Asymmetrical Information - Megan McArdle


Can the Cleveland Clinic Save American Health Care?

Salaried doctors. Great care. The administration used them as a model and a goal for Obamacare. But will it work outside of Cleveland?

"The best doctors are not motivated by money."  It's a common sentiment (this particular example comes from the comments section of Felix Salmon's blog, but you see it anywhere that doctor salaries are discussed.)  

We all seem to believe it.  And yet, we don't act on it.  When we get sick, we don't head straight for the nearest county charitable hospital; we head for the biggest, shiniest facility we can find, one where the doctor's parking lot is full of luxury automobiles.  The implication is that we do think good doctors are attracted to money, same as the rest of us.

Perhaps as a result of this belief, American doctor salaries are some of the highest in the world.  And there's quite a lot of dark suspicion that the desire to get paid drives a lot of bad health care: that doctors do procedures because they are lucrative, rather than because they are best for the patient.  Nor is the suspicion limited to doctors.  Steven Brill's 22,000 word article in TIME points the finger at hospital administrators and device manufacturers, among others.  The gist of the expert consensus is that as long as we're willing to pay for whatever they decide to do, health care is going to cost an exorbitant amount of money.  

But what if we actually did take money out of the equation?  What if we put doctors on salary and substituted relentless cost-control for "anything goes"?  

Well, say critics, probably you'd get what everyone complained about with HMOs: a situation where providers have every incentive to undertreat instead of overtreat.  We didn't like that, either.  

And yet, there are places where it works.  One of them is the Cleveland Clinic, one of the nation's leading hospital centers.   Along with other landmark institutions like Mayo, it was a model for the Accountable Care Organizations provisions in Obamacare--provisions which, it was hoped, would help simultaneously control costs and improve care.  However, so far, all four have declined to participate in the ACO program.  On Friday, I spent two hours talking to Delos Cosgrove, the Cleveland Clinic's CEO, about how they improve outcomes and control costs.  

People are, as Cosgrove noted, "looking to us to be the model of health care delivery going forward".  The Cleveland Clinic is not just a clinic any more, but an integrated network, with, says Cosgrove,  "9 hospitals, 18 family medical centers, 11 CVS pharmacies connected by electronic medical records."  They're also consistently ranked among the top hospitals in the country by US News and World Report.  When technocrats dream of an idealized health care system in an idealized America, this is the vision that dances before their wondering eyes.  

From a planner's perspective, hoever, here is the real central fact about Cleveland Clinic: its doctors are salaried. "I think you have to recognize that people do what you pay them to do," Cosgrove told me. "If you pay doctors to do more of something, then that’s what they’ll do. If you put the emphasis on looking after patients, they'll do that. We said we’re going to take the incentive for the doctors out to do more."

Not that going to the Cleveland Clinic involves taking a vow of poverty, of course. "Our salary model is based on the fact that pediatricians don’t get paid as well as neurosurgeons. So we take the academic salary range for hospitals like our across the country and we fit right into that salary range. As far as salary is concerned, we’re right there with everyone else. How we differ is we make sure that doctors do doctor work. For example, I never sent a bill. I never had to worry about my supplies arriving. I never had to go out and find a nurse—the organization does all that for you."

One of the things that Cleveland "does for" the doctors is control costs--a subject upon which Cosgrove is something of a fanatic. Over the course of the interview, he returned to the topic time and time again: the need to reform doctor's pay so that they don't have incentives to overtreat, and the need for hospitals and other providers to be constantly thinking about how to control costs.

"We’re trying to get towards the maximum value for your health care dollar. Continuously improve the quality of care: get rid of the infections, the complications, prevent readmissions. At the same time we’re trying to improve the quality and reduce the cost."

"That is a national responsibility," he said, with a pause for emphasis. "We’re not going to be fiscally sound unless we can control the cost of health care."

As Cosgrove notes, infection control is one of the areas where there seems to be "low hanging fruit" in the health care system. Hospital acquired infections kill about 50,000 people every year, according to Ramanan Laxminarayan of Resources for the Future. The number who suffer but do not die from infections is multiple times that. Those infections are expensive, and at least somewhat preventable.

Cleveland Clinic achieved a dramatic reduction in its rate of infections in central lines, which are used to give direct access to the veins in seriously ill patients. They got infections down using the sort of techniques laid out in Atul Gawande's Checklist Manifesto. "Standardized approach," says Cosgrove. "Everyone in a cap and gown and this is how you put it in." And because Cleveland is an integrated research hospital system with an intense focus on quality improvement, they have applied this approach to complications throughout the patient experience, not just with doctors.

So for example, everyone leaves the hospital with an appointment already made for a follow-up. And four or five days after they've been discharged, every Cleveland Clinic patient gets a call. How are they doing? Are they taking their pills? If they have a problem, the Clinic tries to arrange outpatient treatment to head off an expensive readmission. Every employee of the Cleveland Clinic--from doctors to parking lot attendants--is referred to internally as a "caregiver" and every service is viewed as an important part of the continuum of care.

It sounds wonderful. It is, in fact, wonderful. But as a policy writer, I have to ask: is it replicable?

Last spring, I interviewed Charles Bosk, a sociologist who specializes in studying the way that doctors and medical systems handle error. "Atul Gawande answers the wrong question," he told me. "It’s not whether checklists are effective. Anybody who has made it through third grade, and/or made it through the supermarket with or without a list, knows that checklists work. What you need is to ask is, 'What would motivate professionals to use checklists?' " Checklists invert the normal doctor/nurse hierarchy, giving nurses the authority to, say, step in and stop the procedure of inserting a central line. That's tough for many doctors to swallow. And hard for nurses to do, unless the culture ensures that they genuinely shouldn't fear later retaliation.

"Even in [Michigan], the most successful project we know of," Bosk noted, "the infection rate was only cut in half and some units didn’t improve at all." Presumably this was not because checklists don't work, but because only some of the units followed procedures to the letter.

Bosk analogized it to handwashing. We know that handwashing saves lives in hospitals. We also know that medical personnel--particularly doctors--don't do it nearly as often as they should. If you interview doctors or nurses, they'll tell you that they wash their hands every single time, or very close to it. They may even believe that they wash their hands every single time. But studies that measure actual handwashing--with spy cameras, or sniffers to test for the residue of hand sanitizer--show shocking rates of noncompliance. The problem isn't developing the right list, or telling doctors that they should wash their hands, or showing them how to wash their hands. It's getting them to do it.

So how does Cleveland get doctors to cooperate with these sorts of initiatives? In part, says Cosgrove, because they involve the employees in the decisions; when they wanted to cut down their cost for replacement hip joints, he got all the surgeons in a room, told them that having too many different joints in circulation was costing the clinic big money, and got them to come to a consensus on which models they should use.

"Our physicians are so engaged in our supply chain that they help negotiate the price down for the things we use," Cosgrove told me, and reeled off a list of examples:

  • When I was the head of surgery, we needed a new heart-lung machine, and we decided there were three models that could work, so we did a reverse auction to get the lowest price.
  • We put price tags on things in the operating room: before you open that $250 set of new sutures, make sure you actually need it.
  • We found out that there’s a lot of redundant tests that are done, or tests that won’t be vital to the patient’s care. We know that there are some things that don’t change. For example, the reticulocyte count can’t change but week to week. So if someone’s ordered a reticulocyte count, you can’t ordered another for a week.
  • Or there are these genomic tests, they cost a fortune, so we said, let’s hire a geneticist, and she can tell us whether that’s really a good test to do. She paid for herself almost immediately.

This is, of course, exactly how we want a health system to function. But how to make it do so? Listening to Cosgrove talk, I also heard a lot about something deeper, and harder to replicate, than meetings or price tags: the culture of the Cleveland Clinic.

"When you talk to doctors at Cleveland Clinic they feel like it’s their Cleveland Clinic—it’s a group practice, not a hospital employing doctors. We’re all part of the team. It’s not a matter of fiefdoms."

Of course, every CEO tells you that. But in the case of the Cleveland Clinic I (mostly) believe it. I've now been visiting companies for ten years as a journalist, and working for companies for a lot longer. And while every company has its petty annoyances and its internal disputes, you can tell when an organization has a great culture. You walk in, or you talk to people, and you can just feel it wafting around you, warm and comforting. The Cleveland Clinic has that feel, even over the phone. Cosgrove acknowledges as much:

"I went to the Cleveland Clinic from Boston for an interview in 1975. Within a half hour I felt that there was a different culture there. I’ve looked at a lot of other places for jobs along the way, and it never had the same feel." He says turnover is just 3.5% a year.

Great institutional cultures can accomplish great things. But in some ways, that's a problem for the rest of us. It's natural to want to emulate the achievements of Cleveland Clinic in our policies. But you can't make a culture out of rules. Culture is an organic outgrowth of an organization's history, it's people, its successes and failures. It cannot be ordered from the top, or nurtured by simply altering the financial incentives. Cosgrove speaks of maintaining the institution's culture in much the way that he talks of maintaining their electronic health records system: a constant process of checking in, re-evaluating, and upgrading.

We can salary doctors and order them to care as much as the Cleveland Clinic doctors and administrators do about continual improvement. We can mandate checklists and appointment follow ups and staffing levels. But we can't make them be the Cleveland Clinic.

When I asked Cosgrove if other hospitals could really emulate the Cleveland Clinic, he said "yes, other people can do it. One of the things that is beginning to drive this is the patient satisfaction scores that is now becoming part of the pay for hospitals."

But he also said "both the incentives and the culture matter. They’re inexorably tied. We’ve gone through a very major cultural renaissance if you will to begin to tie everything together." And he clearly recognizes that on the culture front, the Cleveland Clinic is something special.

Organizations that have that something special end up with a recruiting advantage: they can attract top-flight talent who really meshes with the team, and the institutional philosophy. And they can keep those people. But that can't be a nationwide strategy. It works, at least in part, precisely because these institutions are rare and special.

In the 1970s, we looked at another special health care institution that seemed ripe for replication: Kaiser Permanente. Kaiser (and a bunch of smaller organizations like it) was a completely integrated system: salaried doctors, salaried staff, a continuum of care organized by primary care physicians. It was great. It was also cheap. So we started replicating it.

Only it turned out that when you replicated an HMO without the core cultural committments of a place like Kaiser, it didn't work; Americans hated HMOs with a ferocity that they'd previously applied only to Hitler and the DMV.  In theory, HMOs had incentives to keep you well without overtreatment; in practice critics charged that they had were cherry-picking patients, and using various bureaucratic hassles to keep down the cost of care. And the cost savings didn't really materialize, at least in part because of the legal, political, and market backlashes against the gatekeeper model of patient care.

One telling fact suggests the limits of government reform's ability to induce transformational change: none of the model health systems for ACOs have actually agreed to participate in the program. Cosgrove demurred when I asked him about this, saying only that they'd rather not be early pioneers. But it's a fairly clear signal that whatever Obamacare's ACOs will do, it will not be what the Cleveland Clinic does. It may be closer, and better, than what we have now. But it will not be exactly the same.

And it may be another iteration of the HMO problem: something that works great, and yet, frustratingly, cannot be scaled up (at least not well) much beyond the organizations that are already doing it. There's a real possibility that Cleveland has found the answer to what ails our health system--but that it only works in Cleveland.