“The blood trolley. But I still don’t understand why the hospital doctors didn’t realize the same thing.”
“Because it’s too rare, I’m telling you! Nobody ever thinks of TRALI.”
“‘No one expects the Spanish Inquisition’?”
“No one but the pathologist.”
I still wasn’t ready to go to Hirsch rounds with the case. I had one more call to make first. After I had reviewed the slides and pored over the chart yet again, I picked up the phone and called Doug Blackall. Professor Blackall was head of the blood bank at UCLA Medical Center, and he’d taught me the clinical pathology of blood banking when I was a medical student. I related the whole story. “The boyfriend said she was complaining of back pain, clutching her chest, and saying she was going to die.”
“You know what this is,” Dr. Blackall responded without hesitation.
“It’s TRALI.”
“That’s what it sounds like.”
“I knew it!” I barked a little too loudly. “I just wanted to hear you say it. It’s a diagnosis that usually isn’t made on autopsy.”
“Sounds like you didn’t diagnose it on autopsy. You diagnosed it by chart review, and confirmed it through the lab—all the time at your desk.”
I took the case to three o’clock rounds that same afternoon. Dr. Hirsch was skeptical. “Get the X-rays,” he commanded after I finished presenting. “And talk to the radiologist about it, see if he agrees with you. Then come back to me with it.”
The radiologist’s report, along with copies of the two chest films, arrived several weeks later. The X-rays were astonishing: Taken just twelve hours apart, they looked like before-and-after pictures of fatal lung injury, the latter image a shockingly total whiteout of accumulated fluid. I showed the X-rays to Dr. Hirsch after nine-thirty morgue rounds. “What’s the radiologist’s report say?” he wanted to know.
“Noncardiogenic pulmonary edema,” I read off the paper. “Given the time course, consistent with TRALI.”
Hirsch looked down at me—and cracked a tight smile. “Good case,” he said. It was the highest compliment about my investigative work I ever got from him, and I cherish it to this day.
In the end, that kind word from my boss was the only thing that pleased anyone involved in the Rivera case. The hospital’s blood bank had to go track down the donor and say something like, “We think you might have antibodies in your blood that are harmless to you but would prevent you from donating again, and we need to test your blood.” That sounds a lot better than, “Your bad blood just killed somebody, maybe. Come in and get tested so we can stop you before you kill again.” Once they had collected a blood sample from the donor and tested it, however, the blood bank discovered that the donor’s blood did not contain antigranulocyte antibodies. That meant the culprit in this case was antibodies in Veronica’s own plasma, reacting to the donor’s blood.
So, though Rivera’s death was a therapeutic complication, it was unavoidable—and not due to any error. The only error was on the part of the doctors and nurses at the hospital ward who did not recognize the transfusion reaction and report it to their blood bank, because they assumed she was just another junkie suffering the effects of an overdose. Regardless of whether the hospital missed the diagnosis, there is nothing Veronica’s caregivers could have done differently to prevent her death. Veronica really was anemic. From a clinical standpoint, she clearly needed the blood transfusion. Could she have lived without it? Maybe. Could anyone have foreseen the freak reaction that caused her TRALI? No.
TRALI is irreversible, often misdiagnosed, and kills hospital patients—but lots of things kill hospital patients. I have seen everything from improperly placed intravenous catheters to open-heart surgery lead to deaths that I mannered as therapeutic complications. Even a routine knee operation or cosmetic surgery can be fatal if the patient responds idiosyncratically to the anesthesia. During my two years training as a medical examiner in New York City, I was quick to learn that there is no such thing as “minor” surgery. “Minor surgery is surgery someone else has,” Dr. Hirsch liked to say.
Simon Nanikashvili was a septuagenarian with hardening of the arteries and heart disease. One of his carotid arteries was severely obstructed by cholesterol deposits, impeding blood flow to his brain, and unless he underwent surgery he was nearly certain to have a stroke. A vascular surgeon at Mount Sinai Hospital in Manhattan cleaned the plaque out of the diseased artery, then patched it back together by fashioning a graft out of another blood vessel. Everything seemed to go quite smoothly; Nanikashvili emerged from general anesthesia alert and in good spirits. The next night, however, he awoke with blood gushing out of the side of his neck, soaking through the bandages. His neck was swelling grotesquely, and his blood pressure was dropping. The Code Team arrived to intubate him and his doctors applied pressure to the wound, but Simon Nanikashvili was dead before they could get him to the operating room.
