’s patients was far from complete, but we had overwhelming evidence against him. We presented information on six patients to the elected District Attorney and other chiefs at the office. Any time we explained the case to a fellow lawyer, the reaction was always the same: Their initial skepticism quickly escalated into overwhelming disbelief that a surgeon could have done these things. By the time we got to Patient No. 3, Patient No. 4, and so on, a scary pattern had clearly emerged. The reaction from the office leadership after hearing our pitch was no different: Our elected DA gave the green light to proceed with prosecution.
    We sealed the cases so Duntsch would not know they were coming. He lived in Colorado at the time but would come to Dallas once a month to visit his children. We could not risk scaring him off. We coordinated with the Dallas Police Department to arrest him as soon as a warrant was issued. Sixteen days after receiving approval from our DA, we indicted six cases, unsealed them, and arrested Duntsch in his hotel room. A detective interviewed him just after his arrest, and he insisted the patients were 90 percent better, even though years of civil lawsuits and inquiries by the Texas Medical Board indicated otherwise. The man was clearly as unconcerned and mistaken about his legal predicament as he was about his patients’ injuries.

Discovery
The discovery in this case raised a number of questions. Our own office had done the investigation, so there were no police reports or witness statements to turn over. We sent over 800 subpoenas for medical records and information on the defendant. We obtained thousands of pages of medical records that pertained not only to our named complainants, but also to every patient the defendant had operated on, which raised a number of privacy concerns.
    We were lucky that most of the civil cases had finished; otherwise, it would have been much more difficult to get the information we needed from hospital administrators and staff. The hospitals’ cooperation was instrumental to understanding the administration structure and process of peer review (more on that later); there were also hundreds of nurses, scrub techs, and radiologists to find. We worked with hospital attorneys to narrow down the personnel we needed to talk to, and they in turn set up days where we would just go sit at the hospital and interview person after person.  
    We interviewed hundreds of witnesses. Without knowing the defensive theory in the case, it was hard to discern what information could be construed as Brady material. The offenses occurred prior to the Michael Morton Act, so it technically did not apply to the case. The prosecution team talked about these concerns repeatedly. Given the complexity and sheer volume of the information, we eventually decided the safest course was to turn over everything, including our work product from witness interviews.  

Selecting the jury
Voir dire presented a few complications. First, we were concerned about how much our venire members might know about the case. It had received national attention when the defendant’s medical license was revoked. Plus, D Magazine had published an in-depth article in November 2016 titled, “Dr. Death,” detailing the acts for which we had indicted the defendant. (You can read it by clicking HERE.) Second, we estimated that the trial would take three to four weeks, so we were highly concerned about the number of jurors who would be disqualified because of the time commitment alone. Further, we needed 12 people who could convict a doctor, who were smart enough to follow the intense medical terminology, who could understand and apply the mental states, and who could consider the full range of punishment from probation to life confinement. We used detailed questionnaires to dive through many of these issues.
    We had planned to voir dire a full panel every day after the questionnaires were completed to get as many jurors as we could each day. However, this strategy quickly changed after we were left with only 17 jurors on day one; we lost a majority of the panel on their questionnaire answers—it was to the point where it was not worth doing a voir dire. The parties agreed to wait until we had enough qualified jurors from the questionnaires alone, which took three days of bringing up panels of 75. By the end of day three, we had 62 possible jurors when we did the actual voir dire. The State spent over two hours with the panel discussing the injury to an elderly individual law and whether the panel could convict a doctor for a surgery gone wrong. In the end, the jury consisted of retirees, business people, and medical personnel. The panel was very State-oriented, possibly because so many bad jurors had been struck for cause already. We also had a very good understanding of our jury once the 14 were seated after the onerous selection process.  

