Designing a High-Performing Health Care System for Patients with Complex Needs Ten Recommendations for Policymakers The Commonwealth Fund International Experts Working Group on Patients with Complex Needs Expanded and Revised Edition September 2017 The Commonwealth Fund International Experts Working Group on Patients with Complex Conditions Elias A. Mossialos, M.D., Ph.D., FFPH, FRCP Co-chair London School of Economics and Political Science (United Kingdom) Robin Osborn, M.B.A. Co-chair The Commonwealth Fund (United States) Martin Roland, CBE, DM Editor Cambridge University (United Kingdom) Melinda Abrams, M.S. The Commonwealth Fund (United States) Volker Amelung, Ph.D. German Managed Care Association (BMC) (Germany) Ran D. Balicer, M.D., Ph.D., M.P.H. Clalit Research Institute and Clalit Health Services (Israel) Malcolm Battersby, MBBS, Ph.D., FRANZCP, FAChaM Flinders University (Australia) Gerard Anderson, Ph.D. Johns Hopkins University Bloomberg School of Public Health (United States) Mats Bojestig, M.D. Jönköping County Council and Höglandssjukhuset (Sweden) Reinhard Busse, Dr. med., FFPH Technische Universität Berlin (Germany) Isabelle Durand-Zaleski M.D., Ph.D. Hôpital Henri Mondor (France) Naoki Ikegami, M.D., Ph.D. Keio School of Medicine (Japan) Gregg Meyer M.D., M.Sc. Partners HealthCare (United States) Mark Pearson Organization for Economic Co-operation and Development (France) Mieke Rijken, Ph.D. Netherlands Institute for Health Care Services Research (NIVEL) (The Netherlands) Samir Sinha, M.D., D.Phil., FRCPC Mount Sinai/University Health Network Hospitals and University of Toronto (Canada) Staff Martin Wenzl, M.Sc. Research Director London School of Economics and Political Science Dana Sarnak, M.P.H. Research Associate The Commonwealth Fund 2 Designing a High-Performing Health Care System for Patients with Complex Needs Health care costs are heavily concentrated among people with multiple health problems. Often, these are older adults living with frailty, advanced illness, or other complex conditions. In 2014, the New York–based Commonwealth Fund, a private, independent foundation, established the International Experts Working Group on Patients with Complex Needs through a grant to the London School of Economics and Political Science. The group’s purpose was to outline the prerequisites of a high-performing health care system for “high-need, high-cost” patients and to identify promising international innovations in health care delivery for meeting needs of these patients. Drawing on international experience, quantitative and qualitative evidence, and its members’ collective expertise in policy and program design, implementation, and evaluation, the international working group sought to articulate the principles that underpin high performance for this complex population in health systems around the world. What follows are the group’s top recommendations based on these principles. All 10 present challenges, with some requiring profound paradigm shifts—for instance, away from disease-specific care delivery and toward more patient-centered approaches, or away from the single-provider model and toward cooperation and teamwork. Their implementation, however, has the potential to transform care and quality of life for millions. The selected international models that follow the recommendations represent some of the promising frontline care innovations that illustrate the principles laid out here. RECOMMENDATION 1 Make care coordination a high priority. Because patients with complex needs receive treatment from a wide range of providers, their care often becomes fragmented. This can result in more hospitalizations and lower patient satisfaction. What these patients need is a dedicated person who is responsible for coordinat- ing all their care. This could be the patient’s primary care physician, but increasingly health care organizations are employing staff specifically tasked with coordinating treatment for complex patients. Although better coordination should lead to better care, it will less often save money. It is therefore especially important to identify and remove financial disincen- tives to care coordination. RECOMMENDATION 2 Identify patients in greatest need of proactive, coordinated care. Several methods have been developed to identify patients with complex needs. Generally they use data drawn from medical records, sometimes supplemented by professional judg- ment. The aim is to identify patients at risk for poor outcomes, such as unnecessary hospital admissions, and provided targeted, proactive, team-based care. While a number of validated models exist to predict patients’ health care utilization and costs, individual countries will likely need to adapt these models based on the types of data they have available. www.commonwealthfund.org 3 RECOMMENDATION 3 Train more primary care physicians and geriatricians. In most OECD member countries, the number of subspecialists has increased at a much higher rate than the number of generalists. This trend has led to fragmented care and needs to be reversed. To meet the needs of aging populations, more family physicians and geriatri- cians, in particular, will be needed. Medical school curricula and training programs should be altered to support this shift. RECOMMENDATION 4 Facilitate communication between providers—for example, through clinical record integration. It is important that providers treating a patient with complex needs are able to share import- ant data about that patient; this ensures clinicians have the information they need, when they need it. Ideally, this is accomplished by a single electronic record for all the patient’s medical care. Also critical is good and timely provider communication, including the prompt relay of information to the primary care physician following hospitalization and specialist visits and the sharing of care plans with after-hours and emergency services. RECOMMENDATION 5 Engage patients in decisions about their care. For the patient with multiple health conditions, treatment that adheres to evidence-based guidelines for each individual condition can lead to an unacceptable burden of treatment, adverse interactions between treatments, and risks from polypharmacy. Patients with com- plex conditions need to be part of an open discussion of the benefits and risks of individual treatments. Such a process allows them to bring their own needs, preferences, and hopes into the treatment conversation. RECOMMENDATION 6 Provide better support for caregivers. Elderly people and those with complex needs often receive care from family members and friends. They are usually unpaid and often provide support around-the-clock. Health services need to take steps to identify and support these informal caregivers. Support might include re- spite care to provide relief for caregivers and assistance to help them look after their own health. RECOMMENDATION 7 Redesign funding mechanisms to meet patients’ needs. Current funding mechanisms and payment incentives often exacerbate the problems of frag- mented care. For example, fee-for-service encourages the overprovision of specialist services; capitation- and salary-based payments may lead to undertreatment; and quality incentives tend to prioritize only those aspects of care that are most easily measured. Payments systems for complex patients need to be redesigned so that they reduce barriers to collaboration, adequately compensate for the complexity of cases treated, and incentivize hospitals to work with community providers. 