Personalized Decision Support for Older Patients With Diabetes

In 2003, the first geriatric diabetes care guidelines were published that encouraged older patients and their providers to consider less intensive glucose control goals (HbA1C \<8%) among frail, older patients with limited life expectancy, while continuing to pursue intensive glucose control (HbA1C \<7%) among relatively healthy older patients. The guidelines also emphasized the importance of cardiovascular prevention, encouraged routine screening for geriatric syndromes that can influence treatment decisions (i.e., polypharmacy and falls), and advised providers to acknowledge patients' preferences when making treatment decisions. These guidelines represent a conceptual advance in the care of older diabetes patients; however, there has been little effort to implement and evaluate these recommendations in a practice setting. This may be partially due to the fact that many of the recommendations are difficult to carry out in busy clinical practices without sophisticated decision support tools. Determining whether an older patient will benefit from intensive glucose control is a complex cognitive task requiring simultaneous consideration of multiple, sometimes contradictory, clinical criteria (e.g. advanced duration of diabetes and limited life expectancy). Completing this task accurately may only be possible with computer simulation models. Along with this barrier to implementing care guidelines, there is also no consensus on how to elicit patient preferences in the setting of chronic disease management or how to account for these views in the decision-making process. To overcome these challenges, we developed a web-based Geriatric Diabetes Decision Aid (GDDA) which combines a decision analytic model of diabetes complications with the latest prognostic tools from geriatrics. This personalized decision support tool will encourage the individualization of diabetes care among older patients by educating patients on diabetes, delivering prognostic information to providers, providing personalized data on the risks and benefits of diabetes care to patients and providers, and eliciting the treatment preferences of patients. In this proposed set of studies, we developed the GDDA with the input of patients and providers and assessed its impact through individual interviews. The findings from this series of studies will be important for establishing the feasibility of using the GDDA in practice, and providing estimates of the intervention's effect on processes of care for power calculations for a future large scale randomized controlled trial. This pilot randomized controlled trial will be one of the first trials to formally examine new care recommendations for the growing population of older patients living with diabetes.