GLUCOSE CONTROL IN MEMBERS WITH DIABETES

Better Diabetes Care Through Technology

Knowing the importance of good diabetes care, this plan set up a computerized registry of its members with diabetes in 1986. Evaluation of data from this registry demonstrated that only one third of the plan's members with diabetes were in good control, and that members with diabetes consumed 15% of the medical care budget. These factors led the plan to engage in a quality improvement (QI) activity to improve the care of its population of patients with diabetes.

In 1993, the Diabetes Complications and Control Trial (DCCT) showed the appropriate blood glucose control in patients with type 1 diabetes could reduce the incidence of major complications by up to 50%. (These results have been subsequently supported by the United Kingdom Prospective Diabetes Study [UKPDS] for patients with type 2 diabetes.)

Having a large number of patients with poorly controlled diabetes, the plan knew that the costs of caring for diabetes and its complications would represent a significant investment. Therefore, the plan decided to invest in a technology-driven program to improve the health status of its members with diabetes.

The targeted population included all members with diabetes identified in the plan's computerized registry of members with diabetes. The plan chose performance measures of both process and outcome. The plan decided to look at the process by determining whether members with diabetes were receiving appropriate screening tests for glucose control. The plan defined the screening measure as the percentage of continuously enrolled members with diabetes who had received at least one glycosylated hemoglobin (HbA1c) or fructosamine test during the past 12 months.

It looked at outcome by determining the percentage of members with diabetes screened who were in good or excellent control. Good glucose control was defined as a HbA1c measurement of 7.01 to 8.0, or a fructosamine level of 285 to 325. Excellent control was defined as a HbA1c measurement of less than 7.0 or a fructosamine level of less than 285.

The baseline performance, based on 1994 data, showed the 78% of members with diabetes were screened, and 47% were in good or excellent control.

No benchmark was determined in 1995. The plan later used the 1996 American Diabetes Association (ADA) Provider Recognition Standards to identify a benchmark of having 40% of members with diabetes in good or excellent glucose control. No benchmark was established for the screening measure.

The performance goal the plan set for the initial year was qualitative: to control risk factors for the complications of diabetes mellitus. Measurable goals were adopted for both measures in 1996. These goals were to screen 85% of its members with diabetes, and to have 65% of them in good or excellent control.

A barrier analysis was conducted by a diabetes steering committee. This team included endocrinologists, operations staff, a pharmacist, an internist, a diabetes educator nurse clinician, and a methodology expert. They identified access to current clinical patient data as a key issue for this initiative. Awareness of the importance of accurate and up-to-date patient information led to the initial interventions.

FOUNDATIONAL EFFORTS This plan had already implemented a number of interventions prior to formally beginning this activity. In addition to the diabetes registry and the paper flow sheets, the plan had published educational articles in the provider newsletter, and held continuing medical education sessions on diabetes. Although these activities alone did not result in improved glucose control measures, they laid a good foundation for the stronger interventions implemented during the actual QI activity.

From its barrier analysis, the plan suspected that electronic flow sheets would be more widely utilized than paper flow sheets. Since all members with diabetes were already flagged in the appointment system, staff was able to easily identify patients at every visit. An electronic flow sheet would allow the member's clinical data to be reviewed each time they were seen - regardless of location. Working together with staff from the information services department, the diabetes committee implemented an online results reporting system. A summary screen listed key test results and dates, automatically highlighting abnormal findings.

The plan developed a diabetes care manager network. Diabetes care managers were based in primary care areas of outpatient medical centers.

The plan developed streamlined protocols on diabetes care for care managers. Diabetes care managers initially were trained for six primary care modules. Later, the plan held a three day workshop to train additional care managers.

The plan also continued to distribute quarterly clinician-specific reports on their paneled members with diabetes, an activity which actually started in late 1994. Care managers were not included in the distribution.

The plan distributed guidelines to providers on the use of a new oral hypoglycemic agent. They also decentralized the diabetes education program. This enabled the plan's staff to better support education and management activities via small member groups.

In addition, educational efforts for practitioners were continued. Journal articles covered topics pertinent to the treatment of diabetes and updated clinicians on the progress of the initiative.

Remeasurement, using 1995 data from the diabetes registry and the laboratory information system, showed that 81% of members were screened and 59% were in good or excellent control. Both improvements were statistically significant.

The plan concluded that its interventions, including panel reports, online clinical information, care management, and ongoing educational efforts, contributed to the significant improvement in both measures. It decided to continue its efforts.

INFORMAITON TECHNOLOGY This plan made use of innovative information technology throughout this activity. The computerized diabetes registry, automated appointment system that flagged members with diabetes, electronic flowcharting of test results, and the publication of practice guidelines on an intranet all contributed to the plan's success. These efforts demonstrate how health plans can invest in technology to improve care and reduce costs.

Clinician feedback indicated that the care managers were very helpful in managing complex patients. The care management system appeared to increase the efficiency of the health care delivery system.

Based on the results and the clinician feedback, the plan decided to continue to enhance the care manager program. It help additional training for the care managers to keep them up to date.

The plan's pharmacy department evaluated adherence to the medication guideline developed the previous year. Staff distributed these guidelines more widely by installing them on their intranet Web page.

Remeasurement, using 1996 data, showed a screening rate of 81.5%, not a statistically significant change. However, the percentage of members in good or excellent glucose control increased to 63%, a statistically significant improvement.

The change in glucose control represented an improvement in a true outcome measure of diabetes care. For 1997 the plan raised its goal for members screened to 87%. It kept its goal for members in good or excellent glucose control at 65%.

Based on the success of this program, the plan began to develop and implement a pediatric diabetes program. The plan identified clinician experts to provide leadership and hired a pediatric diabetes coordinator. It developed and implemented a team approach to pediatric diabetes care.

The plan made footcare nurses part of the diabetes care teams and implemented joint training sessions. It also widened the distribution of diabetes panel reports, which provided member-specific information about HbA1c and fructosamine testing, to include all offices and primary care areas. It initiated quarterly reports on measures of glucose control by primary care service area, by facility, and later, by physician.

The plan conducted a survey of all of its members in the diabetes registry. The survey focused on a number of issues, including opportunities for education.

1997 results showed that 86.8% of members were screened and that 66.5% of diabetic members were in good or excellent control, both statistically significant improvements.

The plan has continued its activity. Actions implemented in 1998 focused on enhancements of existing systems, and on updating guidelines for identification, diagnosis, and treatment.