The majority of health care utilization decisions in the US are made by the ~27% of Americans with multiple chronic conditions, comprising 66% of all healthcare spending. Overall, it is increasingly recognized that existing public reports of healthcare provider quality do not support decision making for these persons. In particular, reports for persons with diabetes may be misleading as the large majority of persons with diabetes have at least one other chronic condition and most have several other conditions that may influence decisions about treatment priorities or alter the goals of care. Our primary goal is to examine and ultimately support improvement in the use of public reports of provider quality by persons with diabetes and multiple chronic conditions. Specifically, our aims include: (1) define and determine the prevalence of distinct clusters of persons with diabetes who have similar patterns of chronic conditions, (2) examine whether diabetes quality metrics meet minimum sample size requirements after stratifying by condition cluster, and explore resulting variation in provider performance, (3) explore the extent of care fragmentation for persons with diabetes, focusing on the extent to which care is dispersed among primary and specialty providers and clinics after stratifying by condition cluster, and (4) obtain consumer feedback on the usefulness of stratifying diabetes quality metrics and visits to specialists by condition cluster. We partner with the Wisconsin Collaborative for Healthcare Quality, a consortium of provider groups that is a national leader in the public reporting on ambulatory care.
For Aims 1 -3, three multi-specialty provider groups participating in the Collaborative will provide data from their electronic health records for a total of ~52,000 persons with diabetes. In an innovative approach, we will use latent class analysis to identify distinct clusters of persons with similar patterns of chronic conditions, then use these condition clusters to create subgroups of individuals with similar chronic conditions, and examine (within these subgroups) performance on diabetes quality metrics and the extent to which care is dispersed among primary and specialty care providers and clinics. We will also develop interactive web pages to display diabetes quality metrics and visits to specialists for each individual user according to their cluster of conditions, building o the Collaborative's consumer web site. Finally, we conduct cognitive interviews with ~20 persons with diabetes or persons who care for someone with diabetes to obtain consumer feedback on the usefulness and added value of this information. Major advantages of the proposed study include a direct tie between our research and a nationally-recognized dissemination partner with whom we have previous successful partnership activities, the acquisition of data from a broad range of diverse primary care clinics across the state of Wisconsin, use of a set of well-established definitions for diabetes quality, and use of both quantitative and qualitative research approaches. The proposed research will provide information essential to understanding how to improve public reports on provider quality for persons with multiple chronic conditions.
The majority of health care utilization decisions in the US are made by persons with multiple chronic conditions. It is increasingly recognized that existing public reports of healthcare provider quality do not support decision making for these persons. Our long-term goal is to examine and ultimately support improvement in the use of public reports of provider quality by persons with diabetes and other chronic conditions. To achieve this goal, we examine whether reports on a provider's quality of diabetes care can be personalized by presenting information on groups of individuals with similar chronic conditions and whether persons with diabetes perceive added value in this information.
