Jaan Sidorov, MD, FACP, and Ronald Harris, MD

Abstract
In this article, we review the impact of three under-recognized forces in health care for patients with diabetes mellitus: clinical information systems, consumerism, and managed care. We examine the importance of these trends in the care of this disease and outline the array of services that will be necessary. While the gold standard of care for diabetes has been the Diabetes Control And Complications Trial (DCCT), we believe these three trends will significantly affect how the lessons of the DCCT will be applied to clinical practice.

The practice of medicine is facing revolutionary changes. Advances in information systems, the coming birth of medical consumerism, and the continuing growth of managed care will have significant implications for how physicians conduct the practice of medicine. In this paper, we will review these three important trends and describe their impact. We will then focus on how this will change the medical care of an important chronic illness, diabetes mellitus.

Advances in Information Systems
In the book The Third Wave, futurist Alvin Toffler argues that mankind is entering a new period of history.¹ Before the advent of the Industrial Revolution, the supply of goods and services was largely dependent on individual craftsmen. This was eventually replaced by the technologies of mass production, which reduced the provision of goods and services to their individual parts. This approach is being swept away by the Information Revolution. Thanks to computer-based technology, goods and services can be supplied with variety and costs that cater to individual expectations unmatched by the piecemeal factory approach.

Yet the practice of medicine today is still characterized by one-on-one individual craftsmanship. While patients and their doctors correctly resisted the intrusion of factory style medicine at the bedside, the application of clinical information systems can no longer be ignored. Armed with on-line resources, most physicians will be able to avoid the inefficient duplication of information gathering and quickly access summaries of individual and collective patient data.

This approach to managing patient data may determine that collection of a complete history would be more efficiently performed using a survey completed independent of an encounter with a physician. Once the information is collected, these surveys could be forwarded to a processing center, scanned, and assembled in narrative format for on-line use. Other data, such as the costs and charges for services, scheduling of future medical care, or pharmacy use, can be added to these data, providing a more complete data base for health providers.

Aside from meeting the needs of physicians, these information systems likely will also address patient needs independent of individual physician direction. For example, computer generated “prompts” have already been applied to the improvement of preventive health care.² As patient data bases grow in breadth and sophistication, individualized patient initiatives, such as scheduling mammograms for every woman with a significant family history of breast cancer or arranging for every newly diagnosed congestive heart disease patient to see a dietitian for salt-restricted diet instruction, will occur with efficiency that surpasses the current approach of waiting for a doctor to place the order. If a patient desires additional information, interactive voice technology via a toll-free telephone line or a World Wide Web site may be able to provide detailed updates.

Physicians may scoff or feel threatened by this intrusion into medical practice. However, while it can be argued that one-on-one medical care will never be replaced, the potential utility and cost efficiency of information systems to individually guide patients through the health-care system cannot be ignored. Unlike the factory approach, the Information Age will be an irresistible force in medical care.

The Rise of Consumerism
Lacking any initiatives to counter the dependence on market forces to determine health-care systems planning, it is likely that competition will play a significant role in the evolution of health care in the United States. While this competition has been primarily in the area of price, we believe it will shift to the key issue of quality. Immediately following World War II, the Baby Boomers were born. Our society has watched them move through grade schools, colleges, and into employment, and in this transition this generation has defined the tempo of modern-day consumerist convenience and quality.³ Now that they are middle-aged, this generation is on the verge of needing medical care for conditions that are the inevitable consequence of getting older, such as arthritis, heart disease, and diabetes.

While it is difficult to generalize the attitudes of an entire generation, and even more so to define quality as it relates to health care, it is safe to predict that the Boomers’ perception of quality will determine this aspect of medical care. If the option of choice among competing health-care systems is preserved, Baby Boomers will take their business elsewhere if their perceptions of quality are not met.

The Baby Boomers also are not likely to be passive participants in their health care. As a group, they have been characterized as requiring significantly more information than earlier generations, and as being determined to decide what is best for themselves.³ As this new generation begins to seek medical care, their expectations of individualized attention, convenience, access to information, and activism will accompany them. They are likely to be met by an unprepared health-care system. Few doctors would argue that their waiting rooms and nurses’ stations are arrayed with the same attention to consumer needs so evident in today’s fast food restaurants and automobile show rooms. The physicians and health-care systems that respond to these needs will thrive.

