Managing Rapidly Increasing Complexity

The complexity of health care has increased in multiple dimensions—in the ever-increasing treatment, diagnostic, and care management options available; in the rapidly rising levels of biomedical and clinical evidence; and in administrative complexities, from complicated workflows to fragmented financing. The complexity due to ever-increasing treatment options can be illustrated by the evolution of care for two common conditions—heart disease and cancer. During much of the twentieth century, heart attacks commonly were treated with weeks of bed rest. Today, advanced diagnostics allow for customized treatments for patients; interventions such as percutaneous coronary interventions and coronary artery bypass grafts can reopen blocked vessels and restore blood flow to the heart; and pharmaceutical therapies, such as thrombolytics and beta-blockers, improve survival and reduce the chances of subsequent heart attacks (Certo, 1985; Nabel and Braunwald, 2012). Similarly, five decades ago, breast cancer was detected from a physical exam, and mastectomy was the recommended treatment. Today, multiple imaging technologies exist for the detection and diagnosis of the disease, and once diagnosed, the cancer can be further classified and treated according to genetic characteristics and hormone receptor status (Harrison, 1962; IOM, 2001b; Kasper and Harrison, 2005).

As a result of improved scientific understanding, new treatments and interventions, and new diagnostic technologies, the U.S. health care system now is characterized by more to do, more to know, and more to manage than at any time in history. As one quantification of this increase, the volume of the biomedical and clinical knowledge base has rapidly expanded, with research publications having risen from more than 200,000 per year in 1970 to more than 750,000 in 2010 (see Figure S-1). The result is a paradox: advances in science and technology have improved the ability of the health care system to treat diseases, yet the sheer volume of new discoveries stresses the capabilities of the system to effectively generate and manage knowledge and apply it to regular care. These advances have occurred at the same time as, and sometimes have contributed to, challenges in health care quality and value.

FIGURE S-1

Number of journal articles published on health care topics per year from 1970 to 2010. Publications have increased steadily over 40 years, with the rate of increase becoming more pronounced starting approximately in 2000. SOURCE: Data obtained from online (more...)

Conclusion: Diagnostic and treatment options are expanding and changing at an accelerating rate, placing new stresses on clinicians and patients, as well as potentially impacting the effectiveness and efficiency of care delivery.

Beyond the increasing stores of biomedical and clinical knowledge, changes in disease prevalence and patient demographics have altered the landscape for care delivery. The prevalence of chronic conditions, for example, has increased over time. In 2000, 125 million people suffered from such conditions; by 2020, that number is projected to grow to an estimated 157 million (Anderson, 2010). The role of chronic diseases has changed as the demographics of the population have shifted. In general, the population has gotten older; in the past decade, the portion of the population over age 65 has increased at 1.5 times the rate of the rest of the population (Howden and Meyer, 2011). Almost half of those over 65 receive treatment for at least one chronic disease, and more than 20 percent receive treatment for multiple chronic diseases (Schneider et al., 2009); fully 75 million people in the United States have multiple chronic conditions (Parekh and Barton, 2010).

Managing these multiple conditions requires a holistic approach, because the use of various clinical practice guidelines developed for single diseases may have adverse effects (Boyd et al., 2005a; Parekh and Barton, 2010; Tinetti et al., 2004). For example, existing clinical practice guidelines would suggest that a hypothetical 79-year-old woman with osteoporosis, osteoarthritis, type 2 diabetes, hypertension, and chronic obstructive pulmonary disease should take as many as 19 doses of medication per day. Such guidelines might also make conflicting recommendations for the care. If she had peripheral neuropathy, guidelines for osteoporosis would recommend that she perform weight-bearing exercise, while guidelines for diabetes would recommend that she avoid such exercise (Boyd et al., 2005a). These situations create uncertainty for clinicians and patients as to the best course of action to pursue as they attempt to manage the treatments for multiple conditions.

Conclusion: Chronic diseases and comorbid conditions are increasing, exacerbating the clinical, logistical, decision-making, and economic challenges faced by patients and clinicians.

