The Laws of Medicine

The Laws of Medicine

Field Notes From An Uncertain Science

Book - 2015
Average Rating:
Rate this:
One of the world's premiere cancer researchers reveals an urgent philosophy on the little-known principles that govern medicine--and how understanding these principles can empower everyone.
Publisher: New York : TED Books, Simon & Schuster, 2015
Edition: 1st ed
ISBN: 9781476784847
Characteristics: 70 p. : ill. ; 19 cm


From the critics

Community Activity


Add a Comment
Nov 12, 2019

A short book, surprisingly informative with very interesting examples to explain the information. Siddhartha Mukherjee is an excellent explainer.
Well worth reading.

Nov 09, 2018

This book at 70 pages is a quick and insightful read. The author shares his Laws of Medicine. There are three of them:

1) The first one is the importance of knowing the prevalence rate of whatever disease you are testing for before conducting the test.
2) The second one is to study and pay attention to unusual outcomes.
3) The third one is not to be blinded by bias and dogma no matter how logical they seem.

The first law is about Bayesian statistics. It entails knowing as well as possible the “prior” probability or prevalence rate of a disease within the population you are testing before conducting a diagnostic test. And, the author uses the following example. A disease has a prevalence rate of 1%, and the test you are using has a false positive rate of 9% (or a sensitivity of 91%). The accuracy of the test sounds very good. However, if a patient gets a positive result, there is only a 1%/(1% + 9%) = 10% probability that he actually has the disease. That’s how the author explains it. However, Bayesian statistics is not quite that simple. And, his short cut is inaccurate. For, it to work you have to assume the test false positive rate [or 1 – sensitivity] is equal to its false negative rate [or 1 – specificity]. Yet, most tests have widely diverging sensitivity and specificity. Nevertheless, this omission does not detract from his main point that the prevalence rate greatly influences the overall accuracy of the diagnostic test outcome. On this count the author criticizes the indiscriminate use of the PSA test for prostate cancer that has too high a false positive rate associated with a disease that has a low prevalence rate and is often benign that would not affect the patient if untreated.

The second law of studying outliers is interesting. He mentions numerous counts whereby scientists in different fields and doctors in medicine uncovered true gems and advanced the body of science by studying the unusual outcomes that diverged from normal ranges. However, he does not mention the numerous counts where studying outliers may have been a waste of time as they can often be just a source of Noise.

The third law is to avoid bias and dogma. He mentions the invention of radical mastectomy to eradicate completely breast cancer by removing much extra tissue beyond the breast. This technique was developed by an eminent and forceful surgeon. It made good sense. Everyone believed this surgery was much more successful than normal mastectomy until three large clinical trials confirmed otherwise. In the meantime tens of thousands of women suffered greatly from this overly aggressive surgery for nothing. Now, the author states that “the procedure is rarely, if ever, performed today.”

The author’s concluding phrase is interesting: “the medical revolution will not be algorithmized.” He means that no advance in Artificial Intelligence will substitute for the critical judgment of healthcare scientists and doctors to ensure that we know “prior” probabilities based on human expertise and avoid bias that can be hidden within studies that algorithms rely upon.


Add Age Suitability

There are no ages for this title yet.


Add a Summary

There are no summaries for this title yet.


Add Notices

There are no notices for this title yet.


Add a Quote

There are no quotes for this title yet.

Explore Further

Browse by Call Number


Subject Headings


Find it at CALS

To Top