LESSONS FROM LOCAL ACCESS INITIATIVES: CONTRIBUTIONS AND CHALLENGES Karen Minyard, Deborah Chollet, Laurie Felland, Lindsey Lonergan, Chris Parker, Tina Anderson-Smith, Claudia Lacson, and Jaclyn Wong August 2007 ABSTRACT: Community health initiatives—locally crafted responses to health care access problems—have been steadfast in their efforts to connect uninsured and medically indigent people to health care services and health insurance. These programs assist in outreach, coordinate and integrate care, and help clients use limited resources efficiently. This report offers five case studies of community health initiatives. All five local community initiatives seek to improve access and coverage for those most likely to be uninsured: low-income, nonelderly adults. Some, like Community Health Works in Forsyth, Ga., offer coverage for a limited period of time, often for individuals who seek care after contracting an illness, while others, like Choice Regional Health Network, in Olympia, Wash., manage care for clients with complex needs, chaperoning them through systems they characteristically have trouble navigating. Support for this research was provided by The Commonwealth Fund. The views presented here are those of the authors and not necessarily those of The Commonwealth Fund or its directors, officers, or staff. This and other Fund publications are available online at www.commonwealthfund.org. To learn more about new publications when they become available, visit the Fund’s Web site and register to receive e-mail alerts. Commonwealth Fund pub. no. 1050. CONTENTS About the Authors.......................................................................................................... iv Acknowledgments .......................................................................................................... vi Executive Summary....................................................................................................... vii Introduction .................................................................................................................... 1 Methods .......................................................................................................................... 1 Local Initiative Contributions and Challenges .................................................................. 2 Site Highlights ................................................................................................................. 2 Success Factors and Barriers ............................................................................................. 5 Challenges of Replicating Local Initiatives ....................................................................... 7 Policy Implications .......................................................................................................... 8 Case Studies: Five Local Experiences Near Perfect Replication: Project Access.................................................................. 11 The Power of Local, State, and National Financial Leverage: General Assistance Medical Program (GAMP) .................................................... 16 State Context Matters: Community HealthLink ....................................................... 20 Measuring Outcomes: Community Health Works ................................................... 24 Chaperoning the System: Choice Regional Health Network ................................... 29 Notes............................................................................................................................. 33 LIST OF FIGURES Figure 1 Percent of the Nonelderly Population in the Study States with State Public Coverage or Employer-Sponsored Coverage, 2002.................... 21 Figure 2 Emergency Room Visit Rate by Membership Tenure in Community Health Works....................................................................... 26 iii ABOUT THE AUTHORS Karen J. Minyard, Ph.D., is executive director of the Georgia Health Policy Center at Georgia State University, which focuses on long-term care, child well-being, health philanthropy, rural health, cancer, and access to care for the uninsured. Dr. Minyard is an advocate for restructuring local health care systems to focus on access to care and health status improvements, and an officer on the founding board of Communities Joined in Action, a national partnership dedicated to helping networks achieve health care access. She regularly consults with and presents to groups and organizations that seek to build health and access. Dr. Minyard is the former director of the Networks for Rural Health Program at the Health Policy Center, where she conducted an external evaluation of Georgia Medicaid’s primary care case management program. Her research interests include the study of health care markets; the relationships among identity, image, and willingness to make strategic changes in changing environments; and the role of targeted external facilitation and technical assistance in improving local health systems’ clinical relevance and financial viability. Deborah Chollet, Ph.D., is a senior fellow at Mathematica Policy Research, Inc. in Washington, D.C., and a national expert on state health insurance markets and public/private-sector health care reforms. She has led numerous studies on individual and small-group health insurance markets; the impact of state regulation and policy actions on insurance markets and coverage; the cost, financing, and economic feasibility of major state reforms of public programs and private insurance; and the effects of state high-risk insurance pools on health insurance coverage. Chollet is a senior consultant to the Robert Wood Johnson Foundation State Coverage Initiatives program and a frequent consultant to state officials on health insurance market dynamics and policy options for expanding private health insurance coverage. Laurie E. Felland, M.S., is a health researcher at the Center for Studying Health System Change in Washington, D.C. In her research, she has focused on low-income people’s access to medical care, public insurance coverage, and safety net providers, as well as public health issues. Felland has also worked as a policy analyst at the Massachusetts Division of Health Care Finance and Policy and a redesign analyst at Regions Hospital in St. Paul, Minn. She earned a bachelor’s degree in economics from Carleton College and a master’s degree in health policy and management from the Harvard School of Public Health. Lindsey Lonergan, M.H.S., is a research associate at the Georgia Health Policy Center. In this capacity, she provides technical assistance to communities in the process of building iv networks to improve access and enhance residents’ health status. Her professional interests include Medicaid case management and building leadership capacity in communities. She has been a health policy analyst at the Center for Health Services Financing and Managed Care in the Department of Health and Human Services, where she reviewed Medicaid managed care waivers, demonstration plans, and State Children’s Health Insurance Program state plan amendments; and developed and managed special projects on managed care, managed behavioral health care, the uninsured, cultural competence, and disease case management. She earned a bachelor’s degree in biology from Mercer University and a master’s degree from the Johns Hopkins University Bloomberg School of Public Health. Chris Parker, M.B.B.S., M.P.H., is a research associate at the Georgia Health Policy Center. He has managed research studies related to cancer initiatives, health care access for the uninsured, health care disparities, and public health in Georgia. His earlier experience includes serving as a health care provider and manager for a faith-based clinic and conducting a primary care practice in Kingston, Jamaica. He earned his medical degree at the University of the West Indies and a master’s degree in public health at the Rollins School of Public Health at Emory University. Tina Anderson Smith, M.P.H., is a senior research associate at the Georgia Health Policy Center. Smith oversees the conceptual, resource, and strategic development of the Center’s research, policy, and technical assistance programs. She has also served as director of the Center’s community health systems development program. Smith has been involved in efforts supporting access to care for rural citizens, including Networks for Rural Health, a program to provide technical assistance and facilitation for rural communities attempting to develop and sustain local and regional health networks and partnerships. Her academic and professional experience includes hospital business development, public health program design, grassroots policy research, economic evaluation of existing programs, and community development. She makes national presentations and provides technical assistance to leaders in other states to support rural communities in improving the health of residents. Smith earned a master’s degree in public health from the Rollins School of Public Health at Emory University. Claudia Lacson, M.D., M.P.H., was a research associate and community health systems developer at the Georgia Health Policy Center. She provided technical assistance to state and federally funded networks seeking to improve access to care and health status of populations. Before joining the Center, Lacson worked at the South Carolina Department of Health Migrant Health Program, consulting to state agencies and community organizations on community outreach, cultural competence, and Hispanic health issues. v She also worked as a rural physician in Colombia, where she provided primary care services to medically underserved populations and directed public health initiatives on prenatal care, immunizations, and family planning. (Dr. Claudia Lacson died on July 31, 2004.) Jaclyn Wong is a senior programmer/analyst at Mathematica Policy Research. She has worked on projects targeting a range of social policy issues, with a primary focus on qualitative and quantitative analysis of health policy. Ms. Wong earned a bachelor’s degree in human biology at Stanford University. ACKNOWLEDGMENTS The authors thank the senior staff of The Commonwealth Fund for their creative energy throughout this project. They urged us to focus on beta sites instead of alpha sites and therefore to investigate more systematically how innovations are transferred. We also thank the local and state stakeholders who generously shared their insights, especially: Anne Nelson in Wichita, Kan.; Bob Redford and Kendall Poe in Ratcliff, Ark.; Paula Lucey in Milwaukee, Wis.; Kristen West and Anna Shelton in Olympia, Wash.; and Greg Dent and Shannon Harvey in Forsyth, Ga., for the extensive information they provided and their assistance in coordinating our site visits. Lei Zhang and Karen Fink patiently coded many pages of interview notes, and Lei Zhang investigated and summarized the literature on the replication and transfer of innovation. Vondie Woodbury (director of the Three-Share Plan in Muskegon, Mich., an alpha site), Alan McKenzie (executive director of the Buncombe County Medical Society and founder of Project Access, an alpha site), Ann Wiess and Terri Stoller (Robert Wood Johnson Foundation’s Communities in Charge Program), Marguerite Johnson (Kellogg’s Community Voices Program), and Rick Wilk and Susan Lumsden (HRSA Healthy Communities Access Program) provided advice on the research questions, recommended potential case study sites, and