Diabetes Mellitus Quality Improvement

Diabetes Mellitus Quality Improvement

Behnke, Wendy R

An Initiative by the University of Michigan Hospitals and Health Centers

In December 2001, the Southeast Michigan Health Care Quality Forum and the Michigan Quality Improvement Consortium (MQIC) invited local health systems to participate in a clinical practice guidelines demonstration project designed to improve the clinical management of patients with diabetes mellitus. The project’s focus was on an “all payer performance assessment” and improving physician care processes and practices within the office setting. In January 2003, the University of Michigan Hospitals and Health Centers (UMHHC) chose diabetes management as an Ambulatory Care Quality Improvement initiative and one of UMHHC’s “Quality Pillars” to improve the care of patients with diabetes mellitus. Approximately 4,000 adult patients (age 18 and older) who have been diagnosed with diabetes mellitus and who receive care from the University of Michigan primary care providers are enrolled in this study. The primary goal of this project is to reduce the proportion of patients who have never had their hemoglobin A1c (HgBA1c) levels measured or whose HgBA1c level is > 9.5%. Secondary goals include improving the HEDIS diabetes indicators (see Table 1).

This project combines an audit and feedback system, and the use of a clinical pharmacist or a nurse and pharmacist to assist primary care providers in the management of patients with diabetes mellitus. While both control and intervention sites receive standard audit and feedback (for example, measurement of baseline HgBA1c levels for all patients and provision of this information to all health care providers), only the intervention sites utilize the trained clinical pharmacists and/or nurses for this project.

Project Design

This project focuses on adult patients who have diabetes mellitus and received diabetes care from the University of Michigan’s primary care providers. Patients eligible for the study population included those who were age 18 or older and received primary care at the University of Michigan in 2002. They must also have had an ICD-9 billing code for diabetes mellitus in 2000 or 2001, and received care at one of the University of Michigan’s health centers. Patients were excluded if their billing diagnosis could not be verified through: (1) a review of computerized medical records for hypoglycemic medications; (2) confirmation by having “diabetes” listed in the patient’s problem summary list; or (3) if they did not have a hemoglobin A1c > 9.5%. Because the study was focused on primary care and not specialty care, patients were also excluded if they had been seen in the Endocrinology clinic at any time since 2000.

In order to ensure balanced levels of glycemic control in the intervention and control groups, the health centers were divided by baseline HbA1c levels. all physicians and health centers receive baseline and quarterly reports on their patients as part of the University of Michigan’s Ambulatory Care GUIDES’ feedback program. Intervention nurses and pharmacists also receive electronic and hard copy data for all patients at their health center. For the intensive intervention aspect of the project, Phase I sites were chosen by their readiness to participate (see Table 2). Nurses and pharmacists at the Phase I health centers were trained in diabetes care management and medication protocols developed specifically for this project. If this project is successful in improving care for patients with diabetes, training and implementation at Phase 2 sites will commence.

Project Implementation

A UMHHC Task Force co-chaired by Connie Standiford, MD, an Internist and Medical Director for Ambulatory Care Services, and Wendy Behnke, MSA, MSN, RN, a Health Center Administrator and Regional Nurse Manager for Ambulatory Care Services, began meeting in January 2003 to coordinate the program that was endorsed by multiple UMHHC units (see Table 2). A multidisciplinary committee then met to identify educational needs of the nurses and pharmacists as well as develop plans for the intervention sites. It also obtained approval for nurses and pharmacists to have electronic signature authority, a privilege traditionally granted only to physicians, that allows the nurse or pharmacist to document directly into the electronic medical record so the information is available for physicians and others participating in patient care.

Planning, Education, and Intervention

The committee developed several tools. These included:

* A flowchart summarizing the workflow of physicians, pharmacists, nurses, and data-managers.

* A diabetes care management algorithm consistent with UMHHC diabetes care guidelines.

* Oral agent algorithms.

* Standing lab orders for diabetes management.

Sixteen nurses and pharmacists who work at the intervention clinics received 12 hours of specialized diabetes education (see Tables 3 and 4). The committee also selected videos and tools for use by the nurses, pharmacists, and patients. Nurse and pharmacist training was completed in September 2003. Individual physicians, nurses, and pharmacists then chose which aspects of care on which to focus (such as bringing patients in for a diabetic visit, managing glycemic control or lipids more aggressively), and this is being tracked. All work was endorsed by the Professional Nurse Practice council as well as the diabetes steering committee.

Tracking Changes

Figure 1 shows the proportion of patients with diabetes managed only by primary care seen in at least 2 of the last 3 years with elevated or missing A1c tests at control and intervention sites since January 2003. Data are updated and distributed quarterly. As the project is on-going, no statistical analyses have yet been performed. The lines represent the proportion of patients with either an HbA1c result > 9.5% during 2003 or those who had no test. Although tracking of data is shown since December 2002 and patient specific reports were mailed beginning in July 2003, nurse and pharmacists only completed their training at the end of September 2003. Figure 2 shows that the proportion of those patients with elevated A1 c levels declined by half over a 1-year period.


The diabetes intervention program is still in the initial phases of integration; however, it continues to be rewarding to the clinicians involved. Frequent anecdotal comments are made about the positive and favorable outcomes among the patients who have enrolled. Additionally, several clinicians have noted the reward of being granted a high-level of autonomy in their practice. A repeat analysis of the original study group (n = 495) was planned for June 2004. Additionally, the planning for a quarterly network support meeting is underway. The diabetes tracking graph is under continual process improvement review to make the data user-friendly to the clinicians.

Wendy R. Behnke, MSA MSN, RN

Wendy R. Behnke, MSA, MSN, RA, is a Health Care Administrator/Regional Nurse Manager, University of Michigan Hospitals and Health Centers, Ann Arbor, MI.

Copyright American Academy of Ambulatory Care Nursing Jul/Aug 2004

Provided by ProQuest Information and Learning Company. All rights Reserved