Efforts to reduce the costs of health care, in general, can be expected to have a particular effect on the care of diabetic persons because diabetes is the prototypic chronic disease and accounts for a major portion of health care expenditures. Two comprehensive studies of the economic effect of health care for diabetic persons in the United States have been done. Total health care expenditures in the United States for diabetic persons were $105 billion in 1992, amounting to 14.6% of all health care expenditures in that year [1]. In the same year, the total direct cost of the diabetes component of health care was $45.2 billion, and the indirect cost from lost productivity caused by morbidity and mortality in diabetes was $46.6 billion [2].

Managed Care

Managed care encompasses many forms of health care delivery and financing. For the purposes of evaluating its effect on diabetes, a broad definition of managed care is most useful: "Any arrangement for healthcare in which someone is interposed between the patient and the provider, and has the authority to place restraints on how and from whom the patient may obtain medical and health services, and what services are to be provided in a given situation" [3].

Health maintenance organizations (HMOs) are the purest form of managed care organizations and combine the insurance function and delivery system within a single organization. They offer comprehensive health care services for a predetermined price to a defined panel of enrolled individuals. They organize selected providers to deliver medical services. The broad definition of managed care includes many other kinds of organizations as well. Nearly all government health care financing arrangements manage the care for which they pay, as do most indemnity insurance plans. Managed care plans may pay physicians on a fee-for-service basis, or they may provide a capitated fee for each patient.

Managed care plans are increasing. In July 1994, enrollment in HMOs in the United States was 47.2 million, a 12% increase over the preceding year [4]. As noted in Table 1, 95% of employee health care benefits was under some form of managed care [5]. Private insurance and government health expenditures, which constitute the major components of managed care, are projected to continue to increase at least through the year 2000 [6]. Both Medicare and Medicaid programs are also moving toward managed care arrangements as state and federal governments seek to reduce health care expenditures. By 1994, about 3.1 million Medicare beneficiaries and 7.8 million Medicaid recipients were enrolled in managed care programs [7]. These government programs cover the elderly and indigent and therefore affect large numbers of persons with non–insulin-dependent diabetes mellitus (NIDDM).

How Managed Care Organizations Work

Managed care has two major goals: to control the costs of health care and to provide services of sufficient quality to satisfy enrollees. To accomplish these goals, managed care organizations use several strategies based on both direct control of care and indirect economic incentives. These strategies have critical implications for the services provided to diabetic persons.

Four general categories of strategies are commonly used to manage care. Payment incentives include capitation to encourage the provider to use fewer services and copayments and deductibles to discourage patients from overusing services. Provider designation by managed care organizations seeks to assure that physicians are selected to provide the necessary quality of care at the least cost and includes the use of primary care gatekeepers, limitations on specialty referral, and monitoring of physician practice patterns. Coverage policies control and limit the services for which the managed care organization will pay. Traditional insurance mechanisms include strategies to discourage enrollment of high-risk patients, exclusion of coverage for preexisting conditions, and experience rating.

Do Managed Care Strategies Work?

Several recent reviews have evaluated existing evidence on the performance of managed care as compared with traditional fee-for-service care [5, 8, 9]. Compared with indemnity insurance plans, HMOs have lower hospital admission rates, shorter hospital lengths of stay, the same or more outpatient office visits, and lower use of procedures, tests, and expensive services when alternatives are available [8]. Despite this apparent reduction in use, the evidence that managed care reduces the cost of health care is remarkably inconsistent [5, 10].

Measures of quality of managed care are also mixed. As measured by both process and outcomes criteria, the quality of care is generally comparable in HMO and fee-for-service settings [8, 11, 12]. There is evidence that HMO enrollees consistently receive more preventive tests, procedures, and examinations than fee-for-service patients [8]. However, patient satisfaction with outpatient care is lower in HMOs [13].

Diabetes and Managed Care

There is a paucity of direct studies of the effects of managed care on the care provided to diabetic patients [14, 15], although several promising studies focusing on outcomes measures are currently under way [16-19]. Nevertheless, an analysis of the effects of managed care strategies on physician behavior and on the care of diabetic persons does provide considerable information.

Payment Incentives and Diabetes

Capitation, copayments, and deductibles effectively limit the use of some services, especially inpatient hospital care [8, 20]. Both the Diabetes Control and Complications Trial [21] and a management system for patients with insulin-dependent diabetes mellitus (IDDM) established in Germany [22, 23] used inpatient care as an important element of the initiation of intensified therapy (see "Health Care for Persons with NIDDM: The German Experience"). However, many outpatient self-management training programs for diabetic persons have been shown to be cost-effective [24, 25]. Routine hospitalization for initiation of therapy for patients with either IDDM or NIDDM is probably not appropriate, although some patients clearly require hospitalization to control their diabetes. Payment incentives have the positive effect of appropriately encouraging routine initiation of insulin therapy on an outpatient basis, but they may also adversely affect the care of some diabetic persons who would benefit from inpatient care.

