Tuesday, December 13, 2011

When end of life is not

Twenty years ago, I helped save a man's life.
So begins this New York Times essay by Peter Bach, MD, where he talks about the inadequacy of resource use at the end of life as a policy metric. Now, I am not very fond of policy metrics, as most of you know. So, imagine my surprise when I found myself disagreeing vehemently with Peter's argument. Well, to be fair, I did not disagree with him completely. I only disagreed with the thesis that he constructed, skillfully yet transparently fallaciously (wow, a double adverb, I am going to literary hell!) Here is what got me.

He describes a case of a middle-aged man who was experiencing a disorganized heart rhythm, which ultimately resulted in dead bowel and sepsis. The man became critically ill, the story continues, but three weeks later he went home alive and well. This, Dr. Bach says, is why end of life resource utilization is a bad metric: if this guy, who had a high risk of dying, had in fact died in the hospital, the resources spent on his hospital care would have been considered wasted by the measurement. And I could not agree more that lumping all terminal resource use under one umbrella of wasteful spending is idiotic. Unfortunately, knowingly or not, Peter presented a faulty argument.

The case he used as an example is not the case. Indeed it is a straw man constructed for the cynical purpose of easy knock-down. When we talk about futile care, we are not referring to this middle-aged (presumably) relatively healthy guy, no. We are talking about that 95-year-old nursing home patient with advanced dementia being treated in an ICU for urosepsis, or coming into the hospital for a G-tube placement because of no longer being able to eat or drink. We are talking about patients with advanced heart failure and metastatic cancer, whose chances of surviving for the subsequent three months are less than 25%. And yes, we are also talking about some middle-aged guy with gut ischemia, sepsis and worsening multi-organ failure whose chances of surviving to hospital discharge are close to nil; but in his case, instead of being clear from the beginning, the situation evolves.

So, yes, the costs of end of life care, and specifically hospitalizations, are staggering. But more importantly, among patients with terminal illnesses like metastatic cancer, advanced heart failure and dementia, hospitalizations and heroic interventions at the end of life cause unnecessary pain and suffering, and without much, if any, benefit in return. Their families and caregivers suffer as well, and many studies suggest that these caregivers are not interested in prolonging suffering, provided they are aware of the prognosis. Unfortunately, just as many studies suggest that communication between doctors and patients' families about these difficult issues is less than stellar.

So, let me play the devil's advocate and pretend that I support end of life resource utilization as a quality metric. If I did, I certainly would not be interested in depriving Dr. Bach's middle-aged acutely ill patient of the chance to survive. In fact, my aim would be to make sure that we align resource use with where it can do most good, and turn away from interventions that are apt merely to prolong dying.        


  1. Marya, I agree completely that life support for the elderly and very infirm is costly and (worse) harmful more often than not. But when I read Bach's essay, I didn't construe his intention was to support that sort of use.

    What I got out of it, and why I liked the piece, was that it points to how easy it is to draw conclusions from aggregated data, which leads to loss of valuable information about individuals and, potentially, loss of some (young) lives.

  2. Elaine, I do not doubt that what you say is true, yet there are some very good studies looking at these terminal hospitalizations, and my impression is that the patient Peter describes is not the bulk of the problem in question. Susan Mitchell and Joan Teno published a study in JAGS a few years ago, where they noted that there are nearly 14,000 dementia deaths in hospitals in the US annually. In another study they documented that 20% of all NH patients with terminal dementia get aggressive care, including hospitalizations. I can cite studies in CHF, metastatic cancer and other terminal illnesses. And you know how bad docs are at discussing end of life decisions (not to mention all the studies and surveys that confirm this).
    I agree that these ratings are stupid and disingenuous and say nothing about the care. But I also believe that the argument in the essay is not the right one, or at least it is but a small part of what is wrong with the metric.