Brian Burwell and Sandra Tanenbaum
Managed care financing and delivery models have considerable potential for improving the value and quality of health care and supportive services provided to children and adults with disabilities. Managed care models that encourage flexibility in benefit coverage and which coordinate care across the full spectrum of the insurance benefit package are features that are particularly attractive to persons with disabilities. At the same time, however, managed care incentives to eliminate "inappropriate care" or care that is not "medically necessary" are of great concern to people with disabilities whose experience in obtaining access to needed health care services in the fee-for-service system is already problematic.
Both positive and negative effects of managed care for persons with disabilities are similarly reflected in the limited empirical research that has been conducted to date on the impacts of managed care on disabled populations. Some studies point to improvements in outcomes, while others have found significant reductions in service levels under managed care incentives. In brief, the jury is still out on how managed care models effect the health care status of persons with disabilities, and the challenge to the health care services research community is to monitor the enrollment of persons with disabilities into managed care systems closely, and to identify those factors which contribute to improved and worsened outcomes for these vulnerable populations.
CHILDREN AND ADULTS WITH DISABILITIES: WHO ARE THEY?
Part of the challenge in assessing the impact of managed care on persons with disabilities is that the population of children and adults with disabling conditions is extremely diverse, with broad-ranging differences in both types and levels of impairment. At the same time, managed care models are evolving into a variety of permutations that make the generalizability of managed care impact studies increasingly hazardous. In conceptualizing a research agenda for examining managed care impacts, it is critical that we begin with a fundamental understanding of the defined populations, and how the structure and incentives of managed care models may impact access, cost and quality outcomes for persons with severe and chronic disabling conditions.
Children with Disabilities
National survey data indicate that approximately one in ten children have a "severe chronic illness" (Neff and Anderson, 1995). This estimate obscures dramatic diversity in the characteristics of children with disabling conditions--many children with disabilities have conditions which do not result in health care use or costs significantly higher than the population of children without disabilities, while a significant minority of children with disabilities have severe and multiple conditions that require continuous and expert medical attention. Health care and supportive services for the population of children with special health care needs are also fragmented across a variety of financing and service settings that renders the transition from a fee-for-service framework to a managed care framework operationally cumbersome.
Importantly, within the population of children with high health care needs, there is a subset of children with extremely severe medical conditions that require continuous and highly specialized care. For example, within the target population of SSI children receiving services under the District of Columbia's Managed Care System for Disabled and Special Needs Children Demonstration, a Medicaid Section 1115 waiver program, children with Medicaid expenditures of over $50,000 per year constituted less than three percent of all SSI children in the District in the year prior to implementation of the demonstration program, yet they accounted for about 54 percent of all Medicaid spending for SSI children (Blanchon, 1996).
Childhood disability differs from disability in adulthood in that the nature and extent of the disability frequently changes during the developmental process. Many children experience improvements in functioning as they develop, and the disability may become less limiting with time. Other children with extremely severe medical conditions do not survive childhood at all. Moreover, the health care needs of children with disabilities is confounded over time by the interaction of the disability with the child's normal development, such as the onset of puberty. Consequently, access to appropriate pediatric and adolescent specialists may change frequently during the developmental process.
In regard to accessing health care, parents obviously take an active role in negotiating the health care system for their children. In brief, many parents take on the "coordination of care" role that is generally lacking in the fee-for-service system. Consequently, their interactions with the care coordination function of a managed care system may require a new accommodation of respective roles in managing the care of the disabled child. Managed care organizations are generally not used to the level of advocacy and health care system knowledge exhibited by parents of children with disabilities, and may not know how to positively incorporate that energy and knowledge into their internal care coordination systems.
A common concern of parents is the ability to maintain relationships with pediatric specialists, many of which have developed over the lifetime of the child, once the child is enrolled into a managed care plan. Consequently, in some Medicaid managed care initiatives, states require participating plans to continue to pay for ongoing physician-patient relationships, even if the specialty physician is not otherwise enrolled in the plan. This issue is of obvious concern to plans who feel that they are being paid to manage the care of the enrollee, but may not be given all the requisite tools to do so.