4 Designing a High-Performing Health Care System for Patients with Complex Needs RECOMMENDATION 8 Integrate health and social services, and physical and mental health care. The separation of health and social care fails to recognize some patients’ closely related needs for both types of care. Constrained social service spending may also lead directly to inefficient use of health care resources—for example, when patients are unable to be discharged from the hospital because of a lack of support available in the community. Care for patients with complex needs therefore requires close cooperation between the two sectors. Failure to integrate physical and mental health care also causes problems for patients with complex needs. Care for mental health must be integrated with physical health care, with multidisciplinary teams ensuring that physical and mental health problems are addressed together in a timely fashion. RECOMMENDATION 9 Engage clinicians in change and train and support clinical leaders. Implementing these recommendations will challenge notions of professional autonomy, established beliefs, and engrained ways of working. Clinical leadership is key to delivering successful change, and the clinicians leading change need support from local managers to ensure that local administrative systems and budgetary arrangements do not stifle change. Clinicians may also benefit from formal leadership training and opportunities to meet with peers on a regular basis. RECOMMENDATION 10 Learn from experience and scale up successful projects. Different solutions will suit different environments. Policymakers and health care managers should provide opportunities for sharing experiences and learning from success as well as failure. It is important to understand that successful projects tend to develop iteratively over time—and sometimes over a long period. www.commonwealthfund.org 5 Selected Profiles of Care Models for Patients with Complex Needs CANADA Mount Sinai Hospital Acute Care for Elders (ACE) Strategy LOCATION Toronto, Ontario, Canada losing independence. ACE strategy integrates these interventions to create seamless, interprofessional, YEAR ESTABLISHED 2010 technology-enabled integrated team-based delivery model spanning the care continuum. BACKGROUND Mount Sinai Hospital developed a comprehensive, integrated approach to improve care for INFORMATION SYSTEMS Geriatricized order sets and hospitalized older adults and older adults at high risk care protocols to support safer evidence-based care; of hospitalization, particularly because of functional, tracking systems to monitor flow of ACE patients cognitive, social, or other problems. throughout Mount Sinai Hospital in real time and support timely transfer to ACE unit; secure e-mail OBJECTIVE To improve the delivery and quality of notification and flagging systems to allow primary care, care, patient and system outcomes in all older patients, home care, emergency, and inpatient care providers to and those older patients at especially high risk of poor communicate effectively; and risk identification tools outcomes. (ACE Tracker) to support early identification of high- PATIENTS TARGETED All patients age 65 and older risk patients. admitted with an acute medical condition. High-risk FINANCING AND PAYMENT METHODS patients are identified in emergency department Usual funding through global block payments for (ED) based on having any three or more of: 1) recent hospitals and other community-based agencies. decline in functional abilities; 2) recent change in Physicians paid fee-for-service; other professionals are cognition or behavior; 3) geriatric syndrome (e.g., falls, salaried. Hospital budget structures create incentives to incontinence, acute or chronic pain); 4) complex social reduce admissions and length of stay. No model-specific issues; or 5) Identification of Seniors at Risk (ISAR) incentives. score ≥2. Complementary community-based programs also identify and support high-risk patients. Program EVALUATION METHODS Ongoing quarterly performance enrolled approximately 10,500 patients between 2010 tracking system, using balanced scorecard and regional and 2015. benchmarking to identify areas for improvement. Pre/post implementation comparisons. KEY FEATURES AND INTERVENTIONS ISAR screening for all older ED patients, with additional EVALUATION RESULTS Comparing pre-implementation support from geriatric emergency management nurses. and post-implementation periods, there was 53 percent High-risk medical patients are prioritized to be cared for overall increase in annual admissions of patients age under Acute Care for Elders (ACE) protocol and, when 65 and older within Toronto’s fast-growing population possible, by designated ACE inpatient medical unit. All (due to trend of increasing ED visits). Mount Sinai has older patients have access to hospitalwide consultation maintained region’s lowest admission rate of older liaison services in geriatrics, psychiatry, and palliative patients—25 percent, 18 percent lower than regional medicine and to volunteer-based Hospital Elder Life admission rate. For those admitted to hospital, there Program (HELP). All professionals are educated in was 28 percent decrease in mean length of stay; 13.4 geriatric care. Additional models strengthen community percent decline in readmissions; reduction in “alternate care and improve care transitions; Integrated Client level of care” (“bed blocker”) days per patient of 20 Care Program provides intensive care coordination percent; and increase in patients discharged directly to for targeted high-risk/high-use patients, while home. Average direct cost of care per patient reduced by community outreach teams provide short-term home- 23 percent, and general inpatient medical beds reduced and community-based supports to patients at risk of by 18.2 percent. SOURCES Personal communication with Samir Sinha. S. K. Sinha, S. L. Oakes, S. Chaudhry et al., “How to Use the ACE Unit to Improve Hospital Safety and Quality for Older Patients: From ACE Units to Elder-Friendly Hospitals,” in M. L. Malone, E. Capezuti, R. M. Palmer, eds., Acute Care for Elders—A Model for Interdisciplinary Care (Springer, 2014) S. K. Sinha, J. Bennett, T. Chalk, “Establishing the Effectiveness of an Acute Care for Elders (ACE) Strategic Delivery Model in Delivering Improved Patient and System Outcomes for Hospitalized Older Adults,” Journal of the American Geriatrics Society (2014), 62:S143 6 Designing a High-Performing Health Care System for Patients with Complex Needs Selected Profiles of Care Models for Patients with Complex Needs ENGLAND Early Supported Discharge (ESD) for Stroke Patients LOCATION England Providing rehabilitation in patients’ homes ensures that the process is patient-centered and adapted YEAR ESTABLISHED 1993 to needs of patients and their informal caregivers, thereby increasing self-efficacy and providing a smooth BACKGROUND After a stroke, patients may need transition. prolonged rehabilitation, traditionally provided in inpatient settings. The designers of the early supported INFORMATION