Using Clinical Information Technology in Chronic Disease Care: Expert Workshop Summary Prepared for CALIFORNIA HEALTHCARE FOUNDATION by Robert Mittman August 2004 About the Author Robert Mittman is a technology forecaster and strategist. He leads Facilitation, Foresight, Strategy, a strategy consultancy; and is a senior research affiliate at the Institute for the Future. He can be reached at robert@mittman.org. About the Foundation The California HealthCare Foundation, based in Oakland, is an independent philanthropy committed to improving California’s healthcare delivery and financing systems. Formed in 1996, our goal is to ensure that all Californians have access to affordable, quality healthcare. For more information about CHCF, visit us online at www.chcf.org. ISBN 1-932064-75-3 ©2004 California HealthCare Foundation Contents 4 I. Introduction 6 II. Summary of Key Findings Essential Elements of Chronic Disease Care Disease Registries and Electronic Medical Records as Chronic Care Tools Forecast of Clinical IT for Chronic Disease Care in Three Practice Settings Accelerating Application of Clinical IT to Chronic Disease Care 14 III. Conclusion 15 Appendix: Participant List I. Introduction Workshop participants were CHRONIC CONDITIONS SUCH AS DIABETES, ASTHMA, and coronary artery disease are the major causes of illness, convened to forecast how disability, and death in the United States. In 2000, the medical physician offices, clinics, cost of chronic disease amounted to 75 percent of direct health and group practices will care spending, and the indirect costs of lost work, pain, and suffering were much higher. The optimal management of apply clinical information patients with chronic diseases requires the tracking of patients technology to chronic over time to monitor the progression of the disease, compliance with treatment, and preventive care. disease care. A major focus of the California HealthCare Foundation (CHCF) is to improve chronic disease care by standardizing access to, and improving the use of, clinical information at the point of care. This focus is based on the conviction that access to timely, accurate, well-organized electronic data will improve the quality of care for patients with chronic diseases. Many providers have, over the years, promoted systems to improve care for chronic disease patients, ranging from automated disease registries that combine encounter, lab, and pharmacy data to more functional electronic medical records. Combina- tions of technical, logistical, and resource constraints have slowed implementation and deployment of these systems. Obstacles include: ■ Inconsistent data formats, coding schemes, and timing of information delivery in a range of areas, including outcomes, procedures, lab results, pharmacy orders, etc.; ■ Complexity and cost of matching up information about patients that comes from different sources, given the lack of a uniform patient identifier; ■ Incomplete data in existing electronic health information systems, which limits the ability to create useful, consistent “maps” of chronic disease states and their care; and ■ Limited functionality in existing disease registries. On October 16, 2003, CHCF convened an expert workshop to forecast how physician offices, clinics, and group practices will employ clinical information technology (IT) applications to chronic disease care in the next three to five years. Workshop participants included experts from academic and commercial research organizations, medical groups, integrated health care systems, physician organizations, government, and IT vendors. 4 | CALIFORNIA HEALTHCARE FOUNDATION A list of participants is contained in the Appendix. This report presents a summary of the workshop, including the reporting of findings and the facili- tator’s interpretation and forecasts. It includes a description of how clinical information technolo- gies will contribute to chronic disease care in the future and an analysis of the driving forces and barriers to applying two major forms of clinical IT — disease registries and electronic medical records — to chronic disease care. The report also includes a forecast for the next three to five years of the likely pace and trajectory of the adoption of clinical IT applications in three practice settings: solo and small group practices, large group practices, and community clinics. Using Clinical Information Technology in Chronic Disease Care: Expert Workshop Summary | 5 II. Summary of Key Findings The essential characteristics Essential Elements of Chronic Disease Care needed to support chronic The forecasting session started from the premise that chronic care will be an important factor for improving health care disease care are not all quality and controlling costs. Over time, chronic care will be present in existing clinical supported using a full menu of approaches, including "heroic" providers who keep all the information they need in their IT systems. head; manual, paper-based systems; computerized disease registries; and comprehensive electronic medical records with registry components. Whatever the technology used, effective chronic care must include the following functions: ■ Identify who in the population has the disease; ■ Track one or more outcome measures that indicate how well the disease is being managed (and identify which patients and providers are outliers); ■ Track one or more process measures that indicate whether disease management protocols are being observed (and identify which patients and providers are outliers); ■ Prompt the provider to observe the protocols and to take the required measurements and perform the needed actions; and ■ Provide feedback to the provider, group, or plan about overall performance, with a range in reporting from the individual patient level to the aggregate. The expert group described the essential characteristics of the clinical IT systems needed to support those functions. These characteristics are not all present in existing clinical IT systems, and it