DIABETES CARE MANAGEMENT

Putting All the Pieces Together for Diabetes

Diabetes is one of the most costly and highly prevalent chronic diseases in the United States. Approximately half of the 16 million diabetic Americans are unaware that they have the disease. Diabetes complications cost the country nearly $100 billion a year. [1] Many complications, such as amputations, blindness and kidney failure can be delayed or prevented if they are addressed at an early stage. [2]

At the time this plan started this activity in 1994, diabetes was its eighth most common ambulatory diagnosis, and its sixth most common reason for inpatient admission. Using HEDIS methodology to identify diabetics, the plan had constructed a registry containing the names of approximately 500 patients.

PATIENT REGISTRIES Throughout this activity, the plan relied on a PC-based diabetic registry. This database included member names, demographic information and information on key services. Extracts from the claims and encounters system populated the registry electronically, and laboratory test results were manually entered into the database. This registry became operative during the baseline year of this activity.

The plan felt that great potential existed to reduce the burden of illness borne by suboptimally managed diabetics. Physicians recognized the potential for the improvement in diabetes care. They welcomed any additional resources that would help meet their patients' needs.

Rooted in a primary care staff model, the culture of this organization has been historically oriented toward quality improvement. With strong organizational support for HEDIS, NCQA Accreditation and quality initiatives, the plan's senior management supported an activity to improve the care of all diabetics.

The plan chose five quantifiable performance measures for this targeted population:

• Percent with an annual retinal exam per HEDIS criteria
• Percent with a health education encounter
• Percent with at least one HbA1C test per year
• Percent with two or more HbA1C values per year
• Average HbA1C value among diabetics undergoing testing

The plan used HEDIS specifications and both pharmacy and encounter data to identify diabetic members. The targeted population for most measures included diabetics over the age of 18, but the diabetic eye exam measure was confined to diabetics over age 31. For the three measures related to HbA1c testing, only members enrolled with staff model PCPs were included.

The plan measured baseline data for calendar year 1994, and set performance goals for each of its measures based on these baselines. By 1996, it was able to use NCQA's Quality Compass HMO average for diabetic retinal exams of 41.6 percent as a benchmark.

A project team managed this activity. It was composed of two medical directors, the QI manager, the nursing supervisor; the lab manager, a data analyst and a number of clinicians. Clinician representation changed annually, but it usually included a nutritionist, diabetic nurse educator, two primary care doctors, an optometrist and a physician assistant.

The QI manager and medical director prepared draft lists of barriers to improvement. They took these to the project team for additional brain-storming, and formal identification of root causes.

The lack of outreach efforts for member education and awareness was noted as an important barrier to good diabetic care. The medical record system was not conducive to tracking diabetic services and care.

The plan noted that - since this was its first attempt to develop a systematic, population-based approach to care - one of the most important barriers to improvement was the lack of a staff person to help develop such a program.

HUMAN RESOURCES FOR QI This plan already had many of the resources needed for study design and analysis. Medical management personnel had advanced degrees in epidemiology and statistics. Existing staff could extract encounter information and maintain a PC-based diabetic registry. Hiring an appropriately skilled person to develop a disease management program represented an additional resource. Given the strong organizational culture of quality, approval of this additional position was prompt.

The plan began this initiative by hiring a certified diabetes nurse educator. This person worked to coordinate the care of diabetics, and complemented a newly defined and enlarged role for the nutritionist. The plan also defined specific outcomes of teaching, and designed tools to assess the achievement of these outcomes (included in QP Tool ).

A diabetes flow chart was developed and distributed to all PCPs with a cover memo explaining its use in tracking the care of individual patients (a copy is included in QP Tool ).

The first remeasurement was based on 1995 data:

These data demonstrated some minimal improvement, but the plan was still short of its goals in every area.

The plan felt that its process for scheduling health education appointments might itself be an obstacle. It integrated the diabetic educator's appointment schedule into the primary care appointment system. This allowed a PCP's office to easily schedule an appointment with the educator. It also allowed patients to make appointments with both their PCP and the educator with a single call. An appointment request form made it easier for PCPs to refer to health education.

The roles of the educator and the nutritionist were substantially expanded. The medical director wrote a standing order allowing these individuals to order diabetic lab tests. Another standing order permitted the diabetic educator to modify diabetic medications according to an established protocol. (A copy of the policies, procedures and forms supporting this program are included in QP Tool .)

The plan began to provide feedback to primary care practitioners using lists developed from the diabetic registry. Practitioners could now easily identify diabetics who lacked health education visits or HbA1c tests. The diabetic educator also received a comprehensive listing of these patients. Both the PCPs and the Health Education Department conducted outreach activities.

The diabetic nurse educator adopted a continuing care record developed by a major pharmaceutical company in conjunction with the American Diabetes Association. (A copy is included in QP Tool .)

The second remeasurement was based on 1996 data:

Although the plan saw improvement and a positive trend on all the measures, it felt that problems remained in obtaining optometry exams. To identify specific barriers, it conducted a survey of diabetic members not receiving retinal exams. It found that many diabetic members with no optometry benefit did not realize that a visit to an eye care professional for an annual retinal exam was a covered benefit. As a result, it conducted planwide outreach to over 300 members noncompliant with retinal exams. A cover letter clarified vision care benefits. The plan reserved optometry appointments specifically for diabetic members.

It used reports from its registry to focus attention on diabetics with high HbA1c values. A listing of these patients was sent to each PCP. The patient's diabetic prescriptions and health education histories were included. (A sample report is included in QP Tool .) A copy was also given to Health Education. The plan performed telephonic outreach to these members.

The third remeasurement demonstrated dramatic improvement based on 1997 data:

With the exception of the annual retinal exam, the plan had met all of its performance goals. Statistical tests indicated that all improvements over baseline performance were significant.

The plan has continued to focus on diabetic care. It sent lists of diabetics noncompliant with retinal exam recommendations to primary care providers for direct contact by the doctor or nurse.

Outreach to members has continued. Members now get flu shot reminders in addition to information about diabetes disease management.

The plan added a second diabetes educator.

It has enhanced the reports for physicians, focusing on certain populations, such as diabetics with no HbA1c values in the last 18 months.

The affiliated network has been included in what started as primarily a staff model program, and a network physician was added to the project team.

Data entry of HBA1c and lipid values is now occurring directly in the laboratory, with a PC purchased specifically for this purpose.

The plan continues its efforts, proud of the dramatic and sustained improvement it has achieved in every parameter of diabetic care it has measured.

[1] - National Committee for Quality Assurance, HEDIS 2000 vol 1 (Washington, D.C.: National Committee for Quality Assurance, 2000), 44.

[2] - Ibid.