Trial strategy  
We decided to proceed on the injury to an elderly indictment for two reasons. First, Mary’s case was the best from a factual standpoint because several witnesses in her surgery told the defendant his placement of the spinal fusion hardware was wrong, and the fluoroscopy images taken during the surgery clearly and unequivocally showed it was incorrect. Second, if we could prove that the defendant acted intentionally or knowingly, a first-degree punishment range would be available to the jury at punishment. We could then present several of the injured patients during punishment to—we hoped—get the maximum sentence and spare them from having to testify at numerous trials.
    We knew we needed to find a way to admit some of the defendant’s earlier surgeries to show his knowledge at the time of Mary’s surgery. After much research, we determined that this extraneous evidence should come in under Rule 404(b) and the Doctrine of Chances. Justin Johnson, an appellate prosecutor on the team, drafted a very persuasive motion, which we presented to the trial court at a pretrial hearing. Our argument was that the defendant’s extraneous surgeries were admissible to show intent, knowledge, absence of mistake, and lack of accident. The Doctrine of Chances was a more exciting argument for us, though, because it would allow us to present extraneous surgeries that occurred after our indicted cases. The doctrine recognizes that a series of unusual events that are unlikely to repeat themselves should be admissible so the events are not viewed in a vacuum. We argued that the defendant’s surgeries and their outcomes were so unusual and rare that it is highly unlikely they would have occurred as frequently and as closely together in time unless they were done deliberately. Ultimately, the trial court granted our motion under both theories, giving us leeway to present as many patients as we felt necessary to prove the mental state.  
    Preparing for trial was then a balancing act. How many of the 33 injured patients should we present to the jury? It had to be enough to convince the jury of the defendant’s knowledge and/or intent, but not so many that the jurors would think we were wasting their time or that an appellate court would think we went too far. In addition, every case we presented had to be proven beyond a reasonable doubt, which entailed calling the patient to describe his condition before surgery and injuries after, at least one doctor to explain what the defendant had done wrong to cause the patient’s injury, and any other medical staff that was present and saw something out of the ordinary. Ultimately, we decided the best course of action was to show the jury that the defendant knew, before setting foot into Mary’s surgery, that he had already had catastrophic outcomes in six of his most recent patients.  We decided to hold the later surgeries for punishment.

String of injured patients
To understand the case, you have to understand the timeline of surgeries. The defendant had 11 mediocre surgeries when he first came to Dallas in late 2011. Most of these had a complication or strange occurrence of some sort but were not serious enough to draw attention. Surgery No. 12, on patient Lee Passmore, was when things really started to go wrong. From Passmore forward, almost every patient the defendant operated on was badly maimed.  
    There were a few pieces of Lee Passmore’s story that we wanted the jury to hear. He was the first patient where we could show that Duntsch absolutely knew he had done something wrong because Duntsch had to take Lee in for a second surgery within a week to attempt to correct the problem he had created. (The second surgery did not fix anything.) Dr. Robert Hoyle was the vascular surgeon in Mr. Passmore’s first surgery with the defendant. He testified that Duntsch was doing outrageous things in the surgery—to the point that that Dr. Hoyle had to physically stop him so he would not continue injuring the patient. Dr. Hoyle told Duntsch that he was dangerous, that he was going to hurt somebody, and that Hoyle would never work with him again. Lee currently has a screw poking into his spinal canal that cannot be removed because he might bleed to death. It causes him constant, excruciating pain.
    Barry Morguloff was the defendant’s next patient. Barry walked into the hospital, but he awoke after surgery unable to move his legs. Duntsch labeled Barry a “drug seeker” and refused to administer any medication while he was screaming in pain in the hospital. Barry, in fact, had been sober for years. Eventually, Barry saw another doctor, Dr. Michael Desaloms, who identified what was wrong: Bone fragments were compressing a nerve root and had caused permanent damage. Barry now walks with a cane, is in constant pain, and has a bunch of other medical issues that have developed because of the damage the defendant caused.
    The very next patient was Jerry Summers. Jerry had been the defendant’s best friend since childhood. He moved with Duntsch to Texas and was basically his errand boy. But Jerry had a more sinister side. Various sources told us that Jerry was Duntsch’s drug dealer and that he also sold drugs to high-school students in Plano. The defendant operated on Jerry in February 2012, and Jerry awoke a quadriplegic. The defendant again did not take care of his patient. Instead of immediately taking Jerry back into surgery or sending him for imaging to identify the problem, Duntsch opted to do elective surgery on another patient he had scheduled. Twelve hours later, the defendant came back to Jerry and decided to take him back in for a corrective surgery, but it was too late. Nothing could be done. For days Jerry laid in intensive care, not knowing why he could not move his arms or legs or that his condition was permanent—because the defendant would not tell him so. On day four, Jerry outcried to a nurse that he and Duntsch had done 8-balls of cocaine the night before the surgery. That finally got the hospital’s attention. The defendant was removed from the case, a peer review investigation was started, and the defendant took a forced break from operating.  
    A peer review is a confidential internal investigation conducted by senior physicians to determine the cause of patient injury. It is initiated when there is a serious, unexpected outcome in patient care. The peer review committee reviews the medical records, talks to personnel involved with patient care, and uses experts to analyze whether the physician’s actions met the standard of care. These peers may recommend future discipline against the physician, such as suspension and training.     
    But just a month after the Jerry’s surgery, the defendant was allowed to operate again. He was told to do only very simple surgeries until the peer review on Jerry’s case was finished. His first surgery was on Kellie Martin, and she needed an easy voluntary laminectomy (also known as decompression surgery). During the operation, the defendant punched too far through the spinal anatomy and slashed a major blood vessel. It is a known, but extremely rare, complication of the surgery. All the signs of major internal blood loss were there, but Duntsch ignored them. He denied to everyone in the operating room that there was a problem, and the other medical staff had no way to know what had happened or how to fix it. On her way out of surgery, Kellie was cold and clammy with mottled skin. She arrived at the intensive care unit, where medical staff immediately began resuscitative efforts. They worked on Kellie for three hours, but there was no way to reverse what the defendant had done without his help, and he was denying anything had happened. At the age of 54, Kellie Martin died from the simplest procedure that a spine surgeon performs.
    At this point, authorities at Baylor Scott & White Medical Center in Plano knew they had a problem surgeon. They told the defendant he would never operate there again and started another peer review investigation. But the hospital made a huge mistake it would quickly come to regret: Authorities there allowed the defendant to resign instead of kicking him out of the hospital. This minor technical difference meant that the defendant would not be reported to the National Databank, which is an online database hospitals commonly use to investigate doctors applying to work at their facilities.
    A few months later, the defendant sweet-talked his way into Dallas Medical Center. Dallas Medical Center’s search of the National Databank revealed no red flags, so authorities there gave the defendant temporary privileges right away. In his first week, he operated on Floella Brown, and he cut one of two major arteries supplying blood to the brain and then, in an attempt to stop the bleeding, put too much pressure on the vessel and occluded (blocked) it. It was Floella’s dominant artery, so the occlusion prevented oxygen-rich blood from going to her brain. Again, as usual, the defendant did not tell anyone what had happened during the surgery so doctors could monitor Floella closely. Overnight she had a stroke, and Duntsch did not respond to hospital pages for a couple of hours. When he finally arrived at the hospital, he abandoned Floella. Instead of caring for his dying patient, he decided to start another elective surgery that he could have easily postponed. That other patient was Mary Efurd.
    Hospital officials were exasperated while the defendant was in Mary’s surgery. They asked the defendant multiple times to take care of Floella or transfer her. The defendant wanted to drill a hole in Floella’s head to relieve pressure, but the hospital refused for two very good reasons. First, the defendant was not qualified nor did he have privileges to do any type of brain surgery. Second, the hospital did not have the equipment or personnel for brain surgeries. The only viable option was to transfer Floella immediately, but the defendant refused. For hours Floella waited in a coma, her brain deprived of oxygen, for the defendant to do something. By the time he finally acquiesced and transferred her, she was essentially brain-dead. At about this same time, Mary was waking up from her surgery unable to feel her legs. Dallas Medical Center immediately suspended the defendant’s privileges and called in Dr. Robert Henderson. Dr. Henderson had pioneered modern spinal surgery, and the hospital called him to “salvage” Mary’s back. When he operated on her, he found her condition so appalling that he recorded his revision surgery—he didn’t think anyone would believe what the defendant had done. It was after this surgery that he called the University of Tennessee to see if the defendant was an imposter.

