Analyzing the Impact of Diabetes Self Mangagement Education (DSME) on Diabetes Management from a Policy Perspective
*Names in bold indicate Presenter
A noticeable gap in the literature concerns the nexus between DSME training and over- or under-utilization of medical services, and the temporal impacts of DSME on diabetes self-management practices in addition to glycemic regulation. This paper seeks to address these questions using data from the 2013 Behavioral Risk Factor Surveillance Survey (BRFSS). BRFSS is a telephone-based behavioral surveillance system under the aegis of the Center for Disease Control (CDC). All data are self-reported. The 2013 BRFSS diabetes module was administered in 29 states, as well as the District of Columbia, Guam, and Puerto Rico. Approximately 35,000 individuals (n=34,395) are included in the analysis.
Both linear probability and logistic models will be employed to evaluate the impact of DSME on the utilization of diabetes-related medical services. Other important outcome measures related to diabetes management, including annual eye exams, periodic glycosylated hemoglobin testing, and regular foot exams will be examined according to DSME status. As assignment of treatment is based on the action of the individual diagnosed with diabetes, under the assumption that all will have been made aware of DSME programs at the time of diagnosis, self-selection bias is an important consideration. This will be addressed through inclusion of characteristics likely to affect utilization of diabetes-related medical services and self-management activities. These traits include education, income-level, health insurance status, age, insulin usage, and race/ethnicity.
The policy implications are two-fold. First, the effect of DSME on measures of diabetes management control and utilization can be estimated and employed as comparative tool in assessing the cost of diabetes treatment in those states that do not require DSME as a mandated health benefit or cover diabetes education under Medicaid versus those states that do. Second, by focusing on the time lapse between initial diagnosis/DSME training and diabetes management and utilization measures as captured in the age and age at diagnosis variables in BRFSS, a better understanding of the temporal aspects of DSME on a range of diabetes self-management efforts can be ascertained. This has important implications for policy related to the frequency and duration of on-going diabetes educational efforts following an initial DSME training.