SYSTEMS No specific system. Data on discharge (ESD) model hypothesized that rehabilitation the quality of care are collected through the Sentinel could be more effectively delivered in patients’ homes, Stroke National Audit Programme. shortening length of hospital stays and making rehabilitation more responsive to patients’ needs. FINANCING AND PAYMENT METHODS ESD is financed by Clinical Commissioning Groups (i.e., local payer OBJECTIVE To improve continuity of care by organizations of the National Health Service [NHS]). supporting transition from inpatient to home-based ESD can also be financed from savings from reduced stroke rehabilitation and improve cost efficiency by length of hospital stays. Professionals are salaried shortening length of hospital stays. employees of NHS providers; there are no financial incentives for providers. PATIENTS TARGETED Patients requiring stroke rehabilitation who are sufficiently mobile. EVALUATION METHODS Several randomized controlled trials were published internationally, as well as meta- KEY FEATURES AND INTERVENTIONS Patients are analysis and a cost-effectiveness model. assessed for rehabilitation needs before discharge to set initial objectives and ensure continuity of care. Upon EVALUATION RESULTS Evaluation of the first hospital discharge, patients are visited at home within implementation of ESD in England showed improved 24 hours by the therapy team and receive needed daily patient satisfaction, reduced length of hospital stays, physiotherapy, occupational therapy, and speech and resulted in small cost savings. It did not find therapy for up to six weeks. Other social services are significant differences in health outcomes. Meta- provided as usual. Each patient receives an individual analysis of 14 randomized controlled trials from care plan, which is reviewed at a weekly team meeting. Australia, Canada, Denmark, Norway, Sweden, Thailand, the United Kingdom, and the United States There is variation across England in the composition found a reduction in long-term dependency and and leadership of rehabilitation teams, as well as their admission to institutional care, as well as reducing operational policies and the way in which they interact the length of hospital stay. Meta-analysis also found with referring hospitals during discharge planning. improvements in extended activities of daily living All teams involve stroke specialists, including doctors, scores and patient satisfaction. No significant effects nurses, physiotherapists, and occupational and speech were found in mortality, hospital readmissions, or therapists. Many teams also include or provide access to caregiver-reported health status, mood, or satisfaction. psychologists and social workers. Of the trials that evaluated costs, six found ESD services to show cost savings compared with the control group; one found cost increases. SOURCES N. Chouliara, R. J. Fisher, M. Kerr et al., “Implementing Evidence-Based Stroke Early Supported Discharge Services: A Qualitative Study of Challenges, Facilitators, and Impact,” Clinical Rehabilitation, April 1, 2014 28(4):370–77. R. Beech, A. G. Rudd, K. Tilling et al., “Economic Consequences of Early Inpatient Discharge to Community-Based Rehabilitation for Stroke in an Inner-London Teaching Hospital,” Stroke, April 1999 30(4):729–35. P. Fearon and P. Langhorne, “Services for Reducing Duration of Hospital Care for Acute Stroke Patients,” Cochrane Database of Systematic Reviews, Sept. 12, 2012 (9):CD000443. A. G. Rudd, C. D. Wolfe, K. Tilling et al., “Randomised Controlled Trial to Evaluate Early Discharge Scheme for Patients with Stroke,” BMJ, Oct. 25, 1997 315(7115). O. Saka, V. Serra, Y. Samyshkin et al., “Cost-Effectiveness of Stroke Unit Care Followed by Early Supported Discharge,” Stroke, Jan. 2009 40(1):24–29. www.commonwealthfund.org 7 Selected Profiles of Care Models for Patients with Complex Needs ENGLAND Reconfiguring Stroke Care in London LOCATION London, England Nurses and doctors are trained in a simulation unit. Paramedics also receive training. Units are expected YEAR ESTABLISHED 2010 to engage in regular and continued professional development. BACKGROUND Stroke is the third-highest cause of death and most common cause of adult disability in high- INFORMATION SYSTEMS Hospitals operate their own income countries. Well-organized care by specialized information technology systems. All units participate stroke units can reduce mortality and disability. Poor- in the Sentinel Stroke National Audit Program, the quality stroke care led the London Primary Care Trusts data source for quality of care for stroke treatment in to form a joint committee, supported by a panel of England. expert clinicians, other health professionals, and lay members, to develop evidence-based and centralized FINANCING AND PAYMENT METHODS Hospitals are paid stroke services. through National Health Service case-based payments, at a “best practice” rate for stroke, which includes an OBJECTIVE To improve health outcomes by providing a additional sum for each patient linked to achievement uniform and high-quality standard of care for all stroke of rigorous standards of care. An estimated 9 million patients in London. British pounds in capital investments were made to develop stroke units and an additional 23 million British PATIENTS TARGETED All patients hospitalized with pounds per year were needed to support the model. stroke, except children. EVALUATION METHODS Effects on health outcomes in KEY FEATURES AND INTERVENTIONS Eight specialized London were evaluated, using the rest of England as a hyper-acute stroke units (HASUs) and 24 stroke units control group. Effects on process measures and costs with colocated transient ischemic attack assessment were evaluated in pre–post intervention comparisons. services provide centralized care. HASUs provide faster response times when a stroke is suspected and EVALUATION RESULTS continuous access to specialist care throughout the first 72 hours. Specialized nurses and medical teams assess • Average length of hospital stays and risk-adjusted and treat patients from the time of hospital admission. mortality at three days, 30 days, and 90 days after HASUs are accessible to the entire London population admission were reduced compared to the rest of by ambulance within 30 minutes. England. Stroke units provide ongoing inpatient care as necessary • Cost savings were achieved through lower rates of after 72 hours. All units are staffed by doctors, nurses, admissions to intensive care units, fewer admissions physiotherapists, and occupational, speech, and to long-term nursing home care, and reduced need for language therapists; most also have psychologists. The social supports in the community. model requires regular multidisciplinary team meetings and goal setting. A service manager oversees the unit. • Since its inception, this centralized model of stroke care has been maintained nearly unchanged, with evidence showing sustained high quality of care. SOURCES B. D. Bray, S. Ayis, J. Campbell et al., “Associations Between the Organisation of Stroke Services, Process of Care, and Mortality in England: Prospective Cohort Study,” BMJ, May 10, 2013 346(f2827). R. M. Hunter, C. Davie, A. G. Rudd et al., “Impact on Clinical and Cost Outcomes of a Centralized Approach to Acute Stroke Care in London: A Comparative Effectiveness Before and After Model,” PLoS One, Aug. 1, 2013 8(8):e70420. S. Morris, R. M. Hunter, A. I. G. Ramsay et al., “Impact of Centralising Acute Stroke Services in English Metropolitan Areas on Mortality and Length of Hospital Stay: Difference-in-Differences Analysis,” BMJ, Aug. 5, 2014 349(g4757). Stroke Unit Trialists’ Collaboration, “Organised Inpatient (Stroke Unit) Care for Stroke,” Cochrane Database of Systematic Reviews, Sept. 11, 2013 (9):CD000197. J. M. Wardlaw, V. Murray, E. Berge et al., “Thrombolysis for Acute Ischaemic Stroke,” Cochrane Database of Systematic Reviews, July 29, 2014 (10):CD000213. 