The Phones Are Restless

This month I’ll be spending one afternoon a week at our local psychiatric emergency room which, rather bizarrely, is called the H____ K____ Psychiatric Pavillion, which is an extension of the county hospital. I'm not sure what definition of the word the architects had in mind here. I guess maybe they hoped, through a chain of connotation, to evoke a wholesome county-fair ambience – if so, it didn’t work. The Pavillion is actually a complex of permanent buildings which are locked and lightly guarded. There is an emergency room where voluntarily and involuntarily committed patients are evaluated, and series of locked wards totalling sixty-odd beds where those deemed to be a danger to themselves, a danger to others, or gravely disabled can be admitted for short stays. Most patients are discharged from the emergency room, and a smaller proprotion are admitted. Of these, a very small proportion may end up as long-term patients in one of the four (4) State mental hospitals which survivied Ronald Reagan’s governorship – although most of our patients are “civil” rather than “forensic,” and only two State hospitals even admit the non-criminally insane, and their beds are divided between fifty-eight counties.

I’m only working in the emergency room. It is very unlike a normal emergency room in that there is almost no medical equipment. There is almost nothing at all, really. The main patient care area is a large square atrium, maybe sixty feet on a side. It contains only some built-in benches cushioned with blue mats attached with velcro, a few heavy articulated chairs that can fold down into beds, and a several smaller molded plastic chairs. The unifying principles of design are 1) that everything should be easy to clean and relatively proof against human body fluids, and 2) that nothing should be possible to use as a weapon. There were about twelve patients there today, milling and lolling around, all wearing unisex blue pajamas and yellow socks. All of them will either have committed themselves or, more likely, been brought in against their will on a psychiatric hold. Along the walls of the atrium are a few side-rooms which have only a single bed equipped with restraints in the middle of the floor. The pyschiatrist showing me around took me into one and pointed back the way we had come, above the door, where there’s a ledge that hides some recessed lighting. “If you don’t see the patient through the window in the door,” she says, “They may be on the ledge. In that case it’s probably best not to go in, especially by yourself…” I don't know what the architects were thinking when they designed locked rooms for violent psychiatric patients with ledges over the door, either.


This is a rather chaotic milieu in which to attempt a psychiatric interview, although the psychiatrists who work there have all evolved their own strategies for dealing with it. I tried seeing a patient today over by the phones – I thought this had the advantage of A) being near the nursing station so that I was easily visible to potential help (a doctor who was interviewing a patient in a private room was actually killed at this facility several years ago) B) being near the window, and therefore maybe attenuating the sense of confinement my patient was experiencing and C) being away from the other patients, since the phones are off a corner of the atrium. I didn’t really need to worry about any of this, since my patient wasn’t concerned about being overheard and was completely non-threatening. He had been picked up by the police wandering in the middle of the street trying to touch moving cars and spitting on payphones, with whom, he had apparently said, he was “at war,” but he wasn’t particularly bellicose at the moment - the phones in the Pavillion are free and can only receive calls, which I guess makes them neutral in this particular conflict.


What I didn’t initially realize about my choice of venue was that these are the phones people call to talk to relatives and friends who have been committed, and that they ring in incredibly shrill electronic tones so that they can be heard in the atrium. Moreover, the reception system is rather singular. The first time one rang, I didn’t realize it was the phone and was glancing around nervously in case it was some kind of alarm - but I took a cue from the nurses, who didn’t seem worried. A tall, elderly man in fifties-style glasses with a thick mustache loped around the corner and shot me a resentful look before answering the phone. “Hello, H_____ K_____ Psychiatric Crrrrisis,” he said in a deep and assured voice. He was wearing blue pajamas. He was an inmate. “Ah, yes, hang on a minute.” He padded back over to the atrium and returned with a teenaged boy wearing only an open-backed hospital gown. The boy picked up the phone, and the old man walked back to the atrium.


“Hey look,” said the boy, “I’m really sorry. I just…my mind was just racing, too fast for me to handle. I don’t even remember half the shit I did yesterday.”


Meanwhile my patient was trying to explain that he had been misquoted regarding the payphones by anti-payphone propagandists, and that while he had, for unspecified reasons, destroyed a pumpkin all he was really trying to do was return a lady’s wallet which he had found in the street, even though nobody had ever returned any of the five wallets he had lost over the years.


The phone rang again. The ad hoc receptionist returned. His look said, You know, I’m sitting all the way out there. It wouldn’t kill you to answer the damn phone.


“Hello, H______ K______ Psychiatric Crrrrisis… Well, I don’t know if he’s….Hang on, I’ll check.” He made for the atrium and returned a second later. “Yeah, I’m sorry. He’s a little too out of it to talk to you now. All right. Sorry about that.”


