New Mexico has the nation's lowest per capita income and the highest rate of non-elderly uninsured, at 25.6%.1 Yet the University of New Mexico Health Sciences Center (hereafter “the Center”) competes in Albuquerque, a city with one of the highest penetrations of managed care (64% HMO, 20% PPO) in the United States. These realities necessitate dramatic changes by the Center to sustain its academic and service missions. As the state's largest safety net provider and sole academic health center, it faces a triple threat: (1) managed care organizations siphon off its historic sources of covered lives (Medicaid, Medicare, and CHAMPUS), (2) the number of uninsured patients grows while revenue to care for them shrinks, and (3) federal and state graduate medical education resources (direct and indirect graduate medical education and disproportionate share) decrease at alarming rates.
Major components of the Center include the school of medicine and University Hospital. Over 50% of the operating budget of the school of medicine derives from the faculty's clinical earnings, 33% of which come from Medicaid and Medicare and 38% from health maintenance organization (HMO) and insurance revenues. University Hospital and its affiliated clinics accommodate 100,000 patients, 25% of whom are on Medicaid and 30% of whom are uninsured. Of the latter group, two thirds, or approximately 15,000 to 20,000, are uninsured county residents, sometimes referred to as the “medically indigent” or the “working poor,” who have been deemed eligible for financial assistance by the hospital's eligibility office. The remainder of the uninsured group are “self-pay” patients who have not sought eligibility for an entitlement or assistance program, or patients who live out of the county, or those who are undocumented immigrants.
The hospital receives annual funding from its county's mill levy and from the federal government as “disproportionate share” for the care of uninsured patients. In fiscal year 1997, the totals were $25 million and $12 million, respectively. However, uncompensated care costs to the hospital exceeded $55 million, with an additional $10 million expended in physician costs. At the Center this unmanaged approach to care contributed to hospital losses in excess of $1 million per month at the same time that Medicaid and other revenue margins were narrowing. Without significant intervention, continued financial losses threatened both operational and educational programs.
In response, and despite considerable barriers (see below), the Center invested two years (1995 and 1996) in devising the UNM Care Plan. The Plan calls for reducing costs and improving quality of services by creating a managed care model for the estimated 15,000-20,000 uninsured Bernalillo County residents who met medically indigent criteria and who were seen at the Center. In April 1997, the UNM Care Plan enrolled its first patients. The Plan emphasizes primary care access with its attendant focus on long-term doctor-patient relationships and priority given to preventive services. What follows are a discussion of barriers to the Plan, a description of the Plan, and a review of its successes and failures after two years (July 1997 through June 1999 of operation.
BARRIERS TO IMPLEMENTING A MANAGED CARE PLAN
Primary care utilization
Analysis of utilization patterns of the unmanaged, uninsured population showed that few could identify their own primary care providers (PCPs). Coupled with access barriers to primary care providers, there was little disincentive for uninsured patients to seek care at the UNM urgent care center, the emergency department, or a specialist's office.
Fee-for-service information systems
Patient information systems were designed for the billing purposes of the hospital and physician practice plan rather than for quality management, patient care, or population management.
Accountability for quality and fiscal outcomes
County and federal financing for indigent care went directly to University Hospital without a clear stipulation as to how funds would be spent for their intended purpose. Characteristically, these funds were used to reduce hospital losses for uncompensated care. Providers' departments received a small portion of this total in proportion to their departments' contributions to care for the uninsured. The amount was so small, relative to the compensation the departments received for the care of insured patients, that a natural bias existed against time and money investment in services targeted toward the uninsured. The financial burden on providers imposed by the uninsured generated considerable resentment from many quarters.
Provider mix and location
Hospital and medical resource allocation for patient care favored investment in hospital-based, tertiary care and in procedural specialties over investment in community-based primary care. The allure of higher remuneration in a fee-for-service system led to faculty hiring practices heavily favoring specialists. Only 15% of the Center's physician faculty specialized in primary care, and they were concentrated on campus, with a relatively small number located in neighborhood clinics. This severe specialty and geographic imbalance left the institution ill-equipped to absorb the sharp increase in demand for accessible primary care that was predicted if the institution converted to managed care models.
