Maybe Doctors Don't Die So Differently

John Mandrola, MD


July 07, 2016

I was wrong. Maybe you were too.

After reading Dr Ken Murray's now famous essay, "How Doctors Die," I believed the narrative: of course doctors die differently. We die better.

It makes sense: doctors own inside knowledge; we know death is normal, and we know the limits and dangers of medical care delivered at the end of life. Doctors are informed and empowered patients, and therefore, when it comes to end-of-life care, we would not be bullied by the healthcare system. We would shun futility and value quality over quantity.

New research tells a different story. A group of researchers led by Dr Dan Matlock (University of Colorado) set out to test the hypothesis that US doctors use less healthcare in their last months of life.[1]

What they found is both surprising and informative. Matlock and colleagues used an AMA database to match nearly 10,000 physician decedents to a random sample of more than 190,000 nonphysician decedents. All of these Medicare beneficiaries died between 2008 and 2010. The outcome measures were inpatient care and hospice claims in the last 6 months of life. For inpatient services, they included general hospital admissions and ICU use. The research team adjusted for multiple covariates, including sociodemographic data, comorbidities, and regional variations in healthcare.

The results are easy to summarize: after adjustment, similar proportions of physicians (67%) and nonphysicians (70%) were admitted to the hospital in the last 6 months of life. The mean number of days spent in the hospital in the last 6 months and 1 month of life were nearly identical for physician and nonphysicians. ICU stays in the last 6 months or 1 month of life were not significantly different. Slightly more physicians (46%) than nonphysicians (43%) used hospice services (P < .001).


The first thing to say about this study is that it is retrospective, observational, and based on administrative data. In the discussion section, the authors clearly explain these limitations, including the possibility of unmeasured confounders.

That said, the study included 200,000 patients, and it measured clear outcomes. Admission to the hospital, ICU, or hospice seem like reasonable surrogates for healthcare usage in the last stages of life.

Three thoughts come to mind.

Data Trump Anecdote. I had thought the anecdotes were typical: doctors diagnosed with terminal disease simply close the office and go live their remaining days in peace. We get off the metaphorical train of end-of-life care. We get hospice; we die well. Because we know the deal.

That admittedly biased view stems from more than just a belief in anecdotes: prior surveys of physician attitudes toward end-of-life care supported the notion that physicians would choose less intense care .[2,3] The key word being would, as in the future. The future is easy. Being terminally ill in the present is not easy.

Physicians Are Human. When ill patients ask me, "What would you do if you were me?"I almost always resist the urge. How can I presume to know what it's like to be in their body or to have their disease? Hubris scares me, and therefore I rarely interject my healthy person's point of view. Surely this brand of caution applies to terminal illness.

My wife, Staci, is a palliative care and hospice doctor. We talk often about death. In my confident moments, I think: I'm not afraid of dying. I've had a good life, and I have Staci to look after me. I will not suffer.

But that's a healthy person's view. The ugly path to death, which far too many patients take, is clear in hindsight, but hazy in the future. The train of end-of-life care can accelerate slowly: the chemotherapy that had a small chance of benefit blunts the immune system. Things are okay for a while. Then pneumonia develops, and a Foley catheter leads to a urinary infection. Bed rest and a DVT. Maybe bleeding. A transfusion reaction. Months later, frailty leads to a fall and more surgery.

One could argue physicians should, by virtue of our training, be better able to predict the future. But perhaps it is our experience as healers that nudges us toward optimism rather than realism. How else could we routinely embrace therapies with absolute risk reductions of 1% as if they were, well, necessary?

By email, Dr Diane Meier (Mount Sinai School of Medicine), a professor of geriatrics and palliative medicine, confirmed this view. She wrote, "Doctors are human and share at least as great a fear of death as the general public. We can be control freaks, and I'd argue our fear is greater, as well as our fantasy that we can beat it."

Citing Katie Roiphe's The Violet Hour, a forensic investigation of the deaths of six great writers, Meier's bottom line is this: "There is nothing intellectual about facing death."

Does the US Healthcare System Explain These Findings? Matlock and coauthors write that the "most troubling" explanation for their findings might be that the culture and the fee-for-service system are driving aggressive care in the last 6 months of life. To their credit, they call for future qualitative research into the reasons doctors get aggressive care at the end of life.

Think about what these results say about both our culture and system. If doctors can't stop the train or jump to the tracks, how could a regular person?

Culture is a nebulous term. Recently, I heard an ICU nurse describe weaning a dying patient from a ventilator by saying we were "withdrawing care." She meant no malice, but language is powerful. No. We weren't withdrawing care; we were changing the goals of care—from life prolonging to comfort care. In fact, by unshackling a human from futile therapy, we were caring more.

Another example of culture: a young doctor explains to a family that the strategy of using vasopressors, CPR, ventilators, and dialysis is called "full care." Again, there's no malice, but the words embedded in our language give comfort care lesser status. Is comfort care "half care?" Death denial is strong. It's embedded early.

The way our fee-for-service system exacerbates a death-denial culture needs little explanation. We have incentives and big payments for procedures and little to no coverage for basic supportive services in the home. The recent decision to pay for goals-of-care discussion is a small step in the right direction.


Death is hard. Death will always be hard. This study is important because it tells how much work we have yet to do to bring humanity and caring to the dying. Our broken healthcare system will be far easier to fix than our death-denial culture.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: