Working Stiff

by Judy Melinek, Md

  There is a risk. The carotid endarterectomy itself could dislodge chunks of cholesterol and send them into the brain, causing a stroke. If you decide to undergo a high-risk elective operation like a four-way cabbage, and the carotid endarterectomy is a required precursor, then the possibility of stroke during the endarterectomy is a known risk of the procedure. An operation can be both elective and necessary—“elective” does not mean “optional” in the world of medicine. Elective simply means nonemergent.

  Everything appeared to go smoothly during Patricia Cadet’s carotid endarterectomy and four-vessel CABG. Despite their successful repair of her damaged heart, however, Cadet’s doctors soon discovered that they had arrived at a terrible clinical result. She had suffered a stroke during the operation, and emerged from anesthesia paralyzed on one side and unable to communicate. Her brain started to swell from the inside out, and the damage got worse. After a few days, Patricia lost consciousness, then went into a coma. A little while later her newly repaired heart stopped beating.

  I met Patricia’s brother David in the family room we reserve for identification. He had insisted he needed to see me in person. He was a haggard older man, dressed like a laborer, erect in stature—and angry. “I just don’t understand how a healthy woman . . . ,” he began, and trailed off. “She was joking, happy as can be before she went to surgery. How can she go, just like that?”

  “Did your sister’s doctor explain to you what happened to her?”

  “He was talking, but he didn’t explain. He said something about a clot and something about a stroke, but I didn’t understand what he was saying. And he wouldn’t look me in the eye.” He pointed to his own eyes with the V of two fingers, while keeping them locked on mine.

  I paused to weigh my words. David Cadet distrusted me. I was a young and inexperienced white doctor who was going to lie to him just like the surgeon had. The hospital’s lawyers were circling the wagons. His sister was dead, it was the hospital’s fault, and we were going to cover it up because there was money at stake.

  “First of all,” I said, and met his gaze while I did, “I want you to know I don’t work for the hospital where your sister died, or for any hospital. I don’t have any interest in defending those doctors, Mr. Cadet. I am a civil servant and I am paid to be objective. This is your tax dollars at work.”

  His gaze softened a bit, and he allowed a small smile. “That’s why I came here.”

  “Let me explain to you how my ruling of therapeutic complication in Patricia’s case differs from a death I would rule as natural causes,” I continued. “If the surgery had been an emergency and your sister was rushed to the hospital dying, I would have to blame the thing that brought her there. But this wasn’t an emergency. Since Patricia was undergoing an elective procedure to repair her heart and was otherwise healthy when she went into the operating room—joking and happy, like you said—she would not have died that day but for the surgery. That is what makes her death a therapeutic complication.”

  “Did she die because of the surgery?”

  “Yes,” I said right back. “Your sister’s heart disease was very extensive, and she certainly would have died without the surgery—after a few weeks, or maybe as much as a couple of months. No more than that. But, yes, she would still be alive today if she hadn’t gone into that operating room.” I was willing to confirm the one thing that the hospital’s doctors were not: The surgery had killed the patient.

  David Cadet nodded but didn’t speak. After a moment he got up and headed for the door. “Thank you, Doctor,” he said quietly. Then, before he left, he turned to me again, and this time his eyes showed nothing but sorrow. “You know, I told Tricia everything would be all right when they wheeled her away to the operating room.”

  “I’m sorry,” I said—and guessed I was the first doctor who had dared to.

  I knew Mr. Cadet didn’t have a lawsuit. Even if he was still angry enough to sue, it was unlikely he would find a lawyer willing to take the case. I don’t think that was why Mr. Cadet wanted to speak to me, though. He just wanted someone to give him a straight answer.

  I was taught in medical school, and even more so in my surgery training, to express myself in the passive voice and with clinically specific language. “An embolus was dislodged from the occluding atherosclerotic plaque, leading to an ischemic injury” is medspeak for, “While we were trying to clear out the blocked artery, a piece of the fatty clot came loose and caused a stroke.” It was only during my training by Dr. Hirsch that I weaned myself off medspeak.

