An Interview with Scott Weingarten, M.D.
MANAGED CARE July 1997. ©1997 Stezzi Communications
THOUGHTS FROM A GUIDELINES GURU
This noted internist and lifelong Angeleno has nothing against the art of medicine. He just wants to complement it with the most rigorous science available.
Scott Weingarten, M.D., M.P.H., is director of health services research at Cedars-Sinai Medical Center in Los Angeles, an associate professor of medicine at the University of California–Los Angeles School of Medicine, consulting medical director for PrimeHealth, a managed care network of physicians and health care organizations, and a widely published authority on evidence-based medicine. Recently, he agreed to share his thoughts on a variety of issues with Managed Care.
MANAGED CARE: Your expertise is in the area of practice guidelines and evidence-based medicine, so I'll want to ask you, in the context of managed care, about the effort to standardize medical care — to find what the best and most cost-effective practices are and apply them.
WEINGARTEN: That's fine. We at Cedars-Sinai have also spent a lot of time over the last couple of years applying the principles of evidence-based medicine to disease management programs, measuring whether they actually improve patient care.
MC: Okay, let's start with that. What's your definition of disease management?
WEINGARTEN: There is no gold standard for disease management, but the terms that come up in describing it are comprehensive care, care across the continuum — including both inpatient and outpatient settings. Usually the approach to care is somewhat systematic from a clinical standpoint, and attempts to improve patient outcomes and cost-effectiveness of care. It also involves some form of measurement to document that care has improved.
MC: What is Cedars-Sinai doing with disease management?
WEINGARTEN: We have initiated disease management programs for patients with hypertension and diabetes and for preventive care. And we've studied them rigorously to document whether or not they improve care.
MC: Do they usually turn out to improve care?
WEINGARTEN: If I had to guess, I'd say a large number of programs that are discussed across the country probably have little to no effect. However, with our hypertension disease management program we've been able to document statistically significant improvements in blood pressure control. And for our diabetes program, we've been able to demonstrate statistically significant reductions in glycosylated hemoglobin, which should translate, over the long term, into a reduction in complications from diabetes. And finally, we've been able to document improvements in preventive care, such as mammography rates.
MC: For our primary care physician readers, where is the most reliable place to turn for information on disease management techniques that can and should be put into practice?
WEINGARTEN: There's an awful lot of information flooding the literature at this point. It's hard to pick up a journal in the managed care or medical management field that does not include at least one or two articles on disease management. But many of those articles either talk about the theoretical benefits of disease management or are primarily marketing material for a variety of different organizations. I think a potential filter for primary care physicians who are attempting to discern whether a program is substance or fluff is whether there is scientifically credible information presented that demonstrates improvements in patient care. Doctors should hold these articles on disease management to the same standards they use when they review the clinical literature.
MC: For years, some physicians resisted practice guidelines, calling them "cookbook medicine." Do you find a greater acceptance of guidelines among physicians these days?
WEINGARTEN: I do. And this is not evidence-based, but anecdotal, based on conversations with physicians at Cedars-Sinai and at other organizations around the country. If you can show them, based on scientific evidence rather than someone's opinion, that adoption of guidelines will lead to improvements in patient care, it's much easier to get physician and other clinician buy-in than it was four or five years ago when these concepts were relatively new.
MC: I heard you give a talk in Washington a year and a half ago in which you referred to a study of physician acceptance of different sets of guidelines. One set from the American College of Physicians, you said, was rated twice as highly as another set for the same condition from Blue Cross/Blue Shield. You pointed out that that was remarkable because they were exactly the same guidelines. Doesn't the fact that such broad misunderstanding is possible sort of throw precision out the window when it comes to quantifying physician acceptance of guidelines?
WEINGARTEN: Well, to a certain extent the credibility of the organization sponsoring the guide-lines will always be important to physicians in deciding whether to accept them. But I agree with your implication that it's important for physicians to look beyond the credibility of the sponsoring organization to the process that was used to develop the guidelines and the underlying scientific evidence.
MC: One hears that it's important for guidelines to be developed or at least adapted locally — that the physicians in a given locale or institution ought to have some say about crafting them to meet their own needs. Yet medical conditions don't vary that much from place to place. Is this partly a salesmanship tactic to get physicians to "buy into" the guidelines?
WEINGARTEN: I think guidelines need to be locally processed. There are some genuine issues relating to customizing guidelines for the local culture and resources. For example, the resources that might be available in a large, urban, academic teaching hospital may differ from those available in a small, rural physician organization. An example might be guidelines for patients hospitalized with chest pain. If a cardiac catheterization lab and treadmill laboratory were open seven days a week at one hos- pital and only five days a week at another hospital, clinical pathways in the two institutions might differ.
