Nearly one in three adults has prediabetes, a condition that substantially increases the risk of heart attacks and stroke. The increased cardiovascular risk associated with prediabetes can be effectively managed by lifestyle changes or medication therapy, but recent data shows few prediabetes patients are treated effectively. In this project, we will adapt, implement, and evaluate a proven electronic health record-linked, web-based clinical decision support system to identify patients with prediabetes and provide prioritized treatment recommendations to patients and providers in a rural health system. The results of the project will provide a template for implementation of more efficient and effective rural healthcare and have the potential to substantially and improve cardiovascular quality of care and clinical outcomes of millions of rural Americans with prediabetes.

Nearly one in three adults in rural communities has prediabetes, a condition that increases the risk of heart attacks and stroke but can be managed by use of metformin, lifestyle interventions, and control of major cardiovascular (CV) risk factors. However, current prediabetes care is characterized by: (a) delayed recognition of prediabetes; (b) patient unawareness of effective treatment options for prediabetes; (c) poor control of concomitant major CV risk factors; (d) very low rates of metformin initiation; and (e) low rates of follow up to assess ongoing effectiveness of prediabetes management.1-3 Increased use of electronic health records (EHR) in rural communities now provides a new opportunity to improve awareness and management of prediabetes and to reduce these patients' significant CV risk burden. In this project, we implement and evaluate an EHR-linked, Web-based clinical decision support (CDS) system that identifies patients with prediabetes and provides patients and their primary care providers personalized, evidence-based CDS and follow up to reduce risk of heart attacks or stroke. To accomplish this objective, we randomly allocate 30 primary care clinics with their 450 primary care providers and estimated 17,000 prediabetes patients to one of two intervention arms: Usual Care; or else the prediabetes CDS to optimize management and follow up of prediabetes patients with uncontrolled CV risk factors. Random-effects models assess intervention impact on: (a) American College of Cardiology/American Heart Association (ACC/AHA) pooled CV risk; (b) major CV risk factors (blood pressure, lipids, HbA1c, smoking, and BMI); (c) use of evidence-based drugs, including metformin, and lifestyle interventions to manage prediabetes; and (d) patient and provider satisfaction with the intervention strategy. We also conduct a state-of-the-art cost and a cost-effectiveness analysis of the interventions relative to usual care. The Reach, Effectiveness, Adoption, Implementation, and Maintenance (RE-AIM) framework, supplemented by the Consolidated Framework for Implementation Research (CFIR), is used to assess implementation processes and outcomes in a rural/urban health system.

The results of the project will provide a template for implementation of personalized CDS tools in rural and urban health settings, resulting in more efficient and effective rural healthcare that can be broadly applied across many clinical domains, incorporates patient treatment preferences, and has the potential to substantially improve the quality of CV care and clinical outcomes of millions of Americans with prediabetes residing in medically underserved areas.

Inclusion Criteria:

• Pre-Diabetes Diagnosis or Clinical Lab Values and
• One or more of the following CV risk factors: Current smoking, Hypertension, elevated LDL, or BMI not within normal range, AND
• at least one subsequent primary care visit to a randomized clinic 12 to 24 months after the index visit.

Exclusion Criteria:

• Evidence of Diabetes in the previous 12 months
• Pregnant
• Hospice care or Chemotherapy

The population included in the analyses are those who had at least one follow-up visit and are age 40-75 at the time of the index visit.

Up to 30 months

CDS use was lower than the 80% target rate. Many patients had limited geographic or financial access to diabetes prevention programs. Due to differences in risk equations for patients age 40-75 (American College Cardiology/American Heart Association 10-year) and age 18-40 (Framingham 30 year), and missing lipid data in younger adults, the main analysis focused on age 40-75 years. Many with prediabetes had only modest potential for cardiovascular risk reductions clinicians may not prioritize.