Impact of Health Information Technology Implementation on Diabetes Process and Outcome Measures

Diabetes is a common chronic disease that requires long-term management. Currently, the health care provided to diabetic patients frequently falls short of the “best care” practices established in evidence-based diabetes clinical guidelines. Health information technology has the potential to improve compliance with recommended diabetes-related processes of care and outcome measures. Research has shown that having an electronic health record (EHR) is not enough to increase patient’s attainment of clinical targets: it is the specific functions within it – such as patient identification and tracking systems, problem lists, and electronic visit notes – that lead to improvements.

This project evaluated the impact of an EHR on the quality of diabetes care as measured by compliance with recommended processes of care and patient outcome measures. The EHR has an embedded Diabetes Management Form (DMF), an optional documentation tool that integrates diabetes-specific data review, passive clinical decision support (CDS) (showing recommended targets for key diabetes measures and whether recommended processes of care were due but not prompting any particular action), order entry, and coded data capture capabilities.

The specific aims of this project were to:

• Estimate the impact of an EHR on diabetes outcomes, measured by the proportion of patients meeting the Health Partners Optimal Diabetes Care measure. 
• Estimate impact of an EHR on specific patient outcomes and compliance with recommended process of care related to diabetes. 
• Estimate the prevalence of physician use of the Diabetes Management Form, and the effect of the Diabetes Management Form on patient outcomes related to diabetes as measured by the Optimal Diabetes Care measure. 

The project used an observational study design with primary care practices that underwent a staggered implementation of a commercially available EHR. The primary outcome was meeting diabetes “optimal care” target measures for HbA1c, LDL-cholesterol, blood pressure, not smoking, and documented aspirin use. Compliance was compared between patients exposed and not exposed to the EHR and in a subset of EHR-exposed patients in patients for whom the DMF was used and those for whom it was not.

Patients whose providers were using an EHR were significantly more likely to achieve “optimal care” and were more likely to be in compliance with most process and outcome measures. DMF-exposed patients showed less improvement in attaining targets, but greater improvement in documentation of care. The project team hypothesized that the embedded DMF failed to achieve greater improvement in patient outcomes due to its being passive and physician-focused CDS. Because lifestyle changes are needed to alter a patient’s diabetes-related outcome measures, tools that directly assist patient self-management might be more likely to achieve improvement.