The patient was now explaining to me that all you really need, in the end, to make sure that “you have the four basic food groups, and dabble in drugs and alcohol a little.” I had been struggling to get a word in during his gregarious redaction of the police account, and I jumped in with the one question you’re always supposed to ask if you forget every other question, which in this case was a total non-sequitur and made me sound like an idiot: “Do you ever have any thoughts about harming yourself or other people?"


The boy was still on the phone: “No, they shot me full of something intravenously…No, after the police brought me to the emergency room…I don’t know, I was blacked out.”


“Naw, man,” the patient was saying, “never that, never that…”


The phone rang again, and the receptionist returned. This time he looked really incredulous – You can see me getting up to do this every time, he seemed to say, and yet you sit on your ass asking stupid questions. What’s wrong with you? Answer the phone!


“Hello, H_____ K_____ Psychiatric Crrrrrisis….Yeah….Oh, no, I think he’s been discharged. Yeah, he’s gone. Sorry.”


So next week, I’m not sitting by the phones. Aside from the constant interruption and the distracting conversations, I can’t deal with the guilt-trips. I assume today’s receptionist will be gone by then, but this seemed to be the standard arrangement so I imagine someone will have taken his place, who will probably be equally apalled by my lack of initiative. I’m looking forward to finding out what other administrative functions are open for impromptu staffing by the committed.

The author would like to R. Eley for her awesome techno-skills, which made this post better than it would otherwise have been.

Fabulous Prizes

I have decided to institute a series of competitions in the field of health care. These will be unpublicized and participants will not be made aware that their performance is being monitored, except for unpredictable posts in which I will award unsolicited recognition and, occasionally, cash prizes.

This week's award, granted for the Least Flattering Response to a Conflict of Interest Between Patient Care and Hospital Finances, goes to a local tertiary academic medical center. The winner will remain nameless here, although they will receive a plaque in the mail crediting them, which they may display in the main lobby of their hospital building or some other prominent place.

A month or so ago I saw a new patient, Sra. B, in my clinic. She is an immigrant from Mexico, and she had previously been followed by the transplant surgery unit of the academic medical center I mentioned. She had had utterly intractable diabetes, which had been complicated by end-stage renal failure and a variety of other unpleasant associates of high blood sugar. She is only in her early thirties and has a young child, so clearly this was a disaster. I say that she "had" diabetes, (you don't normally get to say this about people who aren't dead; in general once you've got it, you've got it) because she was the lucky recipient of a combined kidney-pancreas transplant (See fig. 1).

The first kidney/pancreas transplants were performed in the late seventies, and have become more common since although they are by no means considered a standard therapy for diabetes. This is a stunningly ambitious surgery. While kidney transplantation has actually become rather straightforward the pancreas is loathed and feared by everyone, particularly people who have to touch it. It has been described by surgeons as having the consistency of "tissue paper stuck together with spit" and is notorious for its capacity to digest itself and kill its owner in response to the most trivial provocations. It has this propensity because it manufactures the bulk of the enzymes which take the enormous, totally unusable polysaccharides, proteins, and fats we all eat and turn them into small, usable fragments which can be absorbed and turned into fluttering eyelashes and babies' bottoms and other nice things. Clearly, this is a locus of extreme hazard since there isn't really any difference, from the point of view of a proteolytic enzyme, between a fish eye you ate at a cocktail party to demonstrate your gonzo panache and the eye you use to see with. Both, to the pancreas, are offensively baroque, decadent blights which fairly cry out to be reduced to their component macromolecules. Pancreatitis ensues when any part of the pancreas' remarkably fragile system for protecting you from your own concentrated digestive juices is interrupted, which can often happen without any trauma at all. You have a little too much to drink, your calcium level gets a little too high, or you get stung by a scorpion, and all of a sudden your pancreas starts leaking enzymes into its own substance and dissolving itself. Caustic fluids leak out into your abdomen, and produce what is basically a large, internal burn. Pancreatitis can become a self-reinforcing crescendo of inflammation which generalizes to involve the entire body, and it routinely kills people despite maximal medical therapy. The pancreas is the body's self-destruct button.

So just touching this thing is a bit iffy; to take one out of a dead person and stick it in somebody with a defective one in the hopes of curing a lifelong and debilitating affliction takes real chutzpah. And, astonishingly, it works - transplant patients are not only cured of their diabetes and no longer need dialysis (so long as the transplant takes) but often see regression in other complications of diabetes, something which isn't seen to any really exciting extent with any other therapy.