8 Designing a High-Performing Health Care System for Patients with Complex Needs Selected Profiles of Care Models for Patients with Complex Needs FRANCE Personalized Health Plan for Elderly at Risk of Autonomy Loss (PAERPA) LOCATION Nine administrative regions across France. INFORMATION SYSTEMS A secure e-mail system facilitates communication and web-based systems YEAR ESTABLISHED In nine pilot regions across France provide information to patients and professionals. in 2013–14; seven additional regions in 2016 in order Although special legislation permits data sharing among to ensure full deployment on the French territory and members of care teams, medical records are not yet provide PAERPA coverage to a total of 550,000 persons widely shared. BACKGROUND French residents age 75 or older are FINANCING AND PAYMENT METHODS National 9 percent of the population but accrue 22 percent of funding for information systems, coordination units, health expenditures. The High Council for the Future financial incentives, and additional services. Regional of Health Insurance identified several shortcomings in Health Authorities fund pilot projects through social elder care: lack of referral pathways from primary care security and have autonomy in funding local variations. to specialist physicians, burdens on informal caregivers Providers are paid as usual. In addition, an incentive resulting in “respite” hospital admissions, hospitals’ of €100 per elderly patient with a personal care plan lack of attention to geriatric patients’ needs, insufficient is shared between primary care physicians and other availability of health care professionals in nursing involved professionals. Long-term care institutions homes, capacity shortfalls in nursing homes delaying receive €53 per day for temporary residents in transition hospital discharge, regional disparities in the availability out of hospital. of social support and personal care services, and lack of coordination between hospitals and social service EVALUATION METHODS Pre/post comparisons of facilities. process indicators (e.g., number of personal care plans created) are mandatory in local implementations. OBJECTIVE To improve frail elders’ quality of life, better Qualitative and health economic evaluations are under coordinate their care, and reduce caregiver burdens. way at the national level. PATIENTS TARGETED Those age 75 or older who: live EVALUATION RESULTS Not yet available. in long-term care facilities; are admitted to hospital via emergency departments; are frail; take certain prescription drugs; or have one or more chronic condition. Eligibility for a personal care plan is assessed by a primary care physician or care coordinator. Across regions, 6 percent to 14 percent of elders were enrolled. KEY FEATURES AND INTERVENTIONS Some features, including eligibility criteria, vary by region. Common features include systematic identification of those at risk; education for elders and their caregivers; professional education on frail elders’ needs; personal care plans; and interventions to reduce the risk of falls. Integrated health and social services are provided through mobile geriatric teams; respite facilities for informal caregivers; telemedicine; a fast-track application for welfare benefits; and temporary stays in long-term care facilities to facilitate transitions from hospital to home. Nurse coordinators coordinate the work. SOURCES Personal communications with Marie-Aline Bloch, Sebastien Gand, and Elvira Periac. Cour des comptes, Le maintien à domicile des personnes âgées en perte d’autonomie (2016) Y. Bourgueil, J.-B. Combes, N. Le Guen et al., Atlas des territoires pilotes PAERPA—Situation 2012 (IRDES, 2015) Ministère des Affaires sociales et de la Santé, Le dispositif Paerpa 2016, http://social-sante.gouv.fr/systeme-de-sante-et-medico-social/parcours-des- patients-et-des-usagers/le-parcours-sante-des-aines-paerpa/article/le-dispositif-paerpa www.commonwealthfund.org 9 Selected Profiles of Care Models for Patients with Complex Needs ISRAEL Comprehensive Care for Multimorbid Adults Program (CC–MAP) LOCATION Israel Patients are encouraged to involve their informal caregivers. Formal social care services, which are YEAR ESTABLISHED 2012 separately financed and delivered by social welfare services, are not fully integrated. CC–MAP nurses help BACKGROUND Patients with multiple chronic patients access social services. conditions (multimorbidity) require proactive, coordinated care management to effectively manage INFORMATION SYSTEMS Clalit operates an integrated their numerous health conditions. Researchers from information system that centralizes all administrative, Clalit Health Services and the University of Haifa, electronic health, and demographic data. This platform with the support of the Gertner Institute, created the allows for algorithmic identification of high-risk Comprehensive Care for Multimorbid Adults Program patients, sharing of information among providers (CC–MAP) to address this issue. (across primary, specialty, and inpatient services), streamlining care processes, and monitoring outcomes OBJECTIVE CC–MAP aims to improve the quality of care and processes, such as medication adherence and use of and reduce preventable hospital admissions for adult preventive services. Clalit members with multiple morbidities who are at risk for deteriorating health status and incurring high costs. FINANCING AND PAYMENT METHODS CC–MAP is currently cofinanced by Clalit and the Gertner Institute. PATIENTS TARGETED Adults with multiple morbidities, The only additional resource are the CC–MAP nurses, defined as three or more chronic diseases, and who are who are salaried. Primary care physicians continue at risk for deteriorating health status, as defined by a to receive their usual salaried payment. There are no validated risk prediction score in primary care clinics financial incentives for professionals, and participation that serve the largest percent of multimorbid patients. is voluntary by informed consent. KEY FEATURES AND INTERVENTIONS The intervention EVALUATION METHODS Clustered controlled trial. is overseen by CC–MAP nurses, who work with primary care physicians to provide comprehensive EVALUATION RESULTS Preliminary results comparing care management for 100 to 150 of the highest-risk 12-month follow-up of 600 patients in the intervention patients in each targeted clinic. Nurses and primary versus 600 control patients indicated a 40 percent care physicians receive tailored training and have access reduction in hospital days (average of -2.3 days per to a set of supportive practice tools developed for the patient) relative to baseline. Additionally, quality intervention. of chronic care, quality of life, and the performance of daily activities (such as shopping and medication Main components of the intervention include a management) were significantly higher in patients comprehensive assessment of patient and family enrolled in the intervention compared to controls. needs; formulation of a coordinated care plan based on integrated care guideline summaries; an action plan for patients; caregiver support including self-management education; proactive monitoring of patients’ personalized goal attainment; and coordination of care from all providers including follow-up on institutional transitions. SOURCES Authors’ communication with: Efrat Shadmi, Ph.D., Clalit Health Services, Health Policy and Planning, Chief Physician’s Office, Tel Aviv, Israel; and Ran D. Balicer, M.D., Ph.D., Clalit Research Institute, Tel Aviv, Israel. 