The conference was held in Washington, D.C., from 11 to 13 December 1994. The conference format was a series of presentations followed by discussion. After these presentations, the group was divided into three concurrent sessions (fora) to develop the recommendations that appear in this supplement's conclusion (see "Where Do We Go from Here?"). As an integral part of the conference, it was felt that the conference papers, deliberations, and conclusions should be widely disseminated. This supplement seeks to achieve that purpose.

When the conference began, concern existed that the term "intensive management," used to define the aggregate treatment approach used in the DCCT, has come to mean for many practitioners the intensive use of insulin. This perspective is clearly not that of the conference sponsors or participants. Thus, many would substitute the term "comprehensive management" to move away from the belief that the intensive use of insulin is or should be the primary mode of treatment for NIDDM. Thus, we define comprehensive management of NIDDM, as did the DCCT for intensive management of IDDM Table 2, as a comprehensive and interactive array of treatments (Table 3). The goal is to reduce all risk factors in each patient.

After introductory remarks by Dr. Gorden, director of the National Institute of Diabetes and Digestive and Kidney Diseases, Dr. Vinicor presented a thoughtful review of the process of translating the specific findings of a research study to clinical practice [6].

Mechanisms of Diabetes Complications: The Glucose Hypothesis

Dr. Nathan reviewed the data supporting the relation between glucose control (and diabetes duration) and microvascular and neurologic complications and their reduction by improved glucose control (see "The Pathophysiology of Diabetic Complications: How Much Does the Glucose Hypothesis Explain?"). Dr. Klein then presented the epidemiologic data relating the incidence of these complications to glycemic control (see "Relation of Glycemic Control to Diabetic Microvascular Complications in Diabetes Mellitus"). Taken together, these data show that the pathophysiology of these specific complications is the same in IDDM and NIDDM and that the incidence is the same in the two types of diabetes when compared by hemoglobin A_1c levels. Further, lowering blood glucose levels in patients with NIDDM should have benefits similar to those seen in IDDM.

Risks of Intensive Management of NIDDM: The Insulin Hypothesis

The main untoward effects of intensive insulin management reported in the DCCT were an increased incidence of hypoglycemia and weight gain [1]. In addition, at least a theoretical fear exists that insulin itself, either endogenous or exogenous, could be a risk factor for macrovascular disease, the complication most responsible for morbidity and mortality in NIDDM. Drs. Stern and Genuth examined the hypothesis that either endogenous or exogenous insulin might in itself be a risk factor for macrovascular disease. The data presented by Dr. Stern did not support endogenous insulin as a risk factor for macrovascular disease in nondiabetic or diabetic patients. The data suggested that the insulin-resistant state, for which insulin concentrations are a marker, is associated with an increase in cardiovascular risk factors and in the incidence of cardiovascular disease (see "Do NIDDM and Cardiovascular Disease Share Common Antecedents?"). Similar conclusions have been reached by others [11-14].

Dr. Genuth then reviewed the data about exogenous insulin and the development of macrovascular disease from the University Group Diabetes Program (UGDP). He concluded that no such risk was evident in that longterm, multi-site study (see "Exogenous Insulin Administration and Cardiovascular Risk in NIDDM and IDDM"). Abraira and colleagues [15] recently came to the same conclusion, although they recommended further study. In another presentation about insulin and NIDDM, Dr. Henry showed that intensive insulin management in patients with NIDDM who fail oral agent therapy reduces insulin resistance and, thus, suggests that glucose control in itself can improve this aspect of the pathophysiology of NIDDM (see "Glucose Control and Insulin Resistance in NIDDM").

Current NIDDM Practices

A main conference objective was to address what changes in the management of NIDDM were necessary to improve diabetes outcomes. Thus, assessing the current state of diabetes care in the United States was important. Research in this area was summarized by Dr. Harris, who showed that practitioners lacked a systematic approach to diabetes and its complications (see "Medical Care for Patients with Diabetes: Epidemiologic Aspects"). One can imagine several reasons for this lack. Certainly, managing diabetes is a difficult and time-consuming task for already overextended primary care practitioners. Further, primary care practitioners may lack the training and comprehensive resources necessary for intensive management of diabetes. In addition, these efforts are often poorly reimbursed.

The review by Marrero [16], however, brought forth a second and more disconcerting hypothesis. Marrero looked at adherence to standards, including screening for complications—where resources should not be a barrier—and found that screening procedures were done infrequently, particularly for NIDDM [16]. This finding suggests that the physicians surveyed did not feel that NIDDM was as serious a disease as IDDM. Similar findings of low adherence to treatment and screening guidelines have been reported by others in both traditional and health maintenance organization environments [17-19]. An alternative interpretation of these findings is that physicians do not feel that screening for complications is of value because nothing could be done about those complications once discovered. These studies suggest that the data supporting the value of early treatment of diabetes microvascular disease (for example, laser photocoagulation for retinopathy or angiotensin-converting enzyme inhibitors for microalbuminuria) and their implications for clinical practice need to be more widely disseminated to the primary care community [20].

Ongoing and Proposed Trials

Dr. Colwell reported on the Veterans Affairs Cooperative Study on Glycemic Control and Complications in NIDDM (VACSDM)—a pilot study on the feasibility of intensive insulin management of patients with NIDDM who have failed oral agent therapy (see "The Feasibility of Intensive Insulin Management in NIDDM: Implications of the Veterans Affairs Cooperative Study on Glycemic Control and Complications in NIDDM"). These patients were well controlled on insulin with a very low hypoglycemic rate and little weight gain. It is hoped that this study design can be the basis for a longer-term prospective trial, which looks at cardiovascular end points.

