In Brief

The Health Disparities Collaboratives initiative of the Department of Health and Human Services Bureau of Primary Health Care (described in the article on p. 102) has allowed hundreds of health centers around the country to improve process and clinical outcomes for people with diabetes. This article reports on the results of these efforts from facilities participating in the collaboratives.

Part 1 of this report on the diabetes efforts within the Health Disparities Collaboratives (HDCs) (p. 102) described how the Breakthrough Series model,1 the Chronic Care Model,2,3 and the Model for Improvement4 were adapted in Bureau of Primary Health Care (BPHC) facilities to improve process and clinical outcomes in people with diabetes. The results of these efforts, from facilities participating in the collaboratives, are reported herein.

As with any initiative, the proof is in the proverbial pudding, and the diabetes HDCs show impressive results. Each set of teams within each HDC has generated a plethora of data for the required and recommended measures. For this discussion, we will include only those teams that have just completed Phase 1 or are currently in Phase 2, as described in Part 1 of this report.

One of the major paradigm shifts represented by the planned care model is the focus on managing an entire panel of patients, rather than individual patients. This shift requires tracking outcomes for the patient population and analyzing the needs of the population as a whole. The population is constantly changing, however, and the registry size fluctuates as new patients are added to the registry and others are removed due to death or relocation. Thus, the following process and outcome results are annotated with registry size statistics.

  • The proportion of patients having two hemoglobin A1c (A1C) tests within 12 months increased from 25% in March 1999 to 38% in July 2003, while the size of the registry increased by more than 24,000 patients. This was a relative improvement of 152%.

  • During HDC 2000-Diabetes (HDC 2000-DM), the proportion of patients with two A1C tests within 12 months showed even greater improvement, with an increase from 18% in April 2000 to 38% in July 2003, while the size of the registry increased by almost 34,000 patients. This was a relative improvement of 211%.

  • During the same HDC, the proportion of patients with documented self-management goal setting increased from 10% in February 2000 to 38% in July 2003, while the size of the registry increased by more than 63,000 patients. HDC 1999-Diabetes (HDC-1999-DM) participants also achieved a rate of 38% of patients having self-management goal setting, with 25,690 patients in the registry as of July 2003. The HDC 2001-Diabetes (HDC 2001-DM) showed an increase in documented self-management goal setting to almost 40%, while the registry grew from about 1,700 to more than 15,000 patients.

  • During HDC 2001-DM, the proportion of patients with two A1C tests within 12 months increased from 14.1 to 31%, a relative improvement of 220%, while the size of the registry increased by more than 13,500 patients. The average A1C result decreased from 8.6 to 8.1% during this time frame.

  • Results from HDC 2002-Diabetes (HDC 2002-DM) include more data, since additional measures were included to assess cardiac risk reduction:

       ✓ About 51% of more than 8,000 patients have documented self-management goal setting.

       ✓ More than 37% of patients have had two A1C tests within 12 months.

       ✓ The average A1C is 7.98%.

       ✓ About 70% of patients over 55 years of age are taking angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers.

       ✓ About 42% of patients over 40 years of age are taking statins.

Individual health centers vary in the strength of their efforts and, consequently, in their results and success. What is clear from the HDCs’ results is that health centers of varying sizes, serving different patient populations in different regions of the country, have been able to achieve extraordinary results in improving the care of patients with diabetes.

The following two case studies illustrate how individual health centers used HDC participation to facilitate remarkable change in health outcomes and in the delivery of health services.

The Fair Haven Community Health Center (FHCHC) has served a New Haven, Conn., inner-city neighborhood for more than 30 years. The center provides the only primary care in the area for approximately 11,000 patients through its main clinic, three school-based health centers, and a satellite clinic in an elderly housing complex. The patient base reflects the community, with minorities comprising 84% of patients, 65% percent of whom are Latino, and 19% of whom are African American.

Like most community health centers, FHCHC serves patients at high risk for diabetes and poor diabetes outcomes. About 35% of the health center’s patients with diabetes have no medical insurance and thus no means to afford the $800–1,000 monthly costs for pharmacy products, blood tests, and glucose monitoring supplies. Forty percent of patients speak little or no English, and a growing percentage are undocumented immigrants. The neighborhood has few safe places to walk or exercise and lacks markets with affordable, fresh, low-calorie foods.