The trade-off is that transplant recipients have to take potent immunosuppressive drugs for the rest of their lives to prevent their immune system, (our bodies' haematological Gestapo,) from picking up on minor genetic differences in the graft and destroying it. These drugs are not benign; some of them cause birth defects so reliably that it's generally recommended that patients take the oral contraceptive pill while on them, no matter what other form of birth control they use. All of them, by design, suppress your immune system which increases your susceptibility to infection. (The Gestapo metaphor is vitiated by the fact that, unlike Germany under the Third Reich, your body really is constantly beset by malevolent, genetically impure monstrosities who want to infiltrate, usurp, and destroy you. At the cellular level, a paranoid police state actually is usually in your best interests; if your cells are doing nothing wrong, they have nothing to hide...). Some of these drugs have narrow therapeutic indices and need monitoring of their concentration in plasma to insure that they aren't dangerously diluted or toxically concentrated. Laboratory tests indicating kidney and pancreas function need to be monitored regularly to catch any incipient rejection of the transplant early. When a patient who is highly motivated, disciplined, and ready to learn a lot about their condition comes together with a healthcare provider who is ready to make a serious, ongoing commitment to ensuring the best possible outcome for them, this operation can be a stunning demonstration of the god-like power of scientific medicine.

People who perform these surgeries select their patients very carefully to try to maximize the probability that things will go well for them; you do not, obviously, want to do something like this to someone who never takes pills or misses appointments or has no fixed address or is just basically perverse and bloody minded, because you'll end up making things worse than they were before you intervened. Now, instead of diabetes and renal failure, the patient will have diabetes, renal failure, transplant rejection, and various drug toxicities.

Sra. B is covered by a subset of Medicaid which only covers emergencies, and to some extent palliative care, and for which anybody is eligible, regardless of immigration status. It doesn't, however, cover chronic care - like the kind you need to monitor immunosuppression and transplant viability over a lifetime. The academic medical center funded the actual transplant, probably because it was part of a study, or because the kidney-pancreas transplant unit enhances its prestige, or perhaps as a purely charitable gesture. Medicaid definitely does not cover this kind of operation. (Medicaid doesn't even cover insulin dosing pens that make clicking noises so that blind diabetics can dose their insulin accurately.) Moreover, I am sure the transplant was not her idea - she could never have afforded it, and I don't think she even understands exactly what they did to her.

So she had the surgery, and they followed her for a while, but then she lost her emergency Medicaid funding which, in any case, would have eventually stopped paying for her care there. At this point, they were presented with a conflict between the center's financial interest and their duty of care to the patient. And it's because of the choice they made that they win this week's award. They could have said, "You know, we're the ones who suggested that this lady have this surgery, and we did it to her, for free; so really, we've incurred a moral obligation to see it through, regardless of whether we make our money back." This would have been the "right" answer. Instead, the transplant unit told her to arrange further follow-up at my hospital, (which has a mandate to treat all residents of the county regardless of their ability to pay but, glaringly, does not have a transplant surgery unit,) and she found her way into my primary care clinic. They decided that it was more important to use her slot for an insured patient than it was to provide optimal medical supervision of potentially dangerous medical circumstances which, let's remember, they created. The also decided that they were comfortable with the risks they assumed on her behalf by referring her management to someone who has been practicing medicine for exactly one year as of last Thursday, and who (like most of his seniors who have significantly more experience) doesn't know much about the chronic use of dangerous immunosuppressive drugs in transplant surgery patients.

As I mentioned, there are stringent protocols for evaluating people as candidates for transplant surgery - they have to be reliable, easy to locate, reasonably intelligent, compliant with medical instructions, etc., and if they're not they never make it onto the waiting list. This experience has made me think that we should impose similarly stringent criteria on institutions that want to perform this kind of dramatic intervention. For instance, we might require that they state explicitly their understanding that the intervention they propose creates potential dangers which persist for the entire life of their patients; that, given this, they assume a contractual obligation to provide lifelong follow-up care regardless of changes in their patients' ability to pay; and that, should it become necessary, they only transfer their duty of care to facilities staffed by persons of comparable specialist training. This doesn't seem like a lot to ask. After all, it's really just a specific application of a general principle articulated in the Hippocratic Corpus some twenty-five hundred years ago: "to help, or at least do no harm".

