Saturday 14 July 2012

Lessons from history #2: Evidence based blood letting

Objections to evidence based medicine (a scientific approach to medicine) by doctors often sound like this:

“Practicing physicians are unwilling to hold their decisions in abeyance till their therapies received numerical approbation. Further, they are not prepared to discard therapies validated by both tradition and their own experience on account of somebody else's numbers.”

Things have not changed.
These quotes are from the early 1800’s and express the attitude of physicians at that time to studies such as the landmark paper in the American Journal of Medical Sciences from 1836. The article (by PCA Louis) systematically pooled data from many patients and used statistics to look at the effectiveness of blood letting for pneumonia (then standard practice). It found the treatment to be ineffective.*

So did the study change practice?
Well, the demand for leeches took a hammering, but because the doctors saw some people get better after their treatment, they went on believing their eyes and the practice continued for some time. How long? Well you might be surprised to know that blood letting for pneumonia was still included in William Osler’s 1923 textbook, Principles and Practice of Medicine, and in a 1920 textbook on obstetrics by JW Williams from Johns Hopkins, draining 1 litre of blood was suggested as a treatment option for pregnant women with eclampsia.

We need to start discarding therapies that are only validated by tradition and experience, on account of the numbers (for numbers, read 'science').

Lessons from history #1 is here.

The story of the PCA Louis article is told here.

* Remember, when a treatment is ineffective, it does not mean that people do not get better with the treatment (in this case most patients survived), it is just that they are no more likely to get better than if they do not have the treatment.


  1. This was also the topic of Professor Leung's (consultant cardiologist at Liverpool Hospital) presentation at grand rounds this week. Very interesting article! Out of interest, how was blood letting eventually 'discarded'? Was it a randomised clinical trial?

    1. One theory is that blood letting was replaced with more fashionable placebos, like magnetism.

  2. Bloodletting is actually not something that's completely discredited as a medical practice anymore. It's often called "blood donation", which has a little nicer ring to it, but has gained some validity in the recent future.

    Either way, it looks like the same thing is happening with blood transfusion. People are finding that the risks of adding blood can sometimes outweigh the expected benefits. It's amazing that doctors would not consider the data more heavily. Good article!

    1. Thanks. Blood donation is different: that is ostensibly for he benefit of others, not the donor. However, you are correct that blood letting is stil performed. It is used for haemochromatosis and at our institution we have our own leech farm, for use with revascularisation procedures. Not sure of the evidence though ...

  3. Can you help me understand more about cortisone ijection as im about to have one in a month. What will be the side effect positiv/negative.

    1. Sorry, but individual advice is not really possible without knowing the full picture. My other blogs cover the evidence for cortisone injections, particularly in the spine.


Note: only a member of this blog may post a comment.