Introduction and interview by Sarah Klein
The debate over health care reform has focused primarily on changing the financing of care. Much less attention has been paid to the redesign of care delivery. While both are interrelated, a redesign effort is essential to increase quality and lower costs for patients and payers. Indeed, the authors of The Innovator's Prescription: A Disruptive Solution for Health Care contend it is only by modifying the roles of health care providers and simplifying the business models of hospitals and medical clinics that we will increase access for patients, improve our understanding of complex diseases, and achieve the cost savings necessary for U.S. companies to compete in global markets.
The authors—Clayton Christensen of Harvard Business School, D.B.A., the late Jerome H. Grossman, M.D., of Harvard Kennedy School of Government, and Jason Hwang, M.D., M.B.A., of the Innosight Institute—find evidence for this premise in the experiences of the auto, computer, and telecommunications industries, all of which underwent drastic transformations when a simple, affordable, and ultimately superior product challenged the status quo. In each case, the innovative product that turned the industry on its head by lowering costs and increasing consumer access was discounted by industry leaders, who deemed it inferior and lacking in potential for profit. But as the quality of the innovative product improved, consumers who craved convenience at a reasonable price flocked to it. Thus, personal computers took the place of minicomputers (which had overtaken mainframes); inexpensive Japanese imports were swapped for large American sedans; and cell phones replaced pay phones and many landlines.
According to the authors, the innovation that will disrupt health care, enabling cost reductions and increased access to services, is scientific discovery—specifically that which enhances our ability to diagnose disease and identify reliable treatments. Such discoveries—from advances in molecular biology, refinements in imaging technology, and improvements in telecommunications—will make the treatment of diseases and conditions more routine and rules-based, and therefore less prone to error. In so doing, they will create an opportunity to assign responsibility for treatment to less-expensive and more appropriate caregivers, including primary care physicians, nurse practitioners, and physician assistants (rather than specialist physicians). With widely dispersed technology, those caregivers can provide rules-based care in retail clinics or other venues that are convenient and accessible to patients, rather than capital-intensive hospitals and medical practices, which add overhead expense to medical bills.
"Enabling lower-cost venues of care and lower-cost caregivers to do more sophisticated things … is the mechanism by which health care can become affordable and accessible – not by somehow hoping that a miracle will occur and the expensive ones will become cheap," Christensen told attendees of a conference on innovation in health care at the Mayo Clinic last month.
"It is an uncomfortable message for medical doctors in academic centers that health care won't become affordable and accessible by replicating their expertise and intuition, but rather we have to commoditize [their] expertise. There is scientific progress under way that is actually making that happen for the simplest of diseases," Christensen said.
To explain how and where various types of care should be provided, Christensen and his co-authors have divided medicine into three realms, which fall along a continuum. Precision medicine describes care for diseases that can be diagnosed precisely and for which treatments are predictably effective. Strep throat, Gaucher's disease, and bone fractures are examples of diseases or conditions treated with precision medicine. Empirical medicine covers diseases for which treatment outcomes can be described in probabilistic terms. Heart attack and stroke fall into that realm. The last category, intuitive medicine, describes conditions that are diagnosed only by symptoms and treated with therapies of uncertain efficacy. Examples of diseases and conditions that require intuitive medicine include depression, lupus, and multiple sclerosis.
Rules-based precision medicine would be outsourced to nurse practitioners and physician assistants (or generalist physicians, if necessary). There is considerable evidence demonstrating the value of nurse practitioners in delivering basic care, and a growing body of evidence demonstrating the contributions of physician assistants and other licensed independent providers.
Diseases that fall in the most challenging category of intuitive medicine—that is, those for which we have no reliable treatment and which can only be diagnosed by symptoms—should be left in the hands of specialists, who ideally would be working collaboratively in a hospital setting to diagnose and treat the most complicated cases. Such programs exist in limited numbers. They include the National Jewish Medical and Research Center in Denver, which focuses on pulmonary disease, and the Texas Heart Institute in Houston, which focuses on cardiovascular disease.
"These diseases are still in the realm of intuitive medicine because they often arise at the intersection of multiple systems in a body," Christensen says. And having specialists work together to diagnose and treat these cases is likely to shift them from what he calls "unstructured, trial-and-error intuitive medicine" into the realm of pattern recognition and precision medicine.