The Arrival of Managed Care
Managed care organizations (MCOs) continue to grow in size and complexity. Defined as any system that combines the financing and delivery of health-care services,⁴ these organizations will probably be the predominant form of medical insurance for Americans by the end of this decade. Traditional markers of quality in managed medical care appear to be at least as good as those found in traditional fee-for-service settings.^5-8 At the same time, MCOs have effectively limited the rate of increases in costs through a variety of mechanisms, such as the aggressive coordination of health-care services through a primary care physician or provider.^9-13

MCOs are also developing a variety of approaches to integrating quality and costs through disease state management teams, and demand management interventions.^14-16 These programs represent novel approaches to the integrated and longitudinal care of patients with chronic illness. While other aggressive means of cost control (such as precertification, second opinions, capitated arrangements, and case management) will continue, these are now considered a standard piece of managed care. Disease state management seeks to coordinate the spectrum of care from hospital to home, and allocates resources on the basis of need and outcomes rather than the setting or the availability of reimbursement.¹⁵ Demand management is an integrated approach to patient education and empowerment that allows patients to more appropriately access medical care.^17,18 We believe both strategies will be critically important determinants of the Baby Boomers’ perception of quality. We expect to see these approaches to health care grow significantly in MCOs since they appear to have promise for combining patient satisfaction with optimum costs.

The Implications for Diabetes Care
Jerry Garcia, the late guitarist for the rock band The Grateful Dead and an icon of the ’60s, had diabetes mellitus. As his generation grows older, the number of people with non-insulin-dependent diabetes will grow. This view may already be supported by data from the National Medical Expend-iture Survey, which show that the prevalence of diabetes in the United States has grown to 45 per 1000 population.¹⁹

The Diabetes Control and Complications Trial (DCCT) conclusively demonstrated that individuals with diabetes who maintain optimum blood glucose levels have reduced risks for long-term complications.²⁰ The key to this trial was an intense, multifaceted approach of multiple daily injections or subcutaneous insulin infusion; careful regulation of carbohydrate ingestion; monthly clinic visits with weekly telephone contacts; intensive diabetes education and exercise advice provided by diabetes educators; monthly glycosylated hemoglobin determinations, bi-yearly fundus stereo photographs, yearly eye examinations, and yearly microalbuminuria screening; and ready telephone access to a clinical center in case of intercurrent illness.^21-23

However, not every patient with diabetes was eligible to participate in the DCCT. Exclusion criteria included three or more documented episodes of diabetic ketoacidosis requiring hospitalization during the previous year; planned or desired pregnancy; hypertension; more than two episodes of hypoglycemic seizures during the previous year or more than one hypoglycemic episode resulting in cerebral impairment before the development of warning symptoms of hypoglycemia while awake; active coronary artery disease; residence that would prohibit regular visits; and chronic disease that required certain prescription medications for more that 4 months in the previous year.²¹

The improved outcomes of the DCCT were also accompanied by significant iatrogenic complications. Compared to patients in the control arm of the study, the intervention group had three times as many hypoglycemic reactions (61 vs. 19 per 100 patient-years), three times as many episodes of hypoglycemic coma or seizure (16 vs. 5 per 100 patient-years), and were more likely to become overweight (42 vs. 27%).24 This came at considerable financial cost, too: care for the intervention patients cost from about $4,000 (multiple injections) to $5,800 (continuous infusion pump) per year, compared to care for the conventionally treated patients ($1,700).²⁵

Fundamentally, the DCCT was an efficacy trial, which demonstrated that optimum care for ideal patients with diabetes can delay the progression of complications. Little is known about the effectiveness of this approach in the real-world circumstances of managed care, involving patients who would not be candidates for a formal research trial and who would be reluctant to suffer the side effects outlined above. In a competitive environment that hinges on both price and all the dimensions of perceived quality, it is unlikely that the medical directors and financial officers of the nation’s MCOs will commit funding for all of the elements of the DCCT, especially when its effectiveness and impact for the average diabetic enrollee in an MCO is largely unknown.

