Health care doesn’t lack for big ideas, even if their definitional boundaries do sometimes get fuzzy. Take “population-based medicine” and “patient-centered health care,” for example. Both are phrases we hear and read every day, and maybe even believe in. But do they coincide or collide?
“There’s no clash between them — that’s the good news,” says David B. Nash, MD, MBA, dean of the Jefferson School of Population Health at Philadelphia’s Thomas Jefferson University. “People get thrown off by the jargon. The population approach simply enables us to reduce waste and reallocate resources, which we must do because there’s no new pot of gold coming into health care.”
“The devil’s in the details,” counters Richard G. Stefanacci, DO, MBA, chief medical officer of the Access Group and associate professor of health policy at the University of the Sciences just across town. “When you apply today’s population-based medicine at the granular level, it sometimes results in inappropriate care from a patient-centered point of view.”
A fundamental disagreement — or maybe a Philly feud? No, turns out it’s more like what Carl Sandburg had in mind when he wrote the poem “Elephants Are Different to Different People.” The fact is, one’s view of the compatibility of these concepts depends on what one is used to seeing when one gets down to cases.
Population-based medicine is one of managed care’s fundamental notions. The American Medical Association recently defined it as an approach “that allows one to assess the health status and health needs of a target population, implement and evaluate interventions that are designed to improve the health of that population, and efficiently and effectively provide care for members of that population in a way that is consistent with the community’s cultural, policy, and health resource values.”
Most people in health care have long since cast off the fear that this approach implies a cookie-cutter insensitivity to patients’ individual needs. In fact, as the consultant Peter Boland wrote back in the mid-1990s, it “does not detract from individuality but rather adds another dimension, as individuals benefit from the guidelines developed for the populations to which they belong.”
As for patient-centered care, definitions differ — and that’s part of the trouble. But it has been described by recent Centers for Medicare & Medicaid Services (CMS) Administrator Donald Berwick as including “transparency, individualization, recognition, respect, dignity, and choice” in the patient’s health care experience. It is also widely understood to draw on the concept of the patient-centered medical home (PCMH), in which a practice is organized to provide truly coordinated, proactive and therefore cost-effective care.
“We’ve been able to develop tools and processes to identify people who are at risk,” says Stephen Rosenthal of Montefiore Medical Center, about using predictive modeling to manage an urban population.
“The physician is the leader of a team of professionals,” says Stephen Rosenthal, vice president for network management at Montefiore Medical Center in the Bronx, N.Y., and president of the affiliated Care Management Co., describing how the PCMH model works. “They manage the complexity of the medical issues for a population of patients so the physician doesn’t need to have face-to-face contact with each patient on each occasion to satisfy their health needs.”
“They are two sides to the same coin,” says Boland of our two ideals.
Paul Kasuba, MD
“The two concepts do align well,” agrees Paul Kasuba, MD, chief medical officer of Tufts Health Plan in Massachusetts. “Both point toward giving the individual the right kind of care in the right place at the right time. And we need to keep both of them in mind by aligning financial incentives for our providers and our members around similar outcomes in health and health measurement.”
So what’s Stefanacci’s gripe?
For one thing, he is a geriatrician, and older patients tend to be medically complicated. “If you’re doing population-based medicine through a large health plan,” he complains, “your guidelines might tell you, for example, that all nonvalvular fibrillation patients need anticoagulants, and you might ‘ding’ all those practices that don’t have 80 percent of such patients on anticoagulants. But for very frail elderly patients, an anticoagulant may in fact be contraindicated — they could fall and develop a subdural hematoma that would actually cost the patient and the system much more.”
Stefanacci offers another example. From a population-based perspective, a diabetic population should have glycated hemoglobin A1c levels below seven, he says, and doctors may be penalized if a sufficient percentage of their patients aren’t kept to that level. “But when you get down into the weeds and start looking at 85-year-olds in a nursing home, that number may be inappropriate.”
The doctor insists he is not making the old “my patients are sicker” argument with which physicians once resisted practice guidelines and outcomes research. But what has been called the art of medicine, he says, “is really taking all the multiple variables in a patient and making adjustments. Population-based medicine should have enough nuance to apply patient-centered care, but the models that are out there don’t allow that level of variability.”