  Medical examiners have to communicate with the lay public more than we do with other medical professionals, and it’s more important to be understood than to be precise and scientifically accurate. For the death certificate, things are different; there we use precise terminology. Hirsch drummed it into our heads, however, that when talking to a decedent’s next of kin over the phone, or addressing a jury from the stand, the forensic pathologist must seek to use plain speech and eschew condescension. He taught us to use “hardening of the arteries from cholesterol” instead of “atherosclerosis.” Tell the jury the person died of “a heart attack,” not “myocardial infarction.” You can even tell the next of kin that their elderly loved one died of old age. It is more comforting than “presbycardia.”

  In my career straddling medicine and the law I’ve seen bad bedside manner lead to litigation many times. David Cadet had assumed that his sister’s doctors had screwed something up and killed her, and were lying to him to cover their asses. The truth they had failed to make clear, or were afraid to tell him, was that they had cut short Tricia’s remaining days in the effort to extend them.

  Physicians make false assumptions about their patients too. It is especially easy to prejudge a junkie. Doctors across the spectrum of care have plenty of experience with the dysfunctional families and crazy stories that accompany drug abuse and alcoholism. But if you don’t keep a professionally open mind, you might fail to address the real problem. Worse, if you try to treat the medical problem you think you see without fully exploring the differential diagnosis—what Hirsch calls “speculation built on a foundation of assumption”—you can kill your patient.

  Veronica Rivera was a twenty-eight-year-old with a history of alcohol abuse, who arrived on my autopsy table in the early spring of 2002. The investigator’s report said her fiancé had taken her to the emergency room because she was feeling “weak and sick.” Clinicians diagnosed Rivera with anemia and ordered a transfusion of red blood cells. While she was in her ward bed, she suddenly and unexpectedly stopped breathing. A Code Blue team scrambled to Veronica’s bedside with a crash cart and intubated her, but even mechanical ventilation couldn’t keep her alive. Veronica Rivera’s airway was open but her lungs weren’t working, and the amount of oxygen in her blood kept dropping. A urine test registered positive for benzodiazepines, morphine, and methadone. These three drugs are commonly abused, so the chief clinician assumed someone had been sneaking Rivera a fix in the hospital. After a couple of days on the respirator, she was brain-dead. In the medical record her doctor attributed Veronica Rivera’s death to illegal narcotics and double pneumonia.

  When I opened Rivera up, I found the fatty liver of a chronic substance abuser. Her lungs were solid to the touch and showed the signs of ARDS, adult respiratory distress syndrome, the end-stage process of many acute lung diseases. The presence of ARDS at the time of death didn’t tell me much—anyone who goes into respiratory arrest and then goes to an ICU for a few days ends up with crappy-looking lungs. I needed to know what had caused Veronica to stop breathing in the first place. She also had a gray and swollen “respirator brain” with the consistency of pudding, as I had suspected she might. I pulled one undigested pill out of Rivera’s stomach and sent it for toxicology.

  These scant autopsy findings, along with the positive drug screen, pointed to an overdose of narcotics as the most likely cause of Rivera’s respiratory arrest. I was going to have to wait for the definitiv
e toxicology result to come back, and that could take months. In the meantime I had to use other investigative means to figure out exactly which drugs, and how much of them, Rivera had taken. The MLI’s report said that Rivera’s fiancé was at her bedside when she went into arrest, so I suspected he might have given her the drugs. Since he was also apparently the last person to talk to her, calling him was my next step.

  Luis was eager to speak to me and gave me an earful about the terrible care Veronica had received at the hospital. He was certain they had killed her. “I convinced her to go to the emergency room because she was looking real bad. She was so weak. But if I hadn’t brought her to that place, she’d still be alive.”

  We talked about his experience in the ER for a little while before I made my first pass at the tough questions. “Did Veronica take anything while she was in the hospital?”

  “A nurse came by with a couple of horse pills a few minutes before the blood transfusion. That was it.”