Dwindling federal, state, and local revenues to sustain safety net providers caused intense competition between natural community allies. The Center (which is the public academic health center), the local network of community health centers (First Choice Community Health, Inc.), public health clinics run by the New Mexico Department of Health, and others too often found themselves in conflict rather than collaborating around a common health care goal. Services for underserved and vulnerable populations were fragmented in independent “silos” across the community.
CREATING THE PLAN
Because of the barriers just described, a “culture of service” to the uninsured had to be instilled at the institution among providers, administrators, and patients before a model of managed care could succeed for a population previously considered to consist of “charity” patients. A broad, inclusive planning process was adopted to ensure that the needs of major institutional stakeholders were addressed through this service transformation. Representatives of all major components of the Center participated in the two-year planning process. These representatives were from the University Hospital administration, the primary care and specialist leadership of the medical school, the UNM Mental Health Center, and the College of Nursing. External participants in the planning represented important components of the local safety net.
Planning committees learned together about the characteristics and needs of the uninsured population. Many misconceptions were addressed, resulting in a new appreciation for the characteristics of this population. For example, the great majority of uninsured were working but unable to afford health insurance, a substantial percentage were young, single women with children covered by Medicaid, and less than 10% of the patients consumed over 50% of the health care costs. This shared learning and planning process later helped create consensus around strategies for intervention.
As participants in an academic health center that always seemed to play “catch up” in the world of managed care, most of the academicians on the planning committees were enthusiastic about the opportunity to create and study a unique, cutting-edge service model for the largest single group of patients in their health care system.
Four Features of the New Plan
Four features characterize the UNM Care Plan that evolved:
A “health commons.”
Resources of existing safety-net stakeholders in the county are pooled (a “health commons”) to create a managed care program for uninsured patients. For example, a substantial portion of the University Hospital's Bernalillo County funds for indigent care were reallocated to primary care sites and to providers at university-run community clinics and to community clinics managed by First Choice Community Health.
Primary care “gateway” network
A community-based, primary care provider “gateway” is the model's cornerstone. Uninsured Bernalillo County residents apply for the UNM Care Plan at the University Hospital Business and Eligibility Office. The process includes verification of eligibility, education about plan benefits and enrollee responsibilities, and enrollment with a primary care provider (PCP) at a preferred primary care clinic site. PCP options include UNM primary care faculty, residents, nurse practitioners, or physician assistants practicing in one of seven UNM-run clinics located at the university or in an Albuquerque neighborhood. Alternatively, enrollees choose providers at one of five First Choice Community Health, Inc., clinics in Albuquerque. Women have direct access to obstetrician-gynecologists or to nurse midwives for their primary care. PCPs are organized into small practice groups (“pods”).
The Business and Eligibility Office of the hospital enrolls new patients who meet eligibility criteria and also enrolls established patients, previously deemed eligible for the “Bernalillo County Indigent Fund” at their annual renewal of medical indigent status. The latter enrollment mechanism avoided overwhelming the system with a single, mass enrollment of established, medically indigent patients. Eligibility requirements include an income less than 235% of the federal poverty guidelines. The Plan is aligned as closely as possible with Medicaid managed care, offering patients a seamless system whether or not they are enrolled in Medicaid. Each patient receives a UNM Care Plan identification card listing his or her PCP. Enrollees pay no monthly premium, but do pay a $5 or $10 co-pay for primary care visits depending upon income level ($5 if below 185%, $10 if between 185% and 235% of federal poverty guidelines), a $25 co-pay for urgent care visits, and $50 for emergency room visits, to encourage use of the appropriate venue of care.
The benefit package includes (1) choice of one's own primary care provider and designated primary care clinic, (2) reduced out-of-pocket cost of medications if selected from the approved formulary, (3) access to a 24-hour telephone triage system, and (4) services comparable to those available to Medicaid enrollees, except for behavioral health. To align physicians' and patients' behaviors with managed care principles, referral is required for specialist consultation. Prior approval is required for high-cost diagnostic or medical procedures. The measures were designed to reduce inappropriate self-referral to specialists by patients, and unnecessary or duplicate ordering of expensive tests by providers.
