Atul Gawande

gawande_atul.jpg Background Briefing’s latest effort is a lecture to the Commonwealth Club in San Francisco by the journalist, academic and practicing surgeon Atul Gawande. He’s the author of Complications which is a popular book on the science and art of surgery – well worth a squiz if you get the chance to have a look at it. If you don’t you could do much worse than read or download his lecture. Its both fascinating and moving and a good summary of some of the highlights of Complications.

I certainly sat up when I heard him asked how an ordinary person finds a good doctor?

Atul Gawande: We don’t know. I struggle with this all the time. People ask me who should I see? I just got a couple of emails today, can you tell me a good geriatrician in New York City? And I don’t even know how you go about really trying to answer that, except the old-fashioned way: you call around, you talk to a few doctors and you ask them ‘Who’s good? Who’s got a great reputation?’ But of course you would have asked around in Cincinnati and they would have said, ‘Boy, Cincinnati Children’s Hospital is the place you want to be.’

As a surgeon, I do a lot of cancer surgery and I would love to know who really are the best in the kinds of operations I do, who had the lowest complication rates, the highest survival rates? And if I knew that, I would go and watch them and I’d learn from them. Instead what I know are the people who have great reputations, and what I see are people who sometimes seem to really deserve their reputation, I learn a lot from them. But then others who you really wonder where their reputation came from. In some areas, you can go on the internet, cystic fibrosis now for example, posted all of the results of all of their centres in November, in a major experiment, to see whether openness actually is better for the profession. And I’m convinced it’s an experiment worth watching. For about three or four years they have first shared the results among all of the institutions, taking the names off the labels, and letting them see who’s good and who’s not, and then they’d have meetings to share ideas together, and you saw the Bell Curve begin to move. It started rising higher, results started to get better. And the bottom in particular started to lift up, so there was even some narrowing of the curve. But there’s still always going to be the curve. There are a few places that have a greater capacity for learning and changing more quickly and absorbing things. But that openness and that transparency, we’re afraid of it, but it’s the direction I think we have to go.

After I read his book I wrote to Gawande and told him of the Gruen Tender. He was very positive and intrigued by it and we exchanged some emails on who might be interested in taking it further. But we didn’t get far. In any event, I’m not sure how useful Gruen Tenders would be for individual doctors. But they could really come into their own for clinical units in hospitals for the right kinds of procedures where the chances of adverse events are not too small to require huge samples to detect a clear statistical signal and where the event can be contained within a reasonably short time horizon ensuring likely causation. Most of the procedures Gawande mentioned were like this. (Like hernia operations in which the best unit in the US has something like a twentieth of the failure rate as average units!). The program continued:

Woman: Continuing along that same line, how do you find a place with the best outcomes? Because a good record may indicate superior treatment, but it may also indicate acceptance only of the least challenging cases.

Atul Gawande: Yes. One of the great criticisms that come in is whether the data is garbage. When you look at any mix of people, rating how doctors do, you know it’s a mix of sometimes shoddy information, a reputation that somehow is either not founded, or it’s founded well, but then also we’ve seen in New York State for example, that cardiac surgeons are rated there, their information is made public, and there is, even though you can’t seem to find it in the data, people report that sometimes a very sick patient has a hard time finding a doctor who’s willing to take care of them, because if they die, it would be on their statistics.

I want to contrast that a little bit with obstetrics. You know, I think it’s a bad sign when we in medicine start saying we don’t want to measure and let people know how we’re doing, because we’re going to play with the numbers. In obstetrics, people have been reporting maternal death rates, new-born death rates, and APGAR scores, which rate how healthy a child is on a scale of 1 to 10, for four decades, five decades now. And you didn’t see people saying, Oh, I can’t take that unhealthy, complicated, pregnant patient into our program because it’s going to screw up our statistics. We have made it a commitment for a long part of our history that we’re going to make it possible to deliver children anywhere, in any hospital that is taking on obstetrics safely, with excellent results. We have a less than one in 15,000 death rates for mothers today, where just a couple of generations ago, the most common cause of death for women of child-bearing age, was childbirth. You could have people playing with the APGAR scores, faking the numbers a little bit, telling you your child is a ‘Oh your baby’s a 10’ instead of a 5, or something like that. But they recognise that here was a powerful tool, by rating how a child is doing and looking carefully at the real condition, they could realise the right thing to do for that child. If they had a poor score, that meant that they needed to be incubated, needed to be sent to the neo-natal intensive care unit, or you’re in danger of losing them. And vice versa, if they had an excellent score, let the mother hold that baby and begin their life together.

Well this is the opportunity we have in other walks than medicine. To have the information, to know what our outcomes are like, and we know that sometimes they’re complicated. A poor APGAR score can reflect the performance of the team in a child delivery, or a baby that was sick to begin with. But we’ve accepted the responsibility that sick or not, we have to do what we can to save everybody.

You know, we have earnings reports for businesses, and some places like Enron, faked it. But the success of how businesses work depends on having that information be out there, or we wouldn’t have any ability to know who’s doing well or who’s not. And that’s kind of like we are where we are now in medicine. We just don’t know even how I’m doing.

Well Atul, it doesn’t have to be that way. We’ll always have to rely on the goodwill of doctors to some extent, but not for information at least about procedures that have the characteristics I mentioned above. To find the best obstetric unit, you don’t have to worry about the quality of the patients they get. You only have to compare like with like. Conduct a Gruen Tenders to allocate obstetric deliveries and you get a hoard of information from which you can pick which units have the best chances of a trouble free delivery. And you don’t have to rely on the units’ public spiritedness to prevent them from turning away bad risks. They never have an incentive to do that, only to accurately reflect what they think their real chances are when they put their tender bids in on any job.

This entry was posted in Economics and public policy. Bookmark the permalink.
Notify of

Inline Feedbacks
View all comments