author : Atul Gawande

After reading this, my respect for doctors has soared, and I hope I never have to be a patient.

Gawande is a surgeon, and he tells stories of what it's really like: stories of miraculous successes, and heart-rending failures, some due to the skill of the doctors combined with good luck, some due to bad luck. He poses the paradoxical point: we all want to be treated by the best doctors; doctors have to be trained; they get their training by treating people. So, if we want to be treated by well-trained doctors, we have to be treated by trainee doctors too.

His stories of actual patient cases, and how success or failure can rest on the smallest of details, are hair-raising. His stories of how massive improvements could be made with systematic attention to procedure and process (example range from hernias to anaesthesia) are enraging, because these relatively simple ideas could have massive returns if they were applied uniformly, but they aren't. But overall, his stories about people coping with their illnesses are amazing.

A real eye-opener. Even before writing this review, I went off and bought his next book.

Gawande's first book was about why and how things can go wrong in medicine, essentially because it is incredibly complicated and uncertain. His second book brings some hope: here he concentrates on how things can be made better. Essentially the theme is one of process, process, process: a "science of performance". When it is implemented, the results can be spectacular. Gawande gives details as diverse as casualty rates in the Iraq war, polio vaccination in India, the prevalence of Caesarean deliveries, and the treatment of Cystic Fibrosis. However, this barely scratches the surface of what could be done.

He is not arguing against scientific research, just for the extremely cost effective process improvement. But why is this so rarely done? (It's not just in medicine. Take software development, for example. There are (some) companies developing business critical software who don't use even use source code control systems.) Gawande gives an example of why this might be. Take hand washing. After all, the germ theory of disease is well-known, the horrifying death rates from puerperal fever being slashed by hand washing in the mid 1800s is common knowledge, and yet medical staff still don't wash their hands between examining patients. Why on earth not? Well, things are not as simple as they seem. I was surprised to read:

Okay, so "washing hands" in a hospital is a little more involved than I had thought. But inconvenience is not the only problem. There is inertia, too:

So what is to be done? Bringing in changes to procedure requires a culture change: as Gawande points out, no-one could get away without "scrubbing up" for an operation. But culture change is hard. Gawande investigates some examples where things have been successfully changed, and discovers the entirely unsurprising answer: it has been a bottom-up rather than a top-down process. When the front-line practitioners are consulted and asked for suggestions, are involved, things can improve dramatically; when control is imposed top-down, resistance is the common response.

There are other essays in here too, covering things as diverse as the ethics of medical staff attending executions, how much a doctor should be paid, and malpractice suits. In all cases the arguments are thought-provoking, and show how the situation is actually much more complicated and nuanced than the usual sound-bite stories.