In the meantime, oversight of patients suffering from chronic diseases and conditions that require behavioral interventions, such as diabetes, high cholesterol, and obesity, should be assigned to multidisciplinary networks of professionals trained to keep chronically ill patients healthy. Improving the quality of care for those coping with multiple chronic diseases could go a long way toward lowering health care costs. Community-based solutions, including patient support groups, such as dLife.com for patients with diabetes and their families and Alcoholics Anonymous for those fighting alcohol dependence, can be effective as well, especially for chronic conditions that require an extensive behavioral change, but for which the motivation to change can be low.
In the authors' vision, primary care physicians would continue to provide wellness and prevention services and, as technology enables it, would take over testing, imaging, and other services that specialists now provide.
To accomplish this would require significant changes in health care business models. Hospitals that set themselves up as "solution shops," with multidisciplinary teams of specialists who focus on particularly difficult diseases using intuitive medicine, would bill on a fee-for-service basis. Services (typically procedures) provided outside of a hospital after a definitive diagnosis has been made would be paid on a fee-for-outcome basis, because the treatment and its outcome would be more predictable. Services for chronic disease management through online services or other facilitated networks would be paid on a fee-for-membership basis, while firms handling chronic disease management using nurse practitioners or other professionals would be paid through capitation. The authors contend it would be a mistake for providers to attempt to combine these models, because doing so would conflate measurement of value, costs, pricing, and profit. Accurate measures of these items are necessary to make the system more efficient.
Quality Matters asked the authors how their proposal would affect the quality of care. Hwang, one of the authors, provided the responses.
QM: If the innovations you describe in the book occurred, how would the quality of health care improve from the perspective of patients?
Hwang: It's important to recognize that the definition of quality differs depending on the circumstance. For complex conditions, patients want assurance of an accurate diagnosis and effective course of treatment, without the need to visit specialist after specialist over a period of many months to years. These patients will find their demands for better quality met by institutions that have integrated across multiple disciplines to diagnose diseases correctly the first time. Hospitals focused on diagnosing pulmonary diseases or neuropsychiatric conditions are some of the examples emerging today.
QM: What about simple conditions?
Hwang: For those, patients often have a very different definition of quality. Because the care is routine and rules-based, patients' priorities are convenience and affordability. New business models such as retail clinics have been successful because they recognize the changing dimensions of perceived quality.
QM: Will providers view the change in quality differently than patients?
Hwang: Providers may be a little more reluctant to change. The multidisciplinary solution shops described above require historically independent physicians to work in very different ways from which they may be accustomed. Likewise, for routine care, it means accepting that some basic elements of care can be performed as well or even better by providers with less training [than physicians]. Because the highest-cost providers are being disrupted by lower-level providers, they will believe that quality is suffering and will attempt to put up roadblocks against disruption.
QM: In the book, you argue that existing business models for hospitals and physician practices greatly increase the cost of health care by adding overhead charges for services that could be more quickly provided by less-expensive health care providers in less capital-intensive settings, such as retail clinics. What are some of the other benefits of moving care out of traditional physician offices and hospitals?
Hwang: The new business models move care away from centralized facilities like the general hospital and closer to where patients work and live. In addition to lowering overhead costs, there is the element of convenient access. Many of us have experienced the internal debate over whether it is worth taking significant time away from work or from home in order to see a doctor. Making medical care available in on-site clinics at work or in convenient settings like pharmacies and grocery stores will reduce the self-denial of care that exists whenever people are forced to make an inconvenient choice.
QM: Are there downsides to such a redistribution of work—such as the risk that less-experienced practitioners will miss symptoms of more complex problems, especially in patients who've received episodic care or have been medically underserved? If so, what safeguards might be put in place to prevent this from happening?
Hwang: It's important that providers do not practice beyond the limits of their abilities, but if implemented correctly, the risk for negative outcomes from disruptions like retail clinics is less, not greater. The bottom line is that we ought to be encouraging all providers to practice up to the limits of their licensure; when we restrict expansion of duties such that people are practicing too far below their limits, we price patients out of health care. In practical terms, the type of care that ought to be offloaded is the rules-based work—care that is driven by algorithm and routine processes—which does not require the intuition of a higher-cost expert. Simple checklist forms, computer-based decision tools, go/no-go diagnostic tests, and interoperable health records are all tools that already exist to minimize risk in these new settings.
