DIABETES CARE MANAGEMENT

Broad Improvement Across a Set of Comprehensive Measures

Diabetes has been a common topic for quality improvement activities. Plans have worked to improve HEDIS measures, such as the rate of diabetic eye exams. HEDIS 2000 included a comprehensive diabetes measure. As early as 1996, this plan embarked on a quality improvement activity that used a comprehensive set of diabetes performance measures.

Nationally, the number of people diagnosed with diabetes increased fivefold, between 1958 and 1993. [1] According to the ADA, 2,200 people are newly diagnosed with diabetes everyday. [2] The plan noted that the prevalence, clinical and financial impact of diabetes had made it a priority among consumers, employers, and state and federal regulatory agencies.

The plan annually analyzes demographic, encounter and pharmacy data to guide its quality improvement activities. The plan noted that diabetes accounted for a high number of both outpatient and inpatient encounters over the last several years. In 1998 it was the second most common reason for outpatient encounters in both commercial and Medicare populations.

By 1994, the plan had adopted a clinical practice guideline for diabetes. It reported its HEDIS rate for diabetic eye exams. Participating in a statewide effort to improve diabetic eye care, it had already developed a reminder system and standardized chart documentation instrument. Experience with these existing activities convinced the plan that there was room for improvement.

With a significant segment of the plan's population affected by this high volume, chronic disease, the plan decided that quality activities in this area could bring about meaningful improvement in the health status of its membership.

Using HEDIS 3.0 specifications, the plan used encounter and pharmacy data to identify people over age 31 with diabetes. Evidence of diabetes included:

• A prescription of insulin or an oral hypoglycemic agent, or
• An inpatient or ER encounter with an ICD-9-CM code for diabetes, or
• Two ambulatory visits with ICD-9-CM codes for diabetes.

The plan reviewed information from the Foundation for Accountability (FACCT), the Diabetes Quality Improvement Program (DQIP), and the American Diabetes Association (ADA) to select a set of 11 performance measures.

MULTIPLE PERFORMANCE MEASURES How many performance measures should a plan use to guide its quality improvement efforts? The answer is: it depends. Deciding on the number of measures to use is a balancing act. More measures mean the ability to detect a wide range of issues, and to implement a broad range of interventions. Improvement across multiple measures is more meaningful than an increase in a score for an isolated aspect of diagnosis or treatment. On the other hand, too many measures can make the project unwieldy. The process of measurement can consume large amounts of resources, and detract from a focused improvement effort. The tight management of many measures makes this plan's effort all the more impressive.

A random sample of medical records from the targeted population of people with diabetes was used to examine eight measures:

• Percent with an annual retinal exam
• Percent with one of more glycohemoglobin tests
• Percent of those having glycohemoglobin tests showing a level of <8.5 percent
• Percent with an annual screening test for microalbuminuria
• Percent with two or more blood pressure checks per year
• Percent of those with one or more blood pressure checks having a systolic BP <135
• Percent with an annual lipid panel
• Percent of those with an annual lipid panel showing an LDL level <130 mg/dL

A survey mailed to a random sample of the targeted population was used to explore three additional measures:

• Percent of people with diabetes reporting one or more foot exams by a health care provider
• Percent of smokers with diabetes reporting having received smoking cessation counseling
• Percent of people with diabetes reporting having received an annual influenza vaccination

The plan measured baseline performance based on 1996 data. The plan used a benchmark for diabetic retinal exams based on NCQA's 1998 Quality Compass. The remaining benchmarks were determined by examining the best baseline results among sister plans under the same corporate parent. Looking at baseline performance and benchmarks, the plan set its performance goals.

A cross-functional team composed of representatives from QI, health management, pharmacy, network management and member services performed an initial barrier analysis. Subsequently, a physician advisory committee reviewed the analysis. The identified barriers were:

• Missed opportunities for patient screening
• Lack of standardized clinical guidelines
• PCPs unfamiliar with latest treatment recommendations
• Lack of member knowledge about self-management
• Lack of member knowledge about preventive screening services

The plan identified each of these barriers as representing an area of opportunity for improvement.

The plan began by mailing members reminders about the importance of diabetic retinal exams. The plan followed this up by implementing an automated reminder system at participating pharmacies: When a member filled a prescription for insulin or an oral hypoglycemic agent, he/she received a computerized reminder for an eye exam.

Using its provider newsletter, the plan distributed guidelines on diabetic care. It also distributed a flow sheet that outlined the minimum practice recommendations and provided a chart tool for documentation (included in QP Tool ).

The plan initiated a health assessment tool for new enrollees that identified diabetics. This provided valuable information to medical groups about their members' health status and needs during early enrollment period.

The plan participated in a statewide project headed by the Medicare Peer Review Organization (PRO). This project established an additional baseline of conformance to minimal standards of care through a review of a random sample of charts for people with diabetes. The results of this project were shared with physician advisory committees in each market.