is essential that they be integrated over the next several years. The vital characteristics include: ■ Support for a full range of chronic care processes. Clinical IT should apply to all aspects of the continuum of chronic disease care. Specifically, it should support outreach to those with a given disease who are missing aspects of their care. It should operate at the point of care, providing both decision support and documenting the clinical encounter. And it should support population profiling and reporting. ■ A patient-focused and longitudinal approach. Disease- focused IT systems are limited in their capacity and usefulness and might contribute to further fragmentation of 6 | CALIFORNIA HEALTHCARE FOUNDATION care. Especially in complex conditions with disease registries and electronic medical records many co-morbidities, such as diabetes and (EMRs) interact today with chronic care processes, heart disease, a focus on the entire patient is how their functions are likely to expand in the essential. Other efforts to monitor and analyze future, and the barriers each must overcome to the chronic care provided to a given population be a high-functioning chronic care tool. or by a given provider group have captured a snapshot view but have not been able to show Disease Registries the evolution of the population’s care over In general, a disease registry is a system to register time. An effective clinical IT system for and track all cases of a given disease or health chronic care must give providers a complete, condition in a specific population. In the context current picture of the patient’s care and of chronic disease care, a disease registry is “a outcomes. Part of a patient focus includes the computer application for capturing, managing, capacity to capture and represent information and providing access to condition-specific about a patient’s self-care behavior. information for a list of patients to support ■ Application across the continuum of care. organized clinical care.” (Jane Metzger, First Related to the previous points is the require- Consulting Group) ment that clinical IT span the continuum Registries typically aggregate information about of care, from the patient’s home to the lab, cohorts of patients who share a clinical disease or pharmacy, office visits, and other sites of care. demographic characteristic. Ideally they include It should ideally be accessible by and useful electronic feeds from laboratories, pharmacies, to the entire care team and not just a single and clinical encounters, though most in use provider. today do not. ■ Real-world applicability. Clinical IT support for chronic care must work in the real world. Facilitator Comment Specifically, it must, as much as possible, draw The utility of a disease registry for chronic care on existing data flows or integrate easily into depends largely on two factors: the completeness existing workflows. It must be “tunable” to fit and timeliness of its data feeds and the richness different conditions, initiatives, and approaches of its reporting functions. At a minimum, disease to chronic care. It should replicate common registries are used to report retrospectively about features of paper medical records, such as the the treatments the population received. As the face sheet, that provide essential information. reporting capabilities expand, so does the utility of registries for chronic care. Important reporting functions include the capability to: Disease Registries and Electronic • “Slice” retrospective reports by provider, group Medical Records as Chronic Care of providers, patient subpopulation, and other Tools factors; • Produce individual patient reports that can be Clinical IT support for chronic care today is not used during patient encounters at the point yet capable of supporting all the characteristics of care; described in the previous section. Current systems • Link those reports to standards of care in order are evolving from two starting points: electronic to suggest tests and treatments; and disease registries and electronic medical records. • Include selected patient outcomes and other (See Figure 1 on the following page.) Neither clinical markers. fulfills all the criteria of “gold-standard” informa- tion support. This section describes how both Using Clinical Information Technology in Chronic Disease Care: Expert Workshop Summary | 7 Figure 1. The Evolution of IT in Chronic Disease Care 8 | CALIFORNIA HEALTHCARE FOUNDATION Few disease registry systems include all these Simple, stand-alone EMRs are designed for functions today, but they will expand their offer- electronic charting. Their main focus is individual- ings in these directions. Several barriers stand in level information. Many of them have proprietary the way of applying disease registries in chronic data formats and don’t adhere to broader data care. These include: standards. To the extent that they have narrative notes, such as progress reports or operative notes, ■ Patient identification. The absence of a they are text-based and may not be searchable. uniform patient identifier makes it difficult They support very little cross-practice analysis. to aggregate information about a single patient from multiple sources, such as a lab, Few newly installed EMRs are stand-alones pharmacy, emergency room, and possibly anymore; instead, most EMRs offer a more several different clinics. structured electronic record. The structured data formats allow a much wider range of cross- ■ Lack of information standards. There practice data analysis. Many functions that are remains a lack of data format and transmission useful for chronic care, such as the automatic standards for laboratory, pharmacy, and, generation of reminders about patient visits, tests, especially, encounter data. To the extent that and so on, are built into these systems. registries must compile data from several organizations, a substantial cost will be For EMRs to be fully functional for chronic associated with integrating multiple coding