Data-driven SDM to Reduce Symptom Burden in AF

Atrial Fibrillation, Patient Engagement

Shared Decision-Making (SDM), Symptom Burden, Decisional Conflict, Decision Regret, Decision Satisfaction

This study is a single-group feasibility study evaluating decision aid visualizations which display common post-ablation symptom patterns as a tool for shared decision-making. The specific aim of the clinical trial is to evaluate the feasibility of putting the visualizations into clinical practice (n=75). The hypothesis is that patients will report low decisional conflict and decision regret and high satisfaction with their decision about whether to undergo an ablation or not.

Atrial fibrillation (AF) is the most common heart rhythm disorder, and nearly 90% of patients experience symptoms such as shortness of breath that directly impair their health-related quality of life (HRQoL). Catheter ablation is a minimally invasive, surgical procedure that is routinely performed to treat AF and associated symptoms with the goal of improving HRQOL, but also carries potentially serious risks. Shared decision-making (SDM), in which treatment decisions are aligned based on high quality evidence and patient values and goals of care, is a widely encouraged practice for navigating complex healthcare decisions such as these. However, SDM around rhythm and symptom management does not routinely occur due to a lack of detailed evidence about symptom improvement post-ablation, and a lack of decision aids to communicate evidence to patients. The overarching goal of this award is to create an interactive patient decision aid composed of established evidence from clinical trials together with novel "real world" evidence about symptom improvement post ablation mined from electronic health records (EHRs). The investigators propose to use "real-world evidence" drawn from electronic health records (EHRs) to characterize post-ablation symptom patterns, and display them in decision-aid visualizations to support shared decision-making (SDM). In this project, the investigators will first use natural language processing (NLP) and machine learning (ML) to extract and analyze symptom data from narrative notes in EHRs. The investigators will also employ a rigorous, user-centered design protocol created during the Principal Investigator's post-doctoral work to develop decision-aid visualizations. In the clinical trial, the investigators will evaluate the feasibility of implementing these interactive decision-aid visualizations in clinical practice.

United States