Children with disabilities differ from adults with disabilities in one other important respect--children are more likely to receive their health care through a fragmented financing system. Expansions in SSI and Medicaid eligibility for children with disabilities in recent years has meant that there are a growing number of children who have both private health care insurance and Medicaid coverage. Since the Medicaid benefit package is more comprehensive than private health insurance coverage, children and families often use their Medicaid coverage to finance services that are supplemental to their private insurance benefits, particularly home and community-based services and extended therapies. In addition, under the Individuals with Disabilities Education Act, local school systems are required to provide children with disabilities with educationally related services that often extend into the health care arena, particularly in the case of children with severe medical conditions. Lastly, under the Title V Program for Children with Special Health Care Needs, many states provide direct care services to children with disabilities on a categorical basis, not as part of the child's health insurance benefit. Since the implementation of managed care systems generally occurs within payers, not across payers, these multiple financing streams for children with disabilities create special challenges for the managed care marketplace.
Adults with Mental Illness and Substance Abuse Problems
Purchasers of health care services in both the private and public sectors have targeted services to persons with mental illness as prime candidates for managed care financing and delivery initiatives. In the private sector, many large companies have "carved-out" mental health and substance abuse benefits from their mainstream health care benefit programs, and have contracted with specialized vendors to administer these benefits. In the public sector as well, state Medicaid programs are building upon the infrastructure that has developed in managed behavioral health care to similarly "carve-out" at least a subset of mental health and substance abuse-related services covered under their own benefit packages to companies that specialize in the management of these services. On the research side, there is a larger body of research available on the impacts of managed mental health care than on how managed care impacts other services and/or populations.
While there has been significant penetration of managed care systems in the mental health/substance abuse market, it is important to recognize the differences in private and public markets as they relate to persons with mental health and substance abuse problems. In the private sector, purchasers finance mental health and substance abuse care for their employees, retirees and dependents. This population of workers and dependents is predominantly middle class and employed, with the concomitant array of mental health conditions that are most prevalent in this socio-demographic group. Depression and substance abuse disorders are diagnoses of high concern to private purchasers of health care, and the health care benefit programs of employers are structured to maximize value in the early identification and treatment of these conditions, with the objective of sustaining the productivity of their workforces.
In regard to coverage of mental health and substance abuse services for the dependents of employees, the goals are to provide coverage that is sufficiently attractive to recruit and retain a quality workforce (i.e. remaining competitive in the market for qualified workers) while limiting corporate expenditures for mental health and substance abuse care. Coverage of mental health and substance abuse care for adolescents with mental health conditions is often a major benefit issue for employers, since this population includes a subset of persons who account for a high percentage of total expenditures for these services.
In the public sector, the primary population of interest is persons with severe and persistent mental illnesses, particularly persons with disabilities associated with schizophrenia-related disorders. Approximately 30 percent of all adults under the age of 65 receiving SSI benefits, or about 1.5 million persons, qualified for SSI benefits on the basis of a mental disorder other than mental retardation (SSA, 1996). In addition, about 1 million persons with mental disorders received SSDI benefits, and are therefore insured under the Medicare program. As opposed to individuals receiving SSI benefits, persons receiving SSDI benefits have had a sufficient work history to obtain insured status under the Social Security disability system. On the whole, it is therefore reasonable to assume that SSDI beneficiaries have somewhat higher levels of functioning than persons receiving SSI.
Persons with severe and persistent mental illness have a broad range of medical, therapeutic, and supportive care needs, and a key issue in the application of managed care models to this population is what part of the care spectrum should be "managed." Although a number of state Medicaid programs have implemented mental health "carve-out" programs, it is important to recognize that states generally have only "carved-out" acute mental health services under these programs--inpatient care and outpatient follow-up care. Long-term supportive services, such as residential care programs, vocational training, day program services, and intensive case management services, have generally been excluded from the managed care contracts with carve-out vendors. Basic health care services are also usually provided by mainstream plans or the fee-for-service system.
The characteristics of persons with severe and persistent mental illness and their health and supportive service needs forcefully underscore the challenges of applying managed care models to the financing and delivery of services to this population. As a consequence, we are seeing a variety of managed care models emerging. Conceptually, one relatively simplistic way of classifying the service needs of this population is in three broad categories: (1) basic health care needs; (2) mental health-related services needed to deal with acute episodes of mental illness (short-term hospitalization, crisis intervention services); and (3) long-term supportive services intended to maintain individuals in independent or semi-independent community care settings.
