Only one in five adults with newly-diagnosed diabetes in Ontario attended a diabetes self-management education program within 6 months of diagnosis. The remaining patients therefore either received no formal professional self-management education about their new condition, or they received education during other health care encounters such as visits with their family physician, even though the education that can be provided in this context is often inadequate, in part because primary care physicians often misunderstand patients’ informational needs at diagnosis . This low utilization of diabetes self-management education programs occurred despite the fact that these services are available without direct patient charges and often without requiring physician referral, and despite the fact that these services are broadly distributed throughout the province, reducing geographic barriers to access.
Younger patients were more likely to attend a diabetes self-management education program than older patients. This may be due in part to a higher likelihood of type 1 diabetes among young adults with newly diagnosed diabetes. However, the gradient of decreasing utilization with increasing age persisted across all age strata, such that those aged over 80 years were nearly half as likely to attend a self-management education program as young adults. Like advancing age, people with lower socioeconomic status, those who were recent immigrants, and those with mental health conditions or medical comorbidity were less likely to attend a diabetes self-management education program, despite arguably being more in need of self-management support. The observed results are consistent with previous studies, where utilization was found to be lower amongst older adults and those with lower levels of education [22, 23, 27]. A more recent survey of 295 patients with diabetes in Pennsylvania found that only 45% reported having ever received self-management education, and patients with complications or with poor control were no more likely to have attended . A Canadian survey of 781 patients found that demographic factors did not predict utilization of diabetes education . Surveys of health care providers in the United States and Canada have suggested that barriers to utilization include patient unwillingness, programs’ location, languages of service, operating hours, and (in the United States) insurance coverage [30–32]. The disparities in utilization found in this study may also partially explain the observation that patients utilizing education services had lower average costs than non-attendees , since older, poorer and sicker patients who would be predicted to have higher health care costs were also less likely to use such services in the first place. Nonetheless, the presence of these disparities in diabetes self-management education program utilization is noteworthy in a publicly-funded health care system where patients are meant to have equitable access to these services. These populations may face other barriers to access and utilization, such as lower health literacy, less ability to navigate the health care system, or occupational or financial barriers to attending programs.
Patients living in rural areas were markedly more likely to attend diabetes self-management education programs, perhaps because other providers of diabetes care, such as family physicians or endocrinologists, are more difficult to access in rural areas. We had hypothesized that patients who had no family physician visits prior to diabetes diagnosis might have had greater utilization of diabetes self-management education programs as an alternative source of primary diabetes care. Instead, these patients were less likely to have program utilization. Interestingly, patients with many family physician visits prior to diagnosis were also less likely to attend, perhaps because such frequent visits are a marker for other comorbidity or high demand health needs that distract from optimal chronic disease management. We had also hypothesized that patients whose diabetes was diagnosed during a hospitalization might be less likely to attend a diabetes self-management education program, reflecting inadequate attention to chronic disease follow-up upon discharge from hospital. However, attendance did not differ for these patients.
One possible reason for low utilization within the first 6 months after diagnosis could be that wait times for receiving diabetes self-management education services exceeded 6 months. If this were the case, we would expect that examining longer periods of follow-up would lead to ongoing increases in the proportion of patients who attended diabetes education. However, when we reanalyzed the data by lengthening follow-up by one-third to 8 months, the proportion of patients attending a diabetes self-management education program increased by only 1.4%. Hence, it is unlikely that wait times were a significant contributor to low utilization.
This study is unique in a number of ways. It is the first study to examine demographic and clinical predictors of diabetes self-management education program utilization in Canada’s universal health care system, where socioeconomic status and insurance are not the overwhelming barriers to health care utilization that they are in other health care systems. The study shows that even in this universal access context, disparities by age, socioeconomic status and immigration history occur. Virtually all previous studies of diabetes self-management education utilization have relied on small cross-sectional surveys of patients with diabetes; our study examined real-world utilization of self-management education services at a population level in a longitudinal cohort, making it more methodologically rigorous. However, there are some limitations to note. First, although our cohort of patients was linked to large health care databases permitting measurement of many important demographic and clinical predictor variables, other potentially important factors that might be associated with disparities in diabetes self-management education program attendance could not be measured from population-level data, such as education level, employment, health literacy or language proficiency. Second, as noted above, the Ontario Diabetes Database does not distinguish between types of diabetes, so we could not specifically compare predictors of program attendance between type 1 and type 2 diabetes. Third, there are undoubtedly many patients who were referred to self-management education programs who did not ultimately attend. Since our study only captured attendance, these patients would not have been identified. However, our study focused on the outcome measure of program attendance, not on the process measure of receiving a referral to one from a health care provider. Fourth, the influence of variability in the curricula between diabetes education programs could not be evaluated in our study, as these data were not available. However, a Standards Recognition Program for diabetes education has been developed by the Canadian Diabetes Association, so the extent of variability may not be significant. In addition, each program receives at least partial funding from the provincial Ministry of Health, and therefore must adhere to government-mandated standards on curriculum, staffing and access. Finally, our study examined attendance only at formal diabetes self-management education programs. Although these programs are the main sources for self-management education, other sources of self-management education and support (such as an individual nurse working with a primary care physician, or self-management education provided by a pharmacist in a dispensary) could not be captured by this study.
Further research is required to determine what patient, provider and health care system factors contribute to the observed disparity in utilization by vulnerable patients. In addition, interventions to improve utilization of self-management education programs by patients newly-diagnosed with diabetes need to be developed and evaluated.