Diabetes Self-Management in a Community Health Center: Improving Health Behaviors and Clinical Outcomes for Underserved Patients
Self-management is a crucial element of good diabetes care. Several large-scale trials have demonstrated that comprehensive interventions that include self-management can prevent complications from type 11,2 and type 2 diabetes.3 In addition, interventions that promote the adoption of healthy behaviors have been shown to significantly prevent or delay the onset of type 2 diabetes in patients at increased risk for this disease.4 A review and meta-analysis of self-management interventions for diabetes concluded that, although education alone does not lead to improved outcomes, self-management interventions can improve glycemic control.5
However, outside of the research arena, “real-world” settings face challenges when seeking to replicate self-management programs such as those found in clinical trials. Such interventions are resource intensive and not generally designed to meet the needs of patients from underserved populations. Issues such as low literacy, limited English proficiency, poverty, and cultural differences present additional barriers to promoting diabetes self-management.
To combat the growing diabetes epidemic, it is crucially important that such barriers be overcome. Diabetes self-management interventions must be developed and tested to meet the needs of all patients, particularly underserved minority populations. Hispanic and African-American patients in the United States have nearly two times the prevalence of type 2 diabetes as non-Hispanic whites.6-8 Rates for Native Americans are even higher.8 In addition to higher prevalence, ethnic and racial minority patients with diabetes have higher mortality and higher rates of diabetic complications.9 Although the pathophysiology and treatment may be the same for different ethnic and racial groups, differences in behaviors, cultures, and health beliefs have a significant impact on how patients understand their illness and engage in self-management. Programs that account for these differences can improve diabetes outcomes.10-13
An additional barrier to effective diabetes self-management is the high prevalence of comorbid depression, particularly in ethnic and racial minority populations. Diabetic patients have twice the incidence of depression as the general population.14 Patients with both diabetes and depression have worse clinical outcomes,15-17 poor adherence to treatment and self-care guidelines,18 and higher health care utilization.19
The Community Health Center, Inc., (CHC) is the largest federally qualified health center in Connecticut, with primary care centers in cities and towns throughout the state. It developed a comprehensive program to provide self-management education to its underserved, largely Hispanic population of patients with diabetes. With funding and support from the Robert Wood Johnson Diabetes Initiative, a team of nurses, doctors, diabetes educators, and public health specialists engaged in a 4-year quality improvement effort to test and implement a program that would meet the specific needs of its population with regard to language, culture, and literacy level and to address the extremely high prevalence of comorbid depression in patients with diabetes. Self-management was fully integrated into the primary care team, and simple tools were developed to allow all care team members to measure and track patients' self-management goals and their progress toward achieving those goals.
I. Developing a low-literacy, culturally appropriate self-management program
CHC cared for 50,000 patients in 2006, 2,389 of whom had diabetes. These patients are 43% Hispanic/Latino (largely of Puerto Rican descent) and 13% African American. Nearly half speak a language other than English at home. Eighty-eight percent are at or below 200% of the poverty level, and 25% have no medical insurance. The self-management program was developed specifically to meet the needs of this population, particularly with regard to literacy level, language, and culture. Materials were produced in English and Spanish and geared to a fourth-grade reading level, and an initial curriculum was created to be used in a group format. Self-management is often taught in groups, with evidence supporting the benefit of this approach for a variety of chronic illnesses, including diabetes.20-23
Self-management differs from traditional diabetes education in that it focuses less on educational topics and more on actions and behaviors. Therefore, in addition to providing general education on diabetes, each session focused on identifying specific desired behaviors and having patients choose a specific goal that they would adopt. Patients were encouraged to set highly specific goals with a start date and a measurable quantity/outcome whenever possible. In addition, patients were asked to rate on a scale from 1 to 10 their perception of the likelihood of achieving the goal (self-efficacy). Patients with a likelihood of < 7 were encouraged to adjust the goal or adopt a new goal.
The program was modified and improved substantially through a series of PDSA (plan, do, study, act) cycles.
Although the idea of patient-generated goals was appealing and represents an important aspect of the self-management methodology, many patients struggled to come up with individual goals, even after a didactic session. As a result, pamphlets and posters were developed in English and Spanish with common goals that other patients had set (Figures 1 and 2). All patients received these pamphlets at the outset of the program, in an effort to generate ideas and build in their minds the concept of behavior change, action, and self-efficacy. Posters with similar goals were also placed in strategic locations throughout the health center. This encouraged patients to begin thinking about behaviors, goals, and actions they could take to improve their diabetes and were found to be very helpful in this regard.
Repeated emphasis and encouragement, preferably from multiple members of the care team, were needed to promote goal attainment and maintenance over time. Maintaining a log of self-management goals prominently located in each patient's chart allowed each goal to be recorded, tracked, and kept available for various members of the care team to see and review with patients (Figure 3). Repeatedly reviewing goals and reminding patients about them helped maintain gains over time.
Group sessions were informative and fun for many patients, but CHC staff had difficulty recruiting and retaining sufficient numbers of patients to merit the outlay of staff time. In addition, with > 2,000 diabetic patients in need of self-management, the group visit approach simply was not scalable to reach such large numbers of patients. Furthermore, many patients choose not to attend such groups and find the format undesirable. Ultimately, CHC found that more patients could be reached and better outcomes achieved by offering individual sessions with nurses and diabetes educators in addition to group sessions. These sessions used the same materials as those used in the group sessions but were more individualized and gave patients greater flexibility. Ultimately, far more patients chose individual sessions and a much larger proportion of the diabetic population was reached and engaged in self-management.