Dr. Turner presented baseline data on the United Kingdom Diabetes Prospective Study (UKPDS), a prospective trial of metabolic control of patients with NIDDM who use the oral agents glyburide and metformin (recently made available in the United States) and insulin. The experimental group has substantially lower hemoglobin A (_1c) levels, but both groups show an increase in hemoglobin A_1c levels over time. Outcome data will not be available for another 2 years (see "United Kingdom Prospective Diabetes Study 17: A 9-Year Update of a Randomized, Controlled Trial on the Effect of Improved Metabolic Control on Complications in NIDDM"). Dr. Laakso reported on the close correlation between glycemic control and cardiovascular disease in several populations of patients with NIDDM. He concluded that both coronary heart disease and other macrovascular diseases are associated with poor metabolic control, and their incidence increases as hemoglobin A₁ levels increase (see "Glycemic Control and the Risk for Coronary Heart Disease in Patients with NIDDM: The Finnish Studies"). Taken together, these reports strongly suggest that macrovascular disease is increased by hyperglycemia as an independent risk factor but that the benefit of lowering hemoglobin A₁ levels, particularly with insulin, has not yet been established in terms of reducing macrovascular disease.

Dr. Savage then discussed the elements necessary for the implementation of a macrovascular disease prevention trial in NIDDM (see "Cardiovascular Complications of Diabetes Mellitus: What We Know and What We Need To Know about Their Prevention"). In such a trial, control of all known cardiovascular risk factors would need to be a predominant design feature.

Improving Diabetes Care: International Experiences

The next series of presentations concerned the barriers and successes in improving the care of patients with NIDDM in several settings. First, Dr. Hiss presented the results of his longitudinal study with primary care providers in Michigan, and second Dr. Gohdes discussed the experience of the Indian Health Services diabetes program (see "Barriers to Care in NIDDM: The Michigan Experience" and "Improving Diabetes Care in the Primary Health Setting: The Indian Health Service Experience," respectively). Two European examples were then discussed by Drs. Berger and Keen (see "Health Care for Persons with NIDDM: The German Experience" and "Management of NIDDM: The United Kingdom Experience," respectively).

Diabetes Care and Health Systems

The promises and limitations of changes to the health care system for care of patients with NIDDM were discussed next. In her introductory remarks, Dr. Lasker pointed out that we are in an era of limited resources. She suggested that we need to take a public health perspective in approaching NIDDM. Her views were expanded in a recent publication [21].

Improved diabetes care through the use of computers was the topic of Dr. McDonald's presentation (see "The Promise of Computerized Feedback Systems for Diabetes Care"). Computer systems hold great promise for organizing data, reminding physicians of preventive care and surveillance of complications, and even prescribing changes in treatment on the basis of clinical data; however, these systems require the development of precise and scientifically validated care protocols.

Changes in the reimbursement system from traditional fee-for-service to capitated systems are likely to alter the care of chronic diseases where the costs are direct and immediate but the benefits are long term. Dr. Quickel discussed this issue from his perspective as president of the Joslin Diabetes Center (see "Diabetes in a Managed Care System").

Economics of Diabetes Care

Any intensification of diabetes management will have its cost in dollars spent, in manpower diverted, and in the risks inherent in such management. Dr. Javitt showed that the long-range benefit of comprehensive eye care in diabetes outweighs the cost and is commensurate with current preventive interventions (see "Cost-Effectiveness of Detecting and Treating Diabetic Retinopathy"). The health care system, however, is left with the problem of who pays the cost as compared with who benefits from the cost savings. Clearly, both patients and society benefit, but for the 4- or 40-year-old, a 20-year investment may be necessary before the cost savings become evident.

Regenstrief Conference 1995: Conclusions

Three fora were created to address the important questions of the conference, which are outlined in Table 1. The first forum's goal was to define the role of intensive management of NIDDM today; the second forum was to define the important unanswered research questions about the treatment of NIDDM; and the third forum was to define the policy and health financing issues surrounding the management of NIDDM. Each forum made several suggestions.

The first forum suggested that patients with NIDDM would, like patients with IDDM, benefit from reductions in hemoglobin A₁ levels and the consequent lowering of the incidence of microvascular disease (see "Forum One: Current Recommendations about Intensification of Metabolic Control in NIDDM"). Because patients with NIDDM tend to have had diabetes longer at the time of diagnosis and thus tend to be further along in the course of the disease, the potential benefit may be less in these patients than in those with IDDM. On the other hand, the risk of intensive management appears less, at least in terms of hypoglycemia.

The major research issue as determined by the second forum was to define the risks and benefits of intensive management in NIDDM for the prevention of macrovascular disease. These data are particularly necessary for patients who fail diet, exercise, and oral agent therapies and who therefore require insulin. Although most participants felt that insulin in itself was not a risk factor in the development of macrovascular disease, they were concerned about the lack of clear scientific data on the effectiveness of intensive management in reducing macrovascular disease. The major research recommendation of the conference was that a trial should be designed and implemented to answer this question (see "Forum Two: Unanswered Research Questions about Metabolic Control in NIDDM").

In the third forum, the major concern was that the development of cost-competitive, capitated systems creates an environment where the increased cost of intensively and comprehensively managing a patient with NIDDM may put private and public health care organizations at a financial disadvantage. The third forum suggested development of policies that recognize the preventive model as critical, and reward prevention of chronic disease complications as well as comprehensive care of the disease, a great challenge to the health policy community (see "Forum Three: Changes in the U.S. Health Care System That Would Facilitate Improved Care for NIDDM").

Dr. Vinicor: Division of Diabetes Translation, K10, Centers for Disease Control and Prevention, 1600 Clifton Road N.E., Atlanta, GA 30333.