Enabling Uptake of a Registry-Supported Care and Learning System in the United States:

Transcription

1 Enabling Uptake of a Registry-Supported Care and Learning System in the United States: A Report to the Robert Wood Johnson Foundation from Karolinska Institutet and The Dartmouth Institute Research Team (in alphabetic order) The Dartmouth Institute (PI: Eugene Nelson) Paul Batalden, MD; Janet Corrigan, PhD, MBA; Wade Harrison, BA, Carolyn Kerrigan, MD, MCHDS; Greg Kotzbauer, BA, Eugene Nelson, DSc, MPH; Aricca Van Citters, MS; Linda C. Wilkinson, MBA Karolinska Institutet (PI: Staffan Lindblad) Elena Eftimovska, MD, MSc; George Keel, MSc; Christina Keller, PhD; Cristin Lind, BA; Staffan Lindblad, MD, PhD; John Øvretveit, PhD Funding from: The Robert Wood Johnson Foundation. Project officers: Susan Mende, BSN, MPH and Laura Leviton, PhD Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States i

2 Executive Summary The Robert Wood Johnson Foundation provided planning grants to Karolinska Institutet and The Dartmouth Institute to assess the potential value of adapting the Swedish Rheumatology Quality (SRQ) Registry for use in the United States. Experts had suggested that the SRQ provided an innovative, replicable, and spreadable system for improving patient engagement, patient-centered care, health outcomes, practice based learning and improvement, and research. This document summarizes the findings of Karolinska Institutet (Part I) and Dartmouth (Part II) research teams; it also suggests next steps to enable American patients and families, interdisciplinary care teams, researchers, payers and others to benefit from this registry-supported care and learning system approach (Part III). Part I. The Swedish Rheumatology Quality Registry (SRQ) and its Approach to Improving Health The SRQ is a Swedish national quality registry that aims to dramatically improve health for people with chronic disease. More than simply a repository for data, this registry-supported care and learning system enables a synergistic approach to the improvement of health, clinical practice, public reporting, and research. It starts with the patient reporting his or her own symptoms, health, and quality of life prior to a visit. These self-reported data are synthesized and graphically displayed to provide a snapshot of the current health status and a longitudinal image of personal health and treatment trends over time. These snapshots are supplemented by the provider, who adds clinical examination and laboratory data. The result is a clinical decision support tool (a dashboard) that aims to allow the patient and the provider to work together to optimize health according to what matters to the patient. The structured data from each visit is immediately exported to the national registry, thereby facilitating collaboration and leveraging the use of the data for improving patient population health. Twenty years of development have produced measurable improvements in health outcomes. Most notably, the inflammatory disease burden in patients with rheumatoid arthritis (mean C-reactive protein in RA) has been reduced by half in the last ten years, and even more so in clinics taking full advantage of the SRQ approach. Clinical practice improvements include the ability to allow patients to self-manage their disease from home and, according to their disease activity, to schedule a doctor or nurse appointment with significantly reduced wait times. With a modest budget and no mandated implementation, the SRQ has spread to 100% of all rheumatology clinics in Sweden and includes 84% of their patients with RA. It is plausible that short-term savings occur and long term value improves by reducing avoidable use of services, faster adjustment of expensive biologic medications to minimize irreversible damage to joints, and increased confidence of patients to manage this and other co-existing chronic diseases. Recognizing the SRQ s potential for improving value, Porter recommended to the Swedish government that the SRQ approach be used as a model for all Swedish registries. 2 Part II: Enabling Uptake and Adaptation in the United States of Elements of the SRQ Approach A confluence of trends suggests that the time is right for adapting elements of the SRQ approach in the United States. There is fast growing recognition that registries are not just for research, but can provide Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States ii

