Friday, June 20, 2008

Review: How Doctors Think

How Doctors Think (kindle edition) is a book about misdiagnosis. It starts off introducing a woman who was misdiagnosed as being anorexic, but turned out to have a case of celiac disease. Between the misdiagnosis and the proper diagnosis, she suffered for a period of 10 years.

The book covers the three major forms of cognitive errors that lead to misdiagnosis: Anchoring (fixating on a diagnosis that was already provided), Availability (looking at the first thing that comes to mind, rather than searching for more possible diagnosis that might better fit the evidence), and Attribution (having confirmed a problem, attributing all problems to it, rather than considering the possibility of having more than one disease afflicting the patient).

The book presents case study after case study, each covering a specialty in medicine, from Radiology to Surgery and Oncology. Each case study is poignant, interesting in and of itself, and educational. One is reminded time after time that doctors are only human, and that it takes an extra-ordinary one to be capable of diagosing an unusual problem, and that the patient has to take an active role in his care --- Groopman frequently discusses instances where the patient's statement triggers a physician to look deeper or consider other possibilities.

The book also criticizes modern American health-care, where primary care physicians (who do most of the difficult diagnostic work) get paid less than specialist. A whole chapter is also devoted to conflicts of interests between what's good for the doctor financially, and what's good for the patient, including a heart-breaking story about a patient who ignores her oncologist's advice.

Groopman comes across as humble, and all too aware of the limitations of modern medicine. What's interesting to me is that in computer science, the algorithmic/Bayesian approach to diagnostics is the state of the art, while this book is essentially a critique of the approach, and how time spent thinking deeper about the problem when rare cases arise must necessarily supersede the algorithmic approach.

All in all, this is a great book, and highly recommended. If you're stuck on a bug, this is a great book to have by your side to read --- some of the techniques described are great questions to ask as well. I paid full Kindle price for it and have no regrets. If you or a loved one has a difficult medical condition, you owe it to yourselves to read this book and to give a copy to the patient.

I'll close with this passage that illustrates that doctors and software engineers also suffer from the same problem. The top 10% outperforms the median by a huge factor, just like in software engineering:
Potchen’s study using the sixty films was to compare the top twenty radiologists, who had a diagnostic accuracy of nearly 95 percent, with the bottom twenty, who had a diagnostic accuracy of 75 percent. Most worrisome was the level of confidence each group had in its analysis. The radiologists who performed poorly were not only inaccurate; they were also very confident that they were right when they were in fact wrong. “Observers’ lack of ability to discriminate normal from abnormal films does not necessarily diminish their confidence,” Potchen wrote.
Post a Comment