Care delivery also has become increasingly demanding. It would take an estimated 21 hours per day for individual primary care physicians to provide all of the care recommended to meet their patients' acute, preventive, and chronic disease management needs (Yarnall et al., 2009). Clinicians in intensive care units, who care for the sickest patients in a hospital, must manage in the range of 180 activities per patient per day—from replacing intravenous fluids, to administering drugs, to monitoring patients' vital signs (Donchin et al., 2003). In addition, rising administrative burdens and inefficient workflows mean that hospital nurses spend only about 30 percent of their time in direct patient care (Hendrich et al., 2008; Hendrickson et al., 1990; Tucker and Spear, 2006). These pressures are not limited to clinicians; patients often find the health care system uncoordinated, opaque, and stressful to navigate. One study found that for 1 of every 14 tests, either the patient was not informed of a clinically significant abnormal test result, or the clinician failed to record reporting the result to the patient (Casalino et al., 2009).

With specialization, moreover, clinicians must coordinate with multiple other providers; for their health care, Medicare patients now see an average of seven physicians, including five specialists, split among four different practices (Pham et al., 2007). One study found that in a single year, a typical primary care physician coordinated with an average of 229 other physicians in 117 different practices just for Medicare patients (Pham et al., 2009). The involvement of multiple providers tends to blur accountability. One survey found that 75 percent of hospital patients were unable to identify the clinician in charge of their care (Arora et al., 2009).

Conclusion: Care delivery has become increasingly fragmented, leading to coordination and communication challenges for patients and clinicians.

Achieving Greater Value in Health Care

In addition to, and sometimes as a result of, the challenge of complexity, health care quality and outcomes often fall short of their potential. A decade after the IOM (1999) estimated that 44,000 to 98,000 patients died each year from preventable medical errors, recent studies have reported that as many as one-third of hospitalized patients may experience harm or an adverse event, often from preventable errors (Classen et al., 2011; Landrigan et al., 2010; Levinson, 2010). While infections and complications once were viewed as routine consequences of medical care, it is now recognized that strategies and evidence-based interventions exist that can significantly reduce the incidence and severity of such events.

Similarly, medical care often is guided insufficiently by evidence, with Americans receiving only about half of the preventive, acute, and chronic care recommended by current research and evidence-based guidelines (McGlynn et al., 2003). Sometimes this occurs because available evidence is not applied to clinical care, while in other cases evidence is not available.

As a result of all of these factors, the nature and quality of health care vary considerably among states, with serious health and economic consequences. If all states could provide care of the quality delivered by the highest-performing state, an estimated 75,000 fewer deaths would have occurred across the country in 2005 (McCarthy et al., 2009; Schoenbaum et al., 2011).

Conclusion: Health care safety, quality, and outcomes for Americans fall substantially short of their potential and vary significantly for different populations of Americans.

These deficiencies in care quality have occurred even as expenses have risen significantly. Health care costs¹ have increased at a greater rate than the economy as a whole for 31 of the past 40 years, and now constitute 18 percent of the nation's gross domestic product (CMS, 2012; Keehan et al., 2011). The growth in health care costs has contributed to stagnation in real income for American families. Although income has increased by 30 percent over the past decade, these gains have effectively been eliminated by a 76 percent increase in health care costs (Auerbach and Kellermann, 2011). These high costs have strained families' budgets and put health insurance coverage out of reach for many, contributing to the 50 million Americans without coverage (DeNavas-Walt et al., 2011).

In addition to unsustainable cost growth, there is evidence that a substantial proportion of health care expenditures is wasted, leading to little improvement in health or in the quality of care. Estimates vary on waste and excess health care costs, but they are large. The IOM workshop summary The Healthcare Imperative: Lowering Costs and Improving Outcomes contains estimates of excess costs in six domains: unnecessary services, services inefficiently delivered, prices that are too high, excess administrative costs, missed prevention opportunities, and medical fraud (IOM, 2010). These estimates, presented by workshop speakers with respect to their areas of expertise and based on assumptions from limited observations, suggest the substantial contribution of each domain to excessive health care costs (see Table S-1).

TABLE S-1

Estimated Sources of Excess Costs in Health Care (2009).

Although these estimates have unknown overlap, the sum of the individual estimates—$765 billion—suggests the significant scale of waste in the system. Two other independent and differing analytic approaches—considering regional variation in costs and comparing costs across countries—produce similar estimates, with total excess costs approaching $750 billion in 2009 (Farrell et al., 2008; IOM, 2010; Wennberg et al., 2002). While there are methodological issues with each method for estimating excess costs, the consistently large figures produced by each signal the potential for reducing health care costs while improving quality and health outcomes.