As discussed above, most managed care initiatives for persons with severe and persistent mental illness have focused only on the management of one part of the total service continuum, i.e. the management of short-term hospitalizations and outpatient services. Basic health care services and long-term supportive services have, with few exceptions, not been made part of state managed care initiatives, as yet. A major reason for this segmentation of the total benefit package is related to infrastructure issues--states are building upon the infrastructure of managed behavioral health care vendors that have developed from demand created in the commercial marketplace. Another reason for this segmentation relates to the fragmentation of payments sources; Medicaid is generally the primary payer for acute mental health services for this population, while state Departments of Mental Health remain the primary payer for longer-term supportive services.
The limited scope of managed care initiatives for persons with severe and persistent mental illness has created "boundary" issues that affect the operationalization of these programs in critical ways, as well as how this population receives services. One fundamental issue is the boundary between mental health care and basic health care. Does it make sense for persons with severe mental illness to receive their primary health care through one system but have their "mental health" services managed by a separate system? If so, how is medication management coordinated across these dual systems? One major rationale for managed care is to coordinate care across a comprehensive benefit package for an enrolled population, and managed care initiatives which simply mirror the fragmentation of service delivery existent in the fee-for-service system are likely to fall short of this goal.
On the other hand, some state Medicaid programs have "carved-out" mental health services from managed care contracts for basic health care as a means to protect the population from the financial incentives of managed care to reduce services that may not be considered "medically necessary." There is considerable controversy in the commercial insurance market about the "savings" that have been achieved for health care purchasers by behavioral managed care vendors, and whether these savings are affecting mental health outcomes. Thus, in the public sector, mental health carve-outs have been used as a policy tool to protect mental health benefits from the incentives of managed care plans, most of whom have little experience in providing services to persons with severe and persistent mental illness. However, another factor in states' decisions to carve out mental health benefits has been advocacy by the specialized provider systems that serve this population to protect their market share.
Another boundary issue in designing managed care systems for persons with severe and persistent mental illness is whether to combine substance abuse programs with mental health services into an integrated managed care system. Although programmatically, there are strong reasons for bundling mental health and substance abuse benefits for this population in an integrated system, infrastructure issues and provider concerns often act to keep these services unbundled.
A final issue regarding the application of managed care models to persons with severe and persistent mental illness concerns the measurement of plan performance. What measures should purchasers (public or private) use to assess whether plans are doing a "good job?" Persons who support individuals with severe mental illness know that interventions of the highest quality can still lead to undesired outcomes in some individuals, while in other cases, people with mental illness somehow seem to get better or do okay despite inferior care or the absence of care. The relationship between good care and positive outcomes in this population is not straightforward, and the assessment of performance probably needs to measure average outcomes over sufficiently large samples of individuals, wherein the differentiation between inferior and superior care can be more reliability discerned.
Adults with Physical Disabilities and Persons with AIDS
The population of persons with severe and chronic physical disabilities, including persons with multiple sclerosis, cerebral palsy, muscular dystrophy, quadriplegia, and other conditions, encompasses a very broad range of disabilities and impairment levels. Persons with severe physical disabilities are often not well served by the fee-for-service health care system, and many experience the frustration of referrals to multiple specialists without any single physician taking overall responsibility for the oversight of their health care. If the care coordination functions of managed care models truly take hold, then managed care holds some promise for improving access and quality for persons with severe physical disabilities.
However, as with other disabled populations, many people with severe physical disabilities are skeptical that managed care organizations will provide them with access to comprehensive and coordinated medical care. Many worry that managed care organizations will be stringent in the allocation of resources in meeting their medical needs and will perceive them as "undesirable" enrollees, particularly if the cost of their care exceeds the average premium paid by their sponsor, be it an employer, Medicare, or Medicaid. For persons who require highly specialized care, many worry whether managed care plans will deny access to the most qualified specialists, and/or specialists with whom they have developed long-standing relationships.
On the purchaser side, private employers generally place little emphasis on ensuring that covered individuals with severe disabilities are adequately served in the managed care system. The disability programs of employers generally focus on short-term disability issues; the integrated management of their health insurance, workmen's compensation, and disability insurance programs; and rehabilitation initiatives which assist injured workers' to return to work as quickly as possible. The quality of health care provided to persons with severe and chronic conditions is generally not an issue of high concern to most private employers. Furthermore, the assessment of the performance of managed care plans by employers has largely focused on measures that are pertinent to large segments of their covered populations (e.g. prenatal care, immunizations, etc.) rather than on how plans treat individuals with rare conditions.