Sra. B is doing well. I got the transplant unit to give me an official list of her medications and tell me what lab tests I should be monitoring and how often. I got our "Utilization Review" department to pay for her to go see them when one of these tests appeared to be going in the wrong direction, and for her subsequent admission and treatment for transplant rejection. I've managed, through cajoling, wheedling, and appeals to higher principles a follow-up appointment for her there to make sure everything is going all right, although I had to agree to order all lab tests at my hospital and fax them the results. The ignominy of doing their clerical work, I have to say, has tempered some of the boundless pride I felt at the implication that I was transplant surgeon material contained in the original "referral," but if it means they'll pay attention to her, I guess I'll survive. There's an older doctor here who likes to say, of situations like this, "I'm Board-Certified in grovelling, whining, and guilt-tripping."

I'm seeing her on a weekly to monthly basis at the moment. She always comes to clinic with her daughter, who is nine years old and extremely vigilant about her mother's health. She is also incongrously knowledgeable about the workings of the Medicaid program, since she often has to act as a translator. She stays on top of things and remembers to follow questions up when her mother forgets. Somewhat strangely, I think, she wants to be a doctor when she grows up.

Sleeping With The Fishes

El Sueño de la Razon Produce Monstruos

(The Sleep of Reason Brings Forth Monsters)

Francisco Goya

Fish, as is well known, sleep while swimming. This sounds like a really good trick to learn, especially right now.

I wrote something else last night around 3:00AM that I was going to post today, but as I drove home close to noon I realized that what's really on my mind at the moment is sleep, its absence, and how I feel about all of that. I've been provided with a natural opportunity for reflection on this (in my capacity as experiment of nature) in that as I write, I am cementing a strong lead in the race for my Most Extreme Work to Sleep Ratio: out of the last seventy-five hours I have slept about ten. I got up on Tuesday morning, worked a full day in the ICU, then covered the medicine wards until Wednesday morning, when I went back to the ICU to see my regular patients. Then I went home around noon, tried to have a day (not an especially ambitious one, just a day not spent in the hospital) went to bed around eleven. I worked Thursday morning in the ICU, and spent the afternoon in my primary care clinic. I stayed on overnight, again covering the medical wards, and after an eventful morning in the ICU I am now home, coming up on hour seventy-six. My cats have disappeared. I suspect that they have given up on me, and struck out for someplace where the streets are paved with moths and by-catch; a place where with a with a little bit of hard work and a dream, a cat can make something of itself.

The scary thing is that this sort of thing isn't actually that uncommon in the United States. In my residency program it's rare to do every second day on call like this - I actually did this deliberately, in order to avoid a scheduled night on call next week. At some hospitals (although this is increasingly rare) housestaff are deliberately put on call every third night, although usually it's a team of junior doctors. My hospital, with some audacity, leaves it to one intern who has already worked a full day to cover four wards and an intensive care unit. There is also a second or third year resident at night, but they're usually busy helping admit patients from the emergency room or consulting on surgical patients with medical problems. The chain of referral, in any case, starts with the intern who can then decide to call their resident if they feel the situation is beyond their competence. This means the intern's pager goes off, at best, every twenty minutes and at worst almost continuously. This is not compatible with restful sleep. Moreover, (this has happened to me twice now) if you do try to sleep it's always right when you're drifting into REM that you get a call about someone who sounds actually, properly sick and whom you definitely need to go evaluate personally. This is an awful feeling. Not only is it physically unpleasant to be woken up from a fitful nap on a hospital cot by a screeching pager, it's very scary to be asked to make important medical decisions when you are barely awake. I don't even try to sleep anymore, because both of the times I did it I was called from my bed to evaluate somebody who was moving towards septic shock, something that needs to be approached with decisive intelligence even by the well-rested.

A great deal of research has been done on the effects of this kind of sleep deprivation, much of it conducted specifically on doctors in training, since we are an ideal cohort to study, along with PhD students, truck drivers, and soldiers. Overwhelmingly, this research hows that significant decrements in one's ability to think clearly about medicine increase proportionally to the number of hours without sleep. This brings to mind a friend of mine's characterization of sociology as "banal insights into the blatantly obvious". Personally, I think that succinctly describes at least this observation, but in case you're curious one prospective study published in the New England Journal of Medicine in 2004 compared two groups of interns, one working on a call schedule which required thirty-hour shifts every third night and one working on in more conventional shifts. The authors found that the "traditional" group made over fifty percent more "nonintercepted" serious medical errors than the experimental group, and twenty-two percent more serious errors overall - which is to say, the screwed up much more often, and their mistakes were far less frequently caught before they actually damaged someone.