At this level, unnecessary health care costs and waste exceed the 2009 budget for the Department of Defense by more than $100 billion (OMB, 2010). Health care waste also amounts to more than 1.5 times the nation's total infrastructure investment in 2004, including roads, railroads, aviation, drinking water, telecommunications, and other structures.² To put these estimates in the context of health care expenditures, the estimated redirected funds could provide health insurance coverage for more than 150 million workers (including both employer and employee contributions), which exceeds the 2009 civilian labor force.³ And the total projected amounts could pay the salaries of all of the nation's first response personnel, including firefighters, police officers, and emergency medical technicians, for more than 12 years.⁴

Conclusion: The growth rate of health care expenditures is unsustainable, with waste that diverts major resources from necessary care and other priorities at every level—individual, family, community, state, and national.

In sum, as illustrated in Figure S-2, each stage in the processes that shape the health care received—knowledge development, translation into medical evidence, application of evidence-based care—has prominent shortcomings and inefficiencies that contribute to a large reservoir of missed opportunities, waste, and harm. The threats to the health and economic security of Americans are clear, present, and compelling.

FIGURE S-2

Schematic of the health care system today.

Capturing Opportunities from Technology, Industry, and Policy

As noted earlier, new opportunities exist to address the challenges outlined above. Just as the information revolution has transformed many other fields, growing stores of data and computational abilities hold the same promise for improving clinical research, clinical practice, and clinical decision making. In the past three decades, for example, computer processing speed has grown by 60 percent per year on average, while the capacity to share information over telecommunications networks has risen by an average of 30 percent per year (Hilbert and López, 2011). These advances in computing and connectivity have the potential to improve health care by expanding the reach of knowledge, increasing access to clinical information when and where needed, and assisting patients and providers in managing chronic diseases. Studies also have found that using such electronic systems can improve safety—one study reported a 41 percent reduction in potential adverse drug events following the implementation of a computerized patient management system (computerized physician order entry, or CPOE), while another estimated that overall medication error rates dropped by 81 percent (Bates et al., 1998, 1999; Potts et al., 2004). Projections are for 90 percent of office-based physicians to have access to fully operational electronic health records by 2019, up from 34 percent in 2011 (Congressional Budget Office, 2009; Hsiao et al., 2011). Because these capacities are relatively early in their development in the health care arena, there is substantial room for progress as they are implemented in the field. However, multiple nontechnological developments, such as supportive care processes, governance, and patient and public engagement, will be necessary if these technologies are to reach their full potential.

Conclusion: Advances in computing, information science, and connectivity can improve patient-clinician communication, point-of-care guidance, the capture of experience, population surveillance, planning and evaluation, and the generation of real-time knowledge—features of a continuously learning health care system.

In addition to advances in computing and connectivity, new organizational capabilities have been developed in diverse industries to improve safety, quality, reliability, and value. Advances in safety alone, for instance, enabled domestic commercial commuter airlines to report no fatalities from 2007 to 2010 (Bureau of Transportation Statistics, 2011). New capabilities in systems engineering, operations management, and production can be adapted to health care settings to improve performance. In one study, the use of checklists inspired by the aviation industry eliminated catheter-related bloodstream infections in the intensive care units of most hospitals in the study and resulted in an 80 percent decrease in infections per catheter-day (Pronovost et al., 2006, 2009). Commercial strategies to improve the reliability of the delivery of goods and services have potential applicability to health care as well. A pharmacy unit, for example, undertook systematic problem solving and reduced the time spent searching for medications by 30 percent and the frequency of out-of-stock medications by 85 percent (Spear, 2005).

Conclusion: Systematic, evidence-based process improvement methods applied in various sectors to achieve often striking results in safety, quality, reliability, and value can be similarly transformative for health care.

