CHOOSING QUALITY OVER QUANTITY:

[B]efore he even begins to consider heroic medical interventions in the very last stages of life, he spends dozens and dozens of pages pondering assisted living facilities.

When they're bad - and often, he thinks, they are grotesquely bad - they are arranged for methodical bureaucratic routine. Patients will eat at these times, have meds administered at those, will go to sleep at this time, wake up at that. You can think of this routine as cruel indifference or rationalize it as cost-saving. Maybe that's right. But also in the mix is an image of medical care: do what it takes to keep these ailing bodies running.

Keren Brown Wilson's mother, a stroke victim, balked at one such facility.  "Take me home," she pleaded; "Get me out of here." Gawande reports on Wilson's later innovations. The tenants - not patients - of Portland's Park Place "had private apartments with a full bath, kitchen, and a front door that locked (a touch many found particularly hard to imagine). They were allowed to have pets and to choose their own carpeting and furniture. They were given control over temperature settings, food, who came into their home and when." More control than that: "if they wanted not to take certain medications that made them feel groggy; if they wanted to eat pizza and M&M's despite swallowing problems and no teeth and a doctor who'd said they should eat only pureed goop - well, they could."

Gawande reports too on Bill Thomas's innovations, introduced in response to listless despair in upstate New York's Chase Memorial Nursing Home. Missing, Chase thought, "was life itself," even in the presence of all those bodies dutifully inhaling and exhaling, ingesting and excreting. So he proposed bringing in plants, dogs, cats, parakeets, plenty of visiting children. Other staff hesitated: no way, they figured, that state agencies would grant all the waivers needed for this loopy fantasy. But Thomas prevailed. Picture the slapstick comedy: a hundred parakeets got dropped off before the birdcages had been delivered. "The driver ... released them into the beauty salon on the ground floor, shut the door, and left. The cages arrived later that day, but in flat boxes, unassembled." The residents guffawed - and pitched in to assemble the cages, line them with newspaper, catch the birds. Rabbits and hens followed, an after-school childcare program, too.

Thomas detailed the astonishing effects. "People who had been completely withdrawn and nonambulatory started coming to the nurses' station and saying, 'I'll take the dog for a walk.'" And - get this - the patients took fewer drugs, especially fewer "psychotropic drugs for agitation." (Some disturbed and disturbing symptoms are doubtless caused by those shrinking brains and their pathologies. But some are a response to living in a socially sterile setting.) The death rate fell. In an experiment in a Connecticut nursing home, Gawande adds, researchers divided patients into two groups. Patients in one group got a plant to water and a lecture on taking responsibility for their lives. Patients in the other had plants watered for them and listened to a lecture on how staff were responsible for their well-being. Eighteen months later, the first group "proved more active and alert and appeared to live longer."

So it's not just that the mindless bureaucratic routines advance patients' health at the cost of robbing them of autonomy, happiness, engagement with life. It turns out the routines don't even serve patients' health. If you let an ailing elderly patient lock her door, you might delay a nurse's access in an emergency, and that might be a catastrophe. And I suppose parakeets can spread infections, especially when sick tenants are cleaning their birdcages; and anyone who's brought up children know that one of their jobs is to ferry around microbes, so they must have been ferrying microbes into Chase Memorial, too. But it turns out those risks are well worth taking. That there's no tradeoff here between quality of life and life itself might seem too good to be true. It might be true anyway.

By the time Gawande turns to end-of-life issues, the reader is primed to think concretely about quality of life. Not QALYs, the quantitative measure supposed to rationalize precarious and tricky judgments about the attractiveness of various healthcare interventions. I don't know if it's out of contempt, though I kinda hope it is, but anyway Gawande doesn't even mention them. Not the bloated share of GNP we spend on healthcare, the lopsided bit of it we spend on the closing weeks or months of people's lives. Gawande wants to suggest that our feverish campaign to do more or less whatever it takes to keep patients' bodies chugging along is profoundly misguided.

Gawande insists instead that we stop and learn what patients want. Not by having them write out living wills in advance, with a series of answers to often unhelpfully abstract questions. Rather by having gutwrenching conversations with them about their illnesses, their prognoses, the upsides and downsides of various possible courses of treatment. Their answers of course will vary. One seventy-four-year-old emeritus professor was looking at a 20% chance of becoming a quadriplegic if doctors tried to remove a growth in his spinal cord. His daughter talked to him about the options. "Well," he said, "if I'm able to eat chocolate ice cream and watch football on TV, then I'm willing to stay alive. I'm willing to go through a lot of pain if I have a shot at that." After surgery, he developed bleeding in his spinal cord. "The bleeding had already made him nearly quadriplegic, and he would remain severely disabled for many months and likely forever." But if they didn't stop the bleeding, he would die. His daughter asked the surgeons if he'd be able to eat chocolate ice cream and watch football on TV. They said yes; she okayed the surgery.

There's a facile frame to account for this: defer to the patient's autonomy. If people want to live, do what you can; if they don't, well, that's their choice. This isn't making cost-justified QALY calculations. And it's got nothing to do with death panels, government-run or otherwise. Like all other plans to defer to patients' wishes, it won't help with cases where people are suddenly stricken and unconscious. And you might well fret that the week after you've been told you have stage IV cancer might not be the best time to make a sensible decision about what care you do and don't want. But at his most tantalizing, Gawande is suggesting that there's more here than autonomy.

Follow @brothersjudd