Presenting the patients
We presented the cases involving these six patients—Lee Passmore, Barry Morguloff, Jerry Summers, Kellie Martin, Floella Brown, and Mary Efurd—during the guilt-innocence phase of trial. These events clearly demonstrated that the defendant had five horrible outcomes in a row before he stepped into the operating room to perform surgery on Mary Efurd. By this time, he was aware his conduct was reasonably certain to cause her injury, and he should never have put his scalpel to her back. We also argued that his actions in the surgery alone were enough to convict. Every person in the operating room told him that the hardware was not in the bone, and he knew that putting the hardware in the wrong place would cause her injury, but he continued. All in all, we presented 39 witnesses over eight days. It took the jury four hours to find him guilty on the first-degree.
    Unfortunately for the victims, this is not the end of the story. After the defendant left Dallas Medical Center, he went on to two more hospitals, injuring 15 more patients. During punishment, we presented 10 more patients and another 24 witnesses over the course of five days. The jury heard from patients who can no longer feel half their body, who cannot talk from vocal cord damage, who have no control over urinary function, who lost their skin and hair because their bodies were so stressed, who cannot use entire limbs, who have no control over hand movements, who had feeding tubes for months, and the list goes on—all injuries sustained at the defendant’s hands.  
    We finished the case with the defendant’s final surgery on Jeff Glidewell. During the surgery, the defendant biopsied a neck muscle, which he called a tumor, slashed the esophagus leaving an unfixable hole, hit the patient’s vertebral artery causing massive blood loss, stuffed a sponge inside to stop the bleeding, and sewed the patient up with the sponge still inside his neck. The sponge caused an infection that almost killed Mr. Glidewell. The defendant then abandoned him and even refused to return to the hospital after other doctors asked him to (they had discovered the sponge insides Jeff’s chest). Jeff lay in his hospital bed on his way toward death for days until other doctors stepped in to save him. Finally, after weeks of harassment and insistence by multiple doctors, the Texas Medical Board suspended the defendant’s license.

