November 2006

Stephen Granade


The End of Medicine: How Silicon Valley (and Naked Mice) Will Reboot Your Doctor by Andy Kessler

The medical industry is moribund, with rising costs threatening to bring the entire edifice down. Even worse, medical insurance often neglects to pay for yearly physicals and drives doctors to focus on chronic care instead of early detection. What if there were medical technologies that could work the way computers do, becoming more and more powerful and less and less expensive? Could we cure heart attacks, stroke, and cancer? Could we get rid of doctors? Those are the questions Andy Kessler sets out to answer in The End of Medicine: How Silicon Valley (and Naked Mice) Will Reboot Your Doctor. It's a pity that he never really answers those questions, or even presents his quest to answer them in a coherent manner.

Kessler has a background in technology and finance. Following his time at Bell Labs, he was a stock analyst and a hedge fund manager. In this book, he's interested in applying what he learned about the electronics industry to medicine. Specifically, he wants to find the medical industry's answer to Moore's "law," the observation that the number of transistors on an integrated circuit -- and thus, roughly speaking, computer power -- doubles every eighteen months. It's the drum-beat of the semiconductor industry's march, driving costs down and performance up. This scaling of performance does more, though. It opens up new applications and entire new industries. Ten years ago, who would have imagined Orbitz taking the place of so many travel agents?

According to Andy Kessler, the same thing could happen in medicine. More broadly, Kessler wants to get doctors out of the way. They don't scale the way technology scales, you see. Doctors require a lot of specialized training over many years, and cannot be mass-produced. What Kessler's after is a medical industry where the patients have more control and where costs drop every year instead of rising. His panacea is medical scanning, a technology that Kessler hopes will scale like computers and become cheaper and more powerful with each passing year. He's looking to CT scans, computer-assisted gene scanning, computerized analysis of X-rays, and more -- both to detect diseases early and to reduce the need for doctors by letting patients walk into a scanning clinic and get their results themselves.

That's a lovely sentiment, but there are three problems with his approach. One, Kessler claims that the right technology will get rid of doctors, just like it did bank tellers, stock traders, librarians, magazine layout experts, and film editors. The problem is that many of those examples are still going strong. Technology hasn't made them obsolete. It's refocused them, true, but there is still a place for human judgment and intuition in libraries, in magazine layout, and in film editing. Kessler's magic bullet of scanning can reduce or eliminate the number of doctors who pore over X-rays or CT scans, but it won't get rid of our need for expert guidance. Nor will scanning find everything wrong with us, not at this juncture. There is still a need for doctors to diagnose, not to mention to operate when necessary.

Two, Kessler's desire for a democratized medical system ignores the downsides to such an approach. Option paralysis is one such downside. Given enough choices, people often become overwhelmed and are unable to make a rational choice. "Lots of people without engineering degrees figured out how to use PCs, add sound cards, configure DMAs and IRQs, add CD-ROM drives, install software that took days to complete," Kessler says, but lots of people can't do what he's just described! Even if you can, do you really want medicine to be like working with computers? The cost of bad decisions is another downside. If I mis-install computer equipment I'm frustrated for a while, but nothing really bad happens to me. What if I make poor choices about my medical care? Bad decisions can be detrimental over the long haul in ways that aren't immediately obvious. Finally, a democratized medical system can make expert advice harder to come by. Systems that have a lot of choice tend towards a bimodal distribution, with poorer people forced to shoulder the burden of decision-making themselves without access to good support, while richer people buy advice from experts.

Three, doctors aren't even sure scanning will be helpful. Routine whole-body MRIs can turn up scary things like brain tumors that, even if untreated, would never have affected the patient in their lifetime. The closest thing the medical industry has today to routine scanning is a mammogram, and doctors are still debating their effectiveness.

If Kessler has answers to these objections, his surface treatment of the subject and lack of organization hide them. The book is a long series of meandering anecdotes presented in chronological order. Kessler spent a lot of his search flailing about and trying to figure out where to go in his search for computer-like increases in scale. We get to follow him down every abortive side trip. He begins the book by trying to fire his doctor and perform his own cholesterol tests, genetic tests, and more. Then he begins investigating medical scanning, only to wander back to performing his own medical tests. It takes him until about page 70 to really get going, and once he does, he alternates between under-explained technical details and more anecdotes. The whole book is an object lesson in the old science saw that the plural of anecdote is not data.

Moreover, the book's entire flow is choppy. The average chapter is five pages long, with several two- and three-page chapters. Kessler makes no attempt to impose an overarching narrative flow. It's chaotic and dizzying, and in the end is like listening to a Chihuahua who's found the sugar stash.

That's my objection to the writing's large-scale organization. On the small scale, the writing is still sub-par. The text is peppered with sentence fragments, comma splices, and tense shifts. On several occasions I had to read and re-read sentences to understand what was being said. The writing style is apparently meant to evoke the rhythms of Kessler's speech, as if he is sitting beside us and telling us the stories. Unfortunately, the result sacrifices readability for little gain.

That said, the book has a definite and strong personality. It's possible you will enjoy spending time with Kessler, listening to him tell stories. At one point he says, "I'd have made the perfect pacing dad in the waiting room, but the damn emasculation movement of the '80s meant I had to watch all four of my sons' births -- even had to cut the umbilical cords. I haven't looked at my wife the same way since." If that passage or jokes about only drinking on days that end in "y" amuse you, you've got a good chance of enjoying the book's tone.

The End of Medicine claims that advance detection is the next big thing in medicine, and that it'll take better scanning technology to get there. I can certainly believe that. I only wish the book had presented that claim coherently and with evidence to back it up. If you enjoy Kessler's written persona, the book will entertain you. If, however, you're after an in-depth look at medical scanning or insightful analysis of the medical profession, look elsewhere.

The End of Medicine: How Silicon Valley (and Naked Mice) Will Reboot Your Doctor by Andy Kessler
ISBN: 006113029X
368 Pages