For persons with severe disabilities who do not have private insurance and are covered by Medicaid, it appears that mandatory enrollment in some kind of managed care system is increasingly inevitable. With completion of the enrollment of non-disabled Medicaid populations into managed care, states are now focusing their attention on the more difficult challenge of enrolling SSI recipients into managed care (Checkett, 1996). And unlike persons with severe mental illness, mental retardation and/or developmental disabilities, persons with severe physical disabilities generally do not have specific "sponsors" or "programs" within state government whose responsibility it is to look out for their welfare. Just as the needs of persons with physical disabilities often fall through the cracks in the current Medicaid system, there is equal danger that the needs of this population will be largely ignored in the headlong rush to achieve Medicaid savings through managed care approaches.
In contrast, persons with AIDS are receiving special attention in the development of Medicaid managed care models. Led by the model developed by the Community Medical Alliance in Boston, the concept of "specialized health plans" (SHPs) which target a single population type, is now being replicated in other states such as Maryland and New York. Specialized health plans are generally perceived as voluntary alternatives to mainstream managed care plans, rather than mandatory alternatives that persons with certain conditions would be required to enroll in. The development of specialized plans is not totally attributable to demand side factors. Another factor is that specialized provider networks with experience in providing health care services to specific populations want to be able to preserve their "product line" without having to diversify into being mainstream health plans.
The Community Medical Alliance model for managing the care of persons with AIDS places strong emphasis on the substitution of non-institutional care arrangements for institutional care, particularly during the terminal phases of the illness. The recruitment and training of medical care professionals that are committed to the treatment philosophy and culture of the Community Medical Alliance is another key component of the model.
Areas of Commonality Across Populations
Although children with severe disabilities, persons with severe and persistent mental illness, and adults with physical disabilities possess diverse characteristics that raise unique issues in the application of managed care models, there are some common characteristics shared by all of these populations. First, persons with severe disabilities of all types require access to specialty services that may be limited under managed care approaches. Closed panel plans may have few or no physicians with expertise in the care of conditions with low prevalence rates in the general population. Point-of-Service plans may allow enrollees to seek care outside of their networks, but at a higher cost to enrollees, who may have limited financial resources to utilize out-of-network providers.
Second, the health care costs of disabled populations are more predictable than the health care costs of non-disabled populations. Not only are they more predictable at the population level, but also, in many cases, at the individual level. This creates opportunities for health plans to maximize profitability by adopting business strategies to limit the enrollment (or increase disenrollment) of individuals whose health care costs are predictably above the payment rate made to the plan. Risk adjustment strategies which pay plans fairly for the expected costs of persons with disabilities, yet which still reward plans for efficient care, are critical to the application of managed care models to these populations, as well as to ensuring that persons with disabilities are provided quality care by the plans in which they are enrolled (Kronick et al, 1996). However, alternative mechanisms, other than risk adjusted capitated rates, for financially rewarding plans which enroll higher-cost individuals and providing quality services, also need to be explored. Risk-adjusted capitation may prove not to be the best solution to addressing these incentive issues, particularly given the technical and operational challenges of measuring risk and adjusting payments appropriately.
Third, the development of performance measures, which reliably assess the relative performance of plans in providing medical and supportive care to persons with disabilities of all types, is an area that requires extensive work and development. Workable approaches to eliciting the perspective of consumers, many of whom may have disabilities which impede traditional survey methods, is a key issue in the development of such measures.
Fourth, it is frequently the case that people with disabilities are receiving services from multiple payment sources and programs concurrently. The development of managed care models for these populations must respond to a set of needs that are broader than the financing and delivery of medical care. If care for these populations is to be truly integrated, then models need to be developed which consolidate the financing and delivery of health care services, rehabilitative services, long term care services, family supports, respite care, occupational supports, and personal counseling within integrated organizational structures. It may not be necessary for a single organization to possess all of these capabilities, but a managed care approach to these populations must include mechanisms for effectively coordinating the full array of medical and related services that are needed to help persons with disabilities maintain the highest level of independence possible.
What Does the Research Tell Us About the Impacts of Managed Care?
Empirical research which directly measures the health outcomes of persons with disabilities in fee-for-service versus managed care settings is extremely limited, and the research which has been conducted does not paint a consistent picture of the impacts of managed care. Research on the impacts of targeted managed care initiatives seems to paint a more positive picture, while general population studies of managed care impacts are more pessimistic. Also, considerably more research has been conducted of the impact of managed care on mental health populations than on populations with other types of disabilities.