I find it striking that a bunch of highly educated people all of whom had been through medical residency thought that they needed to organize a prospective trial to prove this. Or, to be slightly more cynical, I find it a rather sinister indication of the status quo's tenacity that they should have to marshal this kind of investigative firepower to prove the intuitively obvious point that the best person to handle an urgent situation involving the possibility of death or permanent disability is not the one who's just fallen asleep after, say, twenty hours of continuous work. Not, I would hasten to add, that this study or others like it managed to change much, at least here. The American Council for Graduate Medical Education recently adopted some self-imposed rules (e.g. the 80-hour week, the 30-hour maximum continuous duty period, the requirement of one day off every week averaged over four weeks,) which mildly ameliorated the situation, but are pretty obviously aimed at obviating legal regulation, not at creating a well-rested workforce. I have a friend from medical school now working in England who tells me that for every night he works he gets not only the subsequent day off, but also an entirely separate day, with its attendant night. When I tell American residents about this the responses are usually notable for their creative deployment of astonished profanity.

Last night, at a sleep-hour to wake-hour ratio of around 9:53, I was called to see a patient, (let's call him Mr. I,) in the transitional care unit, (which is the low-calorie version of the intensive care unit,) who was having trouble breathing. Mr. I is a big man in his early forties with severe lung disease and right-sided heart failure, who is trying desperately to get on a heart-lung transplant list. (In fact, he was brought in by an ambulance from the airport where he was about to board a flight to Vanderbilt to be assessed for surgery. I should write Alanis Morisette with this story - maybe the royalties from the resulting hit single would pay for his operation.) Anyway, I went through his chart and lab records in some detail, and by the time I was done writing orders I was feeling pretty good about myself. I had come up with a systematic plan to stabilize him, organ-system by organ-system, and I was also pretty sure I had identified a major flaw in the clinical reasoning that had been applied to the case so far. I realized that I could improve his breathing significantly just by giving him an intravenous infusion of fluid, (this will probably sound stupid if you're used to dealing with heart failure, but in his case it made sense. Full explanation available upon request), without having to transfer him to the ICU or put in a central venous line or intubate him or anything dramatic and unpleasant like that. This isn't, by the way, any kind of comment on the day team - they had many other patients who required an equal amount of attention and the course they pursued was a pretty standard and reasonable one. I just happened to be called at an ebb in the perpetual pager-storm when I could actually concentrate on the case, and I had a fresh pair of eyes and records of everything they'd done and every test result they'd gotten in front of me all at once.

The rest of the night was comparatively reasonable, relative to some of my worse experiences on call. There were periods of, oh, fifteen minutes when my pager didn't go off at all, and when I was called it was mainly for minor things I could handle over the phone. I even had time to do a little reading. I pulled up a few articles on sleep deprivation - it seemed topical. I was not called to see Mr. I again.

When morning finally came I was beginning to flag. I can always tell when physiological exhaustion is setting in because I start losing my short term memory and become increasingly dependent on notes. I wrote up the morning lab reports and overnight vitals for my one ICU patient, came up with a plan of action for her, and wandered down to Grand Rounds, the Friday Department of Medicine morning lecture, to see if some free bagels and omega-3 rich smoked salmon spread would improve my failing higher faculties.

This is where the whole thing devolved into a sort of hallucinatory nightmare.

The Grand Rounds lecture is given in part of the old hospital building where there's no overhead intercom system. This is relevant because that's one of the ways the hospital operators announce medical emergencies, which all members of the ICU team are supposed to attend to whenever they're in the hospital. We were all at the lecture - however, of the four interns, two residents, and one attending physician, least three of us always carry "code pagers" which are little walkie-talkies that are supposed to emit a shrieking noise whenever there's a major medical emergency, (or "code blue") followed by an announcement of its location. Theoretically, then, we shouldn't need to be in earshot of the PA system to know when there's a code. For some reason, which I have insistently pursued without any success, despite enhanced interrogation techniques, there is often a failure to communicate between the operators who announce code blues and the pagers. This fact never loses its novelty for me, since it is transparently imperative that the code pagers work reliably; they are, after all, probably the only pagers in the hospital on which life and death may actually hinge. As you must have already guessed from all the bitter asides, there was a code blue, and not a single one of them went off. We heard about it by word of mouth, which is a totally moronic way to convey this kind of information. Pagers were invented in the first place so that people who need to know things immediately, like doctors and crack dealers, could circumvent exactly this problem.

I have to confess that my first thought was, Shit. There goes my ten O'clock exit visa. My second thought was also Shit, because the code had been called in Mr. I's room. The gentleman I had been feeling so chuffed about having managed in such a systematic and creative way the night before. Mr. I, who I hadn't transferred to the ICU because I'd been confident he would be fine in the TCU. Mr. I, who I was pretty sure just needed a little more intravascular volume to enhance his right ventricular pre-load, and was definitely, definitely, not going to be emergently intubated or go into ventricular fibrillation or otherwise explode. Shit, I thought, I'm a statistic. I haven't slept enough, and I just committed a major, nonintercepted medical error. I screwed up something crucial last night which I would have gotten right if I had been more awake, and now he's going to die. Shit.

