Saturday, September 5, 2009

"End-of-life" discussions and lawyer-client communications

A remarkable article in the NY Times not long ago, by Anemona Hartocollis, "At the End, Offering Not a Cure but Comfort" (August 20, 2009, starting at page A1).

First, it reports that "Dr. Nicholas Christakis, an internist and social scientist at Harvard who has studied end-of-life care, has found that doctors are generally bad at making prognoses. The better they know a patient, the worse they are at prognosticating, possibly, Dr. Christakis has theorized, because they view death as a personal failure. Most predictions are overly optimistic, he has found, and the sicker the patient, the more likely the doctor is to overestimate the length of survival." (Page A16) Without having seen Dr. Christakis' study, it seems to me that another factor here may simply be empathy -- the better the doctors know the patients, the more they feel for them and the more they hope they'll survive and the more their predictions are shaped by their hopes. If accurate predictions are what we want here -- far from a self-evident proposition, as the article reflects -- then this would be a reason for doctors not to know their patients so well. It would, similarly, be a reason for lawyers not to know their clients so well.

Second, the article vividly illustrates the central role that communication without explicit words may play in some end-of-life discussions. The article tells the story of a patient with "an uncertain future" as a result of a recurrence of her cancer. Her doctor asks a range of questions, beginning with "What are your biggest concerns?" and moving on to "What's your understanding of the status of the tumor?" That question leads to an exchange in which the patient, having heard that treatments are available that can achieve some results, says "But there is no way of knowing it, right?" The doctor replies, "There are no established cures. Think very carefully about what treatments you do and don't want to have as these issues arise." The article says that the patient then "fired back with her wish to be 'alive again'"; that is, she replied ""It's more or less, I want to be alive again. Going here, going there. My husband, I want to be able to do things for him." And the doctor then asks, "What gives you strength?" and this question leads to discussion of shopping and the patient's husband teasing her about dressing up to watch TV at home. (The dialogue is recounted twice, with different details, at A1 and A16.)

In all this, the doctor never delivers a prognosis (the article says that he "had been called in to tell her the bad news: she was sicker than she realized, and the progrnosis did not look good"), and in fact the patient dies -- less than four months later -- without ever having been "told ... directy that she was going to die." Why not? The doctor felt that the essential precautions had been taken to protect her, through the appointment of her husband as a health care proxy, and that she had signalled to him that she wanted to "focus on the details of her everyday life, rather than a desire to foresee the future."

It's clearly part of autonomy for a person to know her fate as well as it can be known. But it also seems clearly part of autonomy for a person to decline this knowledge. The problem is that this is knowledge that is hard to explicitly decline when it is explicitly offered. So doctor and patient need to proceed by indirection. The patient does not say, "Don't tell me" and the doctor doesn't say "Would you rather not know?" Instead, the doctor asks the patient what her concerns are, and when the concerns she articulates have to do with everyday living, the doctor infers that she wants to focus on the elements of her life in her remaining time rather than on how much that remaining time is. If that decision does her no concrete harm, there's every reason to honor it -- except that there remains inescapable uncertainty about whether it was in fact her decision.

The article's account of this patient's grim final months seems to reflect that doctors continued to try to fight her cancer, rather than turning solely to "palliative care," though this isn't entirely clear. It would have been much more troubling if they had taken the conversation I described above as an authorization to provide only palliative care thereafter, since that course of action could very easily have done the patient concrete harm, defined as shortening her life. A patient might very reasonably choose that course, preferring not to endure the rigors of heroic treatments, or her health care proxy might make the same decision, but for the doctor to make the decision without the approval of the patient or the person she's chosen to speak for her would be hard to square with autonomy.

But if the doctors continued to fight the cancer, based on the requests of this patient, it must be said that the patient made those requests without explicitly knowing the best available answer to a relevant question, namely how long she had. (Her husband may have known, but the article seems to indicate that he didn't take over decisionmaking until almost the very end.) It's possible to argue that the likely prognostications would have been so uncertain and so error-prone that they wouldn't have had any relevance, but I doubt this. The problem with the information isn't that it would have been totally irrelevant, but that its degree of relevance might have been much less than its impact on the patient and on her thoughts and feelings. The doctor felt the patient was telling him that this was so, and that he was following her guidance. He may very well have been right. But, again, he could not ask her in so many words, nor could she answer in so many words. The critical communication had to take place implicitly and indirectly.

End-of-life discussions are far from the standard fare of doctors or lawyers. But I wonder how many critical pieces of information may also pass, and perhaps must pass, between lawyers and clients by similarly uncharted paths.

It's also hard not to think that the emotions that affect doctors' erroneous prognostications will also affect their assessments of when their patients wish to hear those prognoses -- and that the same will be true for lawyers. The particular palliative care physician at the center of the Times story wears "coolness" as "his armor," and seems disinclined to attempt prognoses in terms of days or weeks anyway, but surely the effort to respond to death with coolness itself affects what a physician hears his patients saying. So would a more emotional engagement with the patients. There is no entirely objective position from which to have conversations like this. Perhaps there is, in truth, no entirely objective position from which to have any conversation with a patient or client.