10 Designing a High-Performing Health Care System for Patients with Complex Needs Selected Profiles of Care Models for Patients with Complex Needs ISRAEL Clalit Readmission Prevention LOCATION All general hospitals and Clalit Health FINANCING AND PAYMENT METHODS Financed by Services primary care clinics across Israel Clalit Health Services. Providers are paid as usual; COC nurses are salaried Clalit employees. COC nurses YEAR ESTABLISHED 2011 represent the main additional investment; the program employs 14 full-time nurses across 27 hospitals. There BACKGROUND Reducing readmissions is a focus of are no additional financial incentives for providers or health care systems worldwide to improve quality of professionals. care and efficiency. Evidence points to the importance of in-hospital interventions that address patient needs EVALUATION METHODS No control group. Ongoing early to prevent unplanned hospital readmissions. quality monitoring (objective and patient-reported) over time provides benchmarks (e.g., readmission rates, OBJECTIVE Develop and implement an ongoing strategy post-discharge primary care visits, patient-reported to prevent 30-day hospital readmissions among high- quality of post-discharge care). risk elderly patients insured by Clalit Health Services. EVALUATION RESULTS Readmissions rate declined by 4 PATIENTS TARGETED All Clalit members, ages 65 and percent to 5 percent on average; up to 15 percent in non- older, admitted to hospitals. A prediction algorithm severely ill patient subgroups. Rate of contacts within (the Preadmission Readmission Detection Model seven days after discharge with primary clinic nurses or PREADM) uses electronic medical record and has increased since the implementation of the program administrative data to derive a risk score and identify to over 85 percent. The PREADM predictive algorithm high-risk patients. exhibits good predictive accuracy. KEY FEATURES AND INTERVENTIONS The PREADM risk score is used by continuity of care (COC) nurses stationed in every hospital in Israel to target high-risk patients. COC nurses provide in-hospital coordination, discharge planning, and coordination with primary care clinic nurses for post-discharge follow-up and monitoring. Electronic messaging between nursing staff in hospital wards and general practices is used to facilitate joint discharge planning. Primary care clinics are responsible for post-discharge follow-up and monitoring, performed by nurses at the clinics according to structured outreach protocols. The PREADM score is used in all primary care clinics to prioritize outreach efforts to high-risk patients within 72 hours of discharge. INFORMATION SYSTEMS Clalit operates an integrated information system that centralizes all electronic health and demographic patient data. This platform allows for identification of high-risk patients, sharing of information among providers, and periodic collection of patient data for monitoring. Additional systems include a platform for automated electronic messaging between hospitals and primary clinics and a post- discharge assessment tool that notifies primary clinics of admissions and discharges and facilitates discharge and post-discharge activities. SOURCES E. Shadmi, N. Flaks-Manov, M. Hoshen et al., “Predicting 30-Day Readmissions with Preadmission Electronic Health Record Data,” Medical Care, March 2015 53(3):283–89. E. A. Coleman, J. D. Smith, J. C. Frank et al., “Development and Testing of a Measure Designed to Assess the Quality of Care Transitions,” International Journal of Integrated Care, published online June 1, 2002. www.commonwealthfund.org 11 Selected Profiles of Care Models for Patients with Complex Needs THE NETHERLANDS U–CARE and U–PRIM LOCATION Primary care centers in Utrecht province INFORMATION SYSTEMS A software application (U– of the Netherlands PRIM) identifies potentially frail older patients using available routine care data in the electronic medical YEAR ESTABLISHED 2009 records of general practices and provides a periodic report to primary care physicians. BACKGROUND Providing improved care for the increasing number of older people with complex care FINANCING AND PAYMENT METHODS Incremental cost needs is a major challenge. A greater focus on proactive of the combined intervention in the initial study was primary care could help older people maintain their approximately 130 euros per patient (100 euros for GP independence and prevent functional decline. payment and 30 euros for U-PRIM software). The initial study was funded by the Netherlands Organisation OBJECTIVE Preserve daily functioning and maintain for Health Research and Development, which also independence in older adults. subsidizes current implementations. No financial PATIENTS TARGETED People aged 60 and older, incentives were provided to GPs. independently living in the community and potentially EVALUATION METHODS Three-arm cluster randomized frail (i.e., those with multiple chronic conditions or controlled trial of U–PRIM and U–CARE vs. U-PRIM vs. taking multiple medications or those who are not usual care with one-year follow-up. Accompanying cost- receiving regular primary care). effectiveness analysis and qualitative surveys of patients KEY FEATURES AND INTERVENTIONS The intervention and providers. uses algorithmic screening of routine primary care EVALUATION RESULTS Small improvement in physical data (U–PRIM), followed by personalized care (U– functioning in both intervention groups compared with CARE). U–CARE is delivered by trained practice nurses the control after one year but no overall benefit of U– in cooperation with general practitioners (GPs), and CARE in addition to U–PRIM. No effects on mortality, includes a frailty assessment of patients followed quality of life, and satisfaction with care. Increased by a comprehensive geriatric assessment for those number of general practice consultations in the U–PRIM identified as frail. Based on the assessments, nurses and U–CARE group vs. other groups, but no effect on create a tailored care plan and provide evidence-based hospital admissions or emergency department visits. geriatric care, care coordination, and follow-up visits. Coordination spans physical and mental health care, The combination of U–PRIM and U–CARE was found to including transitions to and from hospitals. Also have a moderate to high probability of cost-effectiveness includes social care by working