Wednesday, January 20, 2016
If you are in health care, chance are you have come across the essay "How Doctors Die" by Dr. Ken Murray.* If you have not, then you should go read it, because the anecdote likely will resonate if you are an advocate for hospice and palliative care. Essentially, the key point is physicians see so much suffering from unhelpful medical treatments near the end of life for patients, that when a physician encounters a life-limiting illness, they will often forego minimally beneficial treatments.
It makes sense right? As a palliative care doctor, I could imagine myself following a similar path, but what does the research show? Are doctors really in lockstep on this issue? Reflecting on this issue and what I have seen in my peers, everyone seems to agree with this essay when they send it to me, but is that just a confirmation bias? Thankfully more information is emerging to help test this anecdote with data.
This week in JAMA, two research letters examine the locations of death and the intensity of end-of-life care for physicians compared to the general public. In the "Association of occupation as a physician with likelihood of dying in a hospital," Blecker et al, used the National Longitudinal Mortality Study (based on US Census records) to examine nearly 500,000 decedents from 1979 to 2011. The outcome was location of death for four divided groups (physicians, other health professionals, other higher education, all others).
For the total population (over these 30 years), about 40% of people died in a hospital and 72% in any facility (including hospitals.)** Knowing that baseline, and 'knowing' that doctors die differently, what percentages would you guess for physicans? 20% and 50%? 30% and 60%? 38% and 63%? If you guessed the last one you would be right. Compared to all others doctors are less likely to die in a hospital than the 'all others' category (38.3% vs 40.4%, Adj Odds Ratio 1.10). For death in any facility doctors will also be less likely than 'all others' (63.3% vs 72.4%, Adj Odds Ratio 1.34).
Now those numbers are statistically significant, but they do not seem to explain the anecdote quite well. One could make up a number of arguments that could support data like this, which would not include doctors choose less aggressive care at the end of life.
The second letter by Joel Weissman et al, looked at end-of-life resource utilization of Medicare beneficiaries in four states, and compared physicians, lawyers, and the general population. Why physicians and lawyers? Well it isn't about a good joke, it is because they were thought to be similar in education and socioeconomic status. This study is a little bit more telling than the first. The outcomes measured from Medicare claims data was utilization in the last six months of life: surgery, hospice, ICU, hospital death, and expenditures. In 3 out 5 measures (hospital death, surgery, ICU) the physician group leaned towards less health care utilization, but again not by much. The lawyer group was pretty similar to the general population. Interestingly, hospice utilization was not different among the three groups and centered around 45%.***
So after seeing all of this, can we conclude that doctors die differently than everyone else? The statistics are significant, but not really overwhelming. It really isn't a strong enough difference that I would feel compelled to keep the "How Doctors Die" essay as a hallmark of strong differences. Until there is stronger research tying any of this data to the decisions that doctors make, the best I can put my conclusion is: doctors die slightly differently than everyone else, statistically speaking.
*Wouldn't similar rationale hold for nurses to also make different choices at the end-of-life? One study looks at it, but lumps nurses in with many other health professions with less exposure at end-of-life.
**Those numbers have shifted over time as the 2013 Teno Medicare study showed 25% for Medicare decedents in 2009 died in a hospital.
*** Still so many people who die without the support of hospice, it still surprises me.
Christian Sinclair, MD, FAAHPM is the editor of Pallimed, president-elect of the American Academy of Hospice and Palliative Medicine, and a palliative care doctor at the University of Kansas Medical Center.
Image credit: Jared Rodriguez for Truthout.org via Creative Commons