II. Coping with high rates of depression
Initial experience with self-management suggested that comorbid depression was a substantial barrier to patient self-management. Early in the development of the program, staff adopted the Patient Health Questionnaire-9 (PHQ-9) screening tool24 and administered it at intake for all patients participating in the diabetes self-management program. Those who were found to have depression and were not already in treatment were referred for mental health treatment either by onsite mental health staff, primary care providers, or outside community resources. Whenever possible, diabetes and depression treatment was coordinated and integrated. Details of this integrated treatment model for diabetes and depression have been published elsewhere.25 Although the data generated from this project are still being evaluated, preliminary results suggest that patients with depression and diabetes treated in an integrated manner had similar clinical improvements, as measured by hemoglobin A1c (A1C), as those without depression.26
III. Training staff in self-management
Training staff in the methodology of self-management proved more challenging than expected. Medical staff, including nurses, nutritionists, and diabetes educators, tended to lapse into a “didactic mode” and assume a more proscriptive mien unless they received frequent education, support, and review. The project coordinator received “master trainer” instruction in self-management and subsequently became the trainer for CHC employees participating in the project. Nurses from all of CHC's primary care sites then received a half-day session on self-management goal setting. In three sites, nurses received additional follow-up training, which included a review of goals facilitated with patients.
IV. Measuring goal attainment
In addition to tracking specific goals for each patient, project staff noted a need to better record patients' progress and success with attaining specific goals. Collecting information on goal attainment would allow for a more detailed analysis of the program. CHC developed a “self-management attainment score,” which was assigned by the self-management facilitator at follow-up encounters. Staff found that during follow-up, patients who set goals generally clustered into four main groups: those reporting complete success, partial success, limited success, and lack of success at achieving a specific goal. As such, a four-point scale was used to measure goal attainment, with four representing a fully attained goal and one representing failure to attain the goal. As a result, CHC is engaged in a detailed analysis of goal setting, looking at types of goals, predictors of success and failure, and associations with successful goal setting. In addition, the relation between attainment scores and clinical outcomes is being studied in detail.
Participants. In all, 488 patients participated in the diabetes self-management program. Two hundred seventy-seven (63%) listed Hispanic as their ethnicity, and 140 (32%) were monolingual Spanish speakers.
Goal attainment. During the program, patients set a total of 2,139 individual self-management goals. Of those, 1,612 had at least one follow-up assessment and an attainment score assigned. Of the goals that were assessed in follow-up, 836 (52%) achieved an attainment score of 3 or 4 (usually or always) at the first follow-up, and during the entire follow-up, 1,161 (72%) ultimately were attained with a level 3 or 4.
Clinical outcomes. For patients for whom pre- and post-intervention clinical data were available, the mean A1C dropped significantly by 0.9 of a percentage point per year (P < 0.0001). In all, 116 of 263 patients achieved or maintained an A1C ≤ 7.0%. Participant's average LDL cholesterol dropped by 23.3 mg/dl. After participation in the program, 206 (42.3%) achieved a blood pressure < 130/80 mmHg compared with 121 (28.8%) at the outset of the program.
Depression. A total of 414 patients had data collected regarding depression treatment. Patients without current diagnosis and treatment for depression were administered the PHQ-9 depression screen. Two hundred sixty-six (62%) either had an established diagnosis of depression or had a score of ≥ 10 (moderate depression) on the PHQ-9.
Although more detailed analysis of the impact of this program is ongoing, several broad conclusions can be drawn. First, a large cohort of mainly Hispanic patients, a third of whom were monolingual Spanish speakers, were able to participate in a comprehensive self-management program, set and attain specific goals, and improve their glycemic control, blood pressure, and LDL cholesterol. Although not subjected to the rigors of a randomized, controlled trial, these results are sufficiently positive to justify continuing the program, spreading it to additional sites, and initiating a new research project in diabetes disease management that will use a randomized, controlled methodology.
Second, depression was even more common among diabetic patients than expected. Using a simple screening tool, the PHQ-9, was effective for identifying cases. Co-managing depression with diabetes resulted in clinical improvements in the depressed group that equaled those in the nondepressed group, suggesting that such patients can effectively engage in self-management and improve their diabetes control.
Third, recording self-management goals in detail and scoring for attainment of goals provides a valuable tool for collaborative care among different care providers and for more detailed program analysis.
Fourth, experience has shown that training staff in proper self-management education techniques takes time and ongoing monitoring. There is a strong tendency for medical staff to revert back to more didactic modes of education. A method of oversight and frequent review is necessary to maintain fidelity with the self-management model.
The diabetes team was able to use quality improvement methodology to develop and adapt this program to meet the needs of the population being served and in so doing to create a flexible, highly effective intervention. The improvements seen in health behaviors and clinical indexes will serve to reduce the burden of morbidity and mortality and reduce health disparities in a vulnerable group of patients.
Preparation of this article was supported by the Robert Wood Johnson Foundation through its support at Washington University of the National Program Office for its Diabetes Initiative.
Daren Anderson, MD, is the chief medical officer, and Joan Christison-Lagay, MAT-MPH, is the diabetes project director at Community Health Center, Inc., in Middletown, Conn.
- American Diabetes Association