To create PCP access and to redirect uninsured patients from the emergency department, specialty clinics, and inpatient services, University Hospital and the school of medicine hired 12 new primary care physician faculty and five new family nurse practitioners. Clinic hours were extended in two clinics, and the number of clinical sessions per week were increased for most primary care faculty. Hospital staff support was increased in primary care clinics. Because First Choice Community Health, Inc., receives cost-based reimbursement, increased referral of plan patients to it from the university led to increased revenues in its system.
The primary care departments at UNM receive a capitation of $4 per plan member per month (pmpm) for uninsured professional service compensation. This is only around $2 pmpm less than the Medicaid professional primary care services capitation rate. By creating similar primary care professional service capitation rates, the system encourages primary care providers and clinics to expand their panels of plan patients and reduces incentives to exclude uninsured patients in favor of Medicaid or privately insured patients. Specialists were offered no unique financial incentives in the first two years of the plan. They are paid a sharply reduced fee-for-service rate compared with what they had been paid before introduction of the plan.
Investment in decentralized social services
Because social and economic circumstances play a major role in the underlying causes of illness and in patterns of accessing health care among the uninsured, case managers and social workers were moved from inpatient to primary care, ambulatory clinics. The UNM Mental Health Center moved alcohol and substance abuse counselors to two of the larger primary care clinics, while primary care departments hired part-time psychiatrists to supervise behavioral health care delivered by PCP residents, mid-level practitioners, and faculty. The reallocation of social services and case management from an inpatient to an outpatient setting during implementation of the UNM Care Plan demonstrated a willingness by the University Hospital administration to commit to such “front line” services for the uninsured in the hope of realizing a subsequent reduction in costly hospitalizations.
“Seamless” health care delivery
The program allows low-income patients to maintain a consistent health benefit package and primary care continuity as they move onto and off of Medicaid. The University Hospital Business and Eligibility Office can allocate presumptive eligibility for Medicaid or qualify the patient for the UNM Care Plan as appropriate. The model reduces the risk of fragmented care for indigent families in which the children have Medicaid and are cared for in one system while their parents are uninsured and cared for in an emergency room or not at all.
New structures had to be created to manage the Plan. These included expanded quality programs (e.g., specialty referral guidelines and an expanded disease management group that creates “care maps”), member services (e.g., offering panel management, addressing complaints, administering patient satisfaction surveys), medical information systems (e.g., creating relational databases, developing practice-support reports), and utilization management (e.g., conducting prospective and concurrent reviews, handling patient grievances). A medical director was hired to oversee utilization management, including the enforcement of a prior approval process. In addition, a Plan-specific formulary was developed by the HSC Pharmacy and Therapeutics Committee. Concurrently, primary care practice was reorganized with the development of a primary care management team, a primary care oversight body that consisted of each clinic's medical director and clinic manager. The primary care management team is charged with solving operational issues of the diverse primary care sites, and with creating consistency across the primary care clinical enterprise.
University Hospital contracted with Cimarron, Inc., a state-licensed managed care organization, to provide health plan education for the providers, track enrollment and disenrollment, monitor quality-performance measures (e.g., HEDIS 3.0 criteria), and regularly compile financial and resource utilization data for the hospital administration and providers. The data are used to monitor the Plan's effectiveness in such areas as appropriateness of service and cost of care.
Research to Assess the Plan
Data were collected on enrollee age, sex, and race—ethnicity. In addition, rates of patients' enrollment and disenrollment in the Plan and patients' preferences for primary care clinic sites and providers were tracked. Outcomes measures were access to care, utilization of outpatient primary care, specialty, and urgent care/emergency services, hospital admissions, and length of stay. Inpatient and outpatient costs of care were calculated on the basis of average facility and professional costs per hospital day, and on the basis of average facility and professional costs per visit to primary, specialty, and emergency department ambulatory sites.
Data were collected from two populations. One included all Plan enrollees one and two years after initiation of the Plan. The second included a subset of 3,721 enrollees who had been on the Plan continuously for two years (fiscal years 1998 and 1999), and who had been previously enrolled as “Bernalillo County Financial Assistance Patients” during fiscal year 1997, the year before initiation of the Plan, and could thus serve as their own pre- and post-Plan-enrollment comparison group. This group is relatively small compared with the size of Plan enrollees, in part due to the difficulty of identifying which patients were in the system before the Plan began. This group, however, represents a “high-user subset” of all Plan members.