Master et al (1996) describe improved outcomes among persons with severe disability and AIDS in a targeted Medicaid managed care program in Massachusetts. Positive outcomes included increased patient satisfaction, reduced inpatient hospital days, and improved decubitus ulcers and PCP. The study suggests that managed care can improve care for persons with severe disability through the use of innovative providers providing care in innovative settings, relative to the fee-for-service system. The results of this research may be questioned, however, given that the researchers also represent the senior management team of the managed care organization being studied. Similarly, Meyers et al (1987) found improved outcomes from managed care in a population of severely disabled adults in an independent living center, largely associated with increased resource allocation to care provided in the individual's home and centered around the person's individualized needs.
In an 11-year longitudinal study of persons with rheumatoid arthritis receiving care in fee-for-service settings versus prepaid group practice, Yelin et al (1996) found no evidence of differences in either the quantity of health care provided or in health care outcomes on either an annual or long-term basis across the systems of care.
Studies of populations in mainstream managed care plans seem less positive. An analysis of data from the Medical Outcomes Study (Ware et al, 1996) found that while health outcomes for the average patient did not differ between fee-for-service and managed care settings, health outcomes were decidedly poorer for patients who reported ill health at baseline. The study suggests that while managed care plans do quite well in maintaining the health of healthy patients, relative to fee-for-service, that people with higher medical needs fare less well in managed care, due to financial incentives among plans to reduce the level of resources applied to medical interventions. The findings of the Medical Outcomes Study support similar findings by the same research team ten years previously (Ware et al, 1986). Although the population of interest in the Ware study encompassed "chronically ill" persons, not persons with severe disabilities, it is reasonable to generalize the study findings to all populations with higher-than-average medical care needs. In another study of data from the Medical Outcomes Study, Safran et al (1992) found notable differences in dimensions of primary care provided to persons with chronic illness across fee-for-service plans, IPA-model plans, and traditional HMOs, but did not specifically associate these differing primary care paradigms with patient outcomes.
Research on the impacts of mental health managed care models is decidedly richer. The Medical Outcomes Study reported above found superior mental health outcomes in managed care for nonpoverty populations, but inferior outcomes in the poverty group (Ware et al, 1996). Wells et al (1990) found that one managed care network provided less intensive mental health services to their covered population but a higher quantity of services. Lurie et al (1992) found few differences in mental health outcomes among patients served in managed care versus fee-for-service with one exception--persons with schizophrenia showed superior outcomes in a fee-for-service setting. And in a study focusing on a population of persons with depression, Rogers et al (1993) reported that depressed patients declined, on average, in managed care settings, declines that were likely attributable to a drop-off in the prescription of anti-depressant drugs.
Other studies have reported more positive impacts of mental health managed care initiatives. Superior mental health outcomes under managed care, as well as reduced financial impacts on patients, were reported by Babigan et al (1992). Shern et al (1995) also reported greater reductions in problems, fewer unmet needs, and higher adherence to clinical protocols, among mental health clients in a managed care demonstration than in a comparison fee-for-service population.
A few studies have evaluated the impacts of mental health carve-out programs for Medicaid populations, and thus far, have generally reported favorable outcomes. Callahan et al (1995) conducted an evaluation of a Medicaid mental health carve-out in Massachusetts and reported that the carve-out vendor was successful in substantially lowering Medicaid costs for acute mental health services without any overall reduction in quality or access. Christianson et al (1995) also reported significant reductions in Medicaid expenditures for mental health services in the first year of a carve-out initiative, primarily due to reductions in inpatient admissions for mental health treatment, although mental health outcomes were not measured.
Studies that assess the impact of managed care on children with disabilities are very few, although a number of researchers have published on the potential dangers of managed care systems on children with disabilities. Fox et al (1993) reported findings from a survey of parents of children with disabilities, with mixed results. Parents were pleased with the reduced out-of-pocket costs associated with managed care systems, and with improved access to medical services, but at the same time reported increased difficulty obtaining access to specialty services and mental health care. The focus of managed care plans on requiring specialty care interventions to demonstrate rapid improvement was cited as a significant concern, and a barrier to care continuity.
Discussion: Is Managed Care for Children and Adults with Disabilities a Step Forward or a Step Backward?