Across the United States, moreover, there is growing momentum to implement novel partnerships and collaborations to test delivery system innovations aimed at high-value, high-quality health care. In many settings, stakeholders at all levels—federal, state, and local governments; public and private insurers; health care delivery organizations; employers; patients and consumers; and others—are working together with the shared objectives of controlling health care costs and improving health care quality. States ranging from Massachusetts to Utah to Vermont have introduced new initiatives aimed at expanding health insurance coverage, improving care quality and value, and advancing the overall health of their residents. Multiple initiatives by employers, specialty societies, patient and consumer groups, health care delivery organizations, health plans, and others—such as the American Board of Internal Medicine (ABIM) Foundation's Choosing Wisely^® campaign and the Good Stewardship project—are focused on improving the health care system. Other initiatives currently under way range from the Patient-Centered Primary Care Collaborative, which seeks to spread patient-centered medical homes; to community-based initiatives, such as the Aligning Forces for Quality program and the Chartered Value Exchange project; to all-payer databases being established in various states around the country. And drawing on their experiences in improving outcomes and lowering costs through initiatives in their own institutions, a group of health care delivery leaders has developed “A CEO Checklist for High-Value Health Care,” which describes system-change approaches that can be adopted in most health care settings to improve outcomes and reduce costs of care (Cosgrove et al., 2012) (see Appendix B). The convergence of these novel partnerships, a changing health care landscape, and investments in knowledge infrastructure has created a unique opportunity to achieve continuously learning health care.

Conclusion: Innovative public- and private-sector health system improvement initiatives, if adopted broadly, could support many elements of the transformation necessary to achieve a continuously learning health care system.

Generating and Using Real-Time Knowledge to Improve Outcomes

Although unprecedented and increasing levels of information are available in journals, guidelines, and other sources, patients and clinicians often lack practical access to guidance that is relevant, timely, and useful for the circumstances at hand. For example, fewer than half of the clinical guidelines for the nine most common chronic conditions consider older patients with multiple comorbid chronic conditions, even though, as noted earlier, 75 million Americans fall in that category (Boyd et al., 2005b; Parekh and Barton, 2010). In the case of localized prostate cancer, for instance, which treatment works best for a given patient—from watchful waiting, to radical prostatectomy, to radiation and chemotherapy—is unknown. Furthermore, the evidence base for clinical guidelines and recommendations needs to be strengthened. In some cases, 40 to 50 percent of the recommendations made in guidelines are based on expert opinion, case studies, or standards of care rather than on more systematic trials and studies (Chauhan et al., 2006; IOM, 2008, 2011a; Tricoci et al., 2009).

New methods are needed to address current limitations in clinical research. The cost of current clinical research methods averages $15-$20 million for larger studies—and much more for some—yet there are concerns about generalizing study results to all practice conditions and patient populations (Holve and Pittman, 2009, 2011). Given the increasing number of new medical treatments and technologies, the complexity of managing multiple chronic diseases, and the growing personalization of treatments and diagnostics, the challenge is to produce and deliver practical evidence that clinicians and patients can apply to clinical questions.

Conclusion: Despite the accelerating pace of scientific discovery, the current clinical research enterprise does not sufficiently address pressing clinical questions. The result is decisions by both patients and clinicians that are inadequately informed by evidence.

Meeting this challenge will require new approaches for generating clinical evidence that reduce the expense and effort of conducting research and improve the clinical applicability of research findings while retaining the rigorous reliability of the process. The issue is not determining which research method is best for a particular condition, but which method provides the information most appropriate to a particular clinical need. Each study must be well tailored to provide useful, practical, and reliable results for the condition at hand.

Opportunities for achieving these aims leverage the expanded capacity of the digital infrastructure along with new statistical and research techniques. Computational capabilities present promising, as yet unrealized, opportunities for care improvement, while advances in statistical analysis, simulation, and modeling can supplement traditional methods for conducting trials. The application of computing capacity and new analytic approaches enables the development of real-time research insights from patient populations. For example, one study found that real-time analysis of clinical data from electronic health records could have identified the increased risk of heart attack associated with one diabetes drug within 18 months of its introduction, as opposed to the 7–8 years between the medication's introduction and the point at which concerns were raised publicly (Brownstein et al., 2010). Computational capabilities also hold promise for hastening the derivation of important new insights from the care experience. A comprehensive disease registry for heart attack patients in Sweden, for example, has contributed to a 65 percent reduction in 30-day mortality and a 49 percent decrease in 1-year mortality from heart attacks (Larsson et al., 2012).

Conclusion: Growing computational capabilities to generate, communicate, and apply new knowledge create the potential to build a clinical data infrastructure to support continuous learning and improvement in health care.

Harnessing this potential for care improvement will require systematic approaches that address the regulatory, commercial, communications, and technological challenges involved. Results of surveys of health researchers suggest that the current formulation and interpretation of privacy rules have increased the cost and time to conduct research, impeded collaboration, and hampered the recruiting of subjects (IOM, 2009; Ness, 2007). Privacy is a highly important societal and personal value, but the current rules, with their inconsistent interpretation, offer a relatively limited security advantage to patients while impeding the pace and scope of new insights from health research and care improvement.

