The Laws of Medicine:
Field Notes from an Uncertain Science (TED Books), Siddhartha Mukherjee, 2015
“Is medicine a science?”
““Doctors,” Voltaire wrote, “are men who prescribe medicines of which they know little, to cure diseases of which they know less, in human beings of whom they know nothing.””
I read this book because of a convincing TED talk by Mukherjee. He’s a cancer researcher, and he instantly won my respect – especially after Being Mortal and the related TED talk. Here is the starting point of what to review before any class or research into statistics, Bayesian theory, or more general scientific research.
The author says something that fits perfectly with the recent quantum biology book I read: “There are fewer laws in chemistry. Biology is the most lawless of the three basic sciences: there are few rules to begin with, and even fewer rules that are universal.” Indeed, biology is due for a revolution (or two) this century…
1. “A strong intuition is much more powerful than a weak test.”
The first is Mukherjee’s most math-heavy section, but it isn’t overwhelming in the least. It simply shows us that by a few simple tweaks in how we “screen” for who to select for certain medical tests before taking them, we can significantly improve the usefulness of these tests.
2. “Normals” teach us rules; “outliers” teach us laws.
Here is his most scientific section, arguing that we need to understand the outliers in order to truly create – and falsify or prove – good theories.
He also has a good section on the single patient anecdote – and its power when harnessed correctly!
“[W]e still lack a deeper, more unified understanding of physiology and pathology.”
3. “For every perfect medical experiment, there is a perfect human bias.”
Bias, bias, bias. Here Mukherjee begins with the story of the radical mastectomy, and by doing so, he discusses something that’s been below the surface of my thinking about medicine and studies for years. But it’s something I could never really express properly: that the simple act of having a study on health changes the very nature of the outcome. Indeed, most of them in the US and Europe are based on whites and – worse – males. But perhaps equally profoundly,
“[W]hen patients are enrolled in a study, they are inevitably affected by that enrollment. A man’s decision to enroll in a study to measure the effect of exercise on diabetic management, say, is an active decision.
More succinctly:
“The device used to measure the subject transforms the nature of the subject.”
This is one of the few books that it would probably be better to simply review by reading/skimming the entire book than simply this file – especially for the instruction in Bayesian thinking and general probability.