Reaching Down the Rabbit Hole (24 page)

For the moment, there was nothing more to do. The radiologist at the local hospital had written up a report that failed to raise a red flag, and until thirty minutes ago we had relied on that report. Until now, no one at the Brigham had looked closely at the scan. I still thought there was a good chance our radiologist might say, “Oh, yeah, it’s just a blip,” but that was hope staving off conviction. I went home to grab a snack and change my shirt. If they found nothing, then my day was over. Instead, when the phone rang at 4:15 p.m., I found out it was just starting.

“Your patient, Mr. Connaway, has a spinal epidural abscess. It’s massive.”

“Oh, shit!”

The bad news, I told Harry and his wife after my police escort, was that he had an aggressive infection within his spinal column that had begun to put pressure on the spinal cord, cutting off its blood supply. His spinal cord was slowly being strangled. This explained all of his symptoms.

It took me almost eighteen hours to catch a mistake that should have been caught when he was admitted, or even earlier when the ambulance delivered him to his local hospital. Had the technician there moved the image frame even a centimeter higher, no one could have missed it: the lower limit of an enormous abscess running from his neck down to his mid-back. It was all pus, and it appeared as a mere blur in the upper edge of the frame where no one was expecting to see anything. It was missed by a radiologist, it was missed by my junior resident, who possibly didn’t even see it, it was missed by an intern, who did look at it on the previous evening. If someone had told me there was an abnormality, I would have said, “Drop everything. Let’s get moving here!” Instead, the whole thing was now a horrifying mess. Every neurologist should keep epidural abscesses in
mind for this very reason. They’re not common; in fact they are quite rare, but they can be devastating if not caught immediately.

I had to get Connaway to an MRI machine at all costs. We needed to be sure about the abscess and find its extent so that it could be drained by a surgeon. His blood pressure had been fluctuating wildly all day, his fever was going up and down, he had coded twice, he had four tubes infusing fluids and blood-pressure-supporting medication into his arms because he was that sick, almost too sick to be shifted off the bed and onto a stretcher to bring to the machine seven floors down. Harry was on the brink. I said to the techs, “I don’t care. It might kill him, but if we miss this we miss our only shot.” Joelle placed the order. They called for him to come down at 5:30, which would have been okay, but the nurse said, “He’s crashing, we can’t bring him down now. He’ll have another cardiac arrest.” So they bumped him to the end of the list.

Two hours gone since the order was placed, then three hours, and they still couldn’t get him to the scanner. He was too unstable. Four hours, and he was still crashing. I kept insisting that we had to make this happen now. I finally said, “I’ll go down with him. We have to get this scan because I need to tell a surgeon what he has, what its extent is, and where he needs to operate. He might die down there, but we have no other choice.”

So three of us—the nurse, Joelle, and I—brought him down with a cardiac monitor and a bag that we compressed by hand to support his breathing. We did a quick-and-dirty MRI, and saw the whole extent of the abscess: It was about as bad as it gets.

Once we had Harry back in his room and were joined by his wife, I couldn’t dwell on what-ifs. There was still much to be done. “I think, unfortunately, that the infection has been around long enough, including today, that it is at risk of damaging his spinal cord irrevocably. So we need to drain the pus. The abscess needs to come out or he’ll be paralyzed, and wither away, and die. I can’t promise you that he’s not going to be paralyzed, even with an operation.”

It was 7:30 that evening when I called around to my neurosurgical
colleagues. At the Brigham, the responsibility for taking care of spine problems alternates between the neurosurgeons and the orthopedists, both of whom are extremely capable. I called my go-to guy first, a neurosurgeon. I was fairly sure he wouldn’t be anxious to come in and perform a ten-hour procedure overnight. He said, “I’d love to do it, but it’s their turn,” meaning the orthopedists. I then got hold of a senior resident in orthopedics, who tried to talk me out of doing it late at night because, as he said, “What’s the difference between now and the morning?” To which I replied, “It’s now or never. I’m going to transfer him to another hospital if you can’t do it.” Facing that threat, the resident called his attending physician, who is a very good guy, and he came in that evening at 9:30, looked at the film, started the case an hour later, and finished it at 8:00 the next morning. I was waiting for him to come out of the OR.