3 direct, real time support to networks of patients, providers, and researchers. The feed forward use of patient-reported outcomes and clinical data is feasible, has been applied in select patient populations, and has generated improved outcomes and better value. Our scan of the United States health care environment is organized around a conceptual model that supports the creation of a learning health system where patients, providers, and researchers partner to co-produce optimal health and high value care. The model highlights a partnership for co-production that Pa;ent'' &'Family' Learning'Health'System' 'For$More$Effec*ve$Ac*on$by$Pa*ents,$Providers,$and$Researchers$ Feed'Forward' PRO'Data'' Personal$Health$Records$ Pa*ent$Facilitated$Networks$$ Partnership'for' CoKproduc;on' Shared'Informa;on'Environment' Registries$ Op;mal'Health'and'High'Value'Care' for'pa;ents'and'popula;ons' Feed'Forward'' Clinical'Data' Provider'&' Care'Team' Electronic$Health$Records$ Collabora*ve$Improvement$Networks$ $2014$Trustees$of$Dartmouth$College$ and$karolinska$ins*tutet$ is supported by a shared information environment which includes patient registries, patient-centered facilitated networks, collaborative improvement networks, and other technologies (see figure). With the right arrangements, registry-supported care and learning systems can: (a) advance meaningful patient engagement in care, self-management, and research, (b) integrate with electronic health records (EHRs), personal health records (PHRs) and clinical workflows, (c) promote practice based quality improvement by providing data which has often been missing from such activities, (d) enhance specialty board maintenance of certification programs that support both continuous professional development and accountability, and (e) provide a data platform for public reporting and research. In short, we found that registries and collaborative networks of a certain type, such as ImproveCareNow for children with inflammatory bowel disease and the High Value Healthcare Collaborative, can bring together patients, clinical teams, health systems, payers, and researchers in a viable and expandable learning health system that focuses on improving the value of services delivered and the outcomes that matter most to patients. Part III. Recommendations for Uptake of a Registry- Supported Care and Learning System in the United States: Proposed Path Forward We recommend that the Foundation sponsor a replication of the Swedish proof of concept to adapt elements of a registry-supported care and learning system approach in the United States. Key mechanisms for change include: data feed forward systems, patient-centered clinical decision support dashboards, meaningful reports available to patients and providers, patient and clinician networks, and multistakeholder learning collaboratives. We suggest starting two complementary programs to support uptake for children and adults with rheumatoid arthritis and with cystic fibrosis, with sponsorship by the American College of Rheumatology and the Cystic Fibrosis Foundation respectively. We aim to design and implement features of the registry-supported care and learning system approach in diverse rheumatology practices and cystic fibrosis centers within five years of the start date and thereby demonstrate proof of concept for replication. Both programs would include patients and families, interdisciplinary care teams, and researchers in leading the design, initial implementation, and evaluation of these efforts. We will facilitate this process by forming a dedicated design and innovation team and Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States iii

4 will engage key leaders from organizations including Dartmouth, Karolinska, Cincinnati Children s Hospital, the American Board of Medical Specialties, the National Quality Registry Network, and other medical professional societies, insurers and industry technology leaders. Conclusion A registry-supported care and learning system can enable higher value healthcare: (a) in the short term, by helping patients and families to be more competent at self-care and providers to give more appropriate and effective care, thereby reducing unnecessary utilization; (b) in the medium term, by providing comparative data on performance for improvement; and (c) in the long term, by enabling comparative effectiveness research through lower cost, real time, real world patient-reported and clinical data collection coupled with other data sources (e.g., claims data), thereby measuring outcomes and costs over time in relationship to different patterns of care. We believe that prospects are promising for successful adaptation of the SRQ approach in the United States, beginning with two important populations and related organizations with strong national recognition. If successful, this would provide a proof of concept of an innovative model to support learning and new approaches to value creation. Through this model, we aim to create replicable and sustainable learning health systems where patients, providers, and researchers partner to co-produce optimal health and high-value care enriched and supported by innovative patient registries, patientcentered facilitated networks, collaborative improvement networks, secure and efficient real time data flows, and value-based payment programs. Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States iv

5 Table of Contents EXECUTIVE SUMMARY... ii PART I. THE SWEDISH RHEUMATOLOGY QUALITY (SRQ) REGISTRY AND ITS APPROACH TO IMPROVING HEALTH: A SUMMARY REPORT FROM KAROLINSKA INSTITUTET... 1 Development of the SRQ Registry Approach Over 20 Years... 1 Description of the Swedish Rheumatology Quality Registry... 2 Patients and Health Care Providers Perceptions of the SRQ Approach... 3 Future and Vision for the SRQ Approach... 5 Conclusion... 5 PART II: ENABLING UPTAKE AND ADAPTATION IN THE UNITED STATES OF ELEMENTS OF THE SRQ APPROACH: SUMMARY REPORT FROM THE DARTMOUTH INSTITUTE... 6 Context and Origins... 6 Patients, Families, and Facilitated Networks... 7 Patient-Professional Partnerships for Co-production of Care... 8 Providers, Interdisciplinary Care Teams, and Patient Registries... 9 Opportunities, Challenges, and Solutions to Advance the Uptake of the SRQ Approach Conclusion PART III: RECOMMENDATIONS FOR UPTAKE OF A REGISTRY-SUPPORTED CARE AND LEARNING SYSTEM IN THE UNITED STATES: PROPOSED PATH FORWARD Recommendation Program Vision Deliverables Proposed Path Forward: Two Complementary Programs for RA and CF Respectively Execution Strategy, Tasks and Timetable Conclusion: Value Proposition for Key Stakeholders LIST OF APPENDICES Appendix 1.1: The Swedish e-health Landscape Surrounding the SRQ Registry Appendix 1.2: Summary of the Evidence Review of the Swedish Rheumatology Registry-Supported Care and Learning Systems Appendix 1.3: Patient and Provider Perceptions of the Swedish Rheumatology Quality Registry (SRQ): A Summary and Analysis of Interview Studies and 2013 Appendix 2: Environmental Scan and Feasibility Assessment for Enabling Uptake of the Swedish Registry-Supported Care and Learning System Approach in the United States Appendix 3: Letters of Support WORKS CITED Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States v

