Evidence Based Medicine?
Modern medicine’s recent major mantra has been “evidence based.” No more listening to anecdote and poorly performed studies. Now we look at “hard evidence” and properly designed trials to determine how we treat human beings.
Death is a hard end point, one presumes.
While many drug studies are now done to change “risk factors,” the outcry has been to study “real endpoints.”
Let’s include death in that category.
Which is why a recent article by Dr. Lawrence Altman in the New York Times should wake some of us up.
To the false reporting part of death.
Not the ultimate outcome, itself, no. The people in the morgues did not rise up to protest the false attributions of their demise.
Yet death records are used for many purposes, including the major issue of figuring out what causes death. And the studies Altman cites makes clear that death is not the end – just another, incorrect data point.
Things To Write About In NY When You’re Dead
The studies were performed in NY, in hospitals of the sort I used to work in. Have things changed much?
Not as much as they should.
-Over half of physicians reported knowingly reporting the wrong cause of death.
Why? Drop down menus off electronic records don’t allow correct answers. Which can be a real problem when you’re writing notes and ordering treatments (but that’s another story.)
-No one told the residents in training how to mark or perform death records.
Here, tradition continues. One example I remember from internship – in California – was of another intern asked to pronounce someone dead. He promptly walked into the patient's room, declaimed “I pronounce you dead,” and left.
What happened after the residents were trained in creating proper death records?
All the numbers changed.
Heat disease deaths fell by 54% - more effective than even the most efficient CCU.
Influenza and pneumonia deaths tripled – up to 11%. Cancer increased from 11% to 16%.
With training, the causes of death change after death.
But what’s truly disturbing about all these “causes” and presumed “causes”? They’re not accurate, no matter how good future training.
Because the real cause of death is best understood through autopsies.
And they are more and more rarely done.
Why? Cost. Insurers won’t pay. Effort. Trouble with families.
And giant institutional reluctance. Hospitals and doctors often don’t want to know what killed a lot of people.
Because autopsies will show what they missed. And they miss stuff. A lot of stuff.
More than we can know with present day imaging technology.
The Evidence of History
Overall, medicine should not feel so bad. Evidence based studies are not performing well in other societal arenas.
Consider the strange case of Giovanni Patalucci – the Italian Schindler.
Patalucci was police chief in Fiume (Rijeka), a part of the Dalmatian coast which is now Croatian. He saved the lives of 5000 Jews before dying, another victim of the Nazis, at Dachau.
Patalucci was a martyr of the Catholic Church. A film was made of his life. Piazzas throughout Italy are named for him and his heroic resistance. His name bears witness at Yad Vashem in Jerusalem as one of the just. Michael Bloomberg honored him in New York.
All of this was true. Until Patricia Cohen broke the story that:
Patalucci was not police chief but important in enforcing race laws in Fiume.
A higher percentage of Jews from Fiume died than in almost any part of Italy – more than 80%. Many were probably sped on their way by Patalucci.
He was an active fascist and Nazi collaborator, except the Nazis suspected him of embezzled and treason before they killed him.
In other words the evidence – and the history, the story, the heroism – were all wrong.
Fabrications. Aided by a Vatican interested in an Italian hero, a family that wanted a pension, a bishop cousin, the search to find Italian “heroes” who had not played ball with the Germans.
Evidence is only as good as the methods used to obtain it. Much of present “knowledge” is the result of shoddy record keeping, inadvertent survival of written reports, ideological selection, and outright lying.
Reader, beware. All those “pooled analyses” used to define what treatments work and don’t work may have been infected by a host of factors – including unconscious biases – that always afflict clinical studies.
We can learn from history. That what they told us was often wrong.
Even if the records – and the evidence – said it was right.