“There’s an element of truth to that,” Rosenthal concedes. “As an industry, health care isn’t quite there yet. We’ve done a better job with certain segments of the population — diseases such as cancer, for example — than with others.”
Rosenthal doesn’t share Stefanacci’s worry about HbA1c targets that may be inappropriate for the elderly because, as he says, “it’s ultimately at the discretion of the physician.” But he moves closer to the Stefanacci stance when he talks about predictive modeling.
He says his organization has thrived while taking some degree of risk for its challenging urban population for 15 years. “We’ve been able to develop tools and processes to identify people who are at risk and be proactive in managing their care,” he says. “But I’d like to have even better predictive tools, based on data I’m receiving from claim activity, laboratory activity, or other sources, so I can be proactive to the right people.”
For Kasuba of Tufts Health Plan, it’s partly a problem of perceptions. Interventions suggested for someone by population-based data, he notes, “may or may not be perceived by that individual to be appropriate for themselves.” In circumstances of shared decision-making, he says, patient-centered care requires consideration of the patients’ own values.
There are “weeds” aplenty in New York’s Hudson Valley, but Susan Stuard peers into them without seeing any fundamental clash between population-based medicine and patient-centered care. “The medical home concept asks you to do both things simultaneously, and it’s eminently possible,” she says.
Patient-centered medical home models operate on the assumption that population-based medicine and patient-centered care need to be done simultaneously, and that is “eminently possible,” says Susan Stuard, of the Taconic Health Information Network and Community.
Stuard is executive director of THINC — the Taconic Health Information Network and Community, one of the anchor partners in a regional program called the Hudson Valley Initiative.
This group’s declared purpose is to “leverage health information technology, physician practice transformation, and value-based purchasing in pursuit of care delivery that is patient-centered, coordinated, accessible, high-quality, and efficiently delivered through sustainable financial models.”
What makes that mouthful more than a platitude is the region’s unusually advanced level of data-sharing. “In our patient-centered medical home project, we’ve been working with six health plans to create a multipayer claims database to look at costs, utilization, and outcomes,” Stuard says.
Those data are put to work. To hear Stuard tell it — and she’s no abstract theorist — it is precisely population-based data that make the PCMH possible. “Patient-centered care is about having time in the exam room to discuss with patients their choices for health care,” she argues. “How is she going to lose those 25 pounds she needs to lose? Will a man with prostate cancer choose watchful waiting or aggressive treatment when both are reasonable options? And does a woman have questions about her mammogram?” But checking whether she’s had the mammogram shouldn’t occupy the doctor’s time, says Stuard, because in the PCMH model such basics “were already flagged on the way in” — and they are on the checklist, thanks to population-based medicine.
Stuard isn’t saying it’s a cinch. “There’s the stuff you always point to, like, ‘Are my diabetics getting their eye and foot exams every year?’” she notes. “But if you want to get more sophisticated, you really can challenge the capabilities of IT systems — certainly within a stand-alone physician practice, and even within an integrated delivery network. Not everything you want to do for population health is necessarily easy to do.” And that goes double for a health plan, which may have little control over care itself — and may not enjoy the same data-sharing that prevails on the Hudson.
“Health plans’ problem is that in many cases they are disconnected from the delivery of services,” says Rosenthal. “They have a relationship with the member and the employer, but only limited relationships with the doctor and the community.”
Whenever health care ideas seem fuzzy, one can do worse than to pick up the phone and call the San Francisco office of Lucy Johns, an independent consultant doing business as Health Care Planning and Policy. Years ago she developed a course for medical directors called “Population Medicine in Managed Care” that counseled physicians to “merge a population perspective with clinical medicine to create a practice that keeps people out of the health care system when they don’t need to be there.”
“There is no theoretical conflict between the two ideals,” she insists. “The issue is how to incorporate a population health perspective into clinical practice, which we are now referring to as patient-centered medical care.”
She explains with a symbolic visual image: a small circle inside a larger one. The large circle, “population health,” represents the population that CMS and others now would like providers to be accountable for. The smaller circle includes people who are receiving care or who have ongoing health problems.
“The border between the two circles is the prevention of problems and the promotion of health, which evidence proves will keep that population out of your office,” says Johns. “Prevention and promotion take resources, but far fewer than hands-on services. If you’re on a fixed budget of any kind, you’ve got to address population health.”
