Abstract 11918: Improving Quality of Care for Patients With Valvular Heart Disease in a Learning Health System

Abstract

Introduction: Delayed treatment of valvular heart disease (VHD) is associated with higher mortality and adverse event incidence. In a large, integrated health system, we developed and piloted a digital health solution embedded in the electronic health record (EHR) to identify, triage, and streamline evaluation of patients with VHD. Methods: The initiative began May 2021 with agreement of our team members and stakeholders for prioritization of the project and implementation. Structured echocardiographic data were made available in electronic data warehouse. SQL-based disease classification algorithms based on published guidelines were built for aortic stenosis (AS), mitral regurgitation (MR) and tricuspid regurgitation. Echocardiograms with suspected VHD were pushed daily into a reporting dashboard and smartform we built within the EHR. Nurse coordinators responsible for VHD patient care tested the platform, and we iteratively adapted components based on feedback. Performance of AS and MR classification algorithms against historical data were evaluated using kappa statistics. We launched the program in March 2022 and report the initial results. Results: Between August and October 2021, 650 echocardiograms were reviewed by a nurse coordinator. Agreement between the guideline-based algorithms and chart review was moderately high (0.64 for AS and 0.86 for MR). Between March and May 2022, 7180 echocardiograms with suspected VHD were surfaced in the EHR for 6588 patients. Among 1501 patients prioritized for review based on disease severity, 341 patients were referred to the valve clinic for procedural intervention, 783 patients are under watchful waiting, and 377 patients were determined to unlikely benefit from valve interventions. Conclusions: By leveraging data from echocardiogram repositories, it is feasible to implement a VHD quality improvement program within a learning health system. This requires a diverse and interdisciplinary team for development, implementation, governance, and continued iteration of EHR-embedded digital health tools to improve quality of care. While we focused on VHD, our methodology and infrastructure allow for expansion and leverage into other therapeutic domains within cardiovascular disease and beyond.