How Doctors Think

by Jerome Groopman, M.D.

Cover image

Publisher: Houghton Mifflin
Copyright: 2007
ISBN: 0-618-61003-0
Format: Hardcover
Pages: 291

Buy at Powell's Books

This non-fiction exploration of medical reasoning in part follows its title and provides a tour of how doctors diagnosis and treat patients, but its particular focus is on errors. Rather than a general discussion, Groopman's goal is to make doctors more aware of common flaws in reasoning and judgement and make patients more aware of how doctors work so that they can watch out for those problems and try to correct for them. It is interesting, and occasionally even useful, within that scope, but as interesting are the looks inside medicine via frequent anecdotes and the opportunities for generalization to other problems of reasoning.

How Doctors Think starts with an anecdote about a misdiagnosed patient who had been pigeonholed (incorrectly) by her doctors, and Groopman continues to use anecdotes and case studies extensively throughout. He balances them well between successes and failures and makes them entertaining enough to be worth reading just for the stories. He also fills the book with details about the day-to-day pressures and concerns of a practicing doctor, which combined with the anecdotes gives the book another strong appeal. As well as a book about thinking, it's also a profile of a profession: a bit skewed due to its focus on diagnosis but still providing me with several useful details about what doctors face (particularly around patient load).

Groopman talks about a lot of judgement and decision errors, but most of them can be seen as some variation on making up one's mind prematurely. This is an excellent study of attribution error and confirmation bias. Many of the flaws in thinking remind me of essays on security. The problem of primary care physicians and, even more strongly, radiologists in finding the unusual among a blizzard of the normal have significant similarities to the difficulties of airline baggage screening and similar security situations.

Most of the book focuses on errors without a clear fault, errors from emotion or too-quick discarding of outlying possibilities. Groopman also spends some time on more nefarious problems, such as influence by the pharmaceutical industry and some of the pressures of the for-profit aspect of medicine, but I thought this was a weaker section of the book. The anecdotes are just as interesting, and one can cheer for the doctors who stand up to the slimy drug company representatives, but unlike the other sections Groopman offers little new analysis of or conclusions about the problem. Saying that people should have more information, think skeptically about claims, and beware of conflicts of interest is a motherhood-and-apple-pie analysis that is entirely true but offers little in the way of new solutions. I think we've all gotten the message about how much money has corrupted choices about drugs and treatment of aging, but I suppose if you've not already been saturated by this message, this is yet another warning.

Better is his treatment of medical insurance and its influence over the diagnosis and thinking process. This is scattered throughout the book and a concentrated chapter on the topic would have been helpful, but the influence is more subtle and has to be shown from more angles. Groopman spends quite a bit of time on decision-tree models of diagnosis and their deceptive concreteness, which appeal to cost-control and to insurance checkboxes but which narrow one's focus away from unusual but possible cases. One of the more memorable concepts in the book is "zebra retreat," the application of "when you hear hoofbeats, think horses, not zebras" to the extreme of discarding rarer diagnoses that may still be correct.

The overall thesis of this book, insofar as it has one beyond an anecdotal exploration of good and bad decisions, is that too much emphasis on statistics and decision trees is misplaced and attention to one's emotional bias and (most importantly) time spent with the patient to take a complete history and pick up on more subtle cues is the best way of avoiding errors of thinking. Groopman is warning against the limits of our ability to use cut-and-dried science to find solutions to often-fuzzy and mysterious problems of human illness. Unfortunately, he's a little short on concrete ideas for how to improve medicine or how to teach that type of diagnosis, particularly in the face of ever-growing pressure to treat more patients more efficiently.

This is worth reading if you've wondered how doctors get off-track and how the better ones take a step back and find their errors, and even more worth reading if you're interested in a sampling of the situations doctors deal with. It's not a formal or rigorous treatment, but it's more readable because of that and Groopman tells good stories. Recommended for the entertainment value, with some useful insights into both medicine and human reasoning to glean along the way.

Rating: 7 out of 10

Reviewed: 2008-04-08

Last modified and spun 2016-05-30