Instead, we believe MCOs will assemble those elements of the DCCT that intuitively accounted for the greatest patient benefit. These elements include, at a minimum, the intensive patient education, ready face-to-face and telephone access to medical care, and a programmatic automatic approach to preventive measures. These fit well with the approach many MCOs have already begun in disease state management.¹⁵

The DCCT results were widely reported in the lay media and press. Yet, data obtained after publication of the DCCT suggest that physicians relying on the craftsmanship of their office practice have not consistently applied the lessons of this study to their patients with diabetes. One study in an MCO in California found as few as 10% of patients with diabetes received education or dietary counseling, and up to 78% of patients were never referred for an ophthalmologic examination.²⁶ Another report from a similar setting found that less than one-third of patients with diabetes in an urban MCO received assessment for proteinuria.²⁷ In a nationally recognized health maintenance organization, only 27% of patients had two hemoglobin A_1c determinations within a 12-month period.²⁸

We believe these data reflect a continuing reliance on one-on-one medical care in the managed care setting. The typical primary care office visit is prone to a broad agenda, and multiple patient concerns can quickly consume the allotted minutes of a typical appointment. As a result, important, yet less pressing items, such as preventive care or patient education efforts, simply may not occur. In addition, not all patients enrolled in a MCO regularly see a physician, and, therefore, may not have an opportunity to be offered the appropriate array of health services. In other words, while the office visit is an important aspect of medical care, we conclude that if it is the sole point of access to the health-care system, important components of medical care for too many patients with diabetes simply will not occur.

Because MCOs generally restrict access to specialty health-care services,²⁹ one could argue that one response to these lapses should be to allow patients with diabetes to have unfettered appointments with medical specialists. Yet data from the Medical Outcomes Study (MOS) would indicate that office-based primary care physicians perform similarly compared to specialty physicians in avoiding the complications associated with diabetes.³⁰ Furthermore, Medicare data obtained from fee-for-service settings without disincentives for patient self-referral would indicate that diabetes care is not optimum in this environment, either. A report based on the Medicare claims database found that only 16.3% of patients had hemoglobin A1c determinations, 45.9% had an ophthalmologic examination, and few received a cholesterol determination.³¹

While the debate about specialty versus primary care physician services continues, it is likely MCOs will develop a third way. If the challenge is to improve the coordinated delivery of appropriate services above and beyond what currently exists in the managed care and fee-for-service environments, clinical information systems may be used to automatically schedule and coordinate medical services independent of physician involvement. As outlined above, these systems hold great promise for consistently delivering individualized and quality-driven patient interventions.

We propose that the following are the minimum elements that an MCO would want to develop and integrate to create the full spectrum of diabetes care. These elements are based on the DCCT, yet have been arrayed to fit with the realities of managed care. We also believe they are the types of services that would meet the demands of the new consumerist agenda in health care and, lastly, take full advantage of the exciting advances in clinical information systems.

We believe these elements will provide a consistent level of care that would not only improve short-term markers of diabetes care, such as glycemic control or diabetic ketoacidosis, but also yield pay-offs down the line in the prevention of renal disease and retinopathy.

Patient education. The cornerstone of the DCCT was an aggressive, diabetes nurse specialist-coordinated program of education designed not only to deliver knowledge, but also to assess barriers to patient self-management, assist in motivation strategies to overcome the barriers, and empower pa-tients in the principles of self care.^{20, 22, 23} However, the realities of the price of such an approach make it unlikely that diabetes nurse specialists can be placed in every primary care setting in a typical MCO. MCOs are more likely to turn to primary care nurses, physician’s assistants, or nurse practitioners to provide the fundamentals of diabetes education under the coordination of the diabetes nurse specialists supplemented by outreach. At the primary care level, nurse or “mid-level provider” education could be scheduled for one-on-one patient appointments, as well as group classes or by telephone and modem.

Elements of this primary care-based education would include, but not be limited to, diet, medications, glucose monitoring, medication compliance, recognition of hypoglycemic and hyperglycemic symptoms, and the need for complications monitoring. If warranted, this education could assure that patients are familiar with the indications and use of a glucose meter. Patients with special needs could be referred to a diabetes nurse specialist or outreach, or if circumstances warrant, to a specialist. Assuming that consumer-patients will desire increased information and input about the management of diabetes, this expanded availability of nurse educators is clearly a must.

Patient self-report surveys. Patient self-reports have been shown to be a useful adjunct to medical care.32 Validated instruments are becoming available in optically scannable formats for ease of database entry and statistical analysis. Theoretically, information from other MCOs could be used to establish baseline benchmarks for comparison. Continued use of such tools could provide multiple assessments of the status of an MCO diabetic population to assist in MCO clinical programming. These surveys would also be a visible reminder to consumers of the commitment of their physicians and nurses to managing their diabetes.