The plan took advantage of diabetes intervention program developed by its corporate parent to address lifestyle and self-care issues. Members were introduced to the program through a mailing that included an eight-question survey on their readiness to make lifestyle changes and a wallet reminder card for preventive services (included in QP Tool ).

Members returning the questionnaire received a variety of interventions based on their responses, including:

• A diabetes self-care kit consisting of a book, a videotape, a goal planning and log book, and recipe cards
• Lifestyle material for well-managed diabetics
• A free glucometer
• Telephone counseling with lifestyle advice on a regularly scheduled basis (seven to 14 calls per year)
• A stop smoking kit for tobacco users wanting to quit

Plan staff worked with medical groups to implement the program, discussing referral forms and policies. They worked with the groups to implement the free glucometer program.

A variety of provider and member newsletters featured the recommendations for preventive services for diabetic patients.

The first measurement was based on 1997 data:

*Indicates statistically significant improvement from 1996 baseline results.

The plan determined (from survey results) that members had access to care, but preventive care was still not routinely occurring at optimal rates. It presented the remeasurement data to the physician advisory committees for barrier analysis and recommendations. Barriers discussed included:

• PCPs remain unaware of the quality initiative
• Changes in recommendations for urine protein testing have not been adopted
• PCPs are not routinely using the charting tool
• PCPs are unaware of, fail to refer patients to, the smoking cessation program
• PCPs are not aware of all of the people with diabetes on their panels
• PCPs fail to refer for eye exams
• Members need more education related to self-care
• Members lack knowledge about routine preventive services
• Members are reluctant to pay co-payments
• Member do not follow through in scheduling exams
• Vision services are provided through a supplemental benefit; with no referral required for these services, no feedback mechanism exists for PCPs

The plan developed and distributed to practitioners a diabetes disease profile (included in QP Tool ). The profile included:

• A list of all patients with diabetes
• Their patients' ER and inpatient claims histories
• Their patients' complications and comorbidities:
• hypertension
• CHF
• retinopathy
• MI
• renal disease
• Their patients' quality indicators:
• retinal exams
• use of ACE inhibitors by patients with hypertension/CHF
• compliance with medications (s)
• Patients' participation in the diabetes self-care program
• The patient's total cost of care

It developed a similar tool for distribution to participating medical group medical directors. The Medical Directors Profile Report (included in QP Tool ) included:

• A list of each PCP's patients with diabetes
• Number of assigned members per PCP, and percent of members with diabetes
• Number of ER visits
• Number of hospitalizations
• Percent of hypertension/CHF patients on ACE inhibitors
• Percent of patients with comorbidities
• Cost per member

As the plan transitioned to a new eye care vendor, it emphasized the importance of coordination of care with respect to diabetic retinal exams. A screening form was used to transmit information from the eye care vendor to the plan and PCP.

The plan continued and fully implemented many of the interventions it had initiated the year before. It continued its pharmacy-based reminders. It mailed the internally developed health assessment survey to all newly enrolled Medicare members at the beginning of each month. It helped distribute free glucometers by allowing PCPs to call a toll-free number and order up to 25 glucometers for distribution to members.

The plan developed a smoking cessation kit that included video-and audio-tapes, seven booklets, a bag of items to help control the urge to smoke and an evaluation card. It waived co-pays for its smoking cessation programs.

The plan used a variety of mailings and newsletters to members and practitioners to raise awareness about proper diabetic care and the availability of programs for diabetics. It implemented a plan Web site that featured articles on lifestyle and Medicare benefits, and links to resources on diabetes.

The second remeasurement was based on 1998 data as shown in the table below.

Although the plan had not met all of its performance goals, it achieved meaningful improvement in the care of people with diabetes by demonstrating a statistically significant improvement in eight of the 11 performance measures it used.

The plan continued its various interventions. It developed a stratification system to identify people with diabetes at increased risk for complications, based on three risk levels with the following definitions:

• Level I: Oral hypoglycemics; no admissions, comorbidities or risk factors
• Level II: Oral hypoglycemics or insulin; good control; no admissions; no more than one comorbidity or risk factor
• Level III:
• Polypharmy or
• Admission or ER visit in last year or
• More than one comorbidity or risk factor

The plan has conducted additional barrier analyses by surveying noncompliant members and their physicians.

*Indicates statistically significant improvement from 1996 baseline results.

Despite the plan's aggressive interventions, it noted that only 20 to 30 percent of the targeted population is responsive and actively participating. Its current efforts focus on improving the percentage of members participating in the various aspects of the program. By increasing the number of members involved in the comanagement of chronic diseases like diabetes, the plan hopes to make its programs the source of even more meaningful improvement in care for its members.

[1] - Harris, Maureen I. "Summary," chapter 1 of Diabetes in America, vol 2 (Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Disease 1995), 1.

[2] - "Diabetes Facts and Figures," In Diabetes Facts and Figures [page on website of the American Diabetes Association]. Alexandria, VA, 2000 [cited January 30, 2001]. Available from www.diabetes.org/ada/facts.asp; INTERNET.