An Interview With Atul Gawande

Atul Gawande's New Yorker article comparing the medical systems of El Paso and McAllen, Tex., has been a definitional piece in the health reform conversation. President Obama has repeatedly invoked it. Senators have talked about it. The media have embraced it. I spoke to Gawande this afternoon about the fallout from his article, the problem of revenue-driven medicine, and whether a public plan would make a difference. He called from Jordan, where he's helping the World Health Organization implement better surgical protocols. A lightly edited transcript of our discussion follows.

You've gotten some pushback on your article about McAllen, Texas. Today, in fact, some doctors from the area held a press conference rebutting your claims, and you published a blog post re-rebutting theirs. What have you found to be the most convincing counterarguments against your piece?

The three lines of criticism were ones I anticipated or even talked about. The idea that these people in McAllen are unhealthier. The idea that it's all malpractice. The one point I didn't get into was the snowbirds [retirees from colder areas who summer in Texas], but they're not in the spending calculations anyway because Medicare counts them in their home area.

The criticisms I'd been hearing and seeing but that hadn't been going away was pointing out that McAllen is the poorest county in the country. They'd say you couldn't compare it to Mayo. But I didn't. El Paso, which I did compare it to, was the sixth poorest in the United States. They're very closely similar in poverty, in immigration, in physician supply, in rates of disease, and so forth.

That brings up another point. Putting aside the difference between El Paso and McAllen, both are going to have higher costs because their people are poorer. This gets to the whole question of social determinants of health: the fact that you're sicker if you have less income, or education. Do we give that enough attention?

I think the really interesting thing is that even beyond the baseline amount of poorer health associated with lower incomes, there's this whole body of powerful literature showing that levels of inequality are even more highly correlated with poor health. So a place like Texas has poverty, yes, but also huge inequity of income. States with similar poverty but less inequity of income have much better health profiles. I've actually had this on my list to write about. I don't completely understand what it is about inequality that drives that. For instance, there's something protective about rural areas, where there's less inequality, so if you take rural areas and urban areas with similar levels of poverty, the rural areas will be healthier. And people say these areas are more socially cohesive and that's what does it. But how does that make cancer rates lower?

One thing that really struck me about your piece was that you focused much more on the question of care providers than insurance providers. The political conversation tends to do the opposite. Want to talk a bit about that distinction?
I had a hard time connecting the dots. My vantage point on the world is the operating room where I see my patients. And trying to think about whether a public option would change anything didn't connect. I order something like $20,000 or $30,000 of health care in a day. Would a public or private option change that?

People say that the most expensive piece of medical equipment is the doctor's pen. It's not that we make all the money. It's that we order all the money. We're hoping that Medicare versus Aetna will be more effective at making me do my operations differently? I don't get that. Neither one has been very effective thus far.

Do you think much that we're hearing in the political conversation is responsive to the issues you pointed out?

Part of the difficulty is that it's very hard. But you can learn from good hospitals. They do peer review, for instance, and that changes what doctors do in their offices. They blunt the financial incentives in various ways that we haven't studied at all. It's kind of ridiculous that there haven't been very many people putting feet on the ground and studying what the positive deviants are doing. There are hundreds of examples out there. They're not just the Mayo Clinic and not just Grand Junction. Go to Portland, Oregon; Temple, Texas; Pensacola, Florida. These are places that are doing something differently.

But getting there requires a change in local medical cultures and rebuilding local medical systems. All medicine is local just as all politics is local. But let's create a cadre of researchers who go into these communities and figure out what's going on and spread the word.

The Washington debate -- there are smart reasons to think about including a public option in the mix, but we have not been thinking hard enough about how we control costs and make a better system. I think it's achievable in about 10 to15 years, and maybe even faster. I can tell you three things that will transform McAllen overnight. But CBO doesn't score them.

So what are the three?

First, they spend more than $3,500 per Medicare beneficiary on home visits. El Paso is around $800. McAllen is spending more than half what many communities spend on their entire health care expenditures. The doctors there have to disinvest from these home health agencies and come to agreement on when those visits are worth using.

At the end of life, McAllen spends $22 per person on hospice but more than $3,000 on ambulance rides. In a place like Portland it will be more than $400 on hospice and around $500 on ambulances. Increasing use of hospice, offering that as an option and working as a community on how to manage end of life, would be a smart move.

Work on basic cardiac prevention like getting people statin drugs. Most studies have shown you'll lower the cardiovascular disease rate by 25 percent and lower the number of procedures ordered. This was done in by Kaiser of Northern California, and they became the first community I've ever heard of where heart disease stopped being the leading cause of death.

If you took those three things and worked on them for a year, you could go from $15,000 per person per year to less than $10,000.

You brought up medical cultures. And those are hard to change. It's hard to legislate norms. But how do they develop it? How does one city end up with a culture with totally different incentives than another?