Provider Designation and Diabetes

Over 90% of managed care plans assign primary care gatekeepers to oversee patient care [26], and these gatekeepers effectively reduce referrals for specialty care [27]. Although considerable evidence exists that primary care physicians can provide effective care for less complex diabetic patients [28], disturbing evidence exists of deficient knowledge of diabetes among primary care physicians [29]. Unfortunately, no direct studies have been done to compare the quality of diabetes care—as measured by process or outcomes—between primary care physicians and specialists. A recent study [30] of the effect of consultation of endocrinologists and diabetes teams on length of hospitalization of patients who were admitted with the primary diagnosis of diabetes as compared with care provided by general internal medicine teams showed a 56% shortening of length of stay when the diabetes team was involved. Health maintenance organizations often use a lower proportion of endocrinologists than are represented in the national physician supply [31].

The effects of provider designation by managed care organizations will likely adversely affect the care of more complex diabetic patients, especially those with IDDM. On the other hand, the structured nature of HMOs may improve some aspects of diabetes care. The availability of structured programs is higher in HMOs, and poorly controlled, elderly diabetic persons may be more likely to receive appropriate eye care in HMOs [15].

Coverage Policies and Diabetes

Self-management training and nutrition counseling services are inconsistently covered by managed care plans, although there appears to be a recent increase in coverage. Home glucose monitoring devices and strips are covered by about half of the managed care plans of patients at the Joslin Diabetes Center (unpublished data), although several states have now passed statutes mandating that insurers provide this coverage. Although these services, which are essential for the prevention of long-term complications of diabetes, are inconsistently provided, acute care needs and the care of documented complications are generally well covered under managed care.

Insurance Strategies and Diabetes

Care of diabetic persons costs three times as much as that for nondiabetic persons [1]. Risk selection is a standard insurance mechanism that endeavors to insure those who are less likely than average to require services. Although the proportion of diabetic persons among the uninsured is no higher than in the general population [32], HMOs tend to have younger, healthier enrollees, and the proportion of diabetic persons enrolled in HMOs is lower than in the general population. Thus, managed care organizations tend to exclude diabetic persons, although they are apparently able to purchase other forms of health care coverage.

Diabetes is the classic preexisting condition, and the use of preexisting condition exclusions by managed care plans leaves diabetic persons without health insurance for necessary services. Further, experience rating, which adjusts premiums according to the use of health care services, distinctly affects the price paid for care by diabetic persons and their employers.

Other Effects of Managed Care

Structured health care systems generally provide better care for diabetic persons [20, 22, 23], especially those with IDDM. Because HMOs have more control over physician practice patterns, tend to have more sophisticated data support systems, and often use the necessary ancillary staff, they have the potential for better diabetes care if they choose to provide it. Additional clear evidence of economic benefit must be obtained to provide managed care organizations with the incentive to meet this potential.

Managed care and price competition are having an adverse effect on the training of new endocrinologists [33]. Whereas medical specialty fellows in training increased 27% from 1976 to 1988, the number of endocrinology fellows declined by 11% during that time [34]. Evidence exists that some of this change results from perceived differences in economic rewards of the various specialties [35], although declining federal funding of research laboratories likely influences this trend as well. Training a higher proportion of generalists is clearly in the national interest, but a deficiency in the supply of diabetologists is likely to adversely affect the quality of care of patients with complex diabetes.

Conclusions and Recommendations

Managed care is becoming the standard in the health care system of the United States, and its reach is increasing. The strategies of managed care organizations, which are designed to limit costs while assuring adequate quality, carry the risk of diminishing the quality of care available to diabetic persons. However, managed care organizations also have the potential to provide excellent health care services for diabetic persons, if clear evidence is presented that costs can be cut and quality improved. An opportunity exists for the diabetes community to work constructively with managed care organizations to develop methods to control costs and improve outcomes. The following recommendations will help to improve the services available to diabetic persons through managed care:

1. The diabetes community must work with managed care organizations to develop methods of medical service delivery that are cost-effective and result in improved outcomes.

2. The roles of primary care and specialist physicians, nurse-practitioners, teaching nurses, nutritionists, and other members of the diabetes treatment team must be defined so that they can be effectively coordinated within managed care systems to provide the best and most cost-effective care.

3. Existing data must be assembled to document the cost-effectiveness of diabetes care, and it must be put in a form that appeals to managed care organizations.

4. Research on diabetes health care delivery must be done by academic centers and supported by funding from the federal government and managed care organizations.

5. Responsibility to organize, coordinate, and oversee the funding of this effort should be assigned to some arm of the federal government, for example, the Centers for Disease Control and Prevention. A National Diabetes Services Task Force, including representatives from the diabetes community, national diabetes organizations, and managed care organizations, should be formed to advise and promote this endeavor.