identified relevant types of key informants. Editorial support provided by Deborah Lorber. vi EXECUTIVE SUMMARY The United States health care system is plagued with at least three serious problems: rising costs, deterioration of the health care safety net, and inadequate public and private health insurance. With 47 million people uninsured, comparative health statistics that rank the U.S. below other industrialized nations, and wide performance variations within the country, it is clear that bold strategies—requiring public and private resources from national, state, and local levels—are essential. Community health initiatives—locally crafted responses to health care access problems—have been steadfast in their efforts to connect uninsured and medically indigent people to health care services and health insurance. Typically, they unite community leaders, providers, and other key stakeholders, building on good-faith relationships to reduce uncompensated care and support the local safety net. Recognizing that merely referring people with complex medical and social needs to care is often insufficient, these initiatives assist in outreach, coordinate and integrate care, and help clients use limited resources efficiently. Previous research has catalogued and described individual community efforts, evaluated the results of specific funding programs, explored how local efforts can substitute for national or state programs, and examined the role of local efforts in changing national policy. This study offers new insights about community initiatives and the successes and challenges they face. Findings fall into the following three areas: 1) the critical importance of state context; 2) the need for community health initiatives and, paradoxically, the difficulty of sustaining them; and 3) the challenges of replication. The research team developed case studies of five community health initiatives that seek to improve access and coverage for those most likely to be uninsured: low-income, nonelderly adults. Highlights from the five case studies include: • Community Health Works in Forsyth, Ga., has served nearly 4,000 uninsured residents with incomes at or below 200 percent of the federal poverty level since 2001. The program emphasizes appropriate use of services and a rigorous case management element across the continuum of care, and enrolls only residents with any of four specific chronic diseases: hypertension, diabetes, heart disease, or vii depression. Community Health Works estimates that its clients use 40 percent less hospital care and 15 percent less emergency room care than a national control group. • The General Assistance Medical Program (GAMP) in Milwaukee, Wis., served approximately 26,000 county residents in 2004 with incomes less than $902 per month. The program makes services available at 17 clinics (including federally qualified health centers) in 23 sites and 10 local hospitals. It leverages millions of national, state, and local dollars to serve the county’s uninsured. • Choice Regional Health Network in Olympia, Wash., helps people enroll in Medicaid, the State Children’s Health Insurance Program (SCHIP), and the Washington Basic Health Program. The program has enrolled as many as 17,000 local residents since 1996. Ninety-eight percent of its applications result in enrollment (compared with 4% of individuals who enroll on their own) and 96 percent remain enrolled three years later (compared with 40% who enroll on their own). • Community HealthLink’s Health Care Access Program in Ratcliff, Ark., is a network health insurance plan currently serving 120 working uninsured residents with incomes below 300 percent of the poverty level. Employers and employees support two-thirds of the cost of coverage, and HealthLink has developed a subsidy fund to cover the final third. • Project Access in Wichita, Kan., serves uninsured residents with incomes below 200 percent of the poverty level. The program enrolls eligible residents when they seek care for a health problem and links them to a “medical home” for ongoing primary care. The program covers primary care for three months and specialty care for six months. Success Factors and Barriers Context matters. Across the five case studies, it is apparent that state political, economic, and social context matters. Local programs can support or complement state public and private insurance programs, but are unlikely to thrive independently. Community initiatives that do not capitalize on state policies and resources struggle against greater barriers. Sustainable leadership, funding, and evaluation. Despite their value to both individuals and the community as a whole, local initiatives are difficult to sustain. Community leaders identified several organizational attributes as necessary for sustainability: strong, dedicated leadership; funding sources, including provider volunteerism, Medicaid partnerships, and federal grants; and data to evaluate and demonstrates initiatives’ success. viii Challenges of replicating local initiatives. Diffusion of innovation among community health initiatives is more likely when there is extensive face-to-face communication between individuals in the original and replication sites, and when there are contextual and organizational factors that are common to both sites. This research indicates that important contextual factors include strong local leadership, high levels of knowledge among interconnected parties, and a state environment with opinion leaders and change agents who value local innovation. Policy Implications Organizing local resources to contribute to health care access and health status improvement is a critical and often neglected component of the health care system. Local initiatives provide bridges to public and private coverage, create steps to care for those who are not covered, and serve as a vehicle for investment. Because all community health initiatives are, in effect, created by national and state policy, it follows that changes in policy would cause the initiatives to adapt and change. Policy change in the current environment, however, would not eliminate their purpose: to serve low-income residents at the edges of both public programs and private coverage. Some proposals at the national level—in particular, block grants to finance Medicaid— could greatly increase the need for community initiatives if states were forced to respond by narrowing program eligibility. Without greater resources for community initiatives, however, the volunteerism they rely on would be strained and could fray. Other national proposals—in particular, those that offer new opportunities for financing coverage—might be used to provide much-needed support to these programs, if care were taken to define qualified coverage to include that offered through community initiatives. In turn, the initiatives could leverage and amplify the value of those funds. For example, refundable tax credits could be used to buy the coverage offered through these networks and their providers. Community initiatives also might be allowed to qualify as “association health plans” that could enroll any small group that includes a threshold proportion of low-wage workers. Small employers might offer these programs as an option available to low-wage workers or to their entire group. Certainly such proposals would warrant careful review by state insurance regulators, but they may be quite feasible with narrow and strategic changes in regulation and oversight and highly beneficial to workers who otherwise could not afford coverage. ix LESSONS FROM LOCAL ACCESS INITIATIVES: CONTRIBUTIONS AND CHALLENGES INTRODUCTION The United States health care system is plagued with at least three serious problems: rising costs, deterioration of the health care safety net, and inadequate public and private health insurance. With 47 million people uninsured, comparative health statistics that rank the U.S. below other industrialized nations, and wide performance variations within the country, it is clear that bold strategies—requiring public and private resources from national, state, and local levels—are essential. Community health initiatives—locally crafted responses to health care access problems—have been steadfast in their efforts to connect uninsured and medically indigent people to health care services and health insurance. Typically, they unite community leaders, providers, and other key stakeholders, building on good-faith relationships to reduce uncompensated care and support the local safety net. Recognizing that merely referring people with complex medical and social needs to care is often insufficient, these initiatives assist in outreach, coordinate and integrate care, and help clients use limited resources efficiently. Previous research has catalogued and described individual community efforts, evaluated the results of specific funding programs, explored how local efforts can substitute for national or state programs, and examined the role of local efforts in changing national policy. This study offers new insights about community initiatives and the successes and challenges they face. Findings fall into the following three areas: 1) the critical importance of state context; 2) the need for community health initiatives and, paradoxically, the difficulty of sustaining them; and 3) the challenges of replication. METHODS The research team developed case studies of five community health initiatives that intentionally or de facto replicated significant components of previously implemented community initiatives. Over 18 months, the researchers visited each of the five sites: Wichita, Kan., Milwaukee, Wis., Ratcliff, Ark., Forsyth, Ga., and Olympia, Wash. Selected to represent a range of geographic areas and operational models, each of these initiatives provides coverage, access to care, or both to individuals who have difficulty finding or navigating among conventional insurance arrangements and public programs. 