Chi-square goodness-of-fit tests were used to compare outpatient visits, inpatient discharges, and numbers of hospital days across time periods. Expected values were calculated assuming equal distribution across the time periods for these variables. The Wilcoxon rank-sum test was used to compare lengths of stay across time periods.
THE PLAN'S FIRST TWO YEARS
County indigent patients enrolled in the program in the first year (April 1997-March 1998) at an average rate of approximately 1,000 per month and in the second year (April 1998-March 1999) at an average rate of approximately 350 per month. By the end of the second year, enrollment and disenrollment had reached nearly a steady state at just over 13,000 individuals enrolled in the Plan.
Characteristics of the enrollees show a predominance of young adult women (69.5% between ages 19 and 49 and 68.6% female). Of the 51.2% of the enrollees listing their race—ethnicity upon registration, 56.0% were Hispanic, 30.8% were Caucasian, 7.5% were Asian, 4.0% were African American, and 1.7% were Native American.
Seventy-four percent enrolled with PCPs in one of the seven UNM-affiliated primary care clinics (54% with family medicine, 43% with internal medicine, 2% with pediatrics, and 1% with obstetrics—gynecology). Approximately 50% chose faculty, 33% chose residents, and 17% chose mid-level practitioners as their primary providers. Twenty-six percent chose PCPs in one of the five First Choice Community Health, Inc., clinics. In both health care systems, enrollees have most commonly chosen PCPs on the basis of their historic relationships to particular providers or sites, or proximity of the clinics to their homes. Over 80% of UNM Care Plan patients complied with the $5 or $10 co-pay at the time of service for primary care clinic visits.
Average patient wait times for routine appointments with PCPs exceeded 45 days (15 days above New Mexico Medicaid criteria) at the onset of the Plan. After PCP expansion, personnel shifts, and expansion of provider and clinic hours, the average wait dropped to 28 days in the first enrollment year, and rose slightly to 32 days in the second enrollment year.
The patterns of health care service utilization changed after institution of the Plan, both for the enrollees as a whole and for the subset of high users. For the overall Plan, primary care, specialty, and emergency outpatient visits per member per year and inpatient hospital discharges and hospital days per 1,000 enrollees decreased significantly (Table 1). For the “high-user subset,” outpatient primary care and specialty visits increased significantly, emergency visits remained unchanged, while hospital discharges rose then fell significantly by the second year of the Plan operation (Table 2). The hospital days per 1,000 enrollees had decreased significantly by the second year of Plan operation.
An assessment of the impact of the Plan on the cost of outpatient and inpatient care revealed estimated savings of $148 per member per year for the 13,114 enrollees in FY 1999. This savings was due to the marked reduction in the enrollees' total hospital days by the end of the second year of Plan implementation. In spite of the increase in ambulatory care utilization, University Hospital realized $1,940,872 in estimated savings for fiscal year 1999. Because University Hospital is often at admission capacity, the reduced number of admitted uninsured patients can often be replaced with paying patients. Such replacements did occur: after two years of the Plan, the replacement of the unpaid days with paying admitted patients yielded $695,000 in revenues (695 fewer unpaid admitted patients, replaced by paying admitted patients at the medical/surgical reimbursement rate of $1,000 per day).