The surgical site was still closed when Nanikashvili’s body came to my autopsy table. When I opened it, I found a tremendous amount of blood in the narrow space of his neck. There was a half-inch hole in the repaired carotid artery, and I could see what had caused the fatal hemorrhage easily enough: The bright blue Prolene suture on the graft was hanging loose. The seam the surgeon had sewn to hold the patch in place had simply unraveled. After the photographer documented it in situ, I removed Nanikashvili’s neck block and placed the whole thing in a plastic container of formalin about the size of a jelly jar.
I certified the case the same day. The cause of death was bleeding from the surgical wound, and the manner was a straightforward therapeutic complication, because the surgery was elective. Nanikashvili could have waited a week, two weeks, a month. “He was living on borrowed time anyway,” his widow said when I called to relate what I had found at autopsy. He had survived a heart attack three years before and a broken hip a year ago. Either event could have ended his life, but vascular surgery might have extended it considerably—had it been successful.
That summer I fielded several calls from Mount Sinai Hospital’s Department of Risk Management, as they call their lawyers. They were investigating the death of Mr. Nanikashvili, and it wasn’t clear who was at fault. The surgeon was insisting that the suture must have broken, but the suture manufacturer, a company called Ethicon, claimed that the surgeon must have tied it wrong. Both sides wanted to examine the surgical site. With the two institutions girding for battle, my primary role was as the official legal custodian of the body—and the specimen.
Mr. Nanikashvili’s daughter gave me permission to release the sample for testing, so when the day came for its evaluation I signed out that critical piece of neck in its jar of formalin. The bright blue suture was visible floating around in there, still partially embedded in the graft wall. In our office lobby I met the serious, suited men who were going to examine it together. Dr. Patrick Lento, chief of hospital autopsy at Mount Sinai, was there on behalf of the hospital. Representing Ethicon was a retired vascular surgeon, Dr. Thomas Divilio. Also coming along was John Moalli, a PhD in polymer technology from MIT, hired by Ethicon to investigate any reported failures of their products.
Our first stop at Mount Sinai was Dr. Lento’s office, where we looked at a composite electron micrograph of test sutures that had been severed in different ways. Under such high magnification, the suture cut with a scalpel had a sharp-edged, square end. Scissors left a pancaked wedge. These were both easy to distinguish from the test suture that had been torn apart—that one looked like a knob of old candle wax, with frayed strands dribbling off its tip.
Together we considered the in situ autopsy photographs from my case file, and we all agreed that the reconstructed blood vessel certainly hadn’t degenerated on its own. The smooth edges of both the carotid artery and the vein graft showed no tissue tears. Dr. Divilio pointed to the bright blue thread in one picture. “You can see one end of the suture is straight, and the other is twisted like a pig’s tail, right?
There are no knots in sight. If the suture had broken, there would still be an intact knot somewhere.” He wasn’t gloating; still, I wondered what Mr. Nanikashvili’s surgeon, who had insisted the suture must have broken, would say to counter this argument. “More important,” Dr. Divilio continued, “this configuration suggests that the knot was tied wrong.” He looked at Dr. Lento. “Instead of squaring off the knots like a good sailor, the surgeon stacked a series of granny knots on top of one another. When these loops were then placed under load, the stacked throws unraveled, leaving the loose ends we see here.”
Dr. Lento didn’t say anything. He, like all of us, was eager to see the suture with his own eyes. I removed the surgical specimen from its preservative jar and gave it to him. He arranged it under a dissecting microscope and slowly worked the focus until, all of a sudden, the suture’s loose ends popped into view. They had the unmistakably keen “pancake” tip produced by surgical scissors on one end and an angular margin made by a scalpel on the other. The surgeon had sewn together the anastomosis—the critically important seam of the repair graft—and then cut the loose end off the needle with surgical scissors. But because the knots were not tied properly, the whole thing later unraveled—leaving one end of the thread twisted into a corkscrew.