6 Part I. The Swedish Rheumatology Quality (SRQ) Registry and its Approach to Improving Health: A Summary Report from Karolinska Institutet This report summarizes our work to describe and evaluate the Swedish Rheumatology Quality (SRQ) Registry and its approach to improving health as a national effort in Sweden over the past twenty years. This summary builds on a set of appendices that describe our findings and provide important references. Nowadays I work together with my doctor and the computer for every change in the treatment. I can check my chronic disease at home in between visits. Before I go to see my doctor, I measure my own health. The internet support system enables me to contribute to the management of my own disease. I can monitor the effects of drugs, how they help me, and if they failed. I can better understand my doctor s reasoning when we decide on my medication and how I should get the best result. The other day I told my husband it has never been this bad in my disease, but he didn t agree. And we looked together at the data on Multimedia material 1: Karin (English) the internet and realized that I ve been worse before. This helps me get through bad periods by myself since I know I ve managed it before. Karin, RA patient (SRQ 2012). Development of the SRQ Registry Approach Over 20 Years Today, the SRQ is viewed as a model registry for chronic diseases because of its effective approach for improving health. The patient contributes personal data about her health and life situation prior to the clinical visit and thereby shares her concerns with her provider in advance of an encounter. These data provide a visualization of the chronic disease course by tracking evolving outcomes. It displays the patient s response to treatments in full and over time. The displays are then used during the visit by the patient and provider to guide care, and the provider adds clinical and laboratory data to complete the picture. The resulting visual display (i.e., a decision support dashboard) enables shared decision making for evidence-based interventions, such as medication or lifestyle changes, that align with the patient s preferences and choices that can be discussed with her provider. Exporting these data to a central registry forms the foundation for a learning health system giving new knowledge by benchmarking clinics performance, enabling public reporting, and building a database and system for research.1 In fact, Porter believes that the SRQ is pioneering new approaches with patients as partners in measuring the results of their care, and indicates that the SRQ approach represents not only a comprehensive national value measurement system, but should be a central goal for every health care system.2 Despite very limited financial resources and no mandated implementation, the SRQ has been widely adopted by rheumatology clinics in Sweden. Its development over time can be considered a change coming from within clinical care that has leveraged technological opportunities, discursive trends, social and political movements, governmental priorities, and new research streams.3 Thus, we view the SRQ approach as an adaptive system of functional modules, with capabilities extending beyond those for which it was designed. Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States 1

7 Description of the Swedish Rheumatology Quality Registry The Swedish Rheumatology Quality Registry system uses three modules (patient, clinical, and national) that are consistent with health information technology (HIT) and legal infrastructure and act as the essential elements and functionalities benefitting different end-users. (See Appendix 1.1 for a more comprehensive HIT, legal, and organizational description, including screen captures of SRQ modules.) Patient Module: Patient reported outcomes (PROs) relevant for the patient s disease and life situation are captured in the patient module known as PER (Patientens Egen Registrering, or Patient s self Registration) and then visualized to inform and prepare the patient for the clinical visit by showing a graphic display of the disease course and response to treatments. Patients can access the module via a secure national web portal called My Health Care Contacts or through a computer/tablet in the clinic s waiting room. Clinical Module: The clinical module comprises a clinical decision support system and a local SRQ database. It displays the patient data from PER and enables the provider to import these data and to enter their own clinical data, such as clinical examinations and global scoring of disease activity. Data on prescribed medications and current patient outcomes are shown on a dashboard for outcome evaluation and decision-making with the patient. Both physicians and nurses use the clinical decision support during clinical encounters. The clinical module also displays group data to show outcomes for the individual provider s patient panel, or the entire clinic, as a data-based foundation for change initiatives to improve health for patient populations. Box 1 describes how the SRQ has shaped care delivery at four county rheumatology clinics in Gävle. National Module: The data reported from the clinical modules of the rheumatology clinics throughout Sweden flow into a national quality registry that stores individualized data on patients, interventions, and outcomes. The national module includes several feedback systems, displaying data analyses for quality of care assessment, performance benchmarking, public reporting, and industry-commissioned analyses. With IRB approval, data sets can be exported from the national database for research. Box 1: Gävle Rheumatology Clinics: An Example of Innovation Using the SRQ Approach Charged with coordination of rheumatology services across the county of Gävle and its four hospitals in 2008, Dr. Sven Tegmark championed clinical integration of the three SRQ modules. The patients voices are now heard by using the PER module and displaying the PROs and clinical data on the individual disease course as a basis for shared decision making. These data are reported to the national registry as soon as the visit is done. Clinical process improvement: These changes have made it possible to improve clinical processes. Rather than scheduling routine follow- up visits which rarely coincide with disease flare- ups, patients use the patient module to track disease activity from home and request a follow- up when needed (Open access). When the disease is worse the patients are scheduled more frequently to optimize treatments to measurably improve health (Tight clinics). This patient- driven scheduling (Open- Tight clinics) has freed physicians to meet only with patients who need them, and allowed patients asking for an appointment to be seen within two weeks. Outcome improvement: In just one year, the Gävle rheumatology clinics became national leaders. Over the last ten years, the inflammatory disease burden among patients with rheumatoid arthritis has been reduced by half nationally, but in Gävle (3% of the Swedish population) the results are even better with a stable mean below the national value since Cost reduction: By decreasing demand for unnecessary clinical visits, promoting more rational use of expensive biological drugs, and preventing potential adverse drug events, Gävle has also accomplished cost reduction. Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States 2