QM: You've categorized medicine practiced in the U.S. into three types: precision medicine, which describes care for diseases that can be diagnosed precisely and for which treatments are predictably effective; empirical medicine, which covers diseases for which treatment outcomes can be described in probabilistic terms; and intuitive medicine, which describes conditions that are diagnosed only by symptoms and treated with therapies of uncertain efficacy. Do you think the means by which we measure and monitor the quality of care should be revised to reflect the different approaches of each?
Hwang: Yes, and this ties in very closely with how we pay for each of these categories of care. For example, when pay-for-performance initiatives attempt to tackle diseases that still reside largely in intuitive medicine, providers object, because there are too many unpredictable variables beyond their control that may impact quality measurements. However, for precision care and much of empirical medicine, you can start to measure—and pay for—quality outcomes in much more certain terms. In these areas, we will increasingly see reflections and measures of quality that are already common in other industries, but have largely been dismissed in health care—customer experience, quality guarantees, ease of accessibility, and timeliness.
QM: In the book, you note the potential of personalized medicine to enhance the quality of care by identifying how biological and non-biological issues can affect a patient's response to treatment. How would you change the way patient data are collected to ensure that providers and researchers have access to the depth of epidemiological data necessary for such work?
Hwang: I would recommend looking at what companies like 23andMe and PatientsLikeMe are doing, which is essentially building a network of users who very willingly share their health data. In fact, in the case of 23andMe, the users actually pay for the privilege of contributing their information to the community database (of course, they get their personal results in return). It really turns the whole process of patient enrollment and clinical trial management on its head. Empowered patients, acting out of their own self-interest, will self-organize and serve as ready partners in biomedical research.
QM: You recommend that physicians transfer oversight of patients with behavior-intensive diseases such as diabetes to nurses or other networks of professionals. However, recent experience suggests better outcomes when disease management is embedded in or integrated with primary care because patients want to feel that the nurse is part of a team that includes their personal physician. Do you think this patient expectation will change as the job assignments of nurses do?
Hwang: Chronic diseases that require significant behavior changes can be managed effectively by several different entities. It's important to coordinate such care with the remainder of the health care system, but it doesn't mean that this care must be provided by or somehow tied to a primary care physician. This argument is currently being made by professional societies that wish to keep physicians in the central coordinating role of patient care. However, there will be more and more patients who only need a software tool, a nurse-staffed telephone line, or reminders via text messaging to manage their chronic diseases effectively. Personally controlled electronic health records will end up serving the important role of data coordination.
QM: Won't assigning disease management responsibilities to an independent person or agency lead to a loss of longitudinal continuity with a personal physician? Can electronic health records really make up for that lack of relationship?
Hwang: Keep in mind that this same situation has played out over and over again whenever a profession built on long-term trust finds itself getting disrupted. Travel agents, real estate agents, retirement portfolio managers, and stockbrokers have all made similar arguments as they were losing customers to new business models within their respective industries. Likewise, this is an argument often made by primary care societies that have a vested interested in holding on to their current position in the system. This is not to say that primary care physicians are unnecessary, but only that people deserve choices when it comes to care coordination. I believe that electronic health records will play an increasingly important role in this regard.
QM: Do you see any disadvantages (financial or other) to locating retail clinics within primary care physician practices to handle care as part of a team, as Group Health Cooperative in Seattle has done?
Hwang: Group Health Cooperative is among the integrated health systems that I believe can implement and manage retail clinics correctly and for the right reasons. However, the main danger when any existing business model tries, in essence, to disrupt itself is that ultimately its motivations and goals will revert back to serving the parent business. In other words, a health system may not see retail clinics as part of an integrated delivery system, but merely as another source of patient referrals to higher-margin hospital and emergency care. If retail clinics fail to drive revenue to the hospital, they'll get shut down, even if they were profitable and delivering quality care all along, simply because the parent organization will find investments that buttress the hospital more attractive.
