Variations in Medicare Health Maintenance Organizations

With encouragement from the federal government, there has been steady growth in the number of Medicare beneficiaries enrolled in health maintenance organizations (HMOs) during the past 2 decades.¹ The Balanced Budget Act of 1997 has further expanded managed care alternatives through the establishment of the Medicare+Choice program, and HMO enrollment may be one third of the Medicare population by 2005.² However, thus far, the overall assessment about the quality of care in Medicare HMOs has been inconclusive.³

In this issue of THE JOURNAL, Riley et al⁴ report that elderly women with breast cancer were more likely to receive recommended diagnostic and therapeutic interventions in Medicare HMOs compared with similar fee-for-service settings. This study also found substantial performance variations among the HMOs, including both favorable and unfavorable patterns of care. Contributory factors to these performance differences could derive from recent changes in the health care marketplace and varied alliances between HMOs and physicians.

Health maintenance organizations have expanded rapidly in many markets, yet enrollment varies across different geographic areas, especially for Medicare. For example, in 1994, more than 20% of Medicare beneficiaries in San Francisco, Calif, were enrolled in HMOs, while HMO enrollment accounted for less than 1% of the Medicare market in Detroit, Mich,⁵ both of which were sites in the study by Riley et al.⁴ In contrast, payments to Medicare HMOs in Detroit were approximately 50% higher in 1994 than payments to Medicare HMOs in San Francisco. These discrepancies between markets and payments reflect local competitive influences and different arrangements among HMOs, physicians, and hospitals that can generate performance variations.

Health care systems with greater degrees of integration between HMOs and physicians may be more likely to create a coordinated experience for patients compared with more fragmented systems.⁶ The degree of integration in an HMO reflects the type of alliance forged between the health plan and its physicians. However, in recent years, HMOs have been transformed from smaller regional plans into national for-profit companies, some with millions of enrollees.⁷ There has also been a shift away from the group and staff models toward independent practice associations (IPAs) and network models.⁸

The growth of for-profit IPA-model HMOs, as well as the oversupply of physicians and hospital beds, has led to new contracting arrangements among physicians and hospitals. For example, some physicians and hospitals have established common ownership, forming vertically integrated delivery systems. Conversely, some physicians are associating with practice management firms that offer attractive capital financing,⁹ forming horizontally integrated groups for specialty or full-risk contracts. Moreover, other physicians and hospitals have organized through contracting networks to compete and administer HMO contracts, forming virtually integrated delivery systems (eg, physician-hospital organizations). These different business arrangements permit economies of scale among both physicians and hospitals and enable operational efficiencies, the opportunity to bear financial risk, and the potential for innovations in quality.¹⁰

Both the different structures of Medicare HMOs and the diverse contractual arrangements of physicians and hospitals create varied relationships that can foster or hinder initiatives to improve care. Health maintenance organizations that employ physicians (eg, staff model), or delegate financial risk to a vertically integrated delivery system are likely to foster a high level of mutual purpose and vision between the health plan and physicians. Health maintenance organizations that have multiple contracts with independent physicians and group practices (eg, IPA and network models), those that contract with physician practice management companies, or those that delegate risk to a virtually integrated delivery system are susceptible to more fragmented care.

In the study by Riley et al,⁴ treatment patterns indicated higher rates of breast-conserving surgery and more appropriate use of radiation therapy in some HMOs. Practice guidelines may have been implemented by these HMOs to improve quality of care. The effectiveness of practice guidelines depends on the methods used for development and supplementary efforts to encourage their use.¹¹ Shared goals between an HMO and its physicians can expedite the formulation and dissemination of guidelines through facilitation by physician leaders,¹² those physicians who influence decision making in group practices and hospitals. Some HMOs emphasize physician leadership and participation by assembling clinical advisory panels for the review of guidelines. Guideline consistency also can affect implementation. For example, while group- and staff-model physicians are likely to receive a single practice guideline from their HMO for the detection and management of a condition such as breast cancer, independent physicians and practices often contract with multiple HMOs and may receive multiple guidelines for the same condition. Although there may be similarities among the different guidelines, the variety of protocols can complicate the daily routine of a busy practitioner.

The earlier-stage breast cancers in HMOs reported in the study by Riley et al⁴ may have resulted from higher rates of mammography. Preventive services have been shown to be especially prominent virtues of group- and staff-model HMOs.¹³ Some HMOs also provide financial incentives to physicians to improve quality of care. Incentives that focus on patient satisfaction and quality have been shown to engender better professional satisfaction among physicians.¹⁴ However, the financial motivation of the HMO may influence the intent of physician incentives. For example, for-profit ownership of HMOs appears to encourage rewarding physicians for more judicious utilization patterns to reduce costs.¹⁵ Some HMOs delegate full financial risk to physician groups,¹⁶ and the Medicare+Choice program will further expand provider opportunities to assume risk through the establishment of provider-sponsored organizations. Yet, physician groups that bear financial risk often operate like HMOs in their focus on costs rather than quality.¹⁷

Health maintenance organizations that implement clinically relevant data systems also can enhance initiatives to improve quality. In the article by Riley et al,⁴ HMOs with favorable outcomes may have used data on performance to increase rates of mammographic screening by primary care physicians, improve the diagnostic accuracy of mammographers, and expand the use of radiation therapy where appropriate. Electronic medical records are especially effective methods of integrating delivery systems and can vastly improve the coordination of care.¹⁸ Furthermore, clinical information systems can improve quality through audit and feedback of performance data.¹⁹ Nevertheless, while a few HMOs have made sizable investments in information systems (eg, electronic medical records) to further improve the accuracy and breadth of information, most HMOs continue to rely on administrative claims data for quality improvement purposes.

Among the more customary sources of relevant data for quality improvement in HMOs is the Health Plan Employer Data and Information Set (HEDIS), a set of standardized performance measures. However, the availability and accuracy of HEDIS data vary among health plans. For instance, a recent audit of HEDIS for the Medicare program identified frequent problems among the different participating HMOs, such as nonstandardized records, misinterpretation of variables, and incomplete data.²⁰ Furthermore, although HMOs that participate in the Medicare program must report on 32 selected HEDIS measures,²¹ a broader assessment will be required for effective evaluation of performance in care of the elderly, especially for chronic illnesses.

The potential of Medicare HMOs has always been to improve the coordination and quality of care for the elderly. Some staff-model HMOs and other vertically integrated delivery systems have the essential components for potential quality improvements in the care of complicated illnesses for the elderly. However, consolidation of the HMO industry, growth of for-profit plans, and competition in the marketplace have created different organizational structures and risk-sharing arrangements among HMOs and physicians. These varied relationships among stakeholders are likely to further evolve and could have a profound influence on the future performance of HMOs for Medicare.