Conclusion: Regulations governing the collection and use of clinical data often create unnecessary and unintended barriers to the effectiveness and improvement of care and the derivation of research insights.

The current system for capturing and using new knowledge is already flawed and, absent change, is likely to be overwhelmed by the pace of knowledge growth. The diffusion of new evidence can take considerable time; in the case of thrombolytic drugs for heart attack treatment, for example, 13 years elapsed between when they were shown to be effective and when most experts recommended the treatment (Antman et al., 1992). Substantial work is required to identify high-quality evidence that minimizes the risk of contradiction by later studies and is sufficiently robust to provide insight on application to a particular patient's clinical circumstances. This is time-consuming work, which goes on while clinical patterns are being formed.

Realizing the prospect of faster, deeper knowledge bases will require parallel advances in the approaches to gathering and assessing evidence, making evidence-based recommendations, translating those recommendations to practice, and reinforcing their use through relevant policies. Computing capacity can help with assessment as well as dissemination. Technological tools, such as decision support tools that can be broadly embedded in electronic health records, hold promise for improving the application of evidence. One study found that digital decision support tools helped clinicians apply clinical guidelines, improving health outcomes for diabetics by 15 percent (Cebul et al., 2011).

Conclusion: As the pace of knowledge generation accelerates, new approaches are needed to deliver the right information, in a clear and understandable format, to patients and clinicians as they partner to make clinical decisions.

Engaging Patients, Families, and Communities⁵

The structure, incentives, and culture of the health care system are poorly aligned to engage patients and respond to their needs. While clinicians supply information and advice based on their scientific expertise in treatment and intervention options, as well as potential outcomes, patients, their families, and other caregivers bring personal knowledge regarding the suitability—or lack thereof—of different treatments for the patient's circumstances and preferences. Information from both sources is needed to select the right care option, particularly because studies have found that patients and clinicians have differing views on the importance of different health goals and health care risks (Lee et al., 2010a,b). At the same time, it is important to note that patient-centered care does not mean simply agreeing to every patient request. Rather, it entails meaningful awareness, discussion, and engagement among patient, family, and clinician on the evidence, risks and benefits, options, and decisions in play.

Currently, patients often are insufficiently involved in their care decisions. Even when they are encouraged to play a role in decisions about their care, they often lack understandable, reliable information—from evidence on the efficacy and risks of different treatment options to information on the quality of different providers and health care organizations—that is customized to their needs, preferences, and health goals. Fewer than half of patients receive clear information on the benefits and trade-offs of treatments for their condition, and fewer than half are satisfied with their level of control in medical decision making (Degner et al., 1997; Fagerlin et al., 2010; IOM, 2011b; Lee et al., 2011, 2012; Sepucha et al., 2010).

To improve patients' involvement in their care decisions, communication tools need to be developed and customized to patient circumstances. Given the complexity of health care, even highly educated people may have difficulty finding and understanding health information and applying it to their own care or that of their loved ones (IOM, 2004), and those who produce health care information need to consider how that information will be received and used by patients (Maurer et al., 2012). Technology offers opportunities for clinicians to engage patients by meeting with them where they are. These opportunities include improving communications outside of traditional clinical visits by providing new venues for care; assisting patients in managing their own health; and explaining options for shared clinical decisions, a capability that highlights health professionals' need to assume new roles in partnering with patients in the use of reliable online sources of health information (Brach et al., 2012).

Patient-centered care takes on increasing importance in light of research linking such care to better health outcomes, lower costs, an enhanced care experience, better quality of life, and other benefits. Patient and family involvement in health care decisions has been associated in primary care settings with reduced pain and discomfort, faster recovery in physical health, and improvements in emotional health (Stewart et al., 2000). Well-informed patients also often choose less aggressive and costly therapies. For example, it has been reported that informed patients are up to 20 percent less likely than other patients to choose elective surgery (O'Connor et al., 2009; Stacey et al., 2011). Similarly, patient-centered communication in primary care visits has been correlated with fewer diagnostic tests and referrals (Epstein et al., 2005; Stewart et al., 2000), as well as with annual charges in the range of 33 percent lower (Bertakis and Azari, 2011a,b).