“Stone cold killer” email
From one of the civil attorneys, we obtained an email from the defendant to his girlfriend/physical assistant a few days before Mr. Passmore’s surgery. It was a stream of consciousness message obviously written while he was high on cocaine. Its most damning statement was when Duntsch claimed he was ready to “become a cold-blooded killer.” He also wrote, “What I am being is what I am, one of kind, a mother fucker stone cold killer that can buy or own or steal or ruin or build whatever he wants [sic].”
    The only way to authenticate the email was to call the girlfriend, Kimberly Morgan, but she came with problems. She herself was being sued civilly for some of the outcomes of Duntsch’s surgeries under the premise that she knew he was doing drugs but did not report him. When we first spoke with her, she was less than cooperative and the month the trial was set, she was deployed overseas with the Air Force. Ultimately, we were able to develop rapport with her, and she testified via Skype from the Middle East. She ended up being a fabulous witness for the State.

Enabling factors
Many factors enabled the defendant to continue operating even with devastating results. First, the University of Tennessee, as the original gatekeeper, allowed him to leave his fellowship and endorsed his abilities. His supervisors there were well aware that Duntsch had a drug problem. They had sent him to a program after there was a complaint he was under the influence of drugs while at work. After that, he was not allowed to operate again without direct supervision. But still, supervisors signed off on his qualifications. That could be because they had a financial interest in Duntsch’s success:  He owned a valuable patent, and they had started a company with him. But things went south quickly, and the defendant moved to Dallas.
    The second major problem was how the hospitals handled Duntsch. Baylor Scott & White allowed the defendant to resign so he would not be reported to the database and the hospital would not be exposed to lawsuits. Dallas Medical Center gave the defendant temporary privileges without waiting for the peer review file from Baylor. That would have told them about the appalling outcomes in the Summers and Martin cases. Further, Dallas Medical Center did not report the defendant to the databank either. The next two hospitals, Legacy Surgery Center of Frisco and University General Hospital, both had warning that the defendant had massive issues, but they ignored the warnings. In each of these situations, the defendant also manipulated the system. He lied to hospitals about his history and had a lawyer with him every time he negotiated his way in and out of jobs.
    Third, the Texas Medical Board (TMB) failed. Early in our investigation, we approached the TMB for assistance in understanding what went wrong. We knew some investigation had been done because the board had suspended Duntsch’s medical license. But when we contacted the person who actually performed the investigation, she was elusive. We were able to get basic records though a subpoena, but information was clearly missing. It became obvious that the TMB was not going to help us. After sifting through the records we understood why they wanted to stay as far away from the case as possible: It did not look good for TMB either. The medical board had complaints 10 months before the defendant’s last surgery. In the meantime, 20 more patients were injured, and at least seven doctors complained to the medical board about the urgency to stop Duntsch, to no avail.

Finality
In the end, we argued that the defendant had sentenced each patient to a life of pain (or death). Many of his patients need walkers, wheelchairs, and canes. Their pain is constant and takes many forms: burning, stabbing, searing, and throbbing. Some had multiple surgeries in an attempt to fix the damage he caused. His patients collectively lost over 23 liters of blood—the equivalent of more than 11 two-liter Coke bottles—when their surgeries should have resulted in minimal blood loss (around two liters for all patients combined). Many still need more surgery to fix his damage, but they are afraid of doctors now. The defendant did everything wrong a spine surgeon could do, and we argued that he deserved the same lifelong sentence as he gave to his patients. Jurors took only an hour to agree: They gave him life in prison.
    The trial was a cathartic process for many of the patients. Those who filed civil suits never testified, and some did not even have to file a lawsuit to get a settlement, so they never got to tell their story until the criminal case. Even the patients who had been reluctant to testify were glad they did in the end. Once they testified, they were able to listen to others tell their stories too. They began to bond, brought together by the dreadful circumstances. That was one of the most rewarding and unexpected results of the trial.
    The jurors ended up being a fantastic venire. They listened closely throughout the trial and took diligent notes. They clearly comprehended the medicine, evidenced by nodding their heads with understanding when they heard a medical explanation more than once on similar topics. Four of them cried as some of our victims testified. When we spoke to them after the trial, they were interested in the patients they had not heard about and many of the topics that we had decided not to cover in trial. A week after the trial was done, one of the jurors sent us personalized letters to forward to each injured patient.
    Every other possible check in the system had failed, but our 12 jurors finally got it right. By sentencing the defendant to life, they ensured that he can never hurt another patient again.