Our Standards and How We Selected Them

Principles and Individual Standards

For each process and outcome measure, our Diabetes Subcommittee reviewed relevant standards of the NQF, NCQA, VHA-DOD, ADA, the American Medical Association (AMA), and the Agency for Health Care Research and Quality (AHRQ). For several of the measures, different organizations have established different standards (target values) above (or below) which a practice could be recognized for desirable (or undesirable) achievement. In these cases, three principles guided the subcommittee’s deliberations about which target value or standard to choose:

• 1) Favorable framing. We first sought to define each standard in a favorable frame, such that higher levels of reported achievement are associated with a higher proportion of a practice’s patients meeting the standard. This principle guided decisions, for example, to reject a commonly used standard for hemoglobin A1c levels – the percentage of a practice’s patients in “poor glycemic control,” where lower numbers are better – in favor of a standard that reports the percentage of patients in a given practice who have “good glycemic control.” We also report the “percentage of documented non-smokers” instead of the “percentage of smokers” for the same reason.
• 2) Relevant to populations. Second, we sought to define standards that would apply to most all diabetic patients in a population, without need for numerous patient exclusions or extensive statistical risk-adjustments. This principle guided our choice of relatively conservative target values for Hemoglobin A1c and blood pressure because of concerns that seeking more aggressive targets on older patients, or those with multiple co-morbid illnesses regardless of age, might lead to more risk than benefit. At least for A1c, data supporting this concern were recently published by the National Institutes of Health5 as they stopped the aggressive treatment arm in a large clinical trial (called “ACCORD”) due to safety concerns.
• 3) Use of all available evidence: a cost-effective approach. Two of our measures and related standards reflect “hybrid” approaches that allow for desirable results either if a test result is better than a specified target value or if recommended treatment has been prescribed for the underlying medical pathology. We developed two hybrid standards: one for management of “bad” cholesterol – giving credit for meeting low LDL cholesterol targets or treating with LDL-lowering statin drugs – and another for managing potential kidney problems – giving credit for regular monitoring of urine protein levels or treating with certain classes of blood pressure drugs called “ACE inhibitors” and “ARBs”. Both of these hybrid standards are based on “Grade A” evidence as evaluated by the American Diabetes Association2 and simultaneously minimize unnecessary repeated testing if proper treatment has been prescribed.

Based on these principles and the recommendations of national quality organizations, our four process standards and our five outcome standards are summarized in Table 1.

Table 1. Diabetes Process and Outcome Measures and Standards

Hemoglobin A1c Test Performed Control of blood sugar (Glycemic control) is fundamental to the management of diabetes. The Hemoglobin A1c is reflects average blood glucose levels over several months. According to the American Diabetes Association, A1c testing should be performed routinely in all patients with diabetes, at initial assessment and then as part of continuing care. Testing is most helpful when obtained repeatedly (e.g., every three months in those patients in whom treatment is changing or target levels have not been achieved). However, the frequency of A1c testing depends on the clinical situation of the individual patient and the judgment of the clinician. A single level in a year’s time is a relatively low threshold.

Kidney Management Diabetic kidney disease (nephropathy) occurs in 20–40% of patients with diabetes and is the single leading cause of end-stage renal disease (ESRD) that requires dialysis. Low levels of protein (albumin) in the urine (microalbuminuria) is a marker for development of nephropathy in type 2 diabetes. Most importantly, there are interventions that reduce the risk and slow the progression of renal disease. In addition to good control of blood sugar and blood pressure, there is substantial evidence for the benefit of treatment with Angiotensin Converting Enzyme Inhibitors (ACE-Is) or Angiotensin Receptor Blockers (ARBs).

Eye Examination Diabetic eye disease (retinopathy) is a highly specific complication of diabetes and is the most frequent cause of new cases of blindness among adults aged 20–74 years. Most importantly, laser photocoagulation surgery is very effective in preventing visual loss due to diabetic retinopathy. Therefore, early identification is beneficial. In addition, glaucoma, cataracts, and other disorders of the eye occur earlier and more frequently in people with diabetes. Patients with diabetes should have annual dilated and comprehensive eye examination by an ophthalmologist or optometrist.

Pneumonia Vaccination Influenza and pneumonia are common, preventable infectious diseases associated with high mortality and morbidity in the elderly and in people with chronic diseases. Patients with diabetes are at higher risk. Therefore, it is recommended that at least one lifetime pneumococcal vaccine be provided for adults with diabetes.