The scene which evolved when we arrived was horrendous, and only made more so by a persistently farcical which quickly developed. Mr. I was lying on his back on the bed, his eyes rolling dysynchronously. He was being ventilated by an Emergency Department resident with a bag-and-mask apparatus. Apparently, he had been returning from the bathroom and had collapsed on his bed, unresponsive.

Codes are supposed to be swift, decisive, methodical, and run according to evidence-based and generally known algorithms. They are like this to some extent, but usually the first five minutes (which are obviously the most important ones) are pretty chaotic. Everyone within three floors comes running and crowds into the room. The problem is usually that it's not entirely clear who's in command. Technically it's the medical resident on call, but most of our medical residents feel diffident about assuming responsibility because the emergency room residents are much more experienced. The ED residents know this and many of them try to politely defer. The attending usually cuts through the chaos to demand that absolutely imperative things like attaching defibrillators and blood pressure monitors happen, but they also usually try to induce a resident to assume control - so there's often this awkward few minutes where it's not entirely clear who's in charge or what order things are supposed to happen in, and so everything starts happening at once while the attending, the medicine resident, and the emergency resident engage in the medical crisis version of "I insist," "Oh no, I insist," "No, no, after you," "Oh no, I couldn't possibly," etc. The defibrillator is attached, and there's a frenzy of flying needles as people try to insert large-bore lines into any vessel they can see or feel. At the head of the bed there's a running conversation about how the bagging is going and whether we're going to intubate, while at the foot there's an independent debate about the rhythm on the monitor, and somebody is loudly demanding to know where that blood pressure cuff is, while someone else is yelling for a central line kit, and yet another person is trying to figure out if they should be starting chest compressions. Eventually somebody assumes command, and things become slightly more systematic and fall into a recognizable resuscitation algorithm. At this point, people get bored and drift out, and things become slightly more manageable. This code, however, was more hectic than most. First there was an incredibly ambivalent debate about whether or not Mr. I in fact had a pulse to go with the electrical activity on the monitor. Nobody, including me, could be absolutely sure. Then it emerged that the crash-cart, which is supposed to have all the stuff you need in it, was missing all kinds of things - sterlizing scrubs, sutures, etc., and there was a mad exodus of residents pelting down to nursing stations at either end of the building to search desperately for the right kind of suture or an extra arterial blood gas kit. The problem is that all the wards, for mysterious reasons, have their own budgets and are therefore reflexively defensive of their supplies - this habit of mind is so ingrained that even when I sprinted into the nursing station of an adjacent ward to requisition some minor antiseptic products, saying breathlessly, "Quick, it's for the code!" the clerk gave me a look of disgusted persecution and whined, "Don't they have that stuff on TCU?"

And then the room flooded. I got back just as my immediate senior, who had been the resident on-call the night before and slept only a few hours more than I did was putting a central line into Mr. I's femoral vein. The line splits at the end into three tributaries, and you have to fill them with water so that when you inject into them you don't shoot air bubbles into the patient's circulation.

"Flush," she said, "I need a flush."

I grabbed a few syringes of saline and started emptying them into the central line tray so she could draw them up with a sterile syringe. Behind me, someone said "Jesus, what's all this water?"

"It's just saline," I said, "for the line."

"Really?" she replied. Then I realized that somebody had put the intubation tray on the sink, which had pushed the cold water tap on, and that it had been running onto a towel which was blocking the drain for several minutes. We were all standing in an inch of bloody water.

And, naturally, as all this was going on, I was dredging my increasingly dysfunctional memory, trying to fiture out which part of the brilliant plan I had formulated the night before had contained the moronic oversight which was going to turn out to have caused this absurd, deadly situation.

..................................

I like to think that I would still be telling this story had the whole thing actually been my fault. Reviewing what I did afterwards with some older doctors, it appears, mercifully, that I was right; it was a good plan, and it probably averted a slow decline towards respiratory failure. It looks, at this point, like whatever happened was a sudden and unpredictable event, (probably a pulmonary embolism,) which I definitely had nothing to do with and may have even delayed. Mr. I. was stable in the ICU when I finally left around eleven, receiving empiric therapy for his presumed PE. Unfortunately he is too large for any of our scanners, (an increasingly frequent problem, for which the radiology technicians have no sympathy whatever since somebody cajoled them to accept a patient over the scanner table's weight rating and the table promptly broke) so we won't know exactly what happened until (or rather, unless) we can deduce it from his clinical course and some indirect testing.