closely with social in the Netherlands, due to decreased lengths of hospital workers and other professionals, like physiotherapists, and nursing home stays. Providers reported improved occupational therapists, and dietitians. Each nurse is cooperation but also challenges due to time constraints responsible for an average of approximately 70 patients. and a lack of financial compensation. Guidelines for geriatric conditions were developed through literature review and expert consultation. They provide decision support for nurses and are integrated into care plans. Nurses completed 70 hours of training, followed by ongoing support. SOURCES N. Bleijenberg, I. Drubbel, M. J. Schuurmans et al., “Effectiveness of a Proactive Primary Care Program on Preserving Daily Functioning of Older People: A Cluster Randomized Controlled Trial,” Journal of the American Geriatrics Society, Sept. 2016 64(9):1779–88. N. Bleijenberg, V. H. ten Dam, I. Drubbel et al., “Development of a Proactive Care Program (U-CARE) to Preserve Physical Functioning of Frail Older People in Primary Care,” Journal of Nursing Scholarship, Sept. 2013 45(3):230–37. I. Drubbel, Frailty Screening in Older Patients in Primary Care Using Routine Care Data (Utrecht University [Netherlands], 2014). M. Soeters and G. Verhoeks, Analyse belemmeringen structurele bekostiging vier NPO-projecten (ZorgmarktAdvies, 2016). 12 Designing a High-Performing Health Care System for Patients with Complex Needs Selected Profiles of Care Models for Patients with Complex Needs SPAIN Integrated Care Model for Complex Cases and Strategy for Chronic Care LOCATION Valencia region, Spain and family preferences. Other resources may be applied, depending on the clinical and social complexity and YEAR ESTABLISHED 2007 acuity of the case. Primary care physicians and their teams lead implementation of the plan. Both nurse care BACKGROUND Policies were developed nationally and managers remain jointly responsible for monitoring the in the Valencia region to respond to an aging population patient, interacting with professionals, and ensuring and the rising prevalence of chronic disease and to appropriateness of care. reorient health care from acute episodes to chronic disease management. HNCMs and CNCMs attend 100 hours of training and a month of on-the-job training. Other professionals receive OBJECTIVE To improve care for complex patients with ongoing training related to care integration and care for multiple morbidities. complex cases. PATIENTS TARGETED Patients with complex chronic INFORMATION SYSTEMS An information system was diseases or in need of palliative care. Electronic medical implemented in the Valencia region. Each patient records (EMRs) are used to stratify the population has a unique identifier; care providers use the system monthly into clinical risk groups (CRGs) and identify to share patient information through EMRs. Data high-risk patients. generated by hospitals is currently being integrated KEY FEATURES AND INTERVENTIONS Integrates into the information system. The system is also used hospital, primary, and community-based health services, for identifying high-risk patients and monitoring their including hospital-at-home units and social workers, conditions and drug use. under a single management in each of the 24 health FINANCING AND PAYMENT METHODS Financed by departments of the region. Social care, which is financed the region of Valencia through its ordinary health care separately, is not formally integrated. budget. All staff are salaried. There are no financial Newly introduced hospital nurse case managers incentives for providers or staff. (HNCMs) and community nurse care managers EVALUATION METHODS The model was not formally (CNCMs) have joint responsibility for complex cases. evaluated in terms of effectiveness or cost-effectiveness, HNCMs identify complex cases at hospitals and are but pre- and post-trends in outcomes were published. responsible for planning hospital discharge to ensure continuity of care. CNCMs are responsible for organizing EVALUATION RESULTS: Reduced emergency department the collaborative care process in the community visits and hospital admissions. and arranging home care. This process starts with a comprehensive assessment of the patient, his or her current informal care, and the environment. This is conducted by a multidisciplinary team that includes CNCMs. It covers medical conditions, medications, accessibility of the home, hygienic conditions, dependency levels, mental conditions, and use of technology. After mapping patient needs, HNCMs draft a care plan and medication review adapted to patient SOURCES Authors’ communication with Barbabella, Hujala, Quattrini et al.; and Gallud, Soler, and Cuevas. F. Barbabella, A. Hujala, S. Quattrini et al., The Strategy for Chronic Care in Valencia Region (Estrategia para la atención a pacientes crónicos en la Comunitat Valenciana), Spain (ICARE4EU, 2015). J. Gallud, P. Soler, and D. Cuevas, “New Nursing Roles for the Integrated Management of Complex Chronic and Palliative Care Patients in the Region of Valencia (Nuevos perfiles enfermería para el manejo integral de pacientes crónicos complejos y paliativos en la Comunidad Valenciana),” International Journal of Integrative Care, April–June 2012 12(Suppl. 2):e24. www.commonwealthfund.org 13 Selected Profiles of Care Models for Patients with Complex Needs SWEDEN Esther Model LOCATION Jönköping County, Sweden improvement projects in the frontline; they are not paid extra—their work as coaches is seen as part of YEAR ESTABLISHED Late 1990s their jobs. BACKGROUND Elderly patients with complex care needs INFORMATION SYSTEMS A “virtual competence center” receive services from multiple specialists, as well as from is used to transmit knowledge to practitioners along primary care physicians, resulting in fragmented care. the care chain. It was supported by a substantial grant In addition, they may have frequent hospitalizations (12 million kronor in 2006, about $1.5 million) to and receive long-term care services at their home or in provide two years of training for members of the model nursing facilities. in systems-thinking, communication, information technology development, medication management, OBJECTIVE Using the negative experiences of an elderly telephone advice, and documentation. Individual patient, known as “Esther,” the program’s founder Mats professionals can sign up for online workshops on topics Bojestig, began to focus on creating a persona, Esther, such as dementia or palliative care. and asking: “What is best for Esther?” Doing so allows caregivers to focus on the needs, preferences, hopes, and FINANCING AND PAYMENT METHODS Budget was 1.8 concerns of patients. million Swedish kronor ($300,000) in 2011, which covered the salary of the coordinators, education of the PATIENTS TARGETED Elderly patients with complex coaches, and new improvement projects. The current care needs budget comes from the Jönköping County Council KEY FEATURES AND INTERVENTION The Esther and covers meeting expenses and coach education. model uses continuous quality improvement, cross- Coordinators are paid from their home organizations’ organizational communication, problem-solving, and budgets. staff training to provide the best care for elderly patients EVALUATION METHODS Results