Steering committee members included the American Diabetes Association (ADA), the Foundation for Accountability (FACCT), the Health Care Financing Administration (HCFA), and the National Committee for Quality Assurance (NCQA). The original steering committee was subsequently expanded to include the American Academy of Family Physicians, the American College of Physicians, and the Veterans Administration. All organizations were represented by experts in their respective fields: generalist physicians; specialists in diabetes care and education; experts in health services research, epidemiology and statistics, and health plan administration; and others. DQIP is funded by the HCFA portion of the Balanced Budget Act of 1997.¹

The need for a set of standard diabetes care measurements was based on numerous studies and reports in recent years that have indicated that diabetes care throughout the United States is suboptimal. This recognition has resulted in the development of internal accountability measures or "report card" systems by managed care organizations, hospitals, and government agencies with the intent of analyzing diabetes care in their own institutions. Except for HCFA's requirement for managed care plans to submit limited performance measurement data, these measurement systems have been voluntary and administered by nonprofit organizations dedicated to improving the health care provided to Americans (e.g., the Joint Commission on Accreditation of Healthcare Organizations [JCAHO]).

Following a thorough analysis and discussion, the committee formally adopted the original list of seven accountability measures: percentage of patients receiving >1 HbA_1c test/year; percentage with the highest risk glucose level; percentage assessed for nephropathy; percentage receiving a lipid profile once in 2 years; percentage with a low-density lipoprotein cholesterol of <130 mg/dl; percentage with blood pressure of <140/90 mm Hg; and percentage receiving a periodic dilated eye exam, as well as two new measures: those receiving an annual foot examination and those receiving counseling on smoking cessation. In addition, the committee approved three quality improvement measures: self-management education, medical nutrition therapy, and interpersonal care (patient satisfaction).

While all committee members agreed to the adoption of the above, NCQA voiced concern about expanding the number of measurements because of the challenges involved in keeping pace with recording volumes of measurement data in a timely manner. A synopsis of the final version of the DQIP Measure Set follows (see Tables 1, 2, and 3). The measures included in Tables 1 and 2 are usually collected via patient chart or lab audit, while measures in Table 3 are obtained via patient reporting.

A great deal of discussion revolved around the establishment of these measures and the level of care that would be acceptable. The committee recognized that providing optimal care for patients often requires more intensive management than indicated by these measures.

The DQIP committee noted that care should be taken not to confuse measures with standards of care. The standards for diabetes care seek to provide health care professionals who treat diabetes with a means to set treatment goals, assess the quality of diabetes treatment provided, identify areas in which more attention is needed, and define timely and necessary referral patterns to appropriate specialists. They also are meant to help people with diabetes assess the quality of the medical care they receive, develop expectations for their role in their medical treatment, and compare their treatment outcomes to standard goals.²

In comparison, the measures are an indicator or tool used to assess the level of care provided within a system of care to populations of patients with diabetes. Measures do not reflect either the minimal or maximal level of care that should be provided to individual patients with diabetes. For example, if one reviews the ADA position statement "Tests of Glycemia in Diabetes," it may be noted that expert opinion recommends glycohemoglobin testing at least 2 times per year for those meeting treatment goals and more frequently (4 times per year) for those in whom therapy has changed or who are not meeting glycemic goals.³ Yet the DQIP Accountability Set measurement indicates that health care plans will have met the goal if one HbA_1c is drawn per year.

The decision to use the measurements shown here was made for several reasons, including 1) many individuals with diabetes in this country do not receive HbA_1c testing even once annually, thus requesting that the test be drawn 2­4 times annually appears very unlikely; 2) stratifying the population with diabetes on this and other variables is currently not possible; and finally, 3) there are no evidence-based data to indicate the correct number of HbA_1c values that should be collected annually for type 1 or type 2 diabetes. For this and other measures, DQIP recommended that plans and providers be accountable for at least the level of testing or variables as indicated.

Some of the measures have exclusions based on age or co-morbidity or based on results from a previous exam. Most measures apply to people with diabetes between 10 and 75 years of age, regardless of type of diabetes.

The Accountability Set includes measures that are recommended for plan-to-plan or provider-to-provider comparisons and are included in the HEDIS Reporting Set as voluntary in 1999 and mandatory in 2000. Each of the components in this list has undergone multiple studies in a variety of practice settings.

The Quality Improvement Set of measures has yet to be validated for comparing across plans and settings. However, some health care plans may choose to use it for internal quality improvement. The committee plans to field test these in DQIP's second phase and consider them for future versions of the HEDIS Reporting Set.

The Patient-Reported Measures reflect expert consensus and some data recognizing that diabetes is a disease for which 95% of the care lies with the patient. Therefore, the quality and level of education and care provided is critical to patients' ability to care well for their diabetes.

As managed care assumes a greater presence throughout the country, providers will be required to demonstrate that they are providing quality care to the members of the health care plans with which they are associated. There is evidence that if providers collect data on what they do, participate in reviewing the data from their patient populations, and are held accountable for the care given, ultimately care will improve. The measurements recommended by DQIP are a step toward assuring that, at a minimum, a basic level of diabetes care will be provided to an increasing number of health care plan members with diabetes.

Of particular interest to diabetes educators are the Patient-Reported Measures (Table 3). Self-management education and medical nutrition therapy are included here. The research-based Nutrition Practice Guidelines for Type 1 and 2 Diabetes demonstrate the value and positive outcomes that may be achieved with practice guideline care.^4-6 Implementation of these guidelines in your practice setting with appropriate outcomes tracking systems will help to document the critical need for quality nutrition care provided by registered dietitians and other qualified nutrition professionals in any health care plan that truly provides quality care to its members.