Research on the impacts of managed care on children and adults with disabilities is decidedly mixed. The limited body of research published to date seems to suggest that the incentives of capitated financing mechanisms are not, in and of themselves, the primary determinants of outcomes. Rather, the research suggests that operational variables, i.e. how managed care models are applied, are equally important, if not more important, in determining how people with disabilities fare in the managed care world. Of particular interest is the nature of the managed care entity with whom the purchaser has contracted to provide care. Managed care organizations with missions to serve persons with disabilities, and organizations who provide specialized services, appear to achieve better outcomes for persons with disabilities than do mainstream plans which have no special focus on the needs of disabled populations.
As managed care models continue to evolve, and as purchasers increasingly pursue innovative managed care purchasing strategies, it will be increasingly important for researchers to help sort out which managed care models are associated with improved outcomes and reduced costs versus those managed care models which achieve reductions in health care costs only to the detriment of the populations they are intended to serve.
Babigian, Haroutoun M., Olivia S. Mitchell, Phyllis E. Marshall, Sylvia K. Reed. 1992. A Mental Health Capitation Experiment: Evaluating the Monroe-Livingston Experience. Economics and Mental Health, ed. Richard G. Frank and Willard G. Manning, Jr. Baltimore: Johns Hopkins Press, pp. 307-331.
Bachman, Sara, Mimi Bernardin and Brian Burwell. 1996. Integrated Service Systems for Children With Disabilities. Unpublished.
Breakey, William and Donald Steinwachs. Forthcoming. Evaluation Baltimore Medicaid Mental Health Experiment.
Callahan, James J., Donald J. Shepard, Richard H. Beinecke, Mary Jo Larson, Doreen Cavanaugh. 1995. Mental Health/Substance Abuse Treatment in Managed Care: The Massachusetts Medicaid Experience. Health Affairs 14(3): 173-184.
Christianson, Jon B., Willard Manning, Nicole Lurie, Tamara J. Stoner, Donald Z. Gray, Michael Popkin, Sally Marriott. 1992. Utah's Prepaid Mental Health Plan: The First Year. Health Affairs 14(3): 160-172.
Fox, Harriette B., Lori B. Wicks, Paul W. Newacheck. 1993. Health Maintenance Organizations and Children With Special Health Needs: A Suitable Match? AJDC 147 (May): 546-552.
Horowitz, Sarah McCue and Ruth E. K. Stein. 1990. Health Maintenance Organizations vs., Indemnity Insurance for Children With Chronic Illness: Trading Gaps in Coverage. AJDC 144 (May); 581-586.
Kronick, Richard, Tony Dreyfus, Lora Lee, Zhiyuan Zho. 1996. Diagnostic Risk Adjustment for Medicaid: The Disability Payment System. Health Care Financing Review 17(3): 7-34. [Executive Summary]
Kaplan, Sherrie H. and Sheldon Greenfield. 1994. New Statewide Health Reform Initiatives.Milbank Quarterly 72(4): 695-699.
Lurie, Nicole, Ira Moscovice, Jon B. Christianson, Michael K. Popkin. 1992. Does Capitation Affect the Health of the Chronically Mentally Ill? JAMA 267(24): 3300-3304.
McFarland, Bentson H., Douglas A. Bigelow, Jay C. Smith, Mark C. Hornbrook, Ala Mofidi, Patrick Payton. 1995. A Capitated Payment System for Involuntary Mental Health Clients. Health Affairs 14(3): 185-196.
Master, Robert, Tony Dreyfus, Sharon Connors, Carol Tobias, Zhiyuan Zhou, Richard Kronick. 1996. The Community Medical Alliance: An Integrated System of Care in Greater Boston for People with Severe Disability and AIDS. Managed Care Quarterly 4(2): 26-37.
Mechanic, David, Mark Schlesinger, Donna D. McAlpine. 1995. Management of Mental Health and Substance Abuse Services: State of the Art and Early Results. Milbank Quarterly 73(1): 19-55.
Meyers, Allan R., Adrienne Cupples, Ruth I. Lederman, Laurence G. Branch, Marie Feltin, Robert J. Master, Doreen Nicastro, Mary Glover, Denise Kress. 1987. A Prospective Evaluation of the Effect of Managed Care on Medical Care Utilization Among Severely Disabled Independently Living Adults. Medical Care 25(11): 1057-1068.
Neff, John M. and Gerard Anderson. 1995. Protecting Children With Chronic Illness in a Competitive Marketplace. JAMA 274(23): 1866-1869.