8 Integration of the SRQ Registry with Other Health Information Systems When a provider is using an electronic health record (EHR) in the flow of patient care, with a single click she can access the SRQ clinical module showing that patient. The provider does not enter any patient data twice and can enter data either into the EHR or the SRQ clinical module. This eliminates the need for double entry, saves time, reduces medical errors, and enables the use of the registry as a real-time decision support system rather than a retrospective audit of care. Currently the SRQ registry is integrated with one of the six leading EHR systems operating in Sweden. This particular EHR system covers 24% of the Swedish population. 1 Laboratory results are made available directly into the SRQ via the EHR system and, in 2013, a pilot integration pulling data from a research biobank was initiated. (See Appendix 1.1, Section 2.2 for more detailed information on HIT integration work and challenges.) Current Status and Major Accomplishments Among the 104 national quality registers in Sweden, the SRQ registry leads in terms of innovative e- health services, patient-centeredness, decision support functionalities, HIT development, and integration. At present, over 60,000 patients are included in the registry and nearly half of them are actively using the patient module to feed forward PROs to the SRQ registry. Service design methods are being used to encourage more widespread use by patients. Health care providers use the real-time decision support functionality of the SRQ registry in 70% of all reported visits. The registry covers the entire specialty and follows 79 adult and pediatric diagnoses. The first diagnosis was rheumatoid arthritis (RA) and now 84% of patients with anti-rheumatic therapy in Sweden are included in the SRQ registry. Significant numbers of patients have two or more (autoimmune) diagnoses and the registry allows their parallel follow up. Since the risk for cardiovascular co-morbidity is increased in chronic inflammatory diseases, the SRQ clinical module incorporates screening, assessment and referral to primary care or specialized care when the need arises. Since 2012, all rheumatology clinics in Sweden (50 public and 10 private clinics) use the SRQ registry and provide data for common national use. Clinical research based on SRQ data has grown continuously and now generates approximately 50 publications per year on clinical population outcomes, health economics, pharmaco-epidemiology, drug safety, and post-marketing follow up. While there has been a limited amount of published research on implementation and uptake of the SRQ, results from patient and provider interviews and case studies show that many value the SRQ approach, and it is probable that the system helps to lower costs and to achieve better health outcomes. Evidence also suggests that the SRQ allows providers to develop more effective clinical services by using real-time and longitudinal data in consultations with patients, and enables a patient-centered approach to care. The review identifies gaps in research, especially about costs, savings and implementation, and highlights how future research into the Swedish experience could fill these knowledge gaps, and could benefit from the 20 years of Swedish experience and availability of data that is already collected. (See Appendix 1.2 for a full review of published literature on the practice and service improvements associated with the SRQ.) Patients and Health Care Providers Perceptions of the SRQ Approach Although all Swedish rheumatology clinics (public and private) use the SRQ registry modules, not all health care providers and patients are adopters and active users of all of its features. Therefore, it is important to understand the context, values, facilitators, and barriers related to adoption of the SRQ Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States 3

