Assessing the Impact of a Dynamic Chronic Care Registry on Quality of Care

This project utilized a diabetes registry in order to improve care delivery and the quality of care for diabetics at eight federally-qualified community health centers. The team completed a prospective randomized controlled trial of three interventions with nearly 5,500 adults with diabetes. These three interventions included: patient report cards both mailed to the patient quarterly and given to them at their primary care visits; quarterly distribution of standard and enhanced provider performance feedback reports; and point-of-care reminders via an interface to the registry data. Clinical and process outcomes were assessed for each of the three interventions both quantitatively and qualitatively. The main objectives of this project were to:

• Distribute an individualized patient report card to the patient's home on a quarterly basis and the patient at the time of each of their primary care visits.
• Develop and distribute to providers both standard report cards at all eight clinics and enhanced report cards at four of the eight clinics on a quarterly basis.
• Develop a point-of-care interface to the diabetes registry data and make it available to the clinical team at the point of care.
• Evaluate the effects of this three-pronged approach on process, quality, cost, and satisfaction, using both quantitative and qualitative measures of assessment.
• Generalize the application of an integrated diabetes registry to other chronic diseases and to other health care systems.

Despite finding that many providers and the majority of patients thought that the report card was an effective tool, it did not improve outcomes. Distribution of patient report cards to patients at the time of their visits also did not improve intermediate outcomes. However, patients at clinics with patient-level provider performance feedback were found to have better glycemic control (p<0.01). Enhanced provider report cards, which included an additional list of up to 10 patients who met certain preset criteria, cost an estimated $46,000 to create, equivalent to $901 per additional patient controlled for hemoglobin A1c. Providers gave mixed feedback to the reports, reporting frustration over the lack of additional support to follow-up on the findings of the reports. While the point-of-care reminder tool was perceived to be an efficient means for collecting diabetes-related patient information in one place, it was rarely accessed, and access to the report itself was found to be problematic.

A sample of the themes and lessons learned during the project includes: the critical need to obtain provider buy-in; the need for qualitative analysis to fully understand the contextual factors and influences outside of the interventions; and the need to study impacts of interventions in order to determine the best way to allocate resources. The team also noted that the overall assessment of the care of diabetics in the absence of a single variable is difficult. A new variable indicating a combination of levels of glycemic, hypertensive, and cholesterol control would improve this assessment.