MC: Speaking of cardiac matters, a recent issue of an American College of Cardiology newsletter referred to "widely used commercial guidelines that typically oversimplify clinical situations in an effort to cut costs." Do you find that to be a valid critique of many commercial guidelines?
WEINGARTEN: I don't know if it's more valid with commercial or noncommercial guidelines. But I would agree that clinical medicine is complicated, and that guidelines that do not account for the clinical nuances of medicine can potentially compromise patient care. A. Gray Ellrodt, M.D., and I, along with others, did a study [Annals of Internal Medicine 1995;122:277–292] in which we analyzed why physicians rejected a set of guidelines for patients with chest pain. One possibility studied was that the doctors hated guidelines, thought they were "cookbook medicine," and refused to adopt them. Another was that they were able to detect subtle clinical nuances in patients so that in the patients' best interests it was necessary to override the guidelines. We found that overwhelmingly it was the latter — subtle clinical issues had caused the physicians to deviate from the guidelines. What we learned from that is twofold: One, clinical medicine is complicated and it's extremely difficult to develop guidelines that anticipate all clinical situations and all patient preferences. Two, if one achieves 100-percent compliance with certain guidelines that are oversimplified in this way, one can compromise patient care.
MC: But you don't think commercial guidelines are any more suspect? Or, conversely, that guidelines promulgated by a specialty society might take care to protect that specialty's turf?
WEINGARTEN: I'm not sure, and that's why I think the evidence-based part is so critical. It's possible for either a commercial or a noncommercial organization to develop evidence-based guidelines, and also it's possible for just about anyone to develop "junk science" guidelines that are oversimplified and could compromise patient care. In terms of subspecialty societies, I think that if subspecialty consultation is going to be suggested as a part of a guideline, many people will want to see the evidence that such consultations contribute to quality and cost-effectiveness.
MC: What was your reaction to the decision by the Agency for Health Care Policy and Research to get out of the business of preparing its own guidelines?
WEINGARTEN: It did create issues for organizations that had already adopted the AHCPR guidelines, perhaps with the belief that updates would be forthcoming. How will these organizations keep the old guidelines current? The guidelines could go stale without being refreshed with new scientific information. However, I believe the underlying premise is a good one, and will provide the scientific foundation for organizations that are developing their own guidelines.
MC: Plans were recently announced by AHCPR, the American Association of Health Plans and the AMA to develop a national clearinghouse for clinical guidelines. Do you see that as a promising development?
WEINGARTEN: At this point I'm not familiar with the criteria they're going to use to certify guidelines as being of high or low quality. But there is a need for that function if it can be done credibly. A number of organizations are trying to decide which guidelines to use, and it becomes confusing when different guidelines on the same subject conflict. They might welcome an easier way to find out which guidelines are based on the most rigorous analysis of the scientific evidence.
MC: Let's talk for a moment about physician profiling. Just as with guidelines, there was some resistance against profiling for a number of years. Although there was an attempt to adjust for the severity of cases — not to punish one physician for having sicker patients than another — some doctors believed that adjustment wasn't sophisticated enough to make the profiling process entirely fair. Has the process gotten better, and is it more accepted now that it has been around for a few years?
WEINGARTEN: A lot of bright people have been working on the process. They've recognized that physician profiling will lose credibility if one attempts to oversimplify the practice of medicine. Another problem that was identified several years ago is that when you use small sample sizes — for example, to do patient satisfaction surveys — scores can vary widely over time even though the physicians aren't doing anything differently. The science of physician profiling has a long way to go, and it will continue to evolve over the next several decades.
MC: In a recent study, you tracked preventive-medicine interventions such as vaccinations and mammographies. In one group, physicians were profiled and given reports on how they had done; in the other, they were simply given educational presentations about the value of the preventive services. I believe the study showed that the profiling and the reports to physicians improved physician behavior. Is that correct?
WEINGARTEN: Yes. For a number of conditions, although not all, the addition of physician profiling and peer-comparison feedback and academic detailing — a pharmacist who sat down with physicians one on one — did improve care more than education alone.
MC: Help me understand how that happens, especially with regard to your comment a few minutes ago that — in one case, at least — the vast majority of deviations from guidelines were because of clinical subtleties rather than physician resistance.
WEINGARTEN: In this study, by contrast, the preventive-medicine guidelines were well accepted and not controversial. Mammography, smoking-cessation counseling and influenza immunization, for example.
MC: But do you see the troubling implication here that I'm struggling to understand about the physicians who changed their behavior? Here's a group of physicians, and they don't do the right thing from their own basic knowledge. And they don't do the right thing even when they're given special education about it. They only do it when someone comes by with profiling comparisons to whip them into line. What does that say to you?