It is nice to be fairly assured that I'm not (yet) the author of a "nonintercepted" serious medical error - although, like a smoker with unexpectedly normal lung function tests, I have to fight the consequent impulse to assume that I'm somehow an exception, that the formula which relates sleep deprivation to misdiagnosis, inappropriate treatment, and simple mistakes somehow doesn't apply to me. And who knows - maybe if I had slept more in the last few days I would have seen something last night which would have allowed me to predict and avert this.

Fish, as is well known, sleep while swimming. This is because fish are stupid and the scope of their professional lives outside of the closely circumscribed fields of sex and predation is basically insignificant. Also, they have developed such that in order simply not to swim, they have to be dead. Even then there is some gray area. I recently caught a fish which, as is standard practice in these matters, I hit on the back of its head with the blunt side of a hunting knife until it had no apparent reserves of centrally controlled executive or automatic function, even within the stunningly narrow ambit of quotidian fishy activity. Its gills ceased to move. Its cold, clammy little heart lay still in its slimy breast. It's gooey brain ceased to issue programmatic statements on the laying of eggs and eating of bugs. I left it on the river bank and went looking for other places to catch fish, and when I returned, it was gone. I found it after a few minutes of searching (it was a big fish) lying in the shallows on its side, dead, its tail twitching as motor complexes hard-wired into its spinal cord burned through their last reserves of energy.

So bollocks to the fish. I breathe when I sleep, and sometimes I move a limb aimlessly, which is about as impressive as sleeping fish swimming - i.e., not very impressive at all. When I am not sleeping, however, I enjoy my membership in a species of highly complicated mammal, and the privileges which inhere therein. These include, but not limited to the pursuit of elaborate, elegant non-genetic structures of cognition and discourse like music, language, art, and medicine. And if I'm going to participate in these at the level to which I aspire, I need a little more sleep.

Good night.

On Pulling The Plug Right The First Time

Did you ever see Dr. Tetrazzini perform? I say "perform" advisedly, because his operations were performances. He would start by throwing a scalpel across the room into the patient. Then he would make his entrance like a ballet dancer - his speed was incredible.

-William S. Burroughs, Naked Lunch

There's a general and romantic vision of dramatic medical situations which always preserves a starring role for the doctor. The role can be that of the hero (i.e. the deeply sensitive oncologist who shows a patient how through their terminal diagnosis, they can really live for the first time) or the villain (the brutal and insensitive surgeon whose callousness brings the young mother to tears) but it's usually central. However, more often I find myself a bit player in titanic interpersonal dramas. I first started making these supporting-cast appearances in medical student. Some may remember my brief cameo as "Sympathetic Guy in Scrubs #3" at St. Helier Hospital in 2002, where I stood by, having no real idea what to do, as a family watched their grandfather die suddenly of a dissecting aortic aneurysm. I have since made a number of other minor appearances in immense human tragedies, which have garnered little critical attention.

The thing about these roles is that it's important to play them well, because, actually, people do remember them. Just because you're a peripheral figure in, say, they precipitous and unexpected death of a young father doesn't mean that, like every other detail of the situation, your performance isn't irrevocably seared into the memories of his family. Even as a bit player, one owes it to the people one finds oneself in these situations with not to do anything stupid or insensitive that will mar their inevitably photographic and deeply emotional memories of the event.

This can be harder than one might think it would be.

I am currently working in the ICU. Inevitably there is a lot of death around, some of which I actually participate in. I had a patient a little while ago who came in completely comatose and immediately had to be intubated and attached to a ventilator. It turned out that he had had a devastating stroke which had destroyed most of the back half of his brain completely and irrevocably. He was seventy-seven, he had enjoyed a long and happy life, and his family all agreed that the best thing to do would be to "withdraw care" which is our anodyne way of saying "pull the plug."

Once the decision was made, he was moved out of the ICU to a private room on a low-acuity ward. There's something of the knell about these transfers. It's a sort of pragmatic, capitulatory ritual where we admit that death is inevitable and that the dying don't need intensive care. Once we authorize death, the imminent decedent ceases to be a patient and becomes a sort of guest of the hospital for the (brief) period during which his vital signs persist. We sent him up to the ward with the ventilator, however, because the family understandably wanted to be there when we turned it off.