Not all care delivered in the name of patient-centeredness reduces costs or improves outcomes. For example, one study found that patient-centeredness was associated with better outcomes but also higher costs (Bechel et al., 2000). Other studies have yielded mixed results with respect to cost, quality, and value for care models that aim to implement different aspects of patient-centeredness, such as disease management and care coordination programs (Nelson, 2012; Peikes et al., 2009). This may be related in part to the difficulty of identifying what truly constitutes patient-centered care, with well-meaning but poorly informed efforts producing changes that are superficial and adding little value to the experience. In the name of patient-centeredness, for example, some health care organizations have adopted luxury, hotel-like amenities or renovated their facilities. Although some of these initiatives may appeal to patient tastes, they do not achieve the true goals of patient-centered care and may increase costs while not directly addressing the patient's needs, preferences, or goals most important to improving quality, health, and value.

This report builds on the definition of patient-centered care offered in Crossing the Quality Chasm: “providing care that is respectful of and responsive to individual patient preferences, needs, and values and ensuring that patient values guide all decisions” (IOM, 2001a). The concept encompasses multiple dimensions, including respect for patients' values, preferences, and needs; coordination and integration of care; information, communication, and education; physical comfort; emotional support; and involvement of family and friends. This definition provides a framework for care to be fully patient-centered.

Conclusion: Improved patient engagement is associated with better patient experience, health, and quality of life and better economic outcomes, yet patient and family participation in care decisions remains limited.

Given the increasing incidence of chronic diseases, the complexity of modern health care, and the multiple determinants of health, the challenges facing the health care system cannot be met by any individual or organization acting alone. Yet care often is poorly coordinated among clinicians both within and across settings. In one survey, roughly 25 percent of patients noted that a test had to be repeated, often because the results had not been shared by another provider (Stremikis et al., 2011). This inadequate, sometimes absent, continuity of care endangers patients and contributes to system waste. For example, almost one-fifth of Medicare patients are rehospitalized within 30 days, often without seeing their primary care provider in the interim (Jencks et al., 2009). Comprehensive health care also requires accounting for factors typically outside of the traditional health care system. Most determinants of the health status of individuals and populations lie not in health care—medical care accounts for only 10 to 20 percent of overall health prospects—but in such factors as behavior, social circumstances, and environment. Thus, close clinical-community coordination is required to protect and improve health (McGinnis et al., 2002).

Conclusion: Coordination and integration of patient services currently are poor. Improvement in this area will require strong and sustained avenues of communication and cooperation between and among clinical and community stewards of services.

Achieving and Rewarding High-Value Care

Health care payment policies strongly influence how care is delivered, whether new scientific insights and knowledge about best care are diffused broadly, and whether improvement initiatives succeed. Clinicians reimbursed for each service tend to recommend more visits and services than clinicians who are reimbursed under other payment methods. In one study, initiation of encounter- and procedure-based reimbursement for primary care led to an increased number of encounters and procedures, with visits increasing from 11 to 61 percent depending on the specialty (Helmchen and Lo Sasso, 2010). As with most aspects of health care, a variety of financial incentives and payment models currently are in use. However, most of these models tend to pay clinicians and health care organizations without a specific focus on patient health and value, which has contributed to waste and inefficiency. One study found, on average, only a 4.3 percent correlation between the quality of care delivered and the price of the medical service, with higher prices often being associated with lower quality (Office of the Attorney General of Massachusetts, 2011).

Conclusion: The prevailing approach to paying for health care, based predominantly on individual services and products, encourages wasteful and ineffective care.

Given the clear need for change, several health care organizations and health insurers across the nation have been testing new models of paying for care and organizing care delivery. While many individual initiatives have demonstrated success, evidence is conflicting on which payment models might work best and under what circumstances. Yet, it is clear that high-value care—the best care for the patient, with the optimal result for the circumstances, delivered at the right price—requires that payment and practice incentives be structured to reward the best outcomes for the patient.

To transition to a health care payment system that rewards value, assessment techniques are needed to identify and encourage high-value care. In part, this is a clinical effectiveness issue. Unnecessary and marginal treatments and tests have the potential for side effects and harm. But at its core, health care value is a basic representation of the efficient use of individual and societal resources—time, money—for individual and societal benefit. Because measures of value must fundamentally balance the results of care with the costs required to achieve the results, accurate information is needed on the various dimensions of cost, as well as the various dimensions of health—health status, quality of life, quality of care, satisfaction, and population health.