disease care, they need a higher level of integra- formats until standards are established. tion with other data systems. Useful integration includes electronic feeds from lab and pharmacy ■ Ownership and control of registries. Because information systems and integration with other registries must aggregate information from major practice-management functions, such as several sources, there is concern over who owns patient scheduling, billing, and online ordering. and controls the data contained in the registry. EMRs also require higher-level reporting func- Many providers, for example, do not trust tions, including population-level analysis of the health plans that aggregate patient data. records to identify people with a given condition Existing players who wish to establish a who might not already participate in a chronic registry may not be able to persuade all their care program. Further, more advanced disease business partners — let alone their competitors registries have sophisticated, ad-hoc querying. — to participate, and new, specialized registry The addition of this function further enhances organizations might lack local credibility. the utility of EMRs. Electronic Medical Records EMRs are suited in different ways to support Forecast of Clinical IT for Chronic chronic disease care. In contrast to disease Disease Care in Three Practice registries, which are designed expressly for popu- Settings lation health management (and have concomitant The diffusion of clinical IT to chronic disease shortcomings in their use for individuals), EMRs care will not occur all at once and won’t be are oriented initially toward individual patient uniform across different practice settings. As with charting and management. Many EMRs have other technologies, large and sophisticated group structured data elements, which enable sorting practices will have the financial and technical and reporting about populations. But most are resources to make use of advanced clinical IT not primarily designed to support chronic care applications before smaller practices will. Clinics in a population. that receive special funding for capital investments Using Clinical Information Technology in Chronic Disease Care: Expert Workshop Summary | 9 Facilitator Comment Several barriers stand in the way of adoption of EMRs that fully support chronic care. These include: • High cost of purchase and installation. The process of customizing and applying an EMR is extremely difficult and cumbersome. Few organizations have the money and time available to under- take a major EMR installation. Even those that do will face a dilemma: either “pave the cowpaths” by adapting the EMR to current clinical and business practices — even if they are suboptimal — or use the occasion to re-engineer care delivery and business processes, an expensive and risky course. • Complexity. The more fully functional an EMR, the more difficult it is to diffuse in a medical practice. More functions come with complexity and, of course, cost. Training physicians and others to use a new system is difficult, both technically and culturally. That complexity has stopped many EMR installations in their tracks. • Lack of interoperability. Despite efforts to make EMRs interoperable with each other and with other electronic data flows, many barriers remain. Lack of data coding and transmission standards, unique patient identifiers, and lack of incentives to apply standards impede the integration of data from EMRs and other systems, such as pharmacy and laboratory information systems. This limits the applicability of EMRs for chronic care. will be able to install clinical IT sooner than Solo and Small Group Practices those that must fund all capital improvements Solo and small group practices — with a group internally. This section looks at three practice size of up to 10 people — are the least receptive settings: solo and small group physician practices, market segment for clinical IT applications in large group practices, and community clinics. chronic disease management. They have a For each setting, it provides an assessment of the relatively low penetration of EMRs and other driving forces and barriers to applying disease clinical IT. Adoption of applications for chronic registries and EMRs for chronic care, as well as disease management will likely be limited by the a forecast of likely developments. same barriers. Analysis of Solo and Small Group Practices DRIVING FORCES BARRIERS • Growing incentives for use of IT in chronic care: pay • Fragmentation of solo and small group practice market for performance, Medicare, pay for data, malpractice makes sales, marketing, and technical support insurance discounts inefficient—not vendors’ first target market • Physician and practice profiling by payers and • Small practices don’t have the capital resources to invest independent evaluators that rewards consistent in clinical IT or the human resources to evaluate adherence to chronic care guidelines systems and services; they are under financial and time • Interaction with independent practice associations pressures (IPAs) that contract with solo and small group practices • Investments in IT are hard to justify in the economics in the form of technical assistance, group purchasing, of a small practice training, and incentives • Solo and small group doctors are resistant to the • Availability of clinical IT systems through application possible loss of control over data about their practices service providers (ASPs) that don’t require costly purchases but use monthly service fees 10 | CALIFORNIA HEALTHCARE FOUNDATION Facilitator Forecast Solo and small group practices will come under increasing pressure from the outside to adopt clinical IT for chronic disease care, especially as pay for performance and other incentives are adopted by payers. They will have very strong internal incentives, as well, to adopt better practices for chronic disease care because a relatively large share of their practices (estimated at 70 percent