Diabetes was recognized as an important and growing problem at the health center several years ago. A few providers formed a team to promote diabetes education, self-management efforts, and community awareness. The team supported improvement in the management of the diabetes population through provider education and improved standards of diabetes care. Diabetes flow sheets were used in patients’ charts; formal treatment policies and recommendations aligned with the American Diabetes Association (ADA) standards of care were adopted. The team assessed overall care delivery by periodic chart reviews and anecdote. However, they recognized the need to more systematically identify patient groups in need of care and knew that their improvement efforts suffered from an inability to measure performance and progress in meeting treatment goals for patients.

In 2000, FHCHC first participated in the intensive work of improving care for diabetes with the HDC and was provided with a patient registry/database (described in Part 1 of this report). This was the very tool they had wished for to fuel their progress. The care model (also described in Part 1) gave the team a formal framework to describe and expand upon work they were already doing; the improvement model (also described in Part 1) helped them to focus on small rapid change cycles to improve care.

With the patient registry as a tool, FHCHC has been able to work on chronic care delivery with a real time database. The team developed systems using the registry to provide up-to-date patient information and patient trend information, making patient visits more efficient and helping providers better plan care. They have used the registry to extend the delivery system to include other members of the care team, thus allocating responsibility for the multiple aspects of diabetes care. For example, the registry-generated encounter note automatically prompts medical assistants to perform an in-office A1C test if one is due and prompts nurses to provide influenza and pneumococcal vaccines if they are needed.

Furthermore, the health center not only has up-to-date information on and reminders about individual patients, but also can use this information to better manage diabetes care for groups of patients or for their entire population. So, for example, when the rare opportunity of a local smoking cessation program in Spanish became available, the team rapidly generated a report of patients with diabetes who smoked and spoke Spanish. They contacted patients individually and encouraged them to participate at the “invitation” of their provider. As another example, each November, staff members run reports to identify patients who have not yet received influenza vaccinations. These patients are then contacted through letters or phone call reminders and scheduled for appointments.

The changes made by FHCHC led to improvements in process and outcome measures. Figure 1 shows some of the results FHCHC achieved during Phase 1 of the HDC. Several of the graphs include a dashed line representing the goal for that measure.

As the FHCHC team entered Phase 2 of the HDC, it faced the challenge of spreading change to new providers, patients, and new conditions while sustaining changes and continuing improvement. The team used the registry and the information and skills learned during Phase 1 to better implement changes for more providers and patients during Phase 2. For example, in Phase 1, the team included cardiovascular risk reduction measures and lipid management among its goals. It set a demanding target: “70% of patients with diabetes will achieve an LDL cholesterol level < 100 mg/dl,” which was more ambitious than the HDC proposed goal (LDL cholesterol < 130 mg/dl). Using the patient registry, the team members discovered that the lipid-screening rate was only 70% and that just 30% of patients screened had LDL cholesterol levels < 100 mg/dl.

The team then ran reports identifying patients who had not been screened and those whose LDL cholesterol level was > 100 mg/dl. Using these reports, they performed a targeted chart review to identify reasons for poor lipid management. They had predicted that lack of health insurance would be the major factor in poor management, but it found that provider omission and data inaccuracy were equally significant. Armed with this information, the team focused its efforts on these areas, working to improve access of uninsured patients to testing and medications, update providers on current standards of care for lipid control, and improve data accuracy. In addition, the team used the reports to create a treatment and follow-up plan for individual patients and to provide feedback to providers about their patients.

Table 1 lists examples of specific changes the health center made in the different components of the care model during Phase 2.

Using the patient registry, FHCHC reexamines data continuously and closely follows clinical improvements. Figure 2 shows the rates of lipid screening and achievement of LDL cholesterol goals among 600 Fair Haven registry patients with diabetes. Despite increasing enrollment in the registry, the percentage of patients screened for hyperlipidemia (currently about 80%) and achieving the LDL cholesterol goal of < 100 mg/dl (currently > 50%) continue to increase. The ongoing improvement demonstrates the effectiveness of quality improvement efforts to meet the needs of people with diabetes.

In addition to improving process and outcome measures, there is emerging evidence that involvement in an HDC saves money. As shown in Table 2, benchmark data from the South Carolina State Budget and Control Board have documented lower rates of hospitalization, shorter lengths of stay per admission, and significant cost savings for health centers trained in planned care through an HDC, compared with those uninitiated to planned care and other providers.