QM: Your description of facilitated network businesses as a network through which members exchange information or products with one another has many similarities to the definition of accountable care organizations (defined as real or virtual organizations that form relationships between different providers to enhance the coordination of care for a population and for which financial reward is based on group rather than individual performance). What role, if any, might accountable care organizations play in your vision of a restructured health care system?
Hwang: Accountable care organizations will add nothing new unless they are truly free to partner with new business models that are determined to deliver care in very different ways. Otherwise, it's merely the latest term for managed care. The bundled payment model is a good idea, as it shifts responsibility for determining and rewarding value closer to the actual providers of care. However, the same in-fighting across different provider groups will remain and quite possibly intensify as they seek to divide up bundled payments. The best way to bypass this situation is to encourage development of organizations that integrate the financing and delivery of health care—which we called integrated, fixed-fee providers in the book.
QM: Your model of health care delivery would encourage a vertical displacement of jobs, as less-expensive providers take over responsibilities for care traditionally provided by providers at a higher level of pay grade. Do you also expect to see horizontal displacement (e.g., physicians taking over pharmacy responsibilities in remote locations or primary care physicians providing dental sealants to patients)?
Hwang: Yes, as long as it leads to greater access to services that previously did not exist (thereby expanding the market of customers), disruption will occur. At HealthPartners [a Minn.-based not-for-profit HMO], for example, pediatricians have been deeply involved in providing dental sealants to children who otherwise may not have received subsequent care from a dentist. In this situation, expanding the market is not just a power grab, but it benefits everyone involved.
QM: What's the best way to foster an evolution in the roles of medical professionals and paraprofessionals over the next decade?
Hwang: It's futile to expect highly trained, highly skilled professionals like physicians to somehow become less costly versions of themselves, no matter how much you cut their payment rates. It will be up to the institutions that employ physicians to determine when it's possible to disrupt them by using less costly providers. But rather than waiting for hospital nursing and medical residency programs to churn out cheaper graduates, health systems may one day offer medical education themselves to ensure a steady supply of workers who do not require significant re-training. As long as these internally trained providers can perform the work expected by the health system, then a lot of associated costs such as licensure, certification, and accreditation are no longer necessary.
QM: Are there any lessons from other industries with similarly entrenched traditions where this has happened?
Hwang: A similar situation has arisen in the disruption of MBA programs. Rather than continuing to pay the rising salaries of top-tier MBA graduates, many companies have instead started training their existing employees in-house. It is far more cost-effective and ensures that employees receive only the education and training that is really necessary to succeed within the company.
QM: You contend that employers—rather than the government—are in one of the best positions to effect the changes necessary to lower the cost of health care services. Is there an industry (or a type of business) that is likely to pursue such innovation? Should the government encourage employer involvement with incentives?
Hwang: All it takes is a business with enough employees to make it worthwhile to make an alternative investment in employee health care. These innovations are not restricted to any particular industry; even Safeway, which is in an industry with high employee turnover, has found significant value in investing in employee health. Government incentives could help smaller businesses, but innovations among large employers are already happening. Forward-looking companies with successful employee health programs like Perdue Farms, Pitney-Bowes, and QuadMed [see this issue's case study for more detail] may eventually open up their clinics to the general public as a more affordable, higher-quality option. They're simply following the same path that led to Kaiser Permanente's founding many decades ago.
QM: How might scientific discovery lead to innovations in population health and prevention?
Hwang: One of the most critical factors to improving population health is the reduction of existing disparities, and patients that normally cannot afford or access health care will especially benefit from new medical technologies that allow for lower-cost, disruptive models of care. This requires existing institutions that attempt to address these disparities, notably the public health system, to embrace innovations like retail clinics as legitimate partners in reaching these underserved populations. Moreover, armed with new technologies and connected through online facilitated networks, patients will be capable of doing much more for themselves, rather than always relying on a provider they may see for only one or two hours out of the year. This is not only the most expedient way, but it is the only affordable option to ensure that everyone receives appropriate preventive and wellness care.
Related: See Robert H. Brook's commentary in the current issue of the Journal of the American Medical Association, Disruption and Innovation in Health Care
