Health Information Technology in Heart Failure Care (New York)

Patients with heart failure experience 4 million hospitalizations per year in the United States. Of these hospitalizations, approximately three-quarters are for causes other than heart failure. Those hospitalized for other causes are less likely to receive heart failure guideline recommended care than patients who are hospitalized primarily for heart failure, putting them at high risk of postdischarge mortality. Reducing hospitalizations for these patients has become a priority for providers, payers, and policymakers. Clinical decision support systems (CDSSs) have the potential to enhance clinician compliance with guidelines. Despite their utility, CDSSs have not been developed to apply more broadly to patients with heart failure who are hospitalized for other causes. This is a gap for millions of patients who might experience improved outcomes if they were to receive guideline recommended care.

This project will develop and test a CDSS for hospitalized patients who have heart failure, regardless of reason for hospitalization. An electronic health record (EHR)-based algorithm will be developed to identify patients with the condition using a variety of computational techniques including natural language processing. A CDSS tool will then be developed and implemented to support delivery of recommended care.

The specific aims of the project are as follows:

• Develop an EHR-based search method that uses both structured data and natural language processing of unstructured text to identify hospitalized patients with heart failure while they are still in the hospital 
• Develop and test the usability of an EHR-based CDSS to support delivery of evidence-based care to hospitalized heart failure patients using human computer interaction methods
• Assess the effect of an EHR-based CDSS on processes of care for heart failure patients who are hospitalized for other causes

Development will include designing a user interface and creating prompts to providers to follow guidelines at the point of care. The project team will conduct usability testing to inform tool refinements. The tool will be evaluated using a time series analysis. The primary outcomes will be compliance with two quality metrics around assessment of left ventricular (LV) heart function and use of appropriate medications for those with LV dysfunction. Uptake of the CDSS and clinical outcomes of postdischarge rehospitalization and mortality will be measured.