The suture didn’t break. The Prolene material had not failed: The surgeon had. When we examined the suture seams from other parts of the carotid artery graft, we found maybe one square knot in the whole specimen. The surgeon had stacked granny knots in series of four or five, and some of them were unraveling on their own, right before our eyes. One had come completely undone, but the end had not yet pulled through the tissue—an arterial anastomosis hanging by a thread.
I was appalled. The first thing you learn as a medical student going into your surgery rotation is how to tie knots “like a good sailor,” as Dr. Divilio had put it. When I was in medical school I had spent hours at my kitchen table practicing my knots on pigs’ feet from the corner bodega, working the surgical needle and suture until I was dreaming about it at night. This vascular surgeon at Mount Sinai, one of the best hospitals in the world, had never learned to tie knots—and a patient had died after elective surgery as a consequence.
Dr. Lento was as horrified as I was. In order to confirm the diagnosis, we still wanted to see the suture ends from Mr. Nanikashvili’s neck block in the highest possible magnification and resolution, under scanning electron microscopy. The electron micrographs demonstrated even more definitively that the two ends of that string had been cut, not stretched or wrenched apart by force. Their sharp ends looked almost exactly like the test sutures that had been scissored and cut with a knife.
I got a good paper out of the Nanikashvili case, “Postmortem Analysis of Anastomotic Suture Line Disruption Following Carotid Endarterectomy,” coauthored with Pat Lento and John Moalli. It might seem a surprise that the chief of autopsy at the hospital where the surgeon had made this fatal medical error would put his name to a published paper on the subject, but that is one of the things I most love about being a doctor. Your mistakes, or the mistakes of people in your institution, can be used to educate others and to advance science. The upshot of the article, the lesson of the Nanikashvili case, is simple: Be good sailors, ye surgeons.
Sometimes doctors kill patients accidentally. Other times patients succumb to the known risks of a medical procedure. But every once in a while a medical examiner comes across a case of lethal malpractice. Gabriella Alonso was a young woman who got pregnant in 1996 and went to a private clinic in Queens for an elective abortion while she was in the seventh week of gestation. Under monitored anesthesia care, or MAC, she received sedative and analgesic drugs through an IV catheter, and oxygen through a face mask. The sedative rendered her unconscious, and the analgesic dulled the pain of the medical procedure.
Monitored anesthesia care is an intermediate step between a local anesthetic, in which only the area of the body being treated is numbed up, and general anesthesia, in which the patient’s vital functions are placed entirely in the hands of the medical team. Anesthesia is a spectrum, and doctors choose the type depending on the procedure, the patient’s level of anxiety, the type of equipment available, and other factors. General anesthesia can be administered only in a hospital by a medical doctor, but the law in most states allows a specially qualified nurse-anesthetist, working under the supervision of a doctor in an outpatient clinic, to administer MAC.
Gabriella Alonso’s therapeutic abortion was routine and uncomplicated. After finishing the procedure, the gynecologist and his nurse-anesthetist wheeled her into what they called the “recovery room”—really a patient waiting area, with eight easy chairs and no medical equipment. The only staff in the waiting area was an office secretary who answered the phone and processed bills. Several other patients who had also just come out of MAC shared the room with Gabriella, but no nurse monitored them. Dr. Ivan Kovachev operated his clinic with a skeleton crew: the nurse-anesthetist, named Dennis Morton, and a couple of phlebotomists who were trained to draw blood samples. Only Dr. Kovachev and Nurse Morton knew CPR.
According to Dr. Kovachev’s later statements, his patient was conscious when he brought her out of the operating suite but then “fell asleep” in the recovery room, and never woke up. He attributed Gabriella’s loss of consciousness to “too much Brevital,” the anesthetic medication they had administered during MAC. Brevital is a short-acting barbiturate derivative used to place a patient into a state of “twilight sleep” for minimally invasive operations. Barbiturates can kill you. One of their side effects is respiratory depression—they slow your breathing and keep it slow, even if your blood oxygen saturation drops to a critical level. That’s the reason for the enriched-oxygen mask. The person in charge of anesthesia has to keep a constant watch over your respiration rate and level of consciousness during the procedure—and, just as important, afterward. You have to be awake, responsive, and breathing on your own when that oxygen mask comes off.