must be interpreted with with complex care needs. Features of the model include: caution as the program was not designed as a research • A steering committee made up of the community care project; no evaluation specific to Esther was conducted. chiefs from each municipality, the chiefs of geriatrics EVALUATION RESULTS In the Höglandet Hospital where and medicine at the hospital, and the heads of some Esther was implemented: primary care centers. • Four “Esther cafés,” take place in municipalities each • Admissions to the medical department (for all year; these are cross-organizational, multiprofessional patients, not only patients 65 and older) of Höglandet meetings for sharing and learning from the Hospital declined from 1998 to 2013; hospital days experiences of patients who have been admitted to in the medical and geriatric ward declined from 2002 the hospital in the past year and currently receive to 2013. However, similar changes were reported home care or other services. elsewhere in Sweden. • Inter-organizational training workshops on selected • Hospital readmissions within 30 days for patients age topics, including wound healing, palliative care, 65 and older dropped from 2012 to 2014. nutrition, fall prevention, and care planning. • Hospital lengths of stay decreased between 2009 and • Esther coaches: clinical and administrative staff 2014 for surgery and rehabilitation. members (not managers) from all the participating organizations; coaches include nurse assistants, nurses, • Surveys conducted in 2008 and 2011 showed that physical and occupational therapists, social workers, Esthers felt safe and were appreciative of the personal and administrators; coaches are expected to support contact. SOURCES N. Vackerberg, The Esther Approach to Healthcare in Sweden: A Business Case for Radical Improvement (Governance International, 2014). B. H. Gray, U. Winblad, and D. O. Sarnak, Sweden’s Esther Model: Improving Care for Elderly Patients with Complex Needs, case study (The Commonwealth Fund, Sept. 2016). 14 Designing a High-Performing Health Care System for Patients with Complex Needs Selected Profiles of Care Models for Patients with Complex Needs UNITED STATES Commonwealth Care Alliance “One Care” Program LOCATION Massachusetts, United States FINANCING AND PAYMENT METHODS CCA receives a risk-adjusted, per member per month, capitated YEAR ESTABLISHED 2003 blended payment from both Medicare and the state Medicaid program. CCA then bears full financial risk for BACKGROUND Adults under age 65 who are eligible for the total cost of care, including long-term services and both Medicaid and Medicare are a particularly vulnerable supports, acute and postacute care, pharmaceuticals, group, with complex and often overlooked needs. and primary care. Given the complexity and cost of OBJECTIVE To provide enhanced primary care and CCA’s beneficiaries, these payments are substantial: In care coordination for dually eligible Medicare and 2015, CCA received $386 million from the Medicaid Medicaid beneficiaries through multidisciplinary teams and Medicare programs, and $273 million for the 15 that include physicians, nurses, and mental health months ending in December 2014. The state’s Medicaid and geriatric specialists, and to generate savings from contribution ranges from a few hundred dollars per reduced hospitalizations and institutional care. member per month for relatively healthy patients to $9,000 or more for patients with extended stays at long- PATIENTS TARGETED Under Massachusetts’ One Care term care facilities. The base rate for Medicare Part A/B demonstration, Commonwealth Care Alliance (CCA) capitation payments are in the range of $770 to $960 per provides coverage to more than 11,000 dually eligible, member per month. nonelderly beneficiaries—the majority of state of residents enrolled in the demonstration. Roughly 80 EVALUATION METHODS A pre/post study of 4,500 CCA percent have multiple chronic health conditions, mental One Care enrollees, without control group. health problems, or functional limitations due to EVALUATION RESULTS Enrollees had 7.5 percent fewer physical and developmental disabilities. hospital admissions and 6.4 percent fewer emergency KEY FEATURES AND INTERVENTIONS Interdisciplinary department visits than in the prior 12 months and care teams—nurse practitioners, physician assistants, greater use of long-term services and supports. A behavioral health and addiction clinicians, social majority of enrollees said they were satisfied with workers, community health workers, and others— the program. A preliminary analysis found that use assemble around medically complex patients, helping of inpatient facilities and inpatient psychiatric days to identify their unmet medical, behavioral health, decreased. and social service needs and deploying resources using flexible benefits. Individualized care plans, developed by clinicians and members, guide resource allocation for long-term care, durable medical equipment, behavioral health services, and other key components. Care delivery innovations, including a community paramedicine program and community psychiatric respite facilities, shift care from acute settings into the home and the community (where appropriate), at lower cost. INFORMATION SYSTEMS Web-based and shared electronic medical record. SOURCES Commonwealth Care Alliance (2016), www.commonwealthcarealliance.org/about-us/history S. Klein, M. Hostetter, and D. McCarthy, The “One Care” Program at Commonwealth Care Alliance: Partnering with Medicare and Medicaid to Improve Care for Nonelderly Dual Eligibles (Commonwealth Fund, Dec. 2016), http://www.commonwealthfund.org/publications/case- studies/2016/dec/commonwealth-care-alliance. www.commonwealthfund.org 15 Selected Profiles of Care Models for Patients with Complex Needs UNITED STATES The Hospital at Home (HaH) Program LOCATION Originally at Johns Hopkins University; • More recent versions also provide robust follow-up there are now versions of the model at Presbyterian and transitional care through 30 days post-discharge. Hospital (N.M.), Mount Sinai Hospital (N.Y.), Centura Health (Colo.), Trinity Health (Mich.), Kaiser Permanente INFORMATION SYSTEMS None specific to the model, (Calif.), Brigham and Women’s Hospital (Mass.), but hospital systems with electronic medical records Massachusetts General Hospital, Cedars Sinai Medical (like Veterans Affairs and Presbyterian) experience more Center (Calif.), Marshfield Clinic (Wis.), and several seamless transitions. Veterans Affairs Medical Centers across the United States. FINANCING AND PAYMENT METHODS Currently no YEAR ESTABLISHED Mid-1990s payment codes for HaH care in fee-for-service Medicare. Thus, implementation of the model has been limited BACKGROUND Patients often are more comfortable to Medicare managed care and Veterans Affairs health receiving care in a familiar home environment. For the systems. An application has been submitted to the frail and elderly, hospital stays can pose a variety of Centers for Medicare and Medicaid Services Physician- health threats, including delirium, infections, and falls. Focused Payment Model Technical Advisory Committee Hospitals also have high fixed costs. to obtain a payment mechanism for the model in Medicare. OBJECTIVE Provide hospital-level, potentially acute care