Satisfaction with care. Consumers’ satisfaction with their care will be critically important as one gauge of medical care, and be a determinant of the success of an MCO in the marketplace. Patient-driven assessments of medical care in the form of report cards can be regularly collected, placed on line, and disseminated to physicians and other workers and consumers in the MCO. 33 Since the quality of medical care and impact of patient education can be assessed by patient recall,³² these would be a useful information adjunct to the surveys outlined above.

A diabetes database. Disease-specific population data must be warehoused and arrayed to permit ongoing population-wide assessments of the prevalence of health and disease. This obviously steps beyond point prevalence approaches that rely on surveys of patient samples. Rather, this is an ongoing disease-based census.

Given the ease of data entry from multiple sources and the robust capacity of computer hardware, patient identifiers (name, medical record number, MCO member number, and social security number), demographic data, patient- and provider-based assessments, lab information, pharmacy use, appointments, resource utilization, access to case measures, quality of life, functional status assessment, satisfaction rates, employer-based data (such as sick time), patient self-report surveys, and other outcomes can now be stored for easy access and manipulation. This can occur on-line via modem or through a wide-area network. Appropriate security measures such as encryption and “firewalls” (electronic safeguards that prevent unauthorized access to patient data) will be necessary, but are readily available.

How these various data would be integrated remains to be determined, but their importance cannot be underestimated.

These systems can also be used for computer-based “prompts” for physicians and nurses to use as reminders that assist in guideline adherence.² However, these population-based information systems and the physicians they serve will need to go a step further and allow blanket and automatic laboratory appointments or referrals for groups of diabetes patients. They may also allow patients to schedule their own services. Given the robust capacity of these clinical information systems, computer-driven clinical decision rules could arrange routine health-care measures without reliance on a physician order.

Telephone medicine. The DCCT demonstrated that the telephone is an important point of access to medical care in an optimal system of health care for patients with diabetes. This underappreciated tool can be a 24-hour-a-day link between MCOs and their patients. It can serve as a prompt system, act as an additional resource in information acquisition through the technology of interactive voice response,³⁴ and assist in demand management. In the latter instance, patients with diabetes can be given consistent, user-friendly advice about the need to see a physician versus expectant follow-up in the event of illness.

Guidelines. Guidelines will become the standard that patients and health-care providers use to measure and deliver medical care.^35-37 The promotion, evaluation, and testing of adherence to diabetes-specific guidelines are a growing feature of medical practice, and assist in development of additional clinical programming.^38,39 This would be driven not only by patient need, but by consumer-driven ideals of access and convenience.

Disease state management. Disease state management will grow and foster a team approach that incorporates clinical leadership, specialty and primary care physicians, information systems personnel, and nurses with special expertise in the front-line management of diabetes.^40-42 This transcends piecemeal and fragmented care that can occur when patients are limited to a paradigm of primary care versus specialty care only. We believe an approach that combines and balances all the members of the health-care team is far more likely to meet consumer expectations.

Professional education. Assuming an integrated approach to diabetes is best, usual continuing medical education for physicians alone will not be enough to bring about improved outcomes for patients with diabetes.⁴³ Rather, state-of-the-art adult learning strategies designed to provide updates and reminders on the fundamentals of diabetes care will be needed to carry out a broader strategy to improve diabetes care. These can include, but not be limited to, regular one-on-one and group educational programs for physicians, as well as video-conferencing for teaching consultation on individual problems or complicated patients in the primary care setting.

Medications, glucose meters, and test strips. Since it is likely that significant savings for the care of patients with diabetes can be achieved if services are streamlined and appropriately targeted for individuals, MCOs may be willing to fully or partially underwrite the purchase of medications, glucose meters, and glucose test strips. When one considers that out-of-pocket expenses can represent a significant barrier for patients who are candidates for the use of these supplies,⁴⁴ this strategy may increase medication compliance and provide a useful incentive for patient participation in disease state management programs.

Research. By creating an integrated approach to diabetes and the accompanying databases, MCOs will have an important opportunity to develop and refine clinical research capabilities.⁴⁵ Innovative strategies of health care that hold significant promise for improving outcomes in a managed care setting offer an important advance in health services research. The availability of a diabetes-specific database is an important first step in this effort. This may attract partners who wish to co-develop and provide funding for exciting advances in the care of patients with diabetes.

Pharmacy support. Pharmacists are an underutilized resource in the overall care of patients with diabetes.⁴⁶ These professionals are in a unique position to ensure that an updated medication list is compiled and confirmed for on-line use, perform “brown bag” reviews of all medications to assist in detection of polypharmacy,⁴⁷ and provide patients with additional education.⁴⁸ This resource may be welcomed by patients interested in further assistance and education.

