On several measures, the practice surged past national targets for diabetes care and exceeded the median performance of 80 similar practices. Key to the effort at the family medicine residency program of the Washington Hospital in Washington, Penn., were the following elements of active diabetes management:

A chart study revealed that most of the group's roughly 1,000 patients with diabetes were coming into the family medicine office with some regularity but for indications other than diabetes, such as urinary tract infections, colds, and so on. They then skipped routine appointments for their diabetes care, perhaps thinking, “‘Well, I was in the office. I assumed if I had any problem with my diabetes, you would have said something,’” surmised Dr. Minteer, associate director of the family medicine program.

In addition, there was clearly a problem with continuity of care. The chances of seeing one's own physician in a nonroutine visit was only about 60% in the “organized chaos” of the practice, which includes 24 family medicine residents, 14 faculty members, more than 40 nonphysician staff members, and which deals with 58,000 patient visits a year.

To make sure that patients did not slip through the cracks, the program began color-coding all the charts, assigning a bright yellow file folder jacket to any patient with diabetes. “It's a very low-tech thing to do, but it made a lot of difference,” he said. “You know when you're walking down the hall and you see a yellow chart, you should start thinking about diabetes.”

A template for diabetes management was soon integrated prominently into the electronic medical records that are accessed by each staff member who scheduled visits or cared for patients. When a patient called for an appointment of any kind, it was evident if he or she was lagging behind schedule on recommended lab tests, specialty examinations, or any routine element of diabetes care.

Standing orders sent such patients to the laboratory prior to the scheduled appointment, so that up-to-date results would be available to whatever physician the patient was scheduled to see. Any test or examination that was still not current—an ophthalmologic examination, for example—was highlighted on the template in the patient's yellow-jacketed chart.

“This is really powerful, because now, I not only know it's a diabetic, I can look at that sheet, and even if it's not my patient, I can get these things ordered.” The reminders significantly improved physicians' adherence to department guidelines for diabetes care, said Dr. Minteer.

Another important shortcoming identified in the chart review was the lack of follow-up with patients who didn't come in for care of any kind.

A database query identified 469 patients whose most current hemoglobin A1c was greater than 7%, and many of them had not been in for care in more than a year. Of 59 patients with HbA1c measures greater than 10%, 30 were successfully contacted, including 8 who had not been seen in the center for more than a year.

A decision was made in early 2005 to actively pursue these patients, first with a letter and then through follow-up phone calls, to emphasize the need for regular diabetes care. “You have to change the culture,” said Dr. Minteer. “You can't sit by the phone and passively wait for someone to call.” Of the 469 patients, 250 were reached by nurses who had been designated to focus on patients with diabetes.

Appointments were scheduled and barriers overcome. If a patient balked at scheduling an appointment because of a high copay, the nurse or a part-time diabetes educator contacted a social worker and the billing department to arrange for assistance.

The focus on high-risk, elusive patients had the temporary effect of undercutting the practice's on-paper performance measures. The percentage of patients with HbA1c levels less than 7% slipped at first before it improved, Dr. Minteer noted.

“These people hadn't been in the office and were out of control,” he said. By early 2006, however, the percentage of patients meeting the HbA1c goal “was doggone close to the upper confidence level.”

The number of obese patients in the practice appeared to increase, because the information was now being entered into the medical records as never before.

Still, important measures started to improve in a dramatic fashion. For example, the target for LDL cholesterol is now being met in a percentage of patients that exceeds the group's benchmark goal.

Other outcome measures have yet to improve, such as the percentage of patients with diabetes meeting targets for blood pressure. But, “we would expect this not to improve for some time,” said Dr. Minteer.

Essential to the Washington Hospital's success was buy-in from everyone in the family medicine program, Dr. Minteer said. Nine nurses were assigned to become diabetes coordinators, with responsibility for a specific group of diabetes patients in the database. The nurses took seriously their responsibility to contact these patients and improve their care, greeting them when they arrived for appointments and taking pride in the patient's progress.

The hospital's diabetes educator, spent 1 day a week in the primary care office, where she helped keep the program on target and provide feedback about strides that were being made.

In a striking demonstration of the pivotal role of nurses, Dr. Minteer displayed charts that showed sudden dips in positive trends that occurred near the very end of the demonstration project.

The hiccups coincided with budgetary cutbacks that left the practice short of its normal nursing staffing.

“Almost all outcome measures dipped,” he said. “I'm using this as a stick with the CEO. Look, this is what we can do when we have the right kind of staffing, and this is what happens when we don't.”

He also noted that it was interesting that insurance companies, eyeing the program's improving bottom line, have been helpful in assisting the practice to implement changes and track its progress.

As an added bonus, improved performance on process and outcome measures spilled over into better care of nondiabetic patients as well as those with diabetes, Dr. Minteer observed.

“If you do a good job with diabetes, you're probably going to improve your numbers for coronary artery disease, renal disease, hyperlipidemias,” he said. “You're killing four or five birds with one stone.”

Family Practice: Volume 36, Issue 18, Page 1,19 (15 September 2006)

===============================

FACT:
Poor Adherence to Diabetes Meds Linked with High Mortality: Two new studies show that nonadherence to medication is common among recent MI patients and in those with diabetes, resulting in higher mortality. Nonadherence in the MI study resulted in an almost fourfold increase in the death rate in the first year after hospital discharge, while the diabetes sufferers had almost a twofold increase in mortality following noncompliance.
Arch Intern Med 2006; 166: 1842-1847.