Measurement itself is only part of the improvement process. Transparency on results produces data that clinicians can use for improvement initiatives, provides information that patients and consumers can use to select care and providers, and draws attention to high-value health care providers and organizations. Several transparency initiatives have been correlated both with improving performance on those measures reported and with encouraging organizations to undertake improvement activities. Following public reporting of pneumonia care measures, for example, rates of compliance with the measures rose from 72 percent to 95 percent in 8 years (Joint Commission, 2011). Results from another initiative showed that providing financial incentives together with helping clinicians monitor their practice patterns against those of others decreased spending by 2 percent per quarter while improving the overall quality of care (Chernew et al., 2011; Mechanic et al., 2011; Song et al., 2011). While further work is needed to improve the practical implementation of transparency and minimize negative consequences, greater transparency is necessary to provide the information needed to promote continuous learning and improvement.

Conclusion: Transparency of process, outcome, price, and cost information, both within health care and with patients and the public, has untapped potential to support continuous learning and improvement in patient experience, outcomes, and cost and the delivery of high-value care.

Creating a New Culture of Care

Although financial incentives can be important to the pace at which change occurs, they do not operate in a vacuum. The culture of health care is central to promoting learning at every level. Continuous improvement requires systematic problem solving, the application of systems engineering techniques, operational models that encourage and reward sustained quality and improved patient outcomes, transparency on cost and outcomes, and strong leadership with a vision devoted to improving health care processes. The goal is to create continuously learning organizations that generate and transfer knowledge from every patient interaction to yield greater performance predictability and reliability.

As with many other aspects of the health care enterprise, there is great diversity in the organizations that deliver care, from small group practices, to independent practice associations, to individual hospitals, to large integrated delivery systems. Each brings different strengths and weaknesses, and each plays a significant and important role in delivering high-quality, high-value care. Given the dramatic differences in local health care infrastructures, substantial heterogeneity will persist for the foreseeable future. Yet, the need for a new culture of care is common to all types of health care organizations; all need to build their capabilities to continuously learn and improve.

Most vital to building a continuously learning organization is leadership and governance that defines, disseminates, and supports a vision of continuous improvement (Cosgrove et al., 2012). One study found that hospitals ranking in the top 5 percent for heart attack outcomes had a strong leadership and governance commitment to improvement, good communication and coordination, shared values and culture, and experience with problem solving and learning (Curry et al., 2011). An organization's leadership—and that leadership's visible priorities—sets its tone, defines and communicates its goals, motivates its staff, and marshals the necessary resources. By defining and visibly emphasizing a vision that encourages and rewards learning and improvement, leadership at all levels of the organization prompt its disparate elements to work together toward a common end.

If leadership provides the top-down mission of an organization, the organization's culture represents the social scaffolding that empowers system transformation. Organizational culture can encourage strong communication and coordination among clinicians, provide psychological safety that encourages open communication, and support innovation and creativity. This culture of care considers the needs and abilities of individual patients and how they can be engaged as members of the care team. Further, an organization's commitment to teaming, partnership, and continuity is fundamental in fostering a culture of continuous learning and improvement. In a large, multifacility integrated health system, for example, an intervention that focused on teamwork training, coaching, and communication skills saw an 18 percent reduction in annual mortality among participating facilities, with adverse events continuing to decrease, versus only a 7 percent reduction among nonparticipating facilities (Neily et al., 2010, 2011).

Continuous learning requires dedicated learning processes—mechanisms that help the organization constantly capture knowledge and implement improvements. Achieving systems-based problem solving requires an organizational culture that incentivizes experimentation among staff—one that recognizes failure as key to the learning process and does not penalize employees if their experiments are unsuccessful. These processes can take many forms, yet they share certain essential elements: systematic problem solving and experimentation, learning from past experience and from others, and the use of internal transparency as a tool to motivate further improvement. Beyond systems-based problem solving, systems that continuously learn and improve also need to be adept at transferring the knowledge they gain throughout the organization. While each of these factors is important, it is the organization's operational model—the way it aligns goals, resources, and incentives—that makes learning actionable. An organization's operational model can incentivize continuous learning, help control variability and waste that do not contribute to quality care, recoup savings to invest in improving care processes and patient health, and make improvement sustainable.

Conclusion: Realizing the potential of a continuously learning health care system will require a sustained commitment to improvement, optimized operations, concomitant culture change, aligned incentives, and strong leadership within and across organizations.

Foundational Elements

Care Improvement Targets

Supportive Policy Environment