Summary
We propose that the practice of medicine is on the verge of a new era. As managed care systems struggle with the costs and newer definitions of quality in an increasingly consumer-driven environment, clinical information systems and disease management strategies will emerge as central components of medical care. While we have attempted to show how this will have an impact on the care of patients with diabetes mellitus, we believe the implications of this transformation will be important for every aspect of the practice of medicine.

References

¹Toffler A: The Third Wave. Bantam Books, New York, 1980.

²Turner RC, Peden JG Jr, O’Brien K: Patient-carried card prompts vs. computer-generated prompts to remind private practice physicians to perform health maintenance measures. Arch Int Med 154:1957-60, 1994.

³Dychwald K, Flower J: Age Wave: The Challenges and Opportunities of an Aging America. St. Martin’s Press, New York, 1990.

⁴Iglehart JK: The American health care system: managed care. New Engl J Med 327:742-47, 1992.

⁵Ware JE Jr., Brook RH, Roger WH, Keeler EB, Davies AR, Sherbourne CD, Goldberg GA, Camp P, Newhouse JP: Comparison of health outcomes at a health maintenance organization with those of fee-for-service care. Lancet 1:1017-22, 1986.

⁶Greenwald HP: HMO membership, co-payment and initiation of care for cancer: a study of working adults. Am J Pub Health 77:461-66, 1987.

⁷Rossiter LF, Langwell K, Wan TT, Rivnyak M: Patient satisfaction among elderly enrollees and dis-enrollees in a Medicare health maintenance organization. J Am Med Assoc 262:57-63, 1989.

⁸Carlisle DM, Siv AL, Keeler EB, McGlynn EA, Kahn KL, Rubenstein LV, Brook RH: HMO versus fee-for-service care of older persons with acute myocardial infarction. Am J Pub Health 82:1626-30, 1992.

⁹Luft HS: Trends in medical care cost: Do HMOs lower the rate of growth? Med Care 18:1-6, 1980.

¹⁰Newhouse JP, Schwartz WB, Williams AP, Witsberger C: Are fee-for-service costs increasing faster than HMO costs? Med Care 23:960-66, 1985.

¹¹Gabel J, Dicarlo S, Sullivan C, Rice T: Employer sponsored health insurance. Health Affairs Fall:161-75, 1990.

¹²Health care in crisis: are HMOs the answer? Consumer Reports August 1992, p. 526.

₁₃Schwartz WB, Mensdelson DN: Why managed care cannot contain hospital costs without rationing. Health Affairs Summer 1992, p. 100-107.

₁₄Peters AL, Davidson MB, Ossorio RC: Management of patients with diabetes by nurses with support of subspecialists. HMO Pract 9:8-11, 1995.

¹⁵Terry K: Disease management: continuous health-care improvement. Business and Health April 1995, p. 64-76.

¹⁶Vickery DM, Lynch MD: Demand management: enabling patients to use medical care appropriately. J Occup Environ Med 37:551-57, 1955.

¹⁷Spalding J: Taking the lead in disease state management. Family Pract Man April 1996, p. 50-57.

¹⁸Schulman KA, Rubenstein LE, Abernathy DR, Seils DM, Sulmasy DP: The effect of pharmaceutical benefits managers: is it being evaluated? Ann Intern Med 124:906-13, 1996.

¹⁹Rubin RJ, Altman WM, Mendelson DN: Health care expenditures for people with diabetes mellitus, 1992. J Clin Endocrin Metab 78:809A-809F, 1994.

²⁰The DCCT Research Group: The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin dependent diabetes mellitus. New Engl J Med 329:977-86, 1993.

²¹The DCCT Research Group: The diabetes control and complications trial (DCCT). Design and methodologic considerations for the feasibility phase. Diabetes 35:530-45, 1988.

²²Fain JA: DCCT: model of partnership. Diabetes Educ 20:9, 1994.

²³Delahanty LM, Halford BN: The role of diet behaviors in achieving improved glycemic control in intensively treated patients in DCCT. Diabetes Care 16:1453-58, 1993.

²⁴The DCCT Research Group: Adverse events and their association with treatment regimens in the Diabetes Control and Complications Trial. Diabetes Care 18:1415-27, 1995.

²⁵The DCCT Research Group: Resource utilization and costs of care in the Diabetes Control and Complications Trial. Diabetes Care 18:1468-78, 1995.

