Tuesday, June 26, 2012

Causes of death and perception skewing

My first job out of college was working in one of the busiest Emergency Departments in the country.  I learned a lot of interesting things about human behavior there, and some random facts about the way the ED interacts with the government as far as reporting goes.  

One of the smaller parts of my job was making sure the proper reports got filed at the appropriate times, and this included death certificates.  Contrary to what you might think, not many people actually die in the Emergency Department.  Trauma victims almost always have enough time to get to the operating room before they die, and people with more chronic illnesses tend to die in the intensive care units.  Thus, when death certificates come up, most residents have no idea how to fill them out.  I don't remember much about them, but I will always remember one thing: heart failure is NOT a valid cause of death in Massachusetts.  You can put unknown, or heart disease or many many other things, but you can't put heart failure.  The reason?  Everyone dies of heart failure.  If your heart is still beating, you're not getting a death certificate.  

I'm thinking of all this because of a very cool new interactive graph put out by the New England Journal of Medicine about causes of death over the years.  I can only post the static graph, but I suggest you check out the interactive one:
Another list here, comparing 1900 and 2010 directly:
It's interesting to see causes that have dropped due to actual dips (tuberculosis) and those that are not there any more due to medical reclassification (senility).

It's a good study in how medical reporting can change over time for various reasons, and why changes should always viewed from both a broad view as well as up close.

1 comment:

  1. Cancer, Heart Disease, Stroke, Accidents, Cirrhosis. That's about it. And those are going down, but slowly.

    And most of those (plus a few others) are alcohol related. FYI, two of the research-oriented psychiatrists where I work are using two alcohol measures as the most imporatant in terms of health: units per week, and most drinks in one night in the last month.