Comparing the Efficacy of Targeted and Blast Portal Messaging in Message Opening Rate and Anticoagulation Initiation in Patients With Atrial Fibrillation in the Preventing Preventable Strokes Study II: Prospective Cohort Study

• DOI: 10.2196/49590
• Publication Date: 2023-11-24T14:03:27Z
• Authors: Alok Kapoor; Parth Patel; Soumya Chennupati; Daniel Mbusa; Hammad Sadiq; Sanjeev Rampam; Robert Leung; Megan Miller; Kevin Rivera Vargas; Patrick Fry; Mary Martin Lowe; Christina Catalano; Charles Harrison; John Nicholas Cat...; Sybil Crawford; Anne Marie Smith
• Keywords: Patient portal; Stroke; Patient Education

Background The gap in anticoagulation use among patients with atrial fibrillation (AF) is a major public health threat. Inadequate patient education contributes to this gap. Patient portal–based messaging linked educational materials may help bridge gap, but the most effective approach unknown. Objective This study aims compare responsiveness of AF an or message between 2 portal approaches: sending messages targeted at upcoming outpatient appointments 1 week before their scheduled appointment (targeted) versus all eligible blast, regardless scheduling status (blast), different systems: University Massachusetts Chan Medical School (UMass) and Florida College Medicine-Jacksonville (UFL). Methods Using approaches, we sent grouped by high-risk on (group 1), off 2), low-risk who become for future 3). Risk was classified based congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, stroke, vascular disease, 65 74 sex category (CHA2DS2-VASc) score. contained link Upbeat website Heart Rhythm Society, which displays print video about anticoagulation. We then tracked opening, review website, use, administered surveys across approaches sites using Epic Systems (Epic Corporation) electronic record data Google traffic analytics. conducted chi-square tests potential differences proportion opening other evaluation metrics, adjusting confounders. All statistical analyses were performed SAS (version 9.4; Institute). Results 1686 1450 blast messages. Message significantly higher (723/1156, 62.5% vs 382/668, 57.2%; P=.005) trended same anticoagulation; subsequent reviews did not differ approach. More baseline started than (adjusted percentage 9.3% 2.1%; P<.001). Conclusions Patients more responsive terms initiation