WHAT WE LEARNED
Requirements of the Model
Those planning to implement a managed care model for the uninsured should expect to see an early increased demand for ambulatory care, especially primary care services, coupled with a decrease in avoidable hospitalizations. Consistent with our study's outcomes, most studies show an early, but transient, increase in utilization of health services for uninsured patients newly enrolled in a managed care plan. Bogard et al.2 found a 30% increase in outpatient visits during the first year of enrollment of low-income uninsured patients in a health maintenance organization, comparable to our increase of 22%. Martin et al. found a pent-up demand for services by the uninsured new enrollees in Washington State's Basic Health Plan, especially among those without coverage for over five years.3 This is consistent with a study by the Congressional Budget Office, which estimated that the uninsured receive only 60% of the care received by the medically insured.4 Enrollment in primary care continuity failed to decrease re-hospitalization rates in a severely ill veterans' population.5 However, in a study group more comparable to the uninsured population enrolled in the UNM Care Plan, and more consistent with the Plan's effect of reducing hospitalization rates and inpatient costs, access to primary care continuity demonstrated its effectiveness in reducing hospitalizations in a fee-for-service Medicaid population.6
Because social and economic factors play a significant causative role in illness among the uninsured,7,8 a substantial investment in social support services is required to effectively manage the health of an uninsured population. Complementing the important “gateway” role of primary care providers, support services ranging from medical translation and assistance with transportation to referral for job training and literacy programs assume a vital role in improving the health of this vulnerable population. Ready availability of primary care case managers correlates with substantial reductions in emergency room utilization in Medicaid demonstration projects.9
Ease of access to care must be a featured component of any managed care plan for the uninsured. Eighty-four percent of the families of uninsured patients are supported by working adults.10 These adults often work in jobs with little security and few benefits, so that many uninsured employees are reluctant to jeopardize their jobs or forego income during working hours. Access to evening hours and availability of 24-hour telephone triage is thus a major incentive to use clinics rather than emergency rooms.
Limitations of the Model
Several challenges limited the Plan's overall effectiveness in its first two years of existence:
Lack of specialty incentives
While PCPs were capitated, specialists continued to receive reduced fee-for-service for UNM Care Plan patients. With better-aligned specialist and facility incentives, and coordination between primary care and specialist services, the cost savings are likely to be greater than those achieved by capitating the PCPs alone. One sign of reduced overutilization of specialty services was a 20% drop in outpatient specialty care received by Plan members from year one to year two, which occurred with the implementation of prior-approval rules for PCP referrals to specialists.
Lack of a uniform behavioral health plan benefit
While some behavioral health services were integrated into primary care sites, the bulk of mental health resources remain sequestered in a separate UNM Mental Health Center. This fragmented behavioral health from primary care. The urgency to link medical and behavioral health is accelerated by a dramatic reduction in Medicaid patient volume and revenue in the UNM Mental Health Center. External forces will compel a closer linkage between behavioral and medical care. Both the New Mexico Department of Health and the Medicaid behavioral health organizations are pressing for less emphasis on inpatient and residential services and more emphasis on community-based behavioral health services in primary care settings. In the year 2000, several activities will be undertaken: The UNM Mental Health Center will implement a managed system for the uninsured tied to the UNM Care Plan model. Each uninsured behavioral health patient will have his or her PCP assigned at intake so that the PCP will be notified of the encounter. In addition, the Center is working with the city for substance abuse and alcohol detoxification services for this population. Finally, the institution is looking for opportunities to better merge behavioral services with medical services.
A considerable disenfranchised population remains
Many indigent patients have been left out of the Plan. There are an estimated 5,000 uninsured individuals in the county who utilize Center providers who are not enrolled, and another 10,000 patients designated “self-pay.” Most are indigent and, as a group, pay only 10% of what they are billed. It is estimated that 80% of these would qualify for Medicaid or for the UNM Care Plan if they applied. Without enrollment in the Plan, they have no assigned PCP, their care is not managed, and they can generate considerable unnecessary costs to the system. To motivate “self-pay” patients to enroll in the UNM Care Plan, the institution is charging a substantially higher co-pay for services for uninsured, non-Plan patients. Care must be taken to protect those who are ineligible for the Plan, such as patients legally residing in a different county, or undocumented immigrants. The Center is committed to caring for patients regardless of ability to pay, and to reducing obstacles that prevent access to care.
Implications of the Model
Public safety-net providers, including federally qualified health centers, departments of health, public hospitals, clinics, and public academic health centers, face reduced financial support and accelerated privatization of health services. Consequently, they are at risk of reducing services to the uninsured.11 As more states receive waivers to enroll their Medicaid recipients in managed systems, this traditional funding stream for safety-net providers moves to the private sector while the uninsured remain in the public systems. Few states have emulated Oregon and Tennessee in incorporating the uninsured into their Medicaid waiver programs.12 Safetynet providers must identify natural allies and manage health care under pooled budgets.
A compelling community benefit can be realized through the application of efficient population-management principles.13 Unlike for-profit systems, margins realized by public safety-net providers can be reinvested to expand health and social services for the uninsured. In such a public system, the “stockholders” (taxpaying citizens) must be informed of the community dividends that derive from their “investment” with clear, concise reports that herald progress, or illuminate the interventions planned to overcome obstacles.