I spoke with one of his daughters, and she told me that the family were planning to convene for the dread moment at six. Having never done this before, I was a little nervous. My superiors didn't seem worried about my doing this alone - I guess they took the view that, since the whole point was to kill the patient, there wasn't really any possibility of medical error. But, again, these are the kind of scenes that are engraved in peoples minds with ruthless fidelity for the rest of their lives; I didn't want his children's recollection of their father's death to include some bumbling technical snafu perpetrated a cackhanded intern, so I spent my few spare minutes in the afternoon familiarizing myself with the basic maneuvers involved in removing an endotracheal tube without, say, leaving a trail of bloody slime across the patient's face. I talked to our palliative care specialist, who gave me some advice on setting up a morphine drip to prevent any unnecessary discomfort (although I doubt highly if the patient was physically capable of experiencing any) and by the time I arrived on the ward at six I felt relatively well prepared.

I hadn't accounted for two things, however. One was the family. There were at least forty of them. They spilled out into the hall, and occupied every possible cranny of the small hospital room. They all looked grave, and when I squeezed into the room I saw they had put the patriarch's favorite hat on him and sat him up in bed in a crude approximation of dignity. I spoke quietly with the daughter, who told me they were still waiting for a few family members, so I waited in the nursing station, fidgeting with the computer and getting progressively more nervous as more and more family member joined the swelling throng. While they were waiting, they broke into an impromptu collective eulogy in a sort of call and response format: someone would list some of his virtues, or give thanks for one of his qualities or accomplishments, and the group would issue a warm, collective "amen", or "praise Jesus". After fifteen or twenty minutes of this, they broke into song. They were clearly all regular churchgoers, because it wasn't some warbling dirge; without any kind of cue they broke into "Amazing Grace" in what sounded like a 17-part harmony. The strong swell of voices emanated from the room and permeated every corner of the ward. Some of the nurses looked slightly aghast. It was awesome and beautiful, but it did nothing for my stage fright.

After a few more songs, everybody had arrived and I made my way to the bedside with his nurse, literally pushing between people and squeezing into a space which hadn't existed a few seconds earlier by the bed. I explained briefly what I was about to do. Some of them had already started crying. I leaned down, and snipped the plastic bands holding the endotracheal tube in place. I hit a button on the ventilator which silences all alarms for two minutes. A few of the family broke into a sort of free-form, harmonic moaning, which caught on until I was inundated in rich, layered song. I deflated the balloon that held the tube in place internally, and, covering his mouth with some gauze, smoothly drew it out. I suctioned all the gross stuff out of his mouth, and did my best to arrange his face in a manner compatible with dignity. He continued to breath, irregularly but, for the moment, sustainably, as I had hoped he would. I was planning to be a few floors away at the actual moment of death.

This was when I realized that there was something else I hadn't accounted for. My plan had been to turn off the ventilator and simply squeeze back out of the room, leaving the family to their grief. The problem with this excellent scheme was that, I realized, I didn't actually know how to turn the thing off. Modern ventilators are very safe and dependable machines. They have backup batteries so that if the plug is, literally, pulled, they continue functioning as normal. Moreover, they have a number of alarms which are triggered by any worrying interuption in any of the parameters they continuously monitor. Needless to say, if you just pull the tube out they go crazy and make a lot of loud noises which are, by design, impossible to ignore. You can press the button I've already mentioned to silence the alarms for two minutes, but it's time limited so that it's impossible to forget to turn them back on. The only way to actually shut it up is to turn the whole thing off, which (I was rapidly realizing as I scrolled futilely through it's many on-screen menus,) was a much more carefully concealed function than I had anticipated. I hit the two-minute button again, but I knew this could only be a temporary solution - I couldn't stand at the bedside pressing the button every two minutes until everybody left, and I couldn't let their moment of profound loss and grief be turned into awkward bathos by the cacophany of incessant electronic beeping noises which would begin to issue from the machine if I stopped pressing the button. This was exactly the sort of thing I had been afraid of. For their entire lives, these people were going to remember me, with a twinge of puzzled irritation, as that vulture-like doctor who lurked at the head of the bed pointlessly fiddling with the ventilator while their venerable patriarch slowly passed on.

Eventually I gave up. I unplugged the ventilator, unscrewed its hoses from the oxygen supply wrapped its various tubes and cords roughly around it, and slowly, painfully, with much bumping of elbows and scraping of shins, hauled it back through the crowd and out of the room. Then I called the respiratory therapist and, I imagine sounding pretty beleagured and desperate, said, "Hey, I just have a quick ventilator question for you."

My two minutes were up again, and every alarm the thing had came on at once.

"How the fuck do you turn these things off?" I snarled.

(That's it. I'm not telling you how you turn them off. You'll have to figure it out for yourself. Or just just think slightly harder than I did about all the things that could possibly go wrong in such an apparently simple situation.)