Flawed measurement tools don’t mean population health and patient-centered care don’t mesh, she argues. And the need to make the two ideals converge isn’t going away. “It’s going to have to be figured out,” she says. “We are never going to have perfect measurements. But we are always going to have limited resources.” Johns agrees with Rosenthal about health plans’ position. “Talking about population health in the context of health-plan strategy is complicated,” she says, “because health plans only know about people in terms of eligibility and payments. Most of them don’t know anything about health status — though there’s one big exception.”
That exception, of course, is Kaiser Permanente. “They have done a remarkable job, investing heavily in analytics in the past 10 years,” says Boland. “They basically have the whole data sheet right there when the doctor reaches for it — the patient’s latest symptoms, what he or she was in for last time, and the various things that Kaiser as an organization can do for that person.
You can’t do patient-centered care if you don’t have access to evidence about population health, says Sharon Levine, MD, of the Permanente Medical Group of Northern California.
Sharon Levine, MD, a pediatrician who for more than two decades has been associate executive director of the nation’s largest medical group, the Permanente Medical Group of Northern California, says that “Not only do I not think there’s a conflict, I don’t think you can do patient-centered care if you don’t have access to an evidence base about population health.
“Our ability today to measure and monitor the effect of illnesses and interventions gives us information about people who are like the patient,” says Levine. “From that we begin to understand the needs, preferences, and clinical concerns of the patient in front of us. That’s not the end of the story — it’s the beginning of the story.”
The challenge, says Levine, is “our accountability for the stewardship of our patients’ health as well as the most appropriate use of their resources — because the truth is, in health care we are spending other people’s money.” When the country as a whole abandons the “inherently inflationary fee-for-service system,” she says, the practice of medicine nationally will change. “That’s my belief based on practicing in a prospectively paid medical group for 35 years now.”
So — a consensus. By calling population-based medicine itself to account for the lapses in its application to patient-centered care, Stefanacci is looking at the elephant’s underside. He is the contrarian whose apparent partial dissent can be safely folded into the general upbeat view that, though we may not be there yet, we all know where we’re going. Except that — just in time to make things interesting — F. Randy Vogenberg, PhD, RPh, agrees with him.
“Patient-centered care — what the hell is it?” says Vogenberg, a pharmaceutical consultant to employers, hospitals, drug companies, and insurers and a principal in the Institute for Integrated Healthcare. “Ask a thousand people and you get a thousand different answers.
“The devil is in the details,” insists Vogenberg. “I do a lot of health services research, as Stefanacci does. People can say whatever they want. What I look at and what Richard looks at is, what do they actually do?” And in down-to-cases reality, he says, population-based medicine and patient-centered care are indeed not in perfect sync — and may never be.
He warns, for example, of one big devilish detail that is on the rise: the category of “personalized” specialty drugs and biologics, such as the prostate cancer drug Provenge, which require individual formulations because they work differently based on different individuals’ genetic makeup.
Today most of these drugs are costly injectables, but they will increasingly appear in oral formulations — as some multiple sclerosis drugs already do — and will pose tough ethical and logistical questions about allocation of resources. The availability of these biologics, says Vogenberg, threatens to throw a monkey wrench into the whole health insurance business and health reform efforts.
“What we have today is very much population-based,” he says. “What you do for one, you do for the other. And whether it’s a copayment or a deductible, case management or disease management, everybody’s exposed to the same benefit design, the same management approach and strategy from 20th-century insurance risk underwriting. But in the next 5 or 10 years we will need a whole different strategy, because the variation in care and the cost of the components of care will be that much greater.” He doesn’t say just what such a new strategy might be, but is helping to lead an employer-driven biologics initiative that intends to develop one.
Whatever surprises the future holds, knowing everything we can learn about populations would appear to be an always indispensable guide — and the delivery of care an always perfectible process. Meanwhile, maybe the health care community could get together on terminology so that everyone has pretty much the same thing in mind when invoking a favorite phrase. After all, it’s poetry, not business vocabulary, that is supposed to have a thousand meanings.
“I know elephants are good to babies,” says one of the observers in the Sandburg poem, while another insists, “He’s a tough son-of-a-gun outside and I’ll bet he’s got a strong heart.”
In theory, certainly.