IMPACT-AF: System fails to affect time to ED visit for physicians treating AF

CHICAGO — A computer decision support system used by primary care physicians did not have a significant effect on the time to an ED visit for atrial fibrillation or an unplanned CV hospitalization compared with usual care, according to data presented at the American Heart Association Scientific Sessions.

“Nonetheless, incident rate ratios do suggest the potential for marked reductions in events with such tools, but we need better user training and longer duration of use to confirm this,” Jafna L. Cox, MD, FRCPC, FACC, professor in the division of cardiology at Dalhousie University in Halifax, Nova Scotia, said during the presentation.

Researchers analyzed data from 1,145 patients with AF confirmed through ECG or documentation of past diagnosis and management.

Patients were assigned to a clinical decision support system (n = 597; mean age, 73 years; 58% men) or usual care (n = 548; mean age, 72 years; 65% men). The clinical decision support system used in this study digitized and operationalized clinical pathways, evidence-based guidelines and algorithms to provide patients with timely, personalized and evidence-informed care.

“Such systems are designed to monitor patient-specific clinical data, proactively identify potential adverse events and assist physicians in terms of providing diagnostic