It's not just about incentives. The interesting thing to me is not that McAllen is different from elsewhere. it's that El Paso is different than McAllen. They have the incentives to go in that direction! My hypothesis is that communities have local anchor institutions that foster values and norms that make the medical system successful. My sense is that in McAllen it was about a few institutions striking out in different ways that set the norm for what others did.

in the early 90s, McAllen was the same cost as El Paso. Three years later, they jumped into the top 10 or 20 and never really left. The first thing to really leap in price was home health care and it happened, it seems, because a few home health agencies came on the scene and began offering doctors something serious for their involvement: extra salaries as medical directors who don't really do much. McAllen is also on the leading edge of for-profit innovations. They were early with a specialty hearts center, for instance, and cardiovascular operation rates began climbing. Then you had physician-owned imaging centers and physician-owned surgery centers and everything began going up a lot.

What I've observed -- and we really need some real data behind in it -- but in those places with organized systems of care, with lower cost and higher quality, they've had to find some way to blunt the incentives of quantity over quality, or, more bluntly, revenues over patients' needs. Some of them have done it by moving doctors to salaries, like Mayo or Kaiser. But others, like Grand Junction, don't salary. They have fee-for-service. But they have medical groups that make sure that people who see more Medicaid or uninsured patients aren't penalized, and they remove a physician's privilege to practice at the hospital if they don't participate in a strong peer review system. And there are, to be sure, salary hospitals where you see very poor quality of care and people leave at three. But if you are in a fee for service system you have to have a structure around it that helps physicians resist the incentives of revenue-driven medicine.

One thing you sound like you're saying but that isn't often said clearly is that profits matter. Incentives matter. Doctors are not entirely altruistic and solely concerned with patient care. And even if the influence of these incentives is only subtle...

It may not be conscious, but it's not subtle! It's things like whether you decide to take phone calls from patients and instead bring them to your office because you're paid for office visits but not phone calls. When you have a referral whether you worry about losing the referral business if you don't do what the referrer expected you to do. Most of us really do work hard to avoid that conflict. The vast majority of doctors really do try to take the money out of their minds. But to provide the best possible care requires using resources in a way that keeps you viable but improves the quality of care.

An example is electronic medical records system. They've been proven to reduce errors and enhance communication and improve care, but they don't pay. You lose revenue getting that system into place. In another system of incentives, your incentive is to figure out how to get that system going with your colleagues and make that possible. I think I said somewhere towards the end of the piece that to reduce waste and improve quality means organizing ourselves in ways that would sacrifice revenues. And that makes it hard.

You were around in 1994 working for Rep. Jim Cooper, and then in the White House. Do you see any important similarities between now and then? Any important differences?

The big question is what has really changed since 1992 that means anything for reform? Two things that matter on the political side and then a couple on the medical side. The big things on the political side is that we've stretched the economy and our budget much thinner with the damage from health care costs. The sense that both people in business and citizens have is that we're really in danger here. That's not something people felt as palpably in 1993. I think that sense of danger has concentrated attention.

On the medical side, there is a much greater sense of dissatisfaction in our work lives. We're working incredibly hard to make the system work for our patients. But it doesn't work very well for them or us. Some of that is the march of science. The increase in coordination and complexity makes it impossible to do this stuff out of your office. We're not in a system that's well equipped to make this a satisfying way to spend your time.

But at the same time, we are in this moment when we're reinventing medicine. Because of my experience with the World Health Organization, I'm in a lot of countries. And they're experiencing much the same thing. The underlying force here is not the politics but the science. We have 13,000 diagnoses for people. 13,000 ways the human body can fail. We can treat a lot of it. But it's not like penicillin. It's complex stuff. Pathology labs and dozens of medications and treatments. Our health system was built for a 20th century science. I can keep giving you numbers if you want.

I love numbers.

Okay. There are also more than 6,000 drugs and 4,000 types of operations and procedures. That's what a hospital has to be able to manage and a doctor's office has to manage. Our system in my hometown struggles with how to manage that. We're trying to reform a system to be prepared for the 21st century of medicine and we'll be struggling for answers for awhile.

That sounds like more than a single human can handle. Does that mean we're going towards more electronic medicine. Will it be your doctor feeding your symptoms into Google?

I think the extreme complexity of medicine has become more than an individual clinician can handle. But not more than teams of clinicians can handle. And part of what's such a marvel about a place like a Mayo or places like it is that they've been able to get teams of doctors work with nurses and nutritionists to work together. I'm sitting here in Jordan showing nurses and surgeons and anesthesiologists from Yemen and Pakistan how to take a 90-second check before an operation to make sure they have the antibiotics and the blood and they've all agreed on what the operation is. We don't know how to do that culturally in surgery. We're doing dress rehearsals on how to talk to each other. It's hilarious. But when we do it, we not only lower costs, we lower the death rate 40 percent and the complication rate 30 percent. And that's why I think the answers will be there.

Photo credit: Harvard School of Public Health.