Abstract:To assess acceptability and feasibility of incorporating individualized risk prediction into clinical assessment, decision making, and communication of risk of type 2 diabetes and prevention recommendations. We integrated a prediction model into the clinical workflow at a US health care organization. We conducted patient and provider focus groups and pre- and post-dissemination surveys and assessed the effect on referrals to and enrollment in the Diabetes Prevention Program, among 2,775 patients with prediabetes who had primary care visits between May 2018 and December 2020. Among patients with prediabetes seen in primary care during the study period, 79% had a calculation with the risk prediction model completed. After implementation of the risk prediction model, prevention intervention rates increased, with 62.3% of high-risk patients receiving an intervention within 1 year. Used at the point of care during a shared decision-making discussion between the patient and provider, the diabetes risk calculator helped providers prioritize patients for diabetes prevention interventions, facilitated communication, and successfully improved rates of engagement in their care among patients with prediabetes.

Role of personalized predictions for people with prediabetes: disseminating patient-centered estimates of benefit