1 To derive lessons for sustainability and replication, the study team: • described the initiatives’ efforts to increase coverage or access and the impact of these efforts on their target populations; • examined the cost-effectiveness and efficiency of their operations; • identified factors that affected their sustainability and expansion; and • examined how states and communities cooperated to close gaps in funding and access. To conduct this research, the study team obtained feedback about the research design from public and private experts; developed beta-site selection criteria and selected the sites; assembled all available information about each initiative; developed interview protocols; and conducted site visits with 10 to 25 key informants in each site, meeting with a total of 82 informants. Following the site visits, all information was integrated, coded, and uploaded to a qualitative database (ATLAS.ti). Analyses of intent, effectiveness, sustainability, replication, state/community interface, and leadership were developed for each site, and comparative analyses were developed across sites. LOCAL INITIATIVE CONTRIBUTIONS AND CHALLENGES Local initiatives can make several types of contributions to the larger national, state, and local system. These contributions include: chaperoning people through the system and changing patient behavior patterns; drawing resources to the community; improving efficiency; creating innovation; and garnering grassroots support for solving these problems. However, these local contributions do not come without challenges. This research sheds light on the challenges that communities face: context (i.e., the political, economic, and social environments); replication; and the financing, leadership, and evaluation necessary for sustainability. All five of the cases touch on all contributions and challenges to varying degrees; however, there tends to be a dominant theme for each case that serves as an illustration of a contribution or a challenge or both. SITE HIGHLIGHTS 1. Access and Coverage All five local community initiatives seek to improve access and coverage for those most likely to be uninsured: low-income, nonelderly adults. The programs that offer coverage typically provide comprehensive benefits for a limited period of time, often for individuals who seek care after contracting an illness. Other programs manage care for clients with complex needs, chaperoning them through systems they characteristically have trouble 2 navigating. Community leaders see the programs as important stop-gap measures for a failing health care system, and recognize that they cannot address the health care needs of all the uninsured in the community. Examples of these programs include: • Community Health Works in Forsyth, Ga., has served nearly 4,000 uninsured residents with incomes at or below 200 percent of the federal poverty level since 2001. The program emphasizes appropriate use of services and a rigorous case management element across the continuum of care, and enrolls only residents with any of four specific chronic diseases: hypertension, diabetes, heart disease, or depression. Its provider network includes three hospitals, two clinics, nearly 100 physicians, and 21 pharmacies that work with a medication bank to provide access to affordable prescription drugs. There is a waiting list for admission to the program. • The General Assistance Medical Program (GAMP) in Milwaukee, Wis., is a county-operated managed care organization that purchases services for its clients. In 2004, it served approximately 26,000 county residents with incomes less than $902 per month. The program makes services available at 17 clinics (including federally qualified health centers) in 23 sites and 10 local hospitals. GAMP providers accept Medicaid rates. • Choice Regional Health Network in Olympia, Wash., helps people enroll in Medicaid, the State Children’s Health Insurance Program (SCHIP), and the Washington Basic Health Program. The program has enrolled as many as 17,000 local residents since 1996. • Community HealthLink’s Health Care Access Program in Ratcliff, Ark., is a network health insurance plan currently serving 120 working uninsured residents with incomes below 300 percent of the poverty level. Together, employers and their employees who enroll in the program support two-thirds of the cost of coverage. HealthLink has developed a subsidy fund to cover the final third. The program’s provider network includes two tertiary care hospitals, four critical care access hospitals, six primary care clinics, four mental health counseling centers, and 200 medical specialists. Network providers agree to accept Medicare rates and to continue seeing patients whose care may exceed the plan’s reimbursable limit. • Project Access in Wichita, Kan., serves uninsured residents with incomes below 200 percent of the poverty level. The program enrolls eligible residents when they seek care for a health problem and links them to a “medical home” (i.e., a primary 3 care practice that provides them with accessible, continuous, and coordinated care) for ongoing primary care. The program covers primary care for three months and specialty care for six months. It connects more than 600 physicians with local hospitals, six outpatient clinics, 36 dentists, and 69 participating pharmacies. The program served 625 uninsured residents in 2004 and has served nearly 4,500 residents since 1998. 2. Cost and Effectiveness In each community, interviews for this report found that local leaders contend that providing more appropriate care is cost-effective for both providers and the community. The initiatives use various strategies to control cost. These include cost-sharing in the form of modest copayments, administrative fees, or membership dues; and health care providers bearing significant risk in the form of discounted rates or capitated reimbursement. While few of the initiatives had made the necessary investment to develop strong evidence of cost-effectiveness, some were able to demonstrate cost-effectiveness on selected measures or a positive return on investment. For example: • Community Health Works estimates that its clients use 40 percent less hospital care and 18 percent less emergency room care than a synthetic control group developed from national data. 1 Moreover, the probability that a client has an inpatient stay declines the longer the client stays in the program 2 —inpatient stays had decreased by 20 percent in the first six months, and by the time clients were in the program for 24 months, their inpatient stays had decreased by 40 percent. • Choice Regional Health Network claims success in enrolling and retaining eligible adults in the Washington Basic Health Plan. Ninety-eight percent of its applications result in enrollment (compared with 4% of individuals who enroll on their own) and 96 percent remain enrolled three years later (compared with 40% who enroll on their own). For each dollar the participating hospitals have invested to enroll eligible uninsured in state programs, Choice estimates that they have received $20 from increased reimbursement. • GAMP reports it has raised an additional dollar for every local dollar used to pay for the care of individuals enrolled in the program—offering a 100 percent rate of return to local funds, not even considering potential gains made in population health and efficient use of care. 4 SUCCESS FACTORS AND BARRIERS Context Matters At the community level, lack of access to health care coverage takes on personal aspects lacking in the national debate: it affects friends and neighbors struggling either to find care or to provide it. Communities motivated to organize a response to this problem typically look to other successful local initiatives as models. Often, they are unaware of the contextual factors that help or impede the success of the local programs they seek to replicate. For example, a program that builds on private insurance may succeed in a state with a strong culture of employer coverage, but is more difficult in a state where employers are reluctant to offer coverage. Across the five case studies, it is apparent that state political, economic, and social context matters. Local programs can support or complement state public and private insurance programs, but are unlikely to thrive independently. Community initiatives that do not capitalize on state policies and resources struggle against greater barriers. Several measures of state context seem especially useful in differentiating whether a community initiative can survive and succeed, as follows: • Supportive public programs or a strong private insurance base. Programs that leveraged either of these resource were generally larger and more successful. Conversely, it is extremely difficult to succeed by attempting to leverage a weak base. For example, in Olympia, Wash., the Choice Regional Health Network focuses exclusively on enrolling eligible individuals in the state’s arsenal of strong public insurance programs: Medicaid, SCHIP, and the state-funded Washington Basic Health Plan. However, in Ratcliff, Ark., where employer coverage is low statewide, Community HealthLink struggles to encourage employers to offer and contribute to coverage. • State-level vision and supportive programs and policies. The willingness of states to make regulatory exceptions and work cooperatively with initiatives can be essential to their survival. For example, Arkansas exempted the local initiative from state insurance regulation, easing its implementation and lowering start-up costs; Georgia assembled a public/private partnership to give grants and technical assistance for local network development; Kansas extended malpractice insurance to providers in safety-net clinics and placed state program eligibility specialists in the clinics; Washington changed charitable immunity laws and also contracted with Choice to provide outreach and enrollment services; and Wisconsin provided block grants and allowed the local initiative to use disproportionate share hospital (DSH) funds—a payment adjustment under Medicare for hospitals that serve a relatively large volume of low-income patients. 5 • Community and provider culture. Most community initiatives rely heavily on providers and other community members volunteering time and services. Both a strong provider sense of attachment to the community and a sense of a common future are critical. For example, in Milwaukee, GAMP’s network providers accept Medicaid rates; in Ratcliff, Ark., HealthLink’s network providers accept Medicare rates and also take on financial risks for care that exceeds the plan’s reimbursement limits. Sustainable Leadership, Funding, and Evaluation Local initiatives typically offer a bridge to public and private coverage, creating steps to coverage and care for individuals who are eligible but have difficulty in finding coverage, staying insured, or assembling the care they need within available coverage. Local initiatives also offer communities a vehicle for investment in the form of grants, state contracts, and organized volunteerism. Without such initiatives, there may be no entity in the community able to receive or organize these resources. Despite their value to both individuals and the community as a whole, local initiatives are difficult to sustain. Community leaders identified several organizational attributes as necessary for sustainability. These include: • Strong leadership. Successful program directors had a strong business sense, creativity, and dedication. • Funding. Local access initiatives need partnerships that result in payment for services, government funding, and grants. • Evaluating outcomes. It is essential that local access initiatives be able to demonstrate their value over time to a wide array of stakeholders. Strong leaders were able to create programs that were solidly grounded in the needs of the target population. They flexibly adapted to the changing environment and engaged in a continuous blending of programs to shape a complete portfolio connecting their clients to care. In two cases the initiative directors has previously been part of state government. These long and trusted relationships translated into financial contracts between the local initiative and the state. Obviously, funding is essential to the sustainability of these programs. Each relied on provider volunteerism to some degree. Health care providers, too, bear significant risk in the form of voluntary participation, discounted rates, or capitated reimbursement. However, community and program leaders recognized that reliance on volunteerism 6 ultimately would limit programs’ potential for growth. Communities walk a thin line between physicians’ desires to serve and their fear of being taken advantage of by the system. With diminished grant funding and increasing numbers of uninsured, program leaders have fought to maintain services for the uninsured. Four of the five programs may be financially sustainable in the short run, but all recognize that ongoing sources of funds are needed in the long run. Foundations, provider–members (providers who are partners in local initiative and donate their services or take lower payments), Medicaid partnerships, and federal grants have been important sources of funding for these initiatives. For example: • Project Access has built sustainable funding for its $2 million operating budget from diverse sources. Physicians and hospitals provide donated care. Local governments pay for prescriptions, and a local community foundation supports the dental component of the program. An ongoing relationship with the local United Way pays for basic staffing. All these partners are committed to continue. Additional funding has come from a federal grant and a national foundation. • GAMP has funding from a county tax levy ($15.6 million in 2003) and a combination of state and federal government block grants and DSH payments ($33.8 million). The program also receives revenue from the application fee. Finally, community leaders frequently mentioned the importance of having data to evaluate and demonstrate the initiatives’ success. Better information might help the initiatives obtain additional funding, but the organizations typically lack the resources to create data systems. CHALLENGES OF REPLICATING LOCAL INITIATIVES Successful replication of community health initiatives in a larger number of communities could help more people find coverage and help the uninsured obtain care. Previous studies of replication have investigated how innovations become diffused among individuals or organizations. These studies suggest that replication is more difficult when the innovations are complex, the network organizations are complex, and there are differences between the initial (alpha) and replication (beta) sites. Conversely, innovations are most easily transferred when they are simple and benefits are easily observable. Initiatives to improve access are necessarily complex, and their results are not quickly or clearly observable. Therefore, it should not be expected that the transfer of access innovations from community to community would come easily or without careful attention to factors that are known to affect successful diffusion of innovation. 7 Diffusion of innovation among community health initiatives is more likely when there is extensive face-to-face communication between individuals in the alpha and beta sites, and when there are contextual and organizational factors that are common to both sites. 3 This research indicates that important contextual factors include strong local leadership, high levels of knowledge among interconnected parties, and a state environment with opinion leaders and change agents who value local innovation. The organizational factors that appeared to encourage diffusion included strong leadership and creating a new formal provider organization to make decisions about implementing the program. The fact that the extent of replication varied across sites speaks to the complexity of diffusing local access initiatives. In only one site was the replication complete. The Wichita example differed from the others in that it had local and state contexts that were similar to the alpha site, as well as extensive communication and collaboration with the alpha site. Other sites: 1) had a very different state context and little communication with the alpha site; 2) visited other sites with different contexts and implemented only certain aspects of the other programs; or 3) had extensive communication with multiple alpha sites and implemented different facets from different programs? POLICY IMPLICATIONS Organizing local resources to contribute to health care access and health status improvement is a critical and often neglected component of the health care system. Local initiatives provide bridges to public and private coverage, create steps to care for those who are not covered, and serve as a vehicle for investment. They also translate and apply national and state policies, helping local providers understand and utilize these opportunities and enhancing the effectiveness of national and state programs. Local health initiatives take different forms in different communities. These variations are due to differences in local context and the complexity of replicating organizations in different locations. Although it can be difficult to make the general case that they warrant investment of national and state public and private funds, each site in this study—as well as those investigated in prior studies—earnestly believes it makes a difference and some offer intriguing evidence to support these claims. Their importance to the communities that they serve is undeniable; at minimum, each bears witness to the failures of an undifferentiated application of federal and state policy to local communities. Because all community health initiatives are, in effect, creatures of national and state policy, it follows that changes in policy would cause the initiatives to adapt and change. Policy change in the current environment, however, would not eliminate their 8 purpose: to serve low-income residents at the edges of both public programs and private coverage. Some proposals at the national level—in particular, block grants to finance Medicaid—could greatly increase the need for community initiatives if states were forced to respond by narrowing program eligibility. Without greater resources for community initiatives, however, the volunteerism they rely on would be strained and could fray. Other national proposals—in particular, those that offer new opportunities for financing coverage—might be used to provide much-needed support to these programs, if care were taken to define qualified coverage to include that offered through community initiatives. In turn, the initiatives could leverage and amplify the value of those funds. For example, refundable tax credits could be used to buy the coverage offered through these networks and their providers. Community initiatives also might be allowed to qualify as “association health plans” that could enroll any small group that includes a threshold proportion of low-wage workers. 4 Small employers might offer these programs as an option available to low-wage workers or to their entire group. Certainly such proposals would warrant careful review by state insurance regulators, but they may be quite feasible with narrow and strategic changes in regulation and oversight and highly beneficial to workers who otherwise could not afford coverage. 9 CASE STUDIES: FIVE LOCAL EXPERIENCES 10 NEAR PERFECT REPLICATION: PROJECT ACCESS Wichita, Kansas WHAT THEY DID Led by one philanthropically minded, yet entrepreneurial, physician, local leaders carefully copied a program from a similar community to provide primary and specialty care to low- income uninsured residents. The program uses less than $180,000 in administrative costs per year to leverage $5 million in donated services. To date, 5,000 people have been served. Physicians, hospitals, and pharmacies donate care and services. Local government, the United Way, and a local foundation support the program financially. BACKGROUND Project Access in Sedgwick County, Kan., is a community program that provides health care to low-income uninsured people through the charitable contributions of local providers. Project Access provides comprehensive inpatient and outpatient services. Although the program focuses on specialty care for people who are already ill, it also helps uninsured individuals find a medical home by matching them with primary care clinics and practitioners. The program staff coordinates donated services for patients and providers. Approximately 70 percent of physicians in the county participate in the program. Project Access is governed by a 12-member board and supported by committees that guide operations and promote physician participation. Eight staff members work at the Project Access program office, which is housed at the Medical Society of Sedgwick County. These include the program director, an administrative assistant, four patient service coordinators, a secretary, and a prescription service coordinator. The Population Served Project Access is intended to be a short-term safety net for people in need of specialty care, with link to primary care services and a medical home. Enrollees are predominantly female, young to middle-aged (31 to 50 years), and unmarried. They tend to be chronically ill, often with dual-diagnosis conditions and significant health care needs. Although the program initially had no restrictions on immigration status, due to physicians’ frustration with some aspects of serving undocumented immigrants, the program is now limited to citizens and documented immigrants. To enroll in Project Access, an uninsured person must be a citizen or legal immigrant residing in Sedgwick County and earn less than 150 percent of the federal poverty level (FPL). Project Access largely adopted these criteria from the alpha program in Buncombe County. 11 In 1996, the University of Kansas School of Medicine performed a community health assessment to build the case for the program to potential funders. That study estimated the uninsured population in Sedgwick County at 55,000, of which 65 percent to 70 percent were employed and approximately 10,000 were potentially eligible for Project Access. In the first six years of the program, nearly 5,000 patients enrolled. HOW THEY DID IT Project Access is a beta site modeled after a prototype program in Buncombe County, (Asheville) N. C. Buncombe County’s program became operational in 1996. It is a physician-led initiative providing primary and specialty care for the low-income uninsured. Paul Uhlig, a Wichita-based physician, became aware of the Buncombe County federal poverty level and started championing the establishment of a similar program in Sedgwick County. In 1998, Dr. Uhlig mobilized key stakeholder groups—including private and clinic-based providers, city and county managers, and potential funders—to consider ways to replicate the Buncombe model in Sedgwick. He encouraged the Medical Society of Sedgwick County to assume a leadership role and reorganized the Central Plains Regional Health Care Foundation—previously established to address the community’s health needs—to administer the new program. He also convinced representatives from the aircraft industry headquartered in Wichita to provide private jet transportation for local leaders to visit the alpha site. The principal stakeholders convened several times in Asheville, where they learned more about the operational elements of Project Access. In April 1999, leaders from the Buncombe program were invited to make presentations to the many stakeholders in Wichita. Buncombe’s involvement in Wichita’s local strategic meetings and presentations was pivotal in garnering city and county support to help drive the process. Once Sedgwick County decided to replicate the program, local stakeholders purchased technical assistance from the Project Access team in Buncombe, including software and training to track patient flow. To build partnerships, raise funds and complete grant applications, the Wichita initiative used trend data from the Buncombe program documenting its successes. Sedgwick also acquired Buncombe’s “Blue Notebook,” which provided the details of Buncombe’s daily operations, as well as patient referral forms, physician recruitment materials, and other printed materials. Many of the characteristics of Buncombe’s model were replicated exactly in Wichita, especially in the area of program leadership. Some modifications were made in 12 eligibility, support services, and funding. In September 1999, Project Access of Sedgwick County was launched and began enrolling patients from the smallest of the six participating clinics. The second clinic was brought on two months later, followed by the third six weeks after that. The gradual enrollment allowed time to iron out bugs, as patients established their medical homes. All of the six clinics were on board within six to seven months. Innovations are most easily transferred when they are simple, can be implemented quickly, and when benefits are easily observable. However, initiatives such as Project Access are complex, with results generally not quickly or clearly observable. Leaders in Wichita were able to overcome this replication challenge by intensely having ongoing communication with the alpha site, capitalizing on the similarities between the two communities, and leveraging the entrepreneurial and pioneering spirit of the community. Wichita’s experience with Project Access points to a number of key elements for replicating such a program in other communities, including: a strong and involved medical society, influential leadership and a day-to-day administrator, extensive public and private partnerships, and diverse and committed funders. Many communities across the country have expressed interest in replicating Wichita’s program, including four or five in Kansas. CONTRIBUTIONS Key informants overwhelmingly reported Project Access is a valuable program for the low-income uninsured and said that enrollees are receiving needed services. The program has helped enrollees obtain more consistent care, with a focus on education and prevention. Although it is difficult to measure the change in the net cost to the community as a whole, hospital admission data indicate people enrolled in the program are receiving appropriate services at a lower cost. Project Access appears to coordinate and expedite the process of obtaining referrals for patients who need them, which helps individuals gain access to appropriate services before conditions become severe. Without Project Access, the uninsured would rely more on the traditional safety net in Wichita. In addition, attempts to access care from private physicians would be more challenging and disjointed. In the past, primary care clinics spent a lot of time—and experienced great frustration—begging specialty physicians to see patients. Without Project Access, many physicians might again shy away from treating the uninsured for fear of becoming overwhelmed with requests. 13 Project Access enrollees generally have lower physical and mental health functioning scores than the general population. However, enrollees’ health status has shown some improvement and enrollees in case management have demonstrated small, positive changes in control over their health. The program estimates that approximately $180,000 in administrative funding generates $5 million worth of donated services a year. Plus, evaluators report donated services are relatively constant while average cost per patient is decreasing. Furthermore, many key informants pointed to David Rogoff’s algorithm for return on investment, which hypothesizes that the move toward coordinated health care has reduced total health care costs by one-third. CHALLENGES TO FUTURE SUSTAINABILITY Provider and partner contributions are crucial to the long-term sustainability of Project Access. County and city governments and the United Way are committed to providing ongoing funding for the prescription assistance program and basic program operations, and physicians and hospitals seem committed to continue offering donated services to patients. Yet, program funding is not sufficient to meet the needs of all uninsured residents of Sedgwick County, and some of the program’s grant funding will end this year. However, key informants expect that, with continued physician involvement and efforts to better manage high- utilization enrollees and control pharmacy costs, the program will be sustainable as long as necessary. Because grant funding is limited, Project Access is exploring ways to ensure its long-term sustainability by reducing its dependence on grants. Leaders would like to see local businesses contribute to the program in some way and are exploring the possibility of creating a small business insurance model to allow more people eligible for Project Access to have health insurance. Also, given the United Way’s goal of replicating similar programs in 500 communities, Project Access staff expects to generate program revenue by consulting for other communities that wish to create similar programs. Strong leadership is also key to sustaining Project Access. Although the program’s original champion, Dr. Paul Uhlig, moved out of the state shortly after the program’s implementation, leadership appears to be stable and strong. The current director, Anne Nelson, has been engaged with the project since 1999, and received high marks from key informants. The governing body, providers, and partners appear to have ongoing commitments to the program. 14 The Project Access Board would like the state government to provide regular funding for the program and others like it. Ultimately, program stakeholders would like to see state and federal governments directly address the problem of the uninsured through insurance coverage and other access initiatives, making the need for Project Access obsolete. Project Access has weathered a very serious challenge to its sustainability. The program was built on the philosophy of pulling together to provide help to all those who need it, yet one group challenged the system. At its inception, undocumented immigrants comprised 5 percent to 6 percent of the program’s population. However, after approximately two years, their continued eligibility was jeopardized when a few provider groups threatened to withdraw from the program. First, the largest oncology group in the area was providing free pharmaceuticals as well as free health care services to program participants. Because pharmaceutical companies only provided indigent care for legal citizens, this provider group’s drug costs became extremely prohibitive, and the group decided to cease participating as long as undocumented citizens were eligible for services. Second, other providers expressed dismay that some undocumented residents were bringing relatives from abroad to obtain care from Project Access. Third, doctors found this population relatively noncompliant with recommended treatments, with one patient dying as a result of noncompliance. Consequently, undocumented immigrants are no longer eligible for Project Access. 15 THE POWER OF LOCAL, STATE, AND NATIONAL FINANCIAL LEVERAGE: GENERAL ASSISTANCE MEDICAL PROGRAM (GAMP) Milwaukee, Wisconsin WHAT THEY DID A unique partnership of local, state, and federal government; county public health; hospitals; physicians; and clinics turned a $15.6 million local tax into $49.4 million in program funding. As a result, 27,000 of Milwaukee’s uninsured are served each year by a broader, more organized safety net. In this program, participating providers both give and receive; they bear risk for their patients needs when resources run out but receive a new funding stream to serve program enrollees. BACKGROUND The General Assistance Medical Program (GAMP) is a Milwaukee County–administered program designed to provide access to primary and secondary health care services for uninsured residents earning less than $902 per month. The initiative is an update of the county’s hospital-based indigent care program, which was threatened when the hospital closed in 1995. The county now plays the role of purchaser of modified managed care services, rather than a provider of those services. As a part of the program’s primary care emphasis, health centers act as the main gatekeepers for residents who must seek medical services to be eligible for enrollment. Providers are reimbursed at Medicaid rates, with program funding coming from leveraged state contributions, local taxes, and intergovernmental transfers. Services are available at 17 clinics in 23 sites, including federally qualified health centers (FQHCs), and 10 local hospitals. The County Board of Supervisors of Milwaukee County has responsibility for setting program policy and direction. This Board is made up of 19 elected officials who represent supervisory districts and face re-election every two years. Eligibility requirements (intended to simulate the FPL) are set by the Milwaukee County Board of Supervisors and are subject to review each year during budget review. From an operations standpoint, GAMP is administered by the Milwaukee County Department of Health. The program employs more than 20 staff members, some located in the participating clinics. Billing functions are contracted out to a third-party administrative group. THE POPULATION SERVED About 100,000 to 120,000 uninsured people live in Milwaukee County; 60 percent are eligible for GAMP. To be qualified for the program, an applicant must be: a Milwaukee 16 resident for the previous 60 days, ineligible for any other entitlement program or third party public or private insurance, and able to provide a verifiable Social Security number. Unlike traditional insurance, clients do not pre-enroll for coverage; instead, they must be seeking services or treatment due to a medical need. Eligible individuals can enroll in GAMP at any of the contracted community clinics and all Milwaukee County hospital emergency departments. Participating hospitals and clinics have trained financial counselors who assist individuals with their applications. At the time of enrollment, the applicant chooses the community clinic that he or she would like to use for primary care. The client, if approved, must use this clinic and its network for the 6-month eligibility period. At the end of this period, the client must reapply to determine continued eligibility and may do so at any site. HOW THEY DID IT GAMP has total funding of $49.4 million. Of this, $15.6 million is provided from the Milwaukee County tax levy for medical services and $33.8 million is provided from the state and federal government through state block grants and disproportionate share hospital (DSH) payments. A total of $13.5 million of state and county funds are matched with $20.3 million in federal Medicaid DSH payments. This total payment is funneled through DSH eligible hospitals in Milwaukee County to the GAMP program. The county portion of the match is transferred to the state through intergovernmental transfer (IGT). The program also receives funding from the $35 application fee; participants are eligible for six months at a time and the fee is charged with each re-application. Capturing and Retaining Funding Milwaukee County has a long tradition of providing quality health care to residents in need, and has operated a program similar to GAMP since the late 1970s. Historically, the program served as a funding mechanism between the county and the State of Wisconsin to address the costs of providing medical care to the county’s indigent population. These funds were primarily allocated to the John L. Doyne Hospital, which was owned by the county. On December 23, 1995, the county closed this hospital and transferred its assets to the Froedtert Memorial Lutheran Hospital, which continued to serve as the primary provider to GAMP patients for two years. In April of 1998, a restructured GAMP program was approved by the County Board of Supervisors. This program focused on providing access to cost-efficient primary care services. Within this same time period, the state legislature, at the Governor’s request, modified Chapter 49 of the Wisconsin statute so that no level of government was statutorily 17 responsible for providing health care services to indigent populations. After this decision, the state began to provide block grants to counties to provide these services. Milwaukee County is the primary beneficiary of this state funding. Milwaukee County and the State of Wisconsin have leveraged their funding to maximize the amount of DSH funding available to the county through IGT. An IGT may take place from one level of government to another (e.g., from counties to states, or within the same level of government). The federal Medicaid statute expressly recognizes the legitimacy of IGTs involving tax revenues, such as the tax levy imposed by Milwaukee County. The Wisconsin state legislature gave Milwaukee County permissive authority to increase the amount of the IGT to further maximize DSH payments. To the extent that the state is able to justify additional DSH payments to the DSH-eligible Milwaukee County hospitals, additional funding in the form of Medicaid hospital payments will be made. The state legislature remains committed to attempting to increase the amount of funding that supports the program. CONTRIBUTIONS Since 1998, GAMP has leveraged millions of national, state, and local dollars to serve many of the county’s uninsured residents. In addition, as one community key informant reported, “The county administrative team has done a remarkable job to improve efficiencies over the years . . . the GAMP program has been a remarkable investment.” GAMP’s overall administrative costs are 7 percent. It has been able to achieve these low costs in part by outsourcing claims payment, limiting pharmacy contracts to one major vendor, and instituting a formulary. The program has also implemented strategies to change enrollee utilization patterns. These include: a very active patient education program on emergency utilization; a 24/7 nurse call line; disease management programs for enrollees with asthma, hypertension, and diabetes; and enrollee cost-sharing for application ($35) and pharmaceuticals ($1 for generics and $3 for brand). An independent evaluation found that between 1997 and 2000, costs per claim decreased from $260 to $194, inpatient services expenditures decreased 7 percent, and per member per month costs declined. 5 The GAMP structure has effectively distributed responsibility for the uninsured across providers, enabling the 135,000 to 150,000 uninsured in Milwaukee County a broader choice of providers. Before GAMP, the county hospital, which is now closed, had 18 uncompensated care levels of approximately 12 percent, while other hospitals had levels of 1 percent to 2 percent. Now, all hospitals in the community have uncompensated care levels of approximately 6 percent. If GAMP in its current form had not been implemented, the community would likely continue with a costly, struggling public hospital and little community support for bolstering its financial position. Key informants indicate that there would be a dramatic increase in emergency department volume as well. Distributing responsibility for the uninsured across providers has created a strong provider constituency in support of GAMP. GAMP clients reside in all Milwaukee County zip codes, strengthening mainstream political support for the program. Most key informants view the uninsured as more of a community issue because of GAMP, although a few still consider the uninsured and GAMP to be largely below the radar screen of most residents. CHALLENGES TO FUTURE SUSTAINABILITY GAMP officials and community leaders wish to continue the program to improve cost- effective access to primary care for patients who generally rely on hospital emergency departments. Most key informants saw the program as being sustainable over three years, but were uncertain about longer term viability because of the its dependence on state and federal participation. The GAMP program faces challenges going forward. There is uncertainty regarding the local, state, and federal financial partnership that supports the program. Some informants are concerned about the program’s limitations: no active outreach and limited services for mental and dental health. In addition, the $35 application fee is seen as a barrier for some enrollees. Limited funding and growing cost add strain to the safety net and may tip the scale against continued participation for some providers. Shortly after the completion of the GAMP case study, the founding leader of the program retired. Since that time there have been two subsequent leaders. The stability of local collaboratives is often threatened when leadership changes, in part because local organizations are built on a complex nexus of relationships. GAMP was able to use grant funding to commission a cost–benefit evaluation of the program. The information in that evaluation was helpful in building the case for continued support of the program. However, continued evaluation and feedback are needed to sustain the program. Resources and expertise to accomplish ongoing evaluation are often not built into implementation budgets. 19 STATE CONTEXT MATTERS: COMMUNITY HEALTHLINK Ratcliff, Arkansas WHAT THEY DID Community leaders and providers developed a network to serve low-income uninsured residents in their communities. They designed a subsidized health plan and the Arkansas General Assembly enacted legislation to exempt the pilot plan from insurance regulations. The health plan pieced together employer contributions, individual dues (similar to a premium), grant subsidies, pharmaceutical discounts, and provider risk-sharing to offer an affordable product to 181 members during its two-year pilot phase. BACKGROUND Community HealthLink is a capitated, subsidized health insurance plan operated by the Arkansas River Valley Rural Health Cooperative, a nonprofit organization. The program provides comprehensive health care coverage for working uninsured residents with incomes below 200 percent of the federal poverty level (FPL) who live in the three contiguous counties of its service region. It is similar to the three-share approach of Muskegon, Mich., in which employers contribute one-third, employees contribute one- third, and the final third comes from a combination of local and state contribution and federal match. In 2005, Community HealthLink completed a two-year pilot phase. Employers and employees provide two-thirds of the cost of care, while the third share is covered by a subsidy fund set up by the Cooperative. Local primary care providers are reimbursed at Medicare rates and are asked to provide services and assume risk for care in excess of the reimbursable limit. The Cooperative’s health plan contains three elements: health care access, prescription drug assistance, and disease management/health education. The provider network currently includes two tertiary care hospitals, four critical care access hospitals, six primary care clinics, four mental health counseling centers, and 200 medical specialists. The Cooperative is governed by a Board of Directors comprised of representatives and community leaders from each of the counties in the service region. A 12-member staff is led by an executive director and guided by program leads for the health care access program, health education/telehealth services and support/prescription drug assistance services. THE POPULATION SERVED In this sparsely populated rural area of Arkansas, an estimated 4,000 to 6,000 people are eligible for the Cooperative’s health plan. The HealthLink program targets working, 20 uninsured adults with incomes 200 percent below FPL and who are ineligible for Medicaid, The client population is composed of 51 percent adult females, 41 percent adult males, and 8 percent children. The population is 99 percent white. The average age of members is 46 years old, and the average family has is 2.6 members. The average client has an income of 134 percent of the FPL. Enrollees were uninsured an average of 5.5 years before entering the program. HOW THEY DID IT The work in Ratcliff reveals an important lesson: local programs can support or complement state public and private insurance programs, but are unlikely to thrive independently. The three-share program in Ratcliff, Ark., has struggled in an environment of systematically low employer support for private group coverage. In contrast, the three- share model in Michigan was built on a relatively strong base of employer group coverage. The general context of public programs and private coverage for each of the five initiatives is summarized in Figure 1. Figure 1. Percent of the Nonelderly Population in the Study States with State Public Coverage or Employer-Sponsored Coverage, 2002 80 Employer-sponsored coverage (%) Wisconsin Kansas 60 Georgia Washington Arkansas 40 20 0 0 5 10 15 20 25 30 Public program coverage (%) Source: Adapted from P. Fronstin, Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 2003 Current Population Survey, EBRI Issue Brief No. 264 (Washington, D.C.: Employee Benefit Research Institute, 2003), Fig. 18, p. 19. As shown, Arkansas has the lowest percentage of private coverage of any of the study’s states. At the community level, lack of access to health care coverage takes on personal aspects lacking in the national debate: it affects friends and neighbors who are struggling either to find care or to provide it. Communities motivated to organize a 21 response to this problem typically look to other apparently successful, local initiatives as models for how they might proceed. Often, they are unaware of the contextual factors that help or impede the success of the local programs they seek to implement. CONTRIBUTIONS Even though it is a very small program, Community HealthLink has provided potentially life-changing insurance access to people who have lived without it for an average of 5.5 years. Before the program, uninsured people avoided seeking outpatient care, relied on over-the-counter medications when prescriptions were needed, and often landed in the emergency department. The use of strategies like encouraging primary care, tapping into pharmaceutical charity care programs, adding health promotion interventions and utilization controls, and providing disease management, has helped control costs and maintain a sustainable medical loss ratio. Additionally, costs are controlled by limiting emergency department visits to two per year unless the visit results in hospitalization. All of these strategies control costs and change the local culture. Individuals are more likely to be healthy when they have a medical home, instead of using the emergency room; take prescription rather than over- the-counter medications; and have their chronic diseases managed. Community HealthLink also supports providers by paying for care provided on a reduced-fee schedule, equal to Medicare rates. Prior to the program, providers were only paid what they were able to collect from the patients. However, the program is not without risk to providers. Costs are contained through a capitated amount of $10,000 in outlays per enrollee per year. After the capitated amount is reached, any additional costs of care are absorbed by the enrollee’s provider. Local leaders has woven together a network of state forgiveness of traditional insurance requirements, provider discounts and risks, pharmaceutical charities, employer contributions, individual dues and grant subsidies, along with behavior-changing and cost- saving strategies to offer affordable health care coverage in the community. CHALLENGES TO SUSTAINABILITY Community HealthLink is fragile. It struggles to encourage employer to offers and employees to contribute to employee coverage in a state with very low employer- sponsored coverage. The grant funding that provided the subsidy to the insurance product is disappearing. Program leaders are seeking a federally qualified health center (FQHC) in the community, partly to offer access to enhanced dental and mental health services. 