9 approach, as perceived by patients and providers. The following section summarizes these perceptions. (See Appendix 1.3 for more in-depth information on patient and provider perceptions of the SRQ.) Patient Perspective A comprehensive patient-centeredness assessment of Sweden s health care system conducted in 2012 showed that although overall health outcomes in Sweden are considered excellent in global comparisons, the paternalistic model of physician-patient relationship still dominates the health care culture in Sweden. 4 In fact, in a 2011 survey of 11 countries, Swedish patients were among those least likely to be engaged by their regular providers and in shared-decision making with specialists. 5 Despite this challenging environment, the SRQ has a reputation in Sweden for being one of the most forward-thinking registries when it comes to partnering with patients. 4 Patients enter their own data not simply as a method for outsourcing work previously completed by professionals, but as a way to facilitate more active participation in decisions about their care during the visit itself. From the beginning, SRQ developers understood that the approach would only work if it could deliver value to patients at the time of input by feeding-back meaningful information at the same time as the information is fed-forward in the patient module. Additionally, the SRQ was the first registry in Sweden to hire a patient to join their management team, help clinics implement the system, and facilitate education on using the SRQ for patient care and practice improvement. Since 2013, patients have been involved as co-designers during a major SRQ e-health service redesign project, focusing especially on enhancing the utility of the patient self-registration module (PER). In terms of patients perceptions, those who appreciate the SRQ approach do so because they see it as a way to gain more insights and knowledge about their disease, structure their self-reported health data, avoid recall bias, and prepare for shared decision-making during the clinical encounter. Patients who are not enthusiastic about using the patient module indicate that they do not know enough about it and do not understand the medical terminology or the rationale for its use. To enable adoption in the latter group, peer-to-peer education, effective communication, and customization of SRQ services are recommended. Health Care Providers Perspective Health care providers in Sweden face challenges that are similar to those faced by providers in the United States. They work in a demanding, accountable, highly disruptive environment with a plethora of HIT systems that seldom communicate with each other. The administrative load and demand for better outcomes are increasing while the time available for actual clinical care and resources decreases. To ensure utility of the registry, its content and modular development have been driven by rheumatology professionals and designed to enable outcomes management, quality improvement, and frontline research. The health care providers that have adopted the SRQ approach do so because they primarily see new opportunities and special uses. Opportunities include enhancing the clinical encounter and the patientprovider dialogue, providing better visualization and structure of the clinical information for decision support, and justifying the targeted use of expensive biological drugs. On the other hand, providers that are currently SRQ non-adopters perceive the registry as an unnecessary administrative burden, a timeconsuming task, a threat to the provider s professional autonomy, and a disruptor to the traditional Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States 4

10 patient-provider relationship (with the HIT system interfering with this relationship). Both adopters and non-adopters point to the general HIT context as a crucial factor for perceived usefulness and adoption success of the SRQ registry. Sweden has more than 70 different EHR systems. Many of these EHR systems and related HIT applications are poorly designed, not workflow-integrated, and are noninteroperable; they become the provider s worst nightmare, reducing efficiency, adding extra time to record keeping and foreclosing fulfillment of potential benefits. Future and Vision for the SRQ Approach Even after 20 years of significant development, the SRQ approach continues to evolve to meet the vision of radically improved health supported by safe, reliable, high value care. For example, in an effort to care for people with lifestyle-influenced chronic illnesses, the SRQ can now combine patient-reported lifestyle factors with research data accessed from biobank samples (e.g., the patients genes and biomarkers) to optimize treatment according to the most recent evidence. An online screening tool for the public or primary care provider is being tested to identify people at high risk for RA, to help prevent the onset of clinical disease. Another service has been piloted allowing patients to order their own laboratory followup tests for potential drug side effects and to view the results. In late 2014, the SRQ will be the first registry to allow patients to download all of their registry data into the new national personal health record (PHR), HealthForMe (HälsaFörMig). The yearly budget for these new developments and additions to the SRQ currently runs at 6 million USD. Other quality registers in Sweden are now emulating and spreading the SRQ approach to improve health, with support from the Quality Register Centre of Stockholm. Conclusion Our evaluation of the SRQ has shown that although a conventional registry is a critical part of the SRQ approach, a database alone is not sufficient to realize the potential for improvement and additional benefits. The SRQ approach begins with patients investing their data, providers working in new ways, and systems that support both. The result is synergistic: patients can gain a dramatically reduced disease burden and improve their self-management competency, providers can deliver care at the front line in ever smarter ways by joining learning networks to share best practices and afford systematic testing of new innovations, health care leaders can create better value health care, researchers can both get better access to patients data and deliver scientific knowledge useful at the front line, and pharmaceutical companies, device manufacturers, and HIT developers can become more knowledgeable about the most effective use of their products in the marketplace. Finally, the approach can create conditions for decreasing unnecessary variation in care, and increasing the likelihood that patients will receive uniform high value care regardless of their geographic location. Overall, our evaluation suggests that the SRQ: (a) is widely accepted and used by patients and providers, (b) can be adopted in very diverse clinical settings and for scores of diagnoses, (c) can be integrated with EHRs, PHRs and other HIT systems, and (d) provides a proof of concept and lessons that may help other registries to advance health care and self-care, public reporting, value-based payments, and research. Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States 5