When the secretary in the waiting area alerted Dr. Kovachev that Gabriella Alonso did not seem to be breathing, Dr. Kovachev and Nurse Morton performed CPR until an ambulance arrived and rushed her to Elmhurst Hospital. It was too late. The patient had gone into an irreversible coma. She spent the next six years in a persistent vegetative state. When Gabriella finally died in the summer of 2002, it became my task to reconstruct the exact sequence of events of that September day in 1996, and to decide on behalf of the City of New York whether Alonso’s death should be classified as an accident or as a therapeutic complication.
I ventured my early impressions at the three o’clock meeting the day I did the autopsy. “If Elmhurst never did a tox for Brevital, then I’m stuck with ‘prolonged vegetative state due to respiratory arrest following elective termination of pregnancy of seven-week fetus.’ I’m waiting for the medical records from the family’s lawyer, but at this point I’m favoring accident as the manner.”
“Why is this not a therapeutic complication?” Dr. Hirsch asked.
“I don’t see a prolonged vegetative state as an expected complication of an abortion. Something seriously wrong must have occurred, either during anesthesia or post-op monitoring. Plus, the doctor’s own opinion of ‘too much Brevital’ supports an inadvertent administration. That’s an accident, not a therapeutic complication.”
“You find anything interesting on autopsy?” Karen Turi asked.
“Not really—walnut brain and pneumonia,” I replied. Walnut brain is diagnostic shorthand for a brain that has atrophied due to a prolonged vegetative state, but has not rotted like a respirator brain. A walnut brain is shrunken and hard but the same shape and grayish-tan color as a healthy brain. Pneumonia had been the final mechanism of death, a common complication of hospitalization in a prolonged vegetative state. After Alonso died, the state’s Office of Professional Medical Conduct had performed an inquiry. I asked our legal department for a copy of those records.
Six weeks later the pile of pape
rs arrived, and I dove into them at my desk. They revealed several pertinent and troubling things. Dr. Ivan Kovachev attended medical school in Eastern Europe, but had completed only one year of medical residency in this country. He was not board certified in obstetrics and gynecology. On the day Gabriella Alonso came to his clinic, Dr. Kovachev performed seven other abortions over the course of an hour and a half. That works out to an average of eleven minutes per procedure. His own records indicated that each woman overlapped in the recovery room by ten to fifteen minutes. Kovachev’s office was an abortion mill. He was performing assembly-line medicine.
All the regulatory reviewers found “significant deviations from accepted standards of medical care” in Dr. Kovachev’s practice. He was supposed to provide a heart and respiration monitor, a blood pressure cuff, a code cart for emergency resuscitation. The clinic had none of these. The police found expired medications and confiscated several bottles of anesthetic agents that were two years past their use-by date. Nurse Morton’s own notes described patient Alonso as “drowsy” after the procedure, when he left her without any monitoring in the waiting area. Morton then immediately returned to the operating suite and anesthetized the next patient. By the time the billing secretary in the front office noticed that Gabriella wasn’t breathing, her brain might have been starved of oxygen for several minutes—but Kovachev and Morton had to revive the new patient from her anesthesia before they could wheel Gabriella back to the operating suite and start CPR.
Considered together, the medical records and the old police report left me convinced that Alonso’s death was not the result of a bad treatment outcome or of a simple error. A planned medical procedure can be hazardous even when everything goes right; if it ends in the patient’s death, I might manner the case as a therapeutic complication. Doctors make fatal mistakes, and I do rule some of these cases as accidents. But this was worse. Shoddy practice at Dr. Kovachev’s clinic made an avoidable injury into an inevitable one. Medical negligence brought about Gabriella Alonso’s death at the hands of her caregivers. Was this case a homicide?
Working Stiff Page 18