²⁶Peters AL, Ossorio RC, Davidson MB: Quality of care for patients with diabetes in HMO setting in California (abstract). J Invest Med 199 :43 (Suppl 3):256A, 1995.

²⁷Martin TL, Selby JV, Zhang D: Physician and patient prevention practice in NIDDM in a large urban managed-care organization. Diabetes Care 18:1124-32, 1995.

²⁸Hancock NA: The epidemiology of costs of diabetes mellitus. USQA Quality Monitor (published by US Healthcare). Winter 1995, p. 3-8.

²⁹Kranther M: The gatekeeper in managed care. J Healthcare Risk Man 4:8, 1994.

³⁰Greenfield S, Rogers W, Mangotich M, Carney MF, Tarlov AR: Outcomes of patients with hypertension and non-insulin-dependent diabetes mellitus treated by different systems and specialties. J Am Med Assoc 274:1436-44, 1995.

³¹Weiner JP, Parente ST, Garnick DW, Fowles V, Lawthers AG, Palmer RH: A claims-based profile of care provided to Medicare patients with diabetes. J Am Med Assoc 273:1503-1508, 1995.

³²Betz-Brown J, Adams ME: Patients as reliable reporters of medical care process. Recall of ambulatory encounter events. Med Care 30:400-11, 1992.

³³Dearing G: Diabetes disease management. Patients and primary care physicians drive “episodes of care” approach. Outcomes Measure and Man 6:3-4, 1995.

³⁴Harris R, Clarke K, and Schmid L: Twenty-four hour patient access to physician intervention for the control of diabetes: preliminary clinical experiences. Hospital Association of Pennsylvania. 1995 Achievement Award.

³⁵American Association of Clinical Endocrinolo-gists and the American College of Endocrinology: AACE guidelines for the management of diabetes mellitus. Endocr Pract 1:149-57, 1995.

³⁶Miller KL, Hirsch IB: Physicians’ practices in screening for the development of diabetic nephropathy and the use of glycosylated hemoglobin levels. Diabetes Care 17:1495-97, 1994.

³⁷Pearson SD, Goulart-Fisher D, Lee TH: Critical pathways as a strategy for improving care: problems and potential. Ann Int Med 123:941-48, 1995.

³⁸Elford RW, Jennett P, Bell N, Szafran D, Meadows L: Putting prevention into practice. Health Reports 6:142-53, 1994.

³⁹Treter DF, Johnson E: The roadmap for clinical quality. HMO Pract 8:5-9, 1994.

⁴⁰Carney PA, Dietrich AJ, Keller A, Landgraf J, O’Connor GT: Tools, teamwork and tenacity: an office system for cancer prevention. J Fam Pract 35:385-87, 1992.

⁴¹Weinberger M, Kirkman MS, Samsa GP, Shortliffe GA, Landsman PB, Cowper PA,

Simel DL, Feussner JR: A nurse coordinated intervention for primary care physicians with non-insulin dependent diabetes mellitus: Impact on health related quality of life. J Gen Int Med 10:59-66, 1995.

⁴²Nettles AT, Kreitzer MJ: Trends in advanced nursing practice and implications for care of diabetes patients. Diabetes Spectrum 7:344-49, 1994.

⁴³Davis DA, Thomson MA, Oxman AD, Haynes RP: Changing physician performance: a systematic review of the effect of continuing medical education strategies. J Am Med Assoc 274:700-705, 1995.

⁴⁴The Task Force for Compliance: Non-compliance with medications. An economic tragedy with important implications. 2300 N Charles Street, Baltimore, MD 21218.

⁴⁵Briesacher B, Erwin WG: Managed care organizations: the new research partners. Med Interface December 1995, p. 69-77.

⁴⁶Hepler CD: The third wave in pharmaceutical education: the clinical movement. Am J Pharm Educ 51:369-84, 1987.

⁴⁷Colley CA, Lucas LM: Polypharmacy. The cure becomes the disease. J Gen Int Med 8:278-83, 1993.

⁴⁸Hepler CD, Strand LM: Opportunities and responsibilities in pharmaceutical care. Am J Pharm Educ 53 (Winter suppl):75-155, 1989.

Jaan Sidorov, MD, FACP, is the medical director of the Continuous Health Improvement Program at the Geisinger Health Plan in Danville, Pa. Ronald Harris, MD, is the clinical program director for diabetes at Geisinger System in Wilkes-Barre, Pa.