Oliver, Michael. 1996. Understanding Disability: From Theory to Practice. New York: St. Martin's.
Rogers, William H., Kenneth B. Wells, Lisa S. Meredith, Roland Sturm, M. Audrey Burnam. 1993. Outcomes for Adult Outpatients With Depression Under Prepaid or Fee-for-Service Financing. Archives of General Psychiatry 50 (July): 517-525.
Safran, Dana Gelb, Alvin R. Tarlov, William H. Rogers. 1994. Primary Care Performance In Fee-for-Service and Prepaid Health Care Systems. JAMA 271(20): 1579-1586.
Schlesinger, Mark. 1986. On the Limits of Expanding Health Care Reform: Chronic Care in Prepaid Settings. Milbank Quarterly 64(2): 189-215.
Shern, David L., Sheila A. Donahue, Chip Felton, Genevive R. Joseph and Norman Brier. Partial Capitation Versus Fee-for-Service in Mental Health Care. Health Affairs 14(3): 208-219.
Stuart, Mary E. 1994. Redefining Boundaries in the Financing and Care of Diabetes: The Maryland Experience. Milbank Quarterly 72(4): 679-694.
Sturm, Roland, Catherine A. Jackson, Lisa S. Meredith, Winnie Yip, Willard G. Manning, William H. Rogers, Kenneth B. Wells. 1995. Mental Health Care Utilization in Prepaid and Fee-for-Service Plans Among Depressed Patients in the Medical Outcomes Study. Health Services Research 30(2): 319-340.
Sturm, Roland, Elizabeth A. McGlynn, Lisa S. Meredith, Kenneth B. Wells, Willard G. Manning, William H. Rogers. 1994. Switches Between Prepaid and Fee-for-Service Health Systems Among Depressed Outpatients: Results from the Medical Outcomes Study. Medical Care 32(9): 917-929.
Sturm, Roland, Lisa S. Meredith, Kenneth B. Wells. 1996. Provider Choice and Continuity for the Treatment of Depression. Medical Care 34(7): 723-34.
U.S. General Accounting Office. 1996.Medicaid Managed Care: Serving the Disabled Challenges State Programs. July, GAO/HEHS-96-136.
Ware, John, Jr., William H. Rogers, Allyson Ross Davies, George A. Goldberg, Robert H. Brook, Emmet B. Keeler, Cathy Donald Sherbourne, Patricia Camp, Joseph P. Newhouse. 1986. Comparison of Health Outcomes at a Health Maintenance Organization with Those of Fee-for-Service Care. Lancet (May 3): 1017-1022.
Ware, John E., Jr., Martha S. Bayliss, William H. Rogers, Mark Kosinski, Alvin R. Tarlov. 1996. Differences in 4-Year Health Outcomes for Elderly and Poor, Chronically Ill Patients Treated in HMO and Fee-for-Service Systems. JAMA 276 (13): 1039-1047.
Wells, Kenneth B., Boris M. Astrachan, Gary L. Tischler and Jurgen Unutger. 1995. Issues and Approaches in Evaluating Managed Mental Health Care. Milbank Quarterly 17(1): 57-75.
Wells, Kenneth B., M. Audrey Burnam and Patti Camp. 1995. Severity of Depression in Prepaid and Fee-for-Service General Medical and Mental Health Specialty Practices. Medical Care 33(4): 350-364.
Wells, Kenneth B., Willard G. Manning, Jr. and R. Burciaga Valdez. 1990. The Effects of a Prepaid Group Practice on Mental Health Outcomes. Health Services Research 25(4): 615-625.
Wells, Kenneth B., Roland Sturm, Cathy Sherbourne and Lisa Meredith. 1996. Caring for Depression. A Rand Study. (Uncorrected page proofs.) Cambridge: Harvard University Press.
Yelin, Edward H., Martin A. Shearn, Wallace V. Epstein. 1986. Health Outcomes for a Chronic Disease in Prepaid Group Practice and Fee for Service Settings: The Case of Rheumatoid Arthritis. Medical Care 24(3): 236-247.
Yelin, Edward H., Lindsey A. Criswell, Paul G. Feigenbaum. 1996. Health Care Utilization and Outcomes Among Persons with Rheumatoid Arthritis in Fee-for-Service and Prepaid Group Practice Settings. JAMA 276(13): 1048-1053.