  “Veronica didn’t ask . . .” I stopped myself. I needed to take care if I was going to glean information without making accusations. “You didn’t bring any drugs in for her, from outside the hospital?”

  “No.”

  “According to the medical records she had a history of alcoholism. Can you tell me how much she drank?”

  “Not very much. Maybe a glass or two a day. She wasn’t a heavy drinker.”

  Luis was denying his fiancée was an alcoholic, but the dead woman’s liver had told me otherwise. I tacked and fired again. “Did Veronica have a history of recreational drug use?”

  “What do you mean?”

  “Did she use heroin or tranks, downers?”

  “No.” He hadn’t paused or hedged.

  “Was she in a methadone program?”

  “No.”

  “I’m asking you this because Veronica’s urine test came up positive for methadone, morphine, and benzos. Where could she have gotten these drugs?”

  “I don’t know. She didn’t use drugs! I don’t use drugs. I was with her the whole time. I didn’t leave her side.” There was a pause. If this guy was a junkie, he was also one hell of an actor. He was convincing me, and I have been lied to by the best professionals in the field. “Could that drug test be positive because of something they gave her at the hospital?” he asked.

  The hospital doctors wouldn’t have prescribed her any methadone, I knew that much for sure. I flipped through the few pages the hospital had faxed. “The hospital records indicate they only gave her fluids and ordered the blood transfusion before the code.”

  “Maybe it was in the blood they gave her?” Luis speculated. I knew that was unlikely but not impossible. Blood donors are rejected for drug use, but blood banks screen donors with a questionnaire and an interview, not a drug test. “All I know is they said her blood was low, and the nurses came in and hung up a bag of blood and started it going in her arm,” he continued. “I was sitting by her bed, and all of a sudden she jerks, and says she has this horrible back pain, and she wants me to rub her back.”

  An alarm bell went off in my head as soon as I heard him say that. “You’re sure the back pain came after they started the blood transfusion? Not before?”

  “Yeah, right after the nurse left—and really bad. So I’m rubbing her back to try to make her feel better—and then she says she can’t breathe! So I hit the button for the nurse, and they all came in and whisked me out of there. It must have been something in that blood . . . it must have been.”

  “What’s a trolley have to do with her blood transfusion?” T.J. asked, when I told him the story.

  “T-R-A-L-I, a transfusion-related acute lung injury!”

  He remained underwhelmed. “So? What’s that?”

  “Oh, it’s the coolest thing—a once-in-a-career finding! Back pain after a blood transfusion—as soon as he said that, I knew it was TRALI. I can’t prove it, though, not until I get the tox back.”

  “So it could just be a drug overdose?”

  “Yeah, but it isn’t! This is TRALI, I’m telling you. The boyfriend’s story is just too compelling. As soon as I got off the phone with him I called the hospital blood bank to alert them.”

  “Why’s it such a big deal?”

  “TRALI causes flash pulmonary edema, kills people, and there’s no way to reverse it. The doctors and nurses on the ward failed to diagnose it. They didn’t make the connection between the blood transfusion and the respiratory arrest—and that’s a major medical error. Blood banks get shut down by the FDA over stuff like this. And when a blood bank gets shut down, so does the hospital.”

  “Holy shit!”

  “Yeah, I’m sure that’s what the blood bank director was thinking. He has to learn about a fatal transfusion reaction from the medical examiner? That is bad if you’re him.”

  The mechanism of TRALI is poorly understood. We know it has to do with antibodies in either the donor’s or the recipient’s plasma, the fluid medium of blood. Antibodies are specialized proteins that protect you from diseases by causing foreign bacteria or viruses to clump together. The body then mounts an immune response and sends white blood cells to destroy the invading pathogens. When you get a transfusion, your body can be fooled into accepting the outside blood—including the plasma, and all the antibodies in it—as if it were yours. In a small handful of cases, however, either the donor’s blood or your own will contain antigranulocyte antibodies. These proteins cause the white blood cells themselves to clump together. More white blood cells flood the area in an immune response, and the antigranulocyte antibodies cause them to clump up, in a vicious cycle. Your body mounts a haywire attack on its own tissues.