Hemoglobin A1c Value There is considerable disagreement among organizations, governmental agencies, and professional societies about the most appropriate target level for A1c that should be used for public reporting and for quality improvement. The American Diabetes Association recommends <7%, but their guidelines state: “Less stringent goals (than <7%) are appropriate for individuals >65, those with co-morbid conditions, and those with side effects.” These guidelines are similar to those of the Veterans Healthcare Administration and Department of Defense and the American Geriatric Society. We have chosen a level that is the simplest to apply across large populations with varying risks of adverse events from aggressive control. We recognize that the target for an individual patient may differ and depends upon clinical circumstances, physician judgment, and patient preferences.

Blood Pressure and Cholesterol Management Cardiovascular disease is the major cause of illness and death for individuals with diabetes. Hypertension and dyslipidemia (hypercholesterolemia) commonly co-exist with diabetes and are risk factors for cardiovascular disease. Diabetes itself is a risk factor. There is a great deal of evidence to support the benefits of controlling cardiovascular risk factors in preventing or slowing cardiovascular disease in people with diabetes. Work continues on the most appropriate target levels for blood pressure and LDL-cholesterol.

Blood Pressure Randomized clinical trials have demonstrated the benefit (reduction of coronary heart disease [CHD] events, stroke, and nephropathy) of lowering blood pressure to <140 mmHg systolic and <80 mmHg diastolic in individuals with diabetes. Lower thresholds have been recommended by some. The ongoing Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial is designed to determine whether lowering systolic blood pressure to <120 mmHg provides greater cardiovascular protection than a systolic blood pressure level of <140 mmHg in patients with type 2 diabetes.

LDL Cholesterol Management For most patients with diabetes, the first priority of dyslipidemia therapy is to lower LDL cholesterol to a target goal of <100 mg/dl. Multiple clinical trials have demonstrated significant effects of pharmacologic (primarily statin) therapy on cardiovascular outcomes in subjects with coronary disease and for primary prevention.

Body Mass Index Maintaining or achieving an ideal body weight is an integral component of diabetes prevention, management, and self-management education. Thus, weight loss is recommended for all overweight or obese individuals who have or are at risk for diabetes. A BMI below 18.5 is considered underweight, 18.5 to 24.9 is normal, 25 to 29.9 is overweight, and 30 or above is considered obese. Our target BMI of less than 30 is consistent with the recommendations of the American Diabetes Association.

Smoking Status Cigarette smoking contributes to one of every five deaths in the U.S. and is the most important modifiable cause of premature death. Studies of individuals with diabetes have consistently found a heightened risk of cardiovascular disease and premature death among smokers. Smoking is also related to the premature development of microvascular complications of diabetes.

Summary Standards: Principles and Targets

As mentioned above, we constructed two Summary Standards to reflect overall good results: one for our outcome standards, and a second for our process standards. Thus, all practices, and the region overall, are rated on these two Summary Standards, reflecting achievement on the 5 outcome standards and the 4 process standards, respectively.

We created our Summary Standards for three main reasons:

• 1) To Simplify Understanding of Overall Performance. First, we want to make it simpler to understand “overall” performance despite the fact that several components are being measured and reported. Measuring and understanding quality is complicated for all of us!
• 2) To Clearly Distinguish Standards that Depend on Different Factors. Second, we want to distinguish overall “Outcomes” achievement from overall “Process” achievement because success on these different measures is likely to depend on different factors. Process measures mostly depend on what doctors and their practices do. Outcome measures also depend very much on patient self-care, as well as other forces, such as the patient’s available resources – including money for medicines and test strips for monitoring diabetes, insurance coverage, places to exercise, the patient’s ability to read and understand instructions, etc.
• 3) To Provide Focus for Improvement. Third, we want to help our practices and other stakeholders to focus on the “big picture” in their efforts to improve, and not just individual standards. As our results in this report show, everyone has room to improve. Where do we start? Our summary standards can provide some guidance over and above their individual components.

The Summary Standards reflect how each patient of each doctor in each practice is doing on our outcomes and process standards. At the practice or region level, we ask:

Summary Outcome Standard:

How many of our 5 individual standards does each patient meet? We measure our achievement by determining the percentage of patients in each practice who meet at least 4 of the 5 standards listed in Table 1.

Summary Process Standard:

How many of our 4 individual standards does each patient meet? We measure our achievement by determining the percentage of patients in each practice who meet ALL 4 standards listed in Table 1.

We were more “strict” on our achievement target for our Summary Process Standard – requiring all 4 standards to be met – because this standard depends more on the doctor and is more until his/her control. We were more “lenient” in our achievement target for our Summary Outcome Standard – requiring that either 4 or 5 of the 5 standards be met - because this standard depends on the patient and the patient’s resources almost as much as the doctor and the practice.