11 The cumulative effects of the SRQ approach come from an integrated whole system perspective with more properties than may be conceived from studying the particular parts presented in this report. One way to gain similar effects in the United States would be to start with the first building blocks, as we did in Sweden the patients and their providers at the front line and the data they need to improve health. Delivering these data to a common registry is the next step to developing utility for more stakeholders using shared data and developing support systems for improving health and the value of care. Introducing American stakeholders to the adaptive SRQ approach may improve capabilities for both learning and creating novel ways to achieve better health at lower cost in changing and challenging environments. Part II: Enabling Uptake and Adaptation in the United States of Elements of the SRQ Approach: Summary Report from The Dartmouth Institute This section of the report summarizes our work to understand the United States health care context, as it relates to the feasibility of adapting the SRQ approach. A confluence of trends and contextual factors suggest that the time is right for adapting elements of the SRQ registry-supported care and learning system for uptake in the United States. (See Appendix 2 for our complete report.) Context and Origins The Foundation awarded Dartmouth a planning grant to assess the feasibility of adapting the SRQ approach for uptake in the United States. To deliver on the grant we: (a) established a collaborative partnership with Karolinska Institutet s team responsible for developing and evaluating the SRQ, (b) established an Advisory Group to provide input on the project and held meetings in October 2013 and April 2014 (see Appendix 2, Exhibit 2.1, for a list of project advisors), (c) studied the current state of care for chronic health conditions (focusing on rheumatoid arthritis and cystic fibrosis), (d) examined patient registries and other information and communication tools that can support patients and providers and their partnership for the co-production of optimal health and high value care, and (e) identified foreseeable challenges to adapting the SRQ approach and potential solutions for overcoming these challenges. To guide our environmental scan and feasibility assessment, we constructed the model illustrated in Figure 2.1. The model reflects emerging, high impact trends. It aims to build a learning health system where patients, providers, and researchers partner to co-produce optimal health and high value care. It is anchored in the understanding that a patient s health and well-being are interconnected with numerous important factors, wherein interactions with health professionals are but one of many factors that influence well-being. Within this context, the model brings together patients and families (left side) and Figure'2.1:'Conceptual'Model:'Linking'People,'Informa=on,'and'Knowledge'Genera=on'for'BeBer'Health' Aim:$$We$aim$to$build$a$learning$health$ system$where$pa*ents,$providers,$and$ researchers$partner$to$coaproduce$ op*mal$health$and$high$value$care.$ Pa=ent'' &'Family' Learning'Health'System' 'For$More$Effec*ve$Ac*on$by$Pa*ents,$Providers,$and$Researchers$ Feed'Forward' PRO'Data'' Personal$Health$Records$ Pa*ent$Facilitated$Networks$$ Key'Mechanisms:$$We$will$do$this$with$a$registryAsupported$care$and$learning$system$ that$integrates:$data$feed$forward$systems,$pa*entacentered$clinical$decision$support$ dashboards,$meaningful$reports$available$to$pa*ents$and$providers,$pa*ent$and$ provider$networks,$and$mul*astakeholder$learning$collabora*ves$ Partnership'for' CoMproduc=on' Shared'Informa=on'Environment' Registries$ Op=mal'Health'and'High'Value'Care' for'pa=ents'and'popula=ons' Feed'Forward'' Clinical'Data' Provider'&' Care'Team' Electronic$Health$Records$ Collabora*ve$Improvement$Networks$ $2014$Trustees$of$Dartmouth$College$ and$karolinska$ins*tutet$ Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States 6

12 interdisciplinary care teams (right side) and suggests that when they come together (center) they form a partnership for co-producing optimal health and high value care. A shared information environment (lower center) provides value for patients, providers, and researchers by enabling their interaction and offering the ability for 24/7 access to peer support, information, and knowledge generation. For example, certain types of patient-centered facilitated networks can become integrated and integral to the lives of patients and families, whereas certain kinds of patient registries, provider-centered facilitated networks, and collaborative improvement networks can become integrated and integral to the daily workflows of interdisciplinary care teams. The model also shows that patient-reported outcomes (PROs) data and clinical data can use feed forward techniques to enrich the patient-professional partnership and ability to co-produce care. PROs can populate and enhance the provider s electronic health record and the patient s personal health record, and may be used to measure key outcomes in real time and longitudinally. Linking the outcomes data with cost data provides vital information needed to both measure and improve value for individual patients and for populations of patients over time. 6,7 The model suggests that a registry-supported care and learning system approach could, if successful, spin off several powerful benefits: (a) for patients and families (e.g., dashboards for decision support and care monitoring, education and social support, real-time self-management guidance, access to best practices regardless of geography, opportunity to participate in improving care for subsequent patients), (b) for providers and care teams (e.g., dashboards for clinical decision support at the point of care, real-time assessment of quality of care, benchmarking data for practice improvement, data for continuous professional development and maintenance of certification, data for public reporting, participation in a learning health system), (c) for researchers (e.g., a collaboratory 8,9 that includes the research community, typical patients, and regular and diverse medical practices and is supported by a platform for data collection and research on real-world efficacy and safety monitoring), and (d) for payers and consumers by decreasing non-value added services and duplication of tests and workups. The next three parts of this summary expand on the current state of care in the United States, organized by the three main components of the conceptual model: patient and family systems (including patientcentered facilitated networks), patient-professional partnerships for co-production, and provider and care team systems (including patient registries and collaborative networks). We focus primarily on care patterns and trends for rheumatoid arthritis (RA) and cystic fibrosis (CF). (See Appendix 2, Exhibit 2.2, for an overview of care for RA and CF.) Patients, Families, and Facilitated Networks Recognition is growing that people living with a health condition, and their families and friends, are the greatest untapped resource for improving health outcomes and the value of health care. The role of the patient is shifting from being passive to being well-informed and actively involved in shared decisionmaking, self-management, service coordination, and evaluating the benefit of care for themselves in everyday life. People living with complex chronic health conditions often have family and friends that provide physical and emotional support and services (an informal health and social lay care system) that enables the person to care for himself or herself and take full advantage of available health and social services. This represents an often hidden and unpaid care system that is powered by the person with the Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States 7