  This attack is most damaging to capillaries, particularly in the lungs. A frothy fluid (edema) inundates the air spaces in the alveolar sacs, and their outer walls become encrusted with protein deposits, which impede gas exchange. An immune response to a false threat overwhelms your body’s ability to take in oxygen. This can happen very rapidly—instantly in Veronica’s case, if this truly was a case of TRALI. As her lungs filled with fluid and their capillaries crusted over, no amount of mechanical ventilation could push up her oxygen saturation. Veronica’s brain starved for oxygen, and she died.

  A few days later, a copy of Veronica Rivera’s complete medical record arrived at my desk. The emergency room doctors treated her diligently and documented their work well. The Code Blue was done by the book. The ward doctors diagnosed Rivera as anemic based on standard tests of hemoglobin efficacy. But then—once her urine test came back positive for drugs of abuse—everyone, doctors and nurses both, seemed to conclude Rivera was just one more Bronx junkie cluttering up the ICU. No one investigated the reason for her sudden collapse; they all assumed it was due to a drug overdose.

  The chart told me their assumptions were wrong. Two of the drugs present in Rivera’s urine had come off the shelves of the hospital dispensary, not off the street. The benzodiazepine they found is the active chemical in midazolam, a sedative, and the opiate screen was positive due to the painkiller fentanyl. The Code Blue team’s record showed they had administered both of these medications when they intubated Veronica during her respiratory arrest.

  That wasn’t all they got wrong. The culture tests were negative for the bacterial infections that commonly cause respiratory disease, belying her doctor’s speculation of double pneumonia. In fact, tests showed she wasn’t suffering from any sort of infection. “He pulled that pneumonia out of his ass,” I said to Stuart as he and I sat back-to-back doing paperwork in our shared cubicle.

  “Sounds like guesswork, not diagnosis,” he agreed, after I showed him the chart. Stuart had built a career for himself as a laboratory pathologist before he went into forensics, so his opinion in this case carried extra weight.

  “And get a load of this—they did two X-rays, one in the ER when she was first admitted, and one in the ICU after they intubated her.”

  “To check tube placement.”

  “Right. T
he first film was negative. The second, twelve hours later, showed a complete whiteout. Her lungs were full of fluid. The ward doc attributed it to pneumonia.”

  “Not in twelve hours, no way.”

  “Exactly! Thank you!”

  “Congestive heart failure can cause a whiteout like that.”

  “Right, but at autopsy her heart was healthy and normal.”

  “What about anaphylactic shock from an allergic reaction?”

  “I’m waiting to hear back from the lab on a tryptase level,” I replied. “If that comes back normal and the tox is negative, this is TRALI. It’s got to be!”

  Stuart raised a skeptical eyebrow. “I wonder if Hirsch will agree,” he said. He had me there.

  When the toxicology report finally arrived in mid-June, it showed that Veronica had a normal tryptase level, which ruled out an allergic reaction. The OCME tox lab also had a surprise in store for me—and a shock for those clinicians who had declared Veronica Rivera a junkie based on the hospital urine test.

  “The methadone was never there?” T.J. was as astonished as I had been when I saw it in black and white. “How the hell does that happen?”

  “It was a false positive. Urine tests are less reliable than blood tests for screening drugs. They are performed using antibodies that sometimes cross-react. They’re sensitive but not specific, so a small percentage of patients who aren’t intoxicated will have a positive screen. That’s why a urine screen is not admissible in court as a forensic sample. Our labs have to confirm those tests using blood samples.”

  “The boyfriend was telling you the truth.”

  “Yes. Veronica was not a drug addict, and an acute drug overdose definitely did not kill her. She wasn’t even drunk—her blood alcohol was zero. She hadn’t had some random drug allergy either. That leaves only one thing in the differential diagnosis.”