“He’s still paralyzed.”

Three hours later, back at the bedside, I had another talk with Mrs. Connaway, who was understandably shaken, and said that her husband absolutely did not want to be kept alive like this. “Five days ago he was playing tennis,” she said. “If there’s a prospect that he’s going to spend the rest of his life a quadriplegic, we need to let him go.”

Then it was Harry’s turn. He was so sick that it took all of my skill to get a little bit of yes or no out of him by lipreading. The tube in his throat and his by-now complete paralysis made it almost impossible for him to communicate. I had to get very close to him to be sure I had his full attention and that I could interpret his responses, a proximity that would have been inappropriate in any other setting. I leaned steeply over the bed, and brought our faces perpendicular to each other. I could tell he was in pain. His lips were moving, his shoulders barely shrugged. I could sense his animation and agitation. The abscess had penetrated all the way up to the nerves that innervate the neck muscles, so far up that he could only communicate by blinking and barely moving his lips. I spent forty minutes with him,
finally establishing that if he was going to be a tetraplegic, he just wanted it over with. I managed to persuade him to wait a day or two.

On the following day, not only was there no improvement, but when we did another MRI scan it was clear that the spinal cord had been turned to mush, that it had been completely deprived of blood flow because of the infection around it, and he was, metaphorically, cooked. Moreover, he was cooked at a level that’s almost incomprehensible: at the junction of the brain stem coming out of the skull, where it connects to the spinal cord. It doesn’t get any crueler. If it went any higher into the cranium, it would crash the medulla and kill him. That’s probably what was happening to him intermittently as his blood pressure careened from one extreme to the other. The team met and we concurred: Harry Connaway would be irrevocably and completely paralyzed. I would tell his wife, then talk to him in order to get another affirmation that he didn’t want anything more done to keep him alive. I did all of the talking. He responded as best he could, well enough to make it clear that he did not want to continue.

Because one of the results of an acute transection of a spinal cord that high up is an inability to sustain blood pressure, the decision was not about removing him from life support. All we had to do was stop raising the intravenous pressors, the medication that was keeping his blood pressure up. We wouldn’t withdraw anything. When his blood pressure started to decline, we wouldn’t raise it as we had before. Harry understood that. So we watched as his BP declined more and more. Two hours later, he died.

I told Harry’s wife everything, pretty much. I didn’t say that we blew it in quite those brutal terms, but I said he had a very extensive abscess, an aggressive infection, that it had grown and grown despite the surgery, despite the antibiotics, and that his spinal cord was thoroughly destroyed. I told her that she had made the right decision by encouraging us to let him die peacefully, because he would have lived as a quadriplegic on a ventilator, if he had survived at all. He made
clear in his determination that he did not want that under any circumstances, and therefore it was justified to back off and quietly let him die. I told her that. I did not elaborate on the timing that I was still beating myself up about. I did tell her that I wished we could have done better for him, and she was extraordinarily gracious. But I didn’t do what we are told to do, openly acknowledge our mistakes, because although I know now what the mistakes were, I couldn’t then be sure they had altered the outcome. It may well have been too late for the surgery to have made a difference. Nor would a nonspecialist understand the subtlety of how this had played out.

“It would help us very much,” I told her, “to be able to do an autopsy. It could provide closure in a case like this.” She agreed.

As residents learn the practice of medicine in a hospital setting, the last skill they acquire is in many ways the most important one, and the most difficult. They need to master the executive skills required to make happen the things that need to happen for their patients. If you need the country’s best hand surgeon to come in and reattach a hand at two o’clock in the morning, you’d better be able to make that happen. If your patient needs a tumor removed from an optic nerve before she goes blind, you’ve got to be able to make the calls and pull the right strings. And if you need to take your patient to another hospital to make it happen, then so be it. You’ll catch hell for it, but you’ve got to save your guy.

