CORONARY ARTERY DISEASE

Supporting and Rewarding Improved Performance

Coronary artery disease remains the number one killer of both men and women. Every year more than a million Americans suffer heart attacks, and 500,000 die from heart disease. Nearly 50 million Americans have high cholesterol. Reducing cholesterol levels in patients with known heart disease can reduce morbidity and mortality by as much as 40 percent.

With approximately 3 percent to 5 percent of its membership suffering from heart disease, this plan targeted a 25 percent reduction in heart disease events as early as 1994. The plan addressed multiple risk factors with initiatives focusing on aspirin use, weight management, exercise, blood pressure control, and tobacco use. In 1996, the plan added a focus on LDL cholesterol monitoring and management. Its findings identified the need for improvement in this important risk factor for heart disease.

Members with known cardiac disease were targeted. In 1996, approximately 14,000 individuals were identified on the basis of ICD-9-CM diagnosis codes for either inpatient or outpatient care. CPT-4 codes for procedures such as coronary artery bypasses and angioplasties were also used to identify patients.

MISTAKEN IDENTITY This plan recognized that many patients who had visits coded as cardiac disease did not, in fact, suffer from heart disease. Miscoding may occur when patients are being considered for a diagnosis that is not subsequently confirmed. The plan considered changing its specifications to require two or more visits with a cardiac ICD-9 CM code, but discovered that as many as 25% of eligible patients would be dropped with this criterion. Instead, the plan staff decided to enlist practitioners' help in identifying patients with erroneous diagnoses on "at risk" lists.

The plan chose two HEDIS measures. One was a process measure, the other an outcome measure. Using administrative lab data, it looked for the percentage of the target population that had received an LDL screening test in the last 12 months. For patients with an LDL test, the plan looked for a level <130 mg/dL. In plan-owned clinics, it used automated laboratory data to determine LDL levels. For contracted medical groups, a random sample of 80 charts were reviewed per medical group. This sampling strategy was designed to provide confidence levels of + or - 5 percent at the health plan level. (A copy of the medical record abstracting instructions for the activity is included in QP Tool .)

Baseline performance, based on 1996 data, showed an LDL screening rate among identified coronary artery disease (CAD) patients of 37.6 percent. Of those screened, only 55.2 percent demonstrated an LDL level of < 130 mg/dL.

The plan set goals of 60 percent for both the LDL screening rate and for the percentage of the LDL levels below 130 mg/dL in screened patients.

A steering group composed of representatives from QI, chronic disease management, performance improvement, preventive services, MIS, and reporting and analysis examined the baseline data. Although it did not perform a formal root cause analysis, it worked together with medical group leaders to determine potential barriers.

Barriers identified included:

• Need for redesigned (but budget-conscious) patient education programs to support chronic disease management
• Need for coordination of education around each individual patient to replace multiple reminders for a variety of services
• Need to track nutritional counseling and use of statin drugs in CAD patients

The plan began its efforts by revising and distributing a number of clinical practice guidelines. At least 14 medical groups implemented a stable coronary artery disease guideline. At least 30 medical groups implemented a lipid treatment and management guideline.

The plan generated "at risk" lists that identified all members with CAD, and the date of their last LDL level. These "at risk" lists were provided to medical groups. The groups used the lists to reach out to members, including "silent members" who had not accessed medical group services.

An existing program that provided financial recognition to medical groups that achieved superior performance was revised to include the measures of LDL testing and control (details about the program are included in QP Tool ).

FINANCIAL INCENTIVES The health plan reported clinical indicators regularly to its medical groups, helped them with tools and resources to better manage their patients and rewarded them for improved performance. In 1998, eight groups met their goals and were awarded bonuses ranging from $10,500 to $32,000. The clinical indicators themselves were made available to consumers through the health plan Web site.

The health plan hired a staff member to serve as a dedicated cardiac advisor. This individual participated in the revision of guidelines; reviewed, developed, and distributed patient education materials; monitored national trends and best practices; and helped medical groups use the "at risk" lists generated by the plan. The cardiac advisor also developed phone call scripts, template reminder letters and a process to implement these tools.

The plan had developed a telephonic health education service. It implemented a class offering one-on-one counseling to patients with elevated lipid levels.

The plan conducted a number of member education efforts. It offered basic and advanced cholesterol classes. A high intensity risk reduction program was initiated. This program featured intensive instruction in a very low-fat plant-based diet, stress reduction and exercise.

It also distributed bulletins and updates to medical groups on the appropriate utilization of statins.

INNOVATION AND COMMUNITY OUTREACH This plan implemented a "Better Health Restaurant Challenge" program to increase the availability of low-fat foods in community restaurants, and to increase diner selection of "low fat" menu items. Registered dieticians evaluated menu items to ensure they met the criteria for the designation of "low-fat." The plan held a contest for the best tasting, low-fat menu items.

The first measurement, based on 1997 data, documented an LDL screening rate of 42.4 percent. The percentage of screened CAD patients with a documented LDL level of < 130 mg/dL had risen to 63.4 percent. A two-sample z-test demonstrated a p-value of < 0.0001 for each of these measures, indicating statistically significant improvements.

The plan reached its goal for LDL levels, but not for screening rates. Nevertheless, it demonstrated meaningful improvement in controlling this important cardiac risk factor.

The health plan continued its efforts during 1998 and 1999. It learned several lessons, identified new barriers and designed new interventions.

As originally conceived, the high intensity risk reduction program proved too costly to maintain. It underwent revision. High-risk patients received "heart care" cards outlining their medications, lab values and treatment goals (included in QP Tool ).

Some members identified cost as a barrier to health education. The plan implemented a policy of waiving class fees, and partnered with community colleges by providing discounts to members who found the plan's own classes inconvenient to attend. Practitioners can now write a prescription for health education. (A sample prescription is included in QP Tool .)

The plan expanded its phone line offerings to include risk assessment and customized development of self-care plans. Issues can be referred to the PCP based on responses to questions about medical history, medication usage, lab values, lifestyle practices and readiness to change. Options include smoking cessation; and weight, cholesterol and stress management programs.

The plan continued its previous interventions of guideline implementation, member and provider education and financial incentives.

Performance measures have continued to improve. 1998 data showed a screening rate of 52.6 percent and a controlled LDL rate of 67 percent. 1999 data demonstrated further improvement, with a screening rate of 65.3 percent, and a control rate of 71.9 percent. The plan recently added a third measure, looking at the percentage of screened patients with LDL levels of < 100 mg/dL.

The plan is now shifting its efforts from the measurement of discrete measures such as LDL levels in isolation, to monitoring how patients are managed across all indicators for a specific condition. By now measuring what percentage of patients have had all the appropriate test for their clinical situations, the plan hopes to achieve major advances in the health of its population.