WEINGARTEN: Well, I think it's a matter of increased physician awareness of the importance of these preventive care procedures. Let's say a patient comes in who's got severe chronic obstructive pulmonary disease, but presents with a sprained ankle. In the past, many physicians would have focused on treating the acute condition, and perhaps secondarily the COPD, and preventive care for patients with COPD. They may or may not have thought, "In addition, how do I optimize preventive care for this patient who is complaining of a sprained ankle?"
MC: So it was, in a way, part of the transition that managed care is bringing about.
MC: Medical care that looks to maintain health, not just to treat a particular condition in an episodic way.
MC: OK. Let's talk a little about managed care in general. It has certainly taken some lumps in the press in the last few months, with a lot of horror stories and a lot of "How To Deal With Your Penny-Pinching HMO" stories. Do you see trouble ahead for the whole managed care system, or is this just a phase that we're going through?
WEINGARTEN: I think managed care is really in its infancy. With each additional year of experience, hopefully, we'll figure out better ways of delivering care to our patients. In terms of evidence-based guidelines, I believe attempts to make clinical care more systematic will intensify, especially with all the literature that's coming out demonstrating gaps between optimal practice and the care that's provided. With or without managed care, I see efforts like what we're talking about — practice guidelines, clinical pathways and disease management programs — increasing.
MC: So they're not really part of the same package, in terms of philosophy?
WEINGARTEN: I don't see it. Practice guidelines, clinical pathways and disease management programs are attempts to make medical care more systematic. I could see those efforts taking place with or without managed care.
MC: But don't you need some of the systems of managed care in place just to have the data to know what you're doing?
WEINGARTEN: That's a good point. I think managed care and some of the changes that have accompanied it — such as attempts to improve information systems and being accountable for an entire population of patients rather than individual patients — have been helpful in our efforts to implement these programs and understand their effects on patient care. But one could argue that even prior to the advent of managed care, we should have been doing a number of these things.
MC: Whatever environment we have, would it be desirable to have a massive public education program so that people understand that we can't have every service for every patient every time?
WEINGARTEN: Well, I think one of the first things we need to do is figure out what practices really improve the health of the population and of individual patients. Based on the research we're reading these days, some of what we do does not contribute positively to the health of our population of patients.
MC: In addition to the broad-brush criticism of managed care that we've seen lately, there has been criticism of a somewhat more knowledgeable character — for instance, Wall Street Journal reporter George Anders' book Health Against Wealth, in which he talks about hospital contracting. The three predominant considerations in managed care organizations' contracting with hospitals, he says, have been "cost, cost and cost." Do you find this a valid critique?
WEINGARTEN: Yes. Over the last several years, competition has largely been based on cost rather than value. But today there's a silver lining. The public, employer groups and business coalitions are getting nervous about competition based on cost. They're afraid it may someday mean providing the fewest services possible. So there are attempts to develop valid, reliable quality measures. Over the last couple of years I've seen more progress in attempting to base competition on value.
MC: How will the job of primary care physician be different in 20 years from what it is now?
WEINGARTEN: In 20 years, I think primary care physicians will have access to much more sophisticated information technology. The electronic medical record will be a way of life. We will talk about the old days before the electronic medical record and wonder how we kept medical records. There will be real-time clinical decision support. There probably will be a greater use of physician extenders to perform many primary care functions. There will be much greater access to outcomes information about patients and, if I had to guess, I'd say the Internet or some evolution thereof will be used to gather information from patients directly on how they're feeling, what symptoms they're having and how satisfied they are with their medical care.
MC: May I ask what your own health coverage is?
WEINGARTEN: I'm covered by an HMO.
MC: Are you pretty much satisfied with it?
MC: You're an internist. Can you identify a moment in your career when it became clear that you would have a strong interest in the science underlying medical choices and in evidence-based medicine?
WEINGARTEN: No. I can tell you this: It was not well planned out. I became interested in all of the changes that were happening in health care, and at the same time very concerned that some of the directions in which health care could go might not be very good for society and for the patients I was caring for. And rather than complain about these changes, I wanted to try to help shape some of them. So my colleagues and I began to try to understand the way health care was practiced — and, as part of seeking that understanding, began measuring the effects of certain systematic changes in patient care on patients' health. And that's where our interest in outcomes measurement came about.
MC: Does your experience imply a recommendation for other physicians?
WEINGARTEN: I believe very strongly that rather than sitting back and letting health care change without their involvement, and complaining about a number of the changes, physicians and other clinicians should get involved, understand the care that is being provided, propose better solutions and become as active and involved as time permits. It's up to us as a profession to help shape the way care is provided in the next century, rather than let others shape it for us.
MC: Thank you, Dr. Weingarten.