Subscribe Twitter Twitter

Tuesday, February 18, 2014


Imagine if your physician said to you one day, upon diagnosing you had a common cold, that what you need is to bleed. Perhaps a liter of blood would do the trick, he says, or maybe more. You'd be repulsed, wouldn't you? You'd never return to this quack again.

But had this been the recent past, you might have been more receptive. Bloodletting, after all, persisted right up until the early 1900s. It was common treatment for almost all medical ailments. Barbers bled you -- hence those red-and-white poles that have come to signify barber shops. They opened your veins using sharp instruments called lancets. If you were lucky, they used leeches. The practice was so widespread, the French imported 40 million leeches a year in the 1830s for medicinal purposes.

It is well known today that bloodletting is ineffective and in many cases harmful to patients, but it took around 2000 years to discover it. Reading about the old practice today is cringe-inducing. Physicians brought suffering, when they thought they were doing good.

I have often been asked by old friends, since arriving from abroad, what it is exactly that I research as a graduate student, and I often like to start my explanation with bloodletting. Around 1835, Pierre-Charles-Alexandre Louis decided to evaluate the merits of bloodletting. He looked at 77 patients with a case of pneumonia and divided them into two groups: those who were bled 1-4 days after illness and those who were bled late, 5-9 days after. It was important that both groups were as comparable as possible, and were of the same age. He applied statistical analysis to compare outcomes between both groups. And his results were, as he remarked then, "startling and apparently absurd." It seems that bloodletting did not decrease, but increased mortality. 44% of those who bled early died, while only 25% for those who bled late did. Below is his wonderful figure reproduced, which I've taken from the internet:

Today, drugs in the US can no longer be legally sold without having been subjected to clinical trials that establish them to be safe and effective. Evidence plays a critical role.

In the field of development and public policy though, the careful evaluation of programs and policies is now only becoming more commonplace (but mostly with NGOs and not governments). It is an exciting development. Although I doubt that there are many "bloodletting" programs out there, it is not obvious which ones are ineffective and wasteful of resources. The job is to discover these and understand how to make them better. It is an endeavor I find great importance in, and try to contribute to, as a graduate student.


Post a Comment