Of course, I didn’t save my guy. Maybe I could have. But I couldn’t be sure until we had the results of the autopsy. It would have been a lot easier to bury the whole thing and leave a big question mark. But I could not allow that to happen, not only for myself, but because I owed it to Harry. In my mind, I might have made an error, but I could still make something positive come of it. So the next day I took Joelle, Stanley, a junior resident, and a medical student down to the morgue where they could see firsthand what happens when you make a mistake. Hannah, who was on neuropathology rotation, also tagged
along. They all needed to see, as did I, whether we could have saved poor Connaway.

Part of my job is to impress upon the residents and the medical students that they are part of a tradition, and that the history of their profession is not merely of anecdotal interest. Unlike the study of, say, mathematics or physics, where it matters little to the high-level practitioner how Newton “discovered” the calculus or his theory of gravity, the practice of medicine benefits from revisiting the discoveries of the past. For a physician, seeing further means looking
over
the shoulders of giants.

The history of medical discovery is instructive, and no one person embodies that process more than William Osler, a Canadian physician born in 1849, who is considered by many to be the father of modern medicine. It was Osler who created the first medical residency program at Johns Hopkins, and established the foundations of medical training that are still largely in use today. As a result of his example, showing up at an autopsy is Oslerian, using the postmortem as a method of improving your own skills is Oslerian, teaching from that experience is Oslerian.

It may have been Osler who set the standards for deep medical training, but for me, the giant who brought clinical neuropathology to its most refined stage, not just looking at slides under a microscope and deciding whether a tumor is good or bad, but understanding how disease affects the nervous system, was Raymond Adams, who trained me at Mass General. Adams fully understood that he was working in the Oslerian tradition in that he was using pathology as the basic science of neurology, and neuropathology was to be his main vehicle for teaching. He was right, and is still right. To be a truly advanced neurologist, one could argue, you have to know about the sophisticated genetics of neurological disease and the cell biology of neurological disease and the immunology of neurological disease. That is the case for the full-fledged professional. But there is a sweet spot for the clinician, and you don’t have to go quite that far to hit it. That sweet
spot is still this kind of autopsy: the Osler-inspired neuropathology that Raymond Adams perfected. A lot of people might argue with that, but it’s true. Anyone who trained under Adams spent a full year doing neuropathology. This doesn’t happen anymore, but as a resident, I removed almost two hundred brains at autopsies in a single year. We took out the brains and the spinal cords, put them in formalin, waited two weeks, took out the material, diced it up, gave the technicians the slides we wanted to be prepared, and told them what stains were needed. It was a core part of the training.

These days, I occasionally remove a brain for the residents in order to point things out to them when there isn’t a lot at stake. But they don’t get the chance to do it themselves. Very few autopsies are performed today for a variety of reasons, mostly due to a misguided faith in the power of scans, and because Medicare and Medicaid do not pay for them. The hospital has to swallow the cost, and it is a very expensive proposition. And physicians, because they no longer tend to think of the autopsy as the immediate extension of clinical work, generally do not think about asking for permission to perform one. Where it was once a routine thing to do, it is now perceived as ghoulish.

In Osler’s day, a faculty member would not get his hands dirty at an autopsy. The cutting would have been done either by a resident or, in most cases, by an assistant called a
diener
, the German word for servant, or more precisely, a corpse servant. Today, the job is no longer menial; the diener is a trained specialist. From a medico-legal point of view, I wasn’t supposed to interfere in her work, and I didn’t. I asked for a full spinal removal. I needed to see the spinal column, leaving as much of the bone intact as possible. When she laid it out on the table, we could see the large column of pus. Its uppermost extent was clearly visible, as was its direct juxtaposition to the spinal cord. The intensity of the infection had blocked all the blood vessels. That in turn had caused the spinal cord to turn to cream pudding.

It is impossible to say what Harry’s spinal cord looked like when he had arrived at the Brigham, but from the autopsy, it looked like it
had been infarcted days earlier. The evolution of the infection was evident, but its cause would remain a mystery. In most instances of epidural abscess, the origin is an infection elsewhere in the body. Unless you introduce bacteria during an operation directly, it can only reach the spinal column via the bloodstream from somewhere else. The diener and the pathologists looked at every organ, every piece of skin, every orifice, even the anal fissures. There were no sores, no pneumonia, nothing in the heart, no endocarditis of the heart valves. That was odd.