22 However, local providers are opposed to the FQHC, arguing they have the capacity to serve the uninsured. This tension threatens the provider participation in the program, which is central to its success. Even though the state has been supportive by passing legislation that supports the local effort, some state leaders are skeptical about the value of the program. The challenges to sustainability faced by Community HealthLink are not different from those in other communities; however the small size of the community and the difficult state environment make this initiative more vulnerable. 23 MEASURING OUTCOMES: COMMUNITY HEALTH WORKS Forsyth, Georgia WHAT THEY DID A multi-county, multidisciplinary initiative provides care management and access to medical care, services, and medication to uninsured people with hypertension, diabetes, heart disease, or depression, who earn less than 200 percent of the federal poverty level. Because a control group was not available for comparison, the program was evaluated by composing a synthetic sample from the Medical Expenditure Panel Survey (MEPS) data. Patients in the national data set with similar diagnoses, incomes, and other demographics were selected and their hospital utilization patterns and emergency utilization were compared to the patients in the community. Those enrolled in the program had 40 percent fewer admissions and 15 percent fewer emergency room visit than those in the sample. 6 BACKGROUND Community Health Works (CHW) was created in 2001 to more effectively address problems of uncompensated care. The program was designed by a collaboration of five nonprofit hospitals in the seven-county region, two public health and mental health districts, representatives from county governments, the medical school, and representatives from business and civic organization. CHW is designed for people with incomes below 200 percent of the federal poverty level who are uninsured and not eligible for any publicly-sponsored or employer-health insurance. The services address adults between the ages of 19 and 64 with high-risk diagnoses of hypertension, heart disease, diabetes, or depression . CHW is administered by a nonprofit organization and relies heavily on provider volunteerism and hospital leadership. The local care network consists of five hospitals, two clinics, nearly 100 physicians and 21 pharmacies. The CHW Board of Directors is composed of 18 members, representing hospitals, nonprofit groups, mental health agencies, community foundations, health care providers, public health departments, and local government. These individuals are also geographically representative of the CHW service area. CHW has five full-time administrative staff, eight care managers located throughout the service area, and two staff members for data entry. THE POPULATION SERVED Program designers studied the regions’ hospital discharge data to identify the four diagnoses, which are costly to treat in the hospital but could be easily treated with regular access to primary care. It was initially estimated that 6,000 to 7,000 individuals were eligible for the program. To date, nearly 4,000 have been served. 24 The average annual income of the CHW population is $7,000; the average educational level is the 11th grade. Seventy percent of CHW clients are female, 67 percent are African American, 31 percent are white, and 45 percent are employed. Enrollees have an average of three diseases and use five medications. Seventy percent have comorbidities. HOW THEY DID IT In their original plans, the CHW founders made a commitment to evaluation an important aspect of sustainability. Three key components of the evaluation include: health status, utilization, and financial impact of the network’s activities. To measure changes in enrollees’ health and determine the intensity of care management needed, CHW administers questions from the Behavioral Risk Factor Surveillance System (a health survey developed by the Centers for Disease Control) every six months and a health risk assessment every three months. Overall, these assessments find that patients’ conditions stabilize or improve throughout their time enrolled in CHW. Although enrollees typically stay in the program indefinitely, some enrollees’ health improves enough for them to return to work, potentially with health insurance benefits, allowing them to disenroll from CHW and make room for additional clients. Disease management and case management strategies are intended to change the costs of chronic illness, partly by preventing emergency room visits and hospitalizations. CHW focused on these two utilization measures as short-run indicators of the effectiveness of the program. Creating a comparison group to measure the effect of CHW on its members’ use of emergency room visits and inpatient stays proved difficult. The solution was to create a comparison group using data from the Household Component of the 2002 MEPS 7 . The Household Component collects data for a sample of families and individuals across the county, drawn from a nationally representative sub-sample of households that participated in the previous year's National Health Interview Survey. It produce annual estimates for a variety of measures of health status, health insurance coverage, health care use and expenditures, and sources of payment for health services. The comparison group was constructed by identifying uninsured individuals in MEPS, with age and income requirements that would qualify them for enrollment in CHW. The comparison group included only those MEPS respondents who had one of the four diseases: hypertension, diabetes, heart disease, and depression. This 25 comparison group may be healthier than CHW members since MEPS respondents may include individuals with one specific disease, while the average CHW member has at least three comorbidities. CHW focused on annual inpatient stay rates, defined as average number of inpatient stays per individual per year, and average number of emergency room visits per individual per year. CHW members are admitted to the hospital at a rate that is 40 percent lower than the comparison group. They also use the emergency room about 15 percent less frequently. Both of these differences are statistically significant. Over time, it has become possible to examine changes in outcome measures for CHW members. If the combination of case management and disease management approaches is effective, inpatient stays and emergency room visits should decline over time. This trend plays out in CHW data. Sixty-five percent of hospital admissions and 70 percent of emergency room visits by CHW members took place in the first year of membership. Figure 2. Emergency Room Visit Rate by Membership Tenure in Community Health Works Rate 0.05 0.038 0.04 0.028 0.027 0.03 0.024 0.02 0.015 0.01 0 First 6 mos. 6–12 mos. 12–18 mos. 18–24 mos. More than 24 mos. Source: Custer, William. Georgia State University, Institute of Health Administration, 2004. CHW tracks program utilization and costs through an information technology (IT) system developed specifically for the program—an investment that was part of the board’s original sustainability plan. The CHW care management model and IT system have the 26 potential to create savings for other populations and programs, as well as generate additional dollars to sustain and grow CHW. Financial impact was measured in several ways: costs avoided because of the utilization improvements, the value of donated physician visits and reduced medication prices, and the new state and federal money that was brought into the community because of CHW. CONTRIBUTIONS The CHW model has demonstrated desired changes in utilization. The evaluation results show changed behavior in terms of inpatient hospital admissions and ED visits, resulting in subsequent decreased costs, and potentially improved health status. The hospitals perceive their contributions to the network to be cost-effective and physicians appreciate knowing upfront what their charity care will be and being part of a team that seeks to improve the health of patients. CHW has generated $13.6 million in free care to date through a model of case management and patient education. CHW also connects patients to free medications, which help them maintain compliance with their physicians’ health care plans. CHW aspires not to be a solitary program, but to create systemic change in the safety net. To that end, it has brought together leaders across the seven counties and across different types of organizations. For instance, prior to the development of CHW, the county commissioners reportedly had little involvement in health care and the safety net, but they now collaborate with the program and providers. The efforts of CHW have helped create a larger, more collaborative safety net for the uninsured. The safety net has improved dramatically for CHW enrollees. They have a medical home and know who to contact for services. In addition, the CHW model imparts a holistic approach focused on meeting the social need of its clients, such as transportation, child care, and employment search assistance, not just the clinical needs. Once patients are in the CHW system, they are treated for all their health care needs, including preventive screenings. While CHW has clear boundaries for membership, care is not necessarily limited to the four diseases, mostly due to the physicians’ wishes to treat patients in a comprehensive manner. CHW also brings an added benefit by providing a touchstone for local health care providers with a common mission. For example, if a smaller hospital has a 27 pressing need, it may feel more comfortable about contacting larger hospitals for assistance because of the relationships it has developed through CHW. Since 2001, CHW reports it has brought $4.3 million in new state and federal money to the seven counties in which it operates. The program estimates this additional revenue has produced more than $5.3 million in economic development to the area. CHALLENGES TO SUSTAINABILITY CHW has an operating budget of $1.66 million. However, a large portion comes from time-limited grant funding. The network does receive investments from hospital partners, who are firm in their support. The return on investment of the program to the hospitals surpasses the original investments, as CHW reduces emergency room utilization and hospital admissions for its members. While physician participation is not sufficient to allow the program to grow beyond its current scale, continued participation by physicians is essential to ensure sustainability. For the program to expand beyond its current scope, would require a stronger, funded primary care safety net, such as a federally qualified health center and more specialist volunteers. A source of even partial payment for physician participation would ease the burden of physician volunteerism. CHW has a strong and stable board and has weathered a transition in executive leadership over the past two years,. The new director has built on previously brokered relationships and has added strong business, economic development, and local political connections. The state environment is a factor in the success of community programs, including CHW. For example, an administrative change resulted in the loss of an expected partnership with Medicaid and an investment of state dollars. In addition, changes in state reimbursement rates to providers can change the willingness of providers to participate in the program. To increase the network’s flexibility and allow for expansion, CHW leaders are looking for more stable state, federal, and private investments. 