THE “HEALTH COMMONS” AND GOALS OF THE DIFFERENT STAKEHOLDERS
If a “health commons” approach to developing managed care for the uninsured is pursued, the approach will necessarily involve many stakeholders with differnt motivations for participation. We found that the definition of the Plan's success in our county varied with different stakeholders' values and priorities. The “bottom line” for some at the Center was to achieve a profitable margin or at least to reduce the magnitude of financial loss in caring for the uninsured. In the past, expensive specialist and inpatient facility resources were expended for problems that could have been managed by PCPs. By appropriately managing the uninsured, new specialist and facility capacity is created for insured patient visits, consultations, procedures, and admissions.
For others, the “bottom line” is improved access, dignity, and equity in health care for this group. Many providers work in public academic health centers and community health centers because of their commitment to underserved populations. Market-driven managed care widens the disparity in care between the “haves” and the “have nots.” Those with commercial insurance often receive preferential access to care and appointments, while uninsured patients wait for hours in a clinic or emergency department. The UNM Care Plan reallocates public resources to “level the playing field” in the minds of many providers and patients. Uninsured patients now have their own PCPs, and receive appointment times. Their doctors are paid to see them, not asked to donate “charity care.” This adds dignity to services received by indigent patients and prevents a discriminatory, two-tiered system of care.
The success of the UNM Care Plan has hinged on the ability of University Hospital, the school of medicine faculty, and community health centers to collaborate. Each had to put their time and resources at risk. Each saw their self-interest better served by sharing resources, decision making, authority, and responsibility. After two years of operation, University Hospital administrators, the school of medicine faculty, and community health center personnel perceive the UNM Care Plan as a qualified success in reducing costs while assuring higher-quality service.
1. U.S. Bureau of the Census, Statistical Abstract of the United States: 1997. 117th ed. Washington, DC: Government Printing Office: 121.
2. Bograd H, Ritzweller DP, Calonge N, et al. Extending health maintenance organization insurance to the uninsured: a controlled measure of health care utilization. JAMA. 1997;277:1067–72.
3. Martin DP, Diehr P, Cheadle A, et al. Health care utilization for the “newly insured”: results from the Washington Basic Health Plan. Inquiry. 1997;34:129–42.
4. Behavioral Assumptions for Estimating the Effects of Health Care Proposals. Washington, DC: Congressional Budget Office, 1993.
5. Weinberger M, Oddone EZ, Henderson WG, et al. Does increased access to primary care reduce hospital re-admissions? N Engl J Med. 1996;334:1441–7.
6. Gill JM, Mainous AG. The role of provider continuity in preventing hospitalizations. Arch Fam Med. 1998;7:352–7.
7. Lantz PM, House JS, Lepkowski JM, et al. Socioeconomic factors, health behaviors, and mortality. JAMA. 1998;279:1703–8.
8. Lynch JW, Kaplan GA, Salonea JT. Why do poor people behave poorly? Variation in adult health behaviors and psychosocial characteristics by stages of the socioeconomic life course. Soc Sci Med. 1997;44:809–19.
9. Hurley RE, Freund DA, Taylor DE. Emergency room use and primary care case management: evidence from four Medicaid demonstration programs. Am J Public Health. 1989;79:843–6.
10. Rowland D, Feder J, Keenan PS. Uninsured in America: the causes and consequences. In: Altman SH, Reinhardt UH, Shields AE (eds). The Future U.S. Healthcare System: Who Will Care for the Poor and Uninsured. Chicago, IL: Health Administration Press, 1998:25–44.
11. Liska DW, Brennan NJ, Bruen BK (eds). State Level Databook on Health Care Access and Financing. Washington, DC: The Urban Institute Press, 1998.
12. Friedman E. California public hospitals: the buck has stopped. JAMA. 1997;277:577–81.
13. Mirvis DM, Chang CF, Hall DJ, et al. TennCare: Tennessee's answer for health care reform. JAMA. 1995;274:1177–250.
14. Kindig DA. Purchasing Population Health: Paying for Results. Ann Arbor, MI: University of Michigan Press, 1997.© 2000 Association of American Medical Colleges