in a patient’s home EVALUATION METHODS Various, mostly comparing patients in the program with comparable patients who PATIENTS TARGETED Hospital at Home (HaH) treats stay in the hospital for care. patients who would otherwise be admitted as inpatients and who meet validated clinical-appropriateness criteria; EVALUATION RESULTS patients must have housing where care can be provided • Compared to similar hospitalized patients, HaH safely and within 30-minute travel time. patients experience better clinical outcomes including lower rates of mortality, delirium sedative medication KEY FEATURES AND INTERVENTIONS use, and restraints • Robust input from physicians, nurses, and home health • Better satisfaction for patient and family, less caregiver aides, who provide daily and intermittent visits and stress, and better functional outcomes 24-hour coverage; providers also assess risk at home, • Cost savings of 19 percent to 30 percent compared to develop patient-centered care plans, and engage traditional inpatient care patients and family in managing care. • Lower average length of stay • Patient retains an “inpatient” status, with the hospital or health system responsibile for the acute • Fewer lab and diagnostic tests compared with similar care episode; care components such as intravenous patients in acute hospital care treatment, durable medical equipment, oxygen A recent evaluation found that patients in the therapy, skilled therapies, diagnostic tests, imaging Presbyterian HaH program had reduced costs (20% studies, and pharmacy support are provided. compared to traditional care) and equal or better • Coordinated continuum of care is similar to inpatient outcomes than comparable hospital inpatients; HaH care; illness-specific hospital-at-home care maps, with patients had slightly lower hospital readmission and clinical outcome evaluations and discharge criteria as mortality rates and almost 10 percent higher satisfaction used in hospitals. scores than comparable patients. SOURCES B. Leff, “Defining and Disseminating the Hospital-at-Home Model,” Canadian Medical Association Journal, Jan. 20, 2009 180(2):156–57. Johns Hopkins Healthcare Solutions, Hospital at Home Success Stories (Johns Hopkins Medicine, n.d.). B. Leff, L. Burton, S. L. Mader et al., “Hospital at Home: Feasibility and Outcomes of a Program to Provide Hospital-Level Care at Home for Acutely Ill Older Patients,” Annals of Internal Medicine, Dec. 6, 2005 143(11):798–808. Health Policy Monitor, Hospital at Home (Bertelsmann Stiftung, n.d.). S. Klein, M. Hostetter, and D. McCarthy, The Hospital at Home Model: Bringing Hospital-Level Care to the Patient (The Commonwealth Fund, Aug. 2016). L. Cryer, S. B. Shannon, M. Van Amsterdam et al., “Costs for ‘Hospital at Home’ Patients Were 19 Percent Lower, with Equal or Better Outcomes Compared to Similar Inpatients,” Health Affairs, June 2012 31(6):1237–43. 16 Designing a High-Performing Health Care System for Patients with Complex Needs Selected Profiles of Care Models for Patients with Complex Needs UNITED STATES Massachusetts General Care Management Program LOCATION Massachusetts, United States • Comprehensive orientation program for nurse care managers, who also receive training to conduct patient YEAR ESTABLISHED 2006 assessments, create comprehensive care plans, arrange for referrals to various services like transportation, and BACKGROUND To provide an enhanced level of care use information systems. to high-­ isk patients using comprehensive, outpatient, r practice­-based case management. INFORMATION SYSTEMS Real-time messages sent to nurse care managers on patient hospitalizations; OBJECTIVE Improve quality of care and outcomes, electronic medical records and advanced clinical and reduce cost for Medicare beneficiaries, improve the administrative information systems for Massachusetts quality of work life of primary care physicians, and General Hospital providers; clinical dashboards, using attract more physicians to the field of primary care. data from electronic medical records, claims data, and PATIENTS TARGETED Medicare beneficiaries who: 1) enrollment tracking database, allow Mass General receive their care from a Massachusetts General primary Hospital to examine trends in health care utilization care provider, 2) reside in one of five counties in Eastern and outcomes. Massachusetts, 3) do not meet exclusion criteria, and 4) FINANCING AND PAYMENT METHODS Monthly Medicare meet inclusion criteria based on annual health care costs fee of $120 per patient, together with a requirement to and a risk assessment algorithm. achieve savings of at least 5 percent. Savings of less than 5 KEY FEATURES AND INTERVENTIONS percent accrue to Medicare, savings in the 5 percent to 10 percent range go to Mass General. Mass General Hospital • Primary-care based model with reliance on provided physicians with $150 in financial incentive per information technology and real-time data. patient per year to help cover the cost of these activities. • Customized services to fit patients’ needs, including EVALUATION METHODS Independent evaluations by end-of-life management, psychological and social Research Triangle Institute and the Congressional Budget evaluations and interventions, management of home- Office using difference in differences methodology to-hospital transitions, inpatient/outpatient mental compared to control population. health program, and pharmacy consultations. EVALUATION RESULTS Among the 87 percent of • Care managers conduct assessments using a tool eligible beneficiaries enrolled, there was high patient developed by Massachusetts General Hospital that and physician satisfaction; hospitalization rate includes several externally validated instruments; among enrolled patients was 20 percent lower than a questions cover challenges encountered with activities comparison group. In addition, enrolled patients had of daily living, among other topics. lower emergency department visit rates, lower annual mortality, and cost reductions compared with the • Using assessment, case managers develop a care plan comparison group. for each patient in conjunction with the primary care provider and the practice’s clinical team. • Case managers conduct home visits on an as-needed basis. SOURCES T. G. Ferris, E. Weil, G. S. Meyer et al., “Cost Savings from Managing High Risk Patients,” Chapter 9: Care Culture and System Redesign, in The Healthcare Imperative: Lowering Costs and Improving Outcomes: Workshop Series Summary (P. L. Yong, R. S. Saunders, and L. Olsen, eds.), Institute of Medicine Workshop Series: Roundtable on Value and Science-Driven Healthcare (National Academies Press, 2010), 301–10. N. McCall, J. Cromwell, and C. Urato, Evaluation of Medicare Care Management for High Cost Beneficiaries (CMHCB) Demonstration: Massachusetts General Hospital and Massachusetts General Physicians Organization (MGH): Final Report (RTI International, Sept. 2010). L. Nelson, Lessons from Medicare’s Demonstration Projects on Disease Management and Care Coordination, Working Paper 2012-01 (Health and Human Resources Division, Congressional Budget Office, Jan. 2012). C. S. Hong, M. K. Abrams, and T. G. Ferris, “Toward Increased Adoption of Complex Care Management,” New England Journal of Medicine, Aug. 7, 2014 371(6):491–93. www.commonwealthfund.org 17