Although most of the health centers participating in the HDCs focused on diabetes, the crucial elements for better care are similar for all chronic conditions and preventive services. Health centers throughout the country have used the care model to improve treatment of other chronic illnesses or co-morbidities, such as cardiovascular disease, depression, and asthma, and are currently piloting use of the care model for prevention activities. Through partnerships with the Centers for Disease Control and Prevention, the Institute for Healthcare Improvement, the National Institute of Diabetes and Digestive and Kidney Diseases, and the Medstar Clinical Research Center, similar methods are being applied to other conditions in other communities, translating the results of the Diabetes Prevention Program5 into widespread clinical practice.

The process of changing to a planned care approach can be both time-consuming and difficult, but the results are rewarding. By integrating sound management and business practices with the components of the care model described in Part 1 of our report, organizations can demonstrate a business case for health care improvement. Health care leaders and providers need to seize the opportunity to align their work in the organization and thus reduce cost, improve productivity, and enhance revenue, while improving clinical quality and patient satisfaction.

Figure 1.

FHCHC population of focus, year 1.

Figure 1.

FHCHC population of focus, year 1.

Close modal
Figure 2.

Rates of lipid screening and achieving LDL goal of < 100 mg/dl among Fair Haven Community Health Center patients. POF, population of focus (patients included in Phase 1 and Phase 2 efforts); spread, patients included in Phase 2 efforts only.

Figure 2.

Rates of lipid screening and achieving LDL goal of < 100 mg/dl among Fair Haven Community Health Center patients. POF, population of focus (patients included in Phase 1 and Phase 2 efforts); spread, patients included in Phase 2 efforts only.

Close modal
Table 1.

Changes Made by Fair Haven Community Health Center

Changes Made by Fair Haven Community Health Center
Changes Made by Fair Haven Community Health Center
Table 2.

CareSouth Carolina Data Profile*

CareSouth Carolina Data Profile*
CareSouth Carolina Data Profile*

Cindy Hupke, RN, BS, MBA, lives in DeKalb, Ill., and is the IHI’s national director for the BPHC HDCs. Anne W. Camp, MD, is an internist and endocrinologist at the Fair Haven Community Center and a clinical instructor of medicine at the Yale School of Medicine in New Haven, Conn. Roger Chaufournier, MHSA, is president and chief executive officer of Patient Infosystems in Rochester, NY, and serves on the faculty of Johns Hopkins University School of Public Health in Bethesda, Md. Gerald J. Langley, MS, is a statistician with Associates in Process Improvement in Cameron Park, Calif., and senior improvement advisor for the BPHC HDCs. Kevin Little, PhD, is a statistician based in Madison, Wisc., and serves as a BPHC collaboratives technical advisor.

Note of disclosure: The authors are all paid consultants for the HDCs described in this article. Mr. Chaufournier is an employee and board member of Patient Infosystems, which provides disease management services for the health care industry, including the federal government. He is also a stock shareholder in Patient Infosystems, Pharmacia, and Solutia, all of which make products or provide services related to the treatment of diabetes.

It is through the vision and leadership of David M. Stevens, MD, FAAF, that the HDCs were crafted. Dr. Stevens and Tricia Trinite, NP, MSPH, have provided ongoing national leadership and guidance from the BPHC for the HDCs and have been integral in the development of strong partnerships with the Institute for Healthcare Improvement (www.IHI.org), Improving Chronic Illness Care (www.improvingchronic-care.org), and the Centers for Disease Control and Prevention Division of Diabetes Translation (www.cdc.gov/diabetes).

1.
Baker GR: Collaborating for improvement: the Institute for Healthcare Improvement’s Breakthrough Series.
New Med
1
:
5
–8,
1997
2.
Bodenheimer T, Wagner EH, Grumbach K: Improving primary care for patients with chronic illness.
JAMA
288
:
1775
–1779,
2002
3.
Bodenheimer T, Wagner EH, Grumbach K: Improving primary care for patients with chronic illness: the chronic care model, Part 2.
JAMA
288
:
1909
–1914,
2002
4.
Langley J, Nolan K, Nolan T, Norman C, Provost L: The Improvement Guide: A Practical Approach to Enhancing Organizational Performance. San Francisco, Jossey-Bass, 1996
5.
The DPP Research Group: Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin.
N Engl J Med
346
:
393
–403,
2002