13 health problem and his or her social network; it has always been available but often not fully recognized and appreciated by the formal medical care system. It is important to note that patient and family competence and confidence in effective self-management can lower the costs of professionally delivered health care services. As payment systems in the United States move from paying for volume to paying for value, it will be increasingly important for providers and payment systems to promote and incentivize the patient s role in tracking health outcomes, coordinating care, and managing their own health conditions. 10 Another high impact, innovative force the IT-enabled, patient-centered facilitated network is bursting onto the healthcare delivery scene and in some instances has shown powerful results. First described by Stabell and Fjeldstad 11 and subsequently popularized by Christensen, 12 the facilitated network model in health care can offer a set of extremely helpful benefits by connecting people and families managing complex health conditions A facilitated network in health care may provide an array of services: peer support, libraries of general information and evidence-based guidance, answers to specific questions from subject matter experts or others who have the same health problem, a personal health record, and connections between willing patients and researchers for rare conditions. It can and should be cocurated, receiving support from both patients and health professionals to make information as accurate as possible given the state of biomedical science and patient experiential knowledge. Some facilitated networks are fast growing commercial enterprises (e.g., PatientsLikeMe ), some are supported by AHRQ (e.g., CHESS 22 ), and some are just getting started with help from PCORI s funding for patientpowered research networks. 23 (See Appendix 2, Exhibit 2.3, for a description of patient-centered facilitated networks, personal health records, and mobile technologies.) Patient-Professional Partnerships for Co-production of Care The research of Nobel Laureate Elinor Ostrom showed how co-production of goods and services by the supplier/producer and citizen/beneficiary could lead to superior results that were more valued by the people and the community, more sustainable and tailored to the culture, and more economical. 24 She observed that co-production can, under some conditions, produce better outcomes at lower costs than producer or beneficiary dominated production. W. Edwards Deming advanced systems thinking by conceiving of suppliers, producers and consumers as all being part of the same system with the aim of meeting fundamental needs of people by using improvement, innovation, and redesign to improve quality and efficiency of service providers to better meet mankind s needs. 25 The insights of Ostrom, Deming, and many others sparked a growing emphasis on empowering and partnering with patients and their families as a fundamental property of routine, high quality care. Patients, families and the interdisciplinary care teams whom they partner with form a clinical microsystem that puts the welfare of the patient at the center of the system. 26 The aim is to generate productive interactions, either in-person or through virtual care assisted by HIT, which enable coproduction of care to generate optimal outcomes at lower costs, and therefore lead to higher value care. Care is co-produced by sharing feed-forward patient generated and clinical data to assess the patient s condition in real time; engage in shared decision making on care plans; execute care plans with both patients and providers fully engaged; follow-up to observe the impact of the care plan and interventions on important results (e.g., disease status, functional status and well being, risk status, care experience, Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States 8