of a small practice’s business) will consist of patients with at least one chronic condition. There will be an increasing number of success stories about small group practices adopting clinical IT, which will be influential among solo and small group practices. That said, there will be little progress in this market segment in the adoption of clinical IT because the fundamental barriers — fragmentation, poor economics of reaching this segment, and lack of investment and human resources — will be difficult to overcome. Market interest will focus much more on larger group practices, unless or until adoption becomes more widespread, prices drop, or user interfaces are made even more intuitive. Large Group Practices Large group practices, especially those with more than 100 physicians, are the most likely to have Facilitator Forecast clinical IT systems already and to have the most This segment will be very diverse in its capacity to apply them to chronic disease care. application of clinical IT to chronic care. Some practices will adopt these applications rapidly Analysis of Large Group Practices and completely, while others will lag. There will be a split between internalists and externalists. DRIVING FORCES The former will tend to create their own clinical • These practices typically have sufficient capital and IT applications based on existing (often self- qualified IT and clinical professionals to evaluate designed) EMRs. Externalists will look for turn- systems and manage their implementation key applications, including some run by ASPs. • The leaders of some large group practices are willing to innovate and experiment • Many practices already have an existing information infrastructure; chronic care applications are incremental • There are success stories of applications of clinical IT to chronic care in this segment • Large practices are the primary target market for clinical IT vendors • Market incentives, including pay for performance, malpractice incentives, and others, fall heavily on large groups BARRIERS • Chronic disease care might not be the primary application priority for large group practices • IT and clinical professionals’ time and attention are limited; other initiatives might take precedence over chronic care systems • It is difficult to change practices throughout large, complex organizations • The fragmented nature of health care delivery and financing limits the case for return on investment (ROI) for IT; economic justification of these investments is commensurately difficult Using Clinical Information Technology in Chronic Disease Care: Expert Workshop Summary | 11 Community Clinics Accelerating Application of Clinical Community clinics are at once a very congenial IT to Chronic Disease Care and a very hostile environment for applying The expert workshop examined several initiatives clinical IT to chronic disease care. On the one being considered by CHCF and added several hand, they have a strong culture of cooperation, other possible initiatives. The experts agreed that information sharing, and adherence to standard the following types of programs (some information formats, as well as a history of using mentioned in the workshop plus additional regional partnerships and other collaborative programs) could substantially accelerate the resources to accomplish large-scale organizational application of clinical IT to chronic disease care. changes. On the other hand, they lack incentives to change their practices. They also (like other ■ Standard formats for data exchange. practice settings) have a diverse installed base of Initiatives are under way in California to information systems in different clinics. standardize lab and pharmacy data exchange. For example, California HealthCare Analysis of Community Clinics Foundation’s California Clinical Data Project DRIVING FORCES (www.chcf.org/topics/view.cfm?itemID=81398) • Access to collaborative resources, such as regional is a collaborative effort of health plans, partnerships, for best practices and technical assistance provider groups, and laboratories to address • A cultural ethos of information-sharing, along with underlying barriers to the access and use of little resistance to standards for uniformity of data integrated clinical data to support chronic formats disease management. The project focuses on • Salary-based compensation helps limit conflicts of interest about the use of capital resources the development and use of data standards to • Availability of capital from Medicare and other facilitate integration of critical lab, pharmacy, government sources and other clinical information. This effort and BARRIERS others in California should be coordinated • Lack of financial and other incentives, such as pay with similar projects being conducted by the for performance Quality Improvement Organizations (QIOs) • Diverse and fragmented installed base of information of the Centers for Medicare and Medicaid systems across clinics Services (CMS). Further, these initiatives • Little experience with clinical data exchange and should be extended to include other types of IT support clinical data such as radiology. Other approaches to facilitate the availability of clinical data are underway that allow a variety of providers to Facilitator Forecast view patient data housed at other institutions Although some community clinics have applied (i.e. allowing a private physician to view clinical IT to chronic care, their success has been variable. Just as several networks of patient data from a recent hospitalization). clinics have conducted common program and ■ Creating community infrastructures for even IT development in the past, they will health information. The Markle Foundation, need to develop and refine a culture and common framework for thinking about how to in its Connecting for Health program use IT for chronic disease care. Government (www.connectingforhealth.org), is creating policy (federal and state) on grants for capital a roadmap of immediate steps for building improvements and federal loan guarantees are a health information infrastructure and the crucial variable in whether community convening working groups and other forums clinics invest in clinical IT for chronic care. to overcome barriers to implementation. 12 | CALIFORNIA HEALTHCARE FOUNDATION A similar mechanism could be established to ■ Development of approaches to activate help solo and small group practices apply consumers. Knowledgeable, active patients clinical IT to chronic care. and consumers are important for the successful care of chronic disease. The challenge is how ■ Guides to help providers integrate clinical to activate patients. So few consumers have IT and chronic care. Many provider encountered clinical IT systems that it’s organizations are willing to consider clinical IT difficult to generalize about what impact applications, especially when they are replacing they might have. Many excellent consumer existing information systems. Physician Web sites exist to support consumers’ self- practices, especially smaller ones, are often management. Others, such as Healthscope, overwhelmed by the complexity of the choices (www.healthscope.org), help consumers select they face. Information and assistance in two health plans, hospitals, medical groups, and areas would be particularly helpful: (1) a insurance plans and include some information buyer’s guide that shows "trajectories" or about chronic disease care. decision trees leading from different IT starting points through to full use of clinical IT for chronic disease care and (2) material that helps link the changes in IT to practice redesign. CHCF has commissioned a series of publications designed to help prospective buyers of systems for chronic disease care, including a disease registry primer, a product review of computerized registries for chronic disease care, and several reports, including a buyer’s guide, on electronic medical records. ■ Collaborative forums on applying clinical IT to chronic care. One of the advantages that community clinics have is their experience with regional collaboratives for IT and other common issues. Improving Chronic Illness Care (www.improvingchroniccare.org), for example, is a program of the Robert Wood Johnson Foundation (www.rwjf.org) based at the McColl Institute for Healthcare Innovation at Group Health Cooperative in Seattle. ICIC has created a Chronic Care Model, which it disseminates through guides for providers, regional collaboratives, and conferences. Using Clinical Information Technology in Chronic Disease Care: Expert Workshop Summary | 13 III. Conclusion Technologies are getting CLINICAL INFORMATION TECHNOLOGY SYSTEMS WILL be essential to improving the quality and consistency of care better each year and for people with chronic diseases. Getting to the next stage — becoming easier to install, putting these technologies into common use across many learn, and use. practice settings, including solo and small physician groups, large medical groups, and community clinics — will require a substantial investment of time and capital on the part of many in the field. Incentives, such as pay for performance and objective measurement of the quality of care, are increasingly in use to encourage providers to adopt clinical IT. The technologies are getting better each year, offering better functions at lower cost, as well as becoming easier to install, learn, and use. Clinical IT systems will evolve from two main starting points: disease registries whose roots are to track populations with a given disease and the care they receive, and electronic medical records whose purpose is to chart patients’ health status and treatment. Health care providers who wish to investigate clinical IT systems, disease registries, electronic medical records, and other ways of improving care for people with chronic disease, are referred to the many resources cited above. The iHealth & Technology topic list on the California HealthCare Foundation’s Web site (www.chcf.org) includes extensive resources on computerized disease registries, EMRs, and other support for chronic disease care. 14 | CALIFORNIA HEALTHCARE FOUNDATION Appendix: Participant List Forecasting the Adoption of Clinical IT in Chronic Disease Management: An Expert Panel October 16, 2003 California HealthCare Foundation Oakland, California Doug Allen, M.D. Jane Metzger Medical Director Consultant Greater Newport Physicians First Consulting Group Thomas Bodenheimer, M.D. Robert Miller, Ph.D. Clinical Professor Associate Professor University of California, San Francisco University of California, San Francisco Eric Brown Jane Stafford Research Director Program Officer Forrester Research Tides Foundation Melissa Buckley Walter Sujansky, M.D. Consultant Principal Sujansky & Associates, LLC Carol Diamond, M.D. Managing Director Kurt Van Riper Markle Foundation Manager, Drug Information Services Kaiser Permanente, Southern California Region Lisa Dolan-Branton HRSA/BPHC FA C I L I TAT O R Division of Health Center Development Health Center Infrastructure Branch Robert Mittman Facilitation, Foresight, Strategy Jonah Frohlich Manager, Reporting and Analysis C A L I F O R N I A H E A LT H C A R E F O U N D AT I O N Brown and Toland Medical Group S TA F F John Haughton, M.D. Sophia Chang Founder Director, Chronic Disease Care Docsite Jan Eldred Michael Hindmarsh, M.A. Senior Program Officer, Chronic Disease Care Manager, Clinical Improvement MacColl Institute for Healthcare Innovation Sam Karp Group Health Cooperative of Puget Sound Director, Health Information Technology David Kibbe, M.D. Thomas Lee Director, Health Information Technology Consultant, Health Information Technology American Academy of Family Physicians Claudia Page Mark Leavitt, M.D., Ph.D. Program Officer, Health Information Technology VP, Clinical Initiatives GE Medical Systems Thomas McDonald, M.D. Internal Medicine Department Palo Alto Medical Foundation Using Clinical Information Technology in Chronic Disease Care: Expert Workshop Summary | 15