28 CHAPERONING THE SYSTEM: CHOICE REGIONAL HEALTH NETWORK Olympia, Washington WHAT THEY DID Choice Regional Health Network began as a seven-hospital response to the threat of a hostile takeover by a for-profit hospital and has transformed into a vehicle to chaperones clients with complex needs through systems of care and coverage they characteristically have trouble navigating. Using a variety of funding sources including membership dues and fees paid by the hospitals, Medicaid match, federal grant programs, and private foundation grant programs, Choice has put together programs to serve 17,000 people. BACKGROUND Choice uses a multi-pronged approach to improve access to care for uninsured individuals, with incomes at or below 250 percent of the federal poverty level (FPL) who reside in a five-county service area. Choice enrolls eligible individuals in state-sponsored programs or links them to donated or discounted local provider services. The program benefits from collaboration with three hospitals, 11 outpatient clinics and federally qualified health centers (FQHCs), and hundreds of physicians. The Choice Board consists of 10 members representing hospitals, physicians, public health agencies, and communities. A three- member executive committee makes administrative decisions related to personnel and reviews financial reports. Choice has approximately 20 staff members and has operated with a budget of approximately $1.6 million for each of the past three years. The staff guide various projects, serve as “geo-leads” (i.e., enrollment specialists for a certain geographic area in the network’s region) in communities, and act as enrollment specialists. THE POPULATION SERVED The target population is the estimated 93,000 individuals who are uninsured, have incomes below 250 percent of the FPL, and live in the five-county service area. Choice has provided access to health services or coverage to 20 percent of the low-income uninsured in the five-county region. Most clients are under 39 years of age and most of the adult clients are employed in low-wage jobs that offer unaffordable coverage or no coverage at all. Almost half the clients have incomes below 65 percent of the FPL. Thirty-one percent of clients are Latino, compared with five percent of the region’s total population. The pharmacy assistance program connects Choice clients to pharmaceutical companies’ free or reduced-price drug programs. Most prescription drugs received by clients are for chronic conditions, including those associated with cardiac, mental health, diabetes, and asthma diagnoses. During the first five months of its pharmacy assistance 29 program, Choice assisted with applications for pharmaceutical products having a market value of more than $11 million. HOW THEY DID IT Choice board members are using the following question to guide the organization’s operations: “How can we plan together to reallocate the health care resources available; to improve access to primary care; improve the care patients receive; and, ultimately, improve the health status of the people who live in our region?” Several programs help them accomplish this mission. Through its Regional Access Program, Choice has developed extraordinary capacity and experience in enrolling—and keeping enrolled—those eligible for Washington’s various programs. Data from the state indicate 98 percent of Choice-assisted applications result in enrollment, compared with 40 percent when people attempted to enroll on their own. Further, 96 percent of the persons enrolled via Choice were still enrolled up to three years later, compared with 40 percent of clients who enrolled independently. 8 The Regional Access Program is staffed by counselors/enrollment specialists called access coordinators. When enrollment in public programs is not an option, access coordinators chaperone clients through the system to get needed health and social support. This model attempts to increase the use of some services (e.g., primary care and prescriptions) and reduce others (e.g., emergency departments). Access coordinators describe their clients as being in a severe cycle of poverty, which is difficult to break. The Access Coordinators work to break powerful behavior patterns to get people engaged in their own care. Geo-leads work in eight Choice communities to develop community collaboration, weave together available services for the uninsured, and build capacity where none exists. In the PharmAssist program, access coordinators help clients enroll in assistance programs offered by pharmaceutical companies. They also work to connecting clients to medical homes and care management services to change utilization patterns. Choice also offers Tu Salud, a program targeted to Latino clients with limited English proficiency. In addition to aggressive enrollment and chaperone programs, Choice is building other programs to strengthen the safety net or increase insurance capacity. One program is focused on building and organizing faith-based capacity (i.e., programs organized by churches, synagogues, etc.) and developing a community-based system to allow information sharing across multiple agencies. Another seeks to replicate the Project Access 30 concept, using physicians as volunteers to increase safety net capacity. Yet another program replicates the Muskegon, Mich. three-share model—in which employers contribute one-third, employees contribute one-third, and the third comes from a combination of local and state contribution and federal match. This model is designed to help small businesses and their employees have access to affordable health insurance. CONTRIBUTIONS Choice links low-income people to existing public insurance programs and also helps people find needed providers and services. Through its emphasis on medical homes and care management, the program changes utilization patterns. It is estimated Choice’s case management and care coordination efforts reduce the annual cost of care per client from an average of $4,000 to an average of $3,000 for a total savings of $3.5 million. 9 Choice helps shore up the financial stability and increase the capacity of the safety net directly through programs and indirectly through relationships. Of all Choice activities to date, the Regional Access Program appears to have the most significant positive impact on the safety net. Community informants report that the faith-based dental and community clinics have grown with the help of revenues received via Choice. Choices’s relationships with policymakers helped bring in the self-sustaining Sea Mar FQHC to the Olympia area. This allowed the hospital in Sea Mar to close a comparable clinic, that costs $100,000 to run annually. Choice also helped alleviate private physicians’ concerns about the potential for the FQHC to compete with their practices. Overall,$160,000 in member dues generated $410,000 in grants and Medicaid matching funds. With this total investment of $570,000, hospitals were estimated to receive $2.5 million in additional to reimbursement from patients who became insured. The member hospital return on investment has steadily increased to 20:1. In addition, low-income people have insurance and better access to care, and Choice has helped communities build their own programs. Given continued shortfalls in the safety net, Choice is broadening its reach to more community partners and exploring ways to generate more savings to cover more people. CHALLENGES TO FUTURE SUSTAINABILITY Indicators show the safety net is crumbling faster than Choice activities can repair it. The Choice director, Kristen West, reported she is hearing “more sad stories and fewer happy endings.” Obtaining access to certain services such as mental health care and substance abuse services is still considered extremely difficult. Several large grants are coming to an 31 end, and the financial futures of the smaller hospitals may interfere with ability to continue to pay dues. The executive leadership of Choice provides a driving force for the organization. The executive director has garnered and maintained statewide political interest and state agency engagement, ensuring some program stability. This leadership is both a benefit and a liability, with so much of the program’s success tied to one person. Choice has evaluated some components of the program and published some findings in reports, however, it has been more focused on building and delivering services than on measuring its direct impact on utilization and cost-effectiveness. 32 NOTES 1 Internal evaluation by Community Health Works, 2005. 2 Ibid. 3 E. M. Rogers, Diffusion of Innovations, 5th ed. (New York: The Free Press, 2003). 4 The Healthy New York program, a reinsurance program operated by the State of New York, uses such a formula to define small-group eligibility for Healthy New York coverage. 5 Internal review by administration of General Assistance Medical Program. Reported to the Milwaukee County Board of Supervisors, 2004. 6 Internal evaluation by Community Health Works, 2005. 7 For a full description of MEPS go to: http://www.ahrq.gov/data/mepsix.htm. 8 Report to the board of directors, Choice Regional Health Network, 2004. 9 Ibid. 33 RELATED PUBLICATIONS Publications listed below can be found on The Commonwealth Fund’s Web site at www.commonwealthfund.org. States in Action: A Bimonthly Look at Innovations in Health Policy. Newsletter. Closing the Divide: How Medical Homes Promote Equity in Health Care—Results from the Commonwealth Fund 2006 Health Care Quality Survey (June 2007). Anne C. Beal, Michelle M. Doty, Susan E. Hernandez, Katherine K. Shea, and Karen Davis. Aiming Higher: Results from a State Scorecard on Health System Performance (June 2007). Joel C. Cantor, Cathy Schoen, Dina Belloff, Sabrina K. H. How, and Douglas McCarthy. Pay-for-Performance in State Medicaid Programs: A Survey of State Medicaid Directors and Programs (April 2007). Kathryn Kuhmerker and Thomas Hartman. Health Care Opinion Leaders’ Views on Priorities for SCHIP Reauthorization (April 2007). Katherine K. Shea, Karen Davis, Anne Gauthier, Rachel Nuzum, Barry Scholl, and Edward L. Schor. State Strategies to Expand Health Insurance Coverage: Trends and Lessons for Policymakers (January 2007). Alice Burton, Isabel Friedenzohn, and Enrique Martinez-Vidal. Creating Accountable Care Organizations: The Extended Hospital Medical Staff (December 5, 2006). Elliott S. Fisher, Douglas O. Staiger, Julie P. W. Bynum, and Daniel J. Gottlieb. Health Affairs Web Exclusive (In the Literature summary). State Policy Options to Improve Delivery of Child Development Services: Strategies from the Eight ABCD States (December 2006). Neva Kaye, Jennifer May, and Melinda Abrams. 34