14 self-management competence, direct and indirect costs, etc.); and revise future care plans based on observed results, biomedical science, and the patient s values and preferences. What we see emerging in the SRQ approach, and in some places in the United States, is the intelligent design of clinical programs to leverage microsystem models and methods to promote effective coproduction of care. 6,27 Some examples of this thoughtful redesign of routine care are happening at the Dartmouth Spine Center (e.g., shared decision making and outcomes tracking), 28 at Geisinger (e.g., Patient Centric Electronic Redesign (PACER) for feed-forward use of patient-reported outcomes in rheumatology), and at Ryhov Hospital in Sweden (e.g., self-care renal-dialysis). 33,34 All of these programs are using co-production principles with the goal of measurably improving outcomes while reducing the costs required to produce needed services. Of special interest is the work of Margolis and his colleagues at Cincinnati Children s Hospital who have developed the Collaborative Chronic Care Network (C3N) approach for children and their families living with inflammatory bowel disease. 35,36 C3N represents an innovative learning health system model that includes: (a) patients and families living with a condition, (b) interdisciplinary care teams that serve this population, and (c) researchers focused on improving care for the condition. Evolving from ImproveCareNow, an established clinical registry and collaborative network of pediatric specialists, the model now incorporates a facilitated network approach, is incorporating patient-reported outcomes, uses an enhanced registry to enable the co-production of care by patients and interdisciplinary care teams, and can pull data directly from the EPIC EHR. (See Appendix 2, Exhibit 2.4, for a description of programs that foster co-production of care.) Providers, Interdisciplinary Care Teams, and Patient Registries Medical practice is becoming a team sport characterized by interdisciplinary staff physicians, nurses, mental health and allied health professionals aiming to coordinate evidence-based, patient-centered care across the continuum and preparing for a shift in payments away from volume of services provided to value of outcomes produced. 37,38 This shift in payment systems has important implications for the program that will be recommended in Part III of this report. Interest in patient registries and clinical collaborative networks is exploding and their development should take into account the trends noted above towards team care across the continuum and accountability for outcomes and costs. Some registries and their associated collaborative improvement networks (e.g., Society of Thoracic Surgeons, Northern New England Cardiovascular Group, Cystic Fibrosis Foundation) are well known and have made great progress at measurably improving outcomes, generating research, and publishing results There are assorted types of registries, with some focused on research and process/outcomes tracking and with others focused on improvement through sharing performance data and best practices. 45,46 Many registries are local or regional, but more recently there is a trend toward national registries (such as those sponsored by professional societies). 47 Most registries focus on a specific patient population defined by a single condition, which provides focus and a community of practice, but may jeopardize understanding whole person care. Of special interest is the High Value Healthcare Collaborative (HVHC), which includes 20 health systems, serves 100 million patients, and focuses on six high impact conditions. 48 Coordinated by Dartmouth, the HVHC aims to identify care models that Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States 9

15 produce the best outcomes at lowest costs and to spread this knowledge to its members and the public. It uses claims and administrative databases that cover all patients and all conditions and is beginning to use PROs and feed forward techniques, along with shared decision making programs, to demonstrate better value care for the six conditions. Other important developments have catalyzed great interest in using patient registries to improve health care and outcomes; these include: (a) the work of the American Board of Medical Specialties (ABMS) and the American Board of Internal Medicine (ABIM) to make participation in certified registries a mechanism to support continuous professional development and attain maintenance of certification (MOC), (b) new federal policy to enable qualified clinical data registries to meet requirements for public reporting (e.g., PQRS), and (c) the formation of the National Quality Registry Network (NQRN) to promote the growth of effective registry principles and practices. (See Appendix 2, Exhibit 2.5, for a description of registry and collaborative improvement network activity in the United States.) Features of the SRQ approach are of interest to American providers, patients, and registry leaders. For example, leaders of the American College of Rheumatology (ACR) are prepared to adapt their registry to incorporate elements of the SRQ approach as they seek to integrate data collection with workflows and populate the registry from data pulled from EHRs. The ACR s evolving registry includes over 500 providers and enables practice improvement, continuous professional development, and PQRS reporting. Likewise, the Cystic Fibrosis Foundation (CFF) is prepared to adopt key elements of the SRQ, including using feed forward patient generated data in the flow of care. The CFF has a well-established web-based registry and collaborative improvement network that engages over 120 interdisciplinary CF centers. Opportunities, Challenges, and Solutions to Advance the Uptake of the SRQ Approach Table 2.1 summarizes a selection of opportunities to improve care, challenges associated with these opportunities, and solutions developed over time by the SRQ that might be used to overcome challenges. In some instances, solutions are directly applicable, while in others they may need to be adapted to fit the United States context. For example, the SRQ approach provides solutions that can support the ability of: Patients and families to benefit from shared decision making, self-management support, and receiving evidence-based care no matter where they live; Interdisciplinary care teams to benefit from real time decision support, based on feed forward tracking of treatments and outcomes; Researchers to gather accurate, longitudinal data that reflect the care processes, outcomes and costs of care based on typical patients served by mainstream clinical programs; Purchasers of care and the public to know the outcomes and costs of care produced by different providers caring for similar patient populations; and Practices to join a learning system based on meaningful comparative data and learning from every patient, reducing unnecessary variation, and rapidly introducing new knowledge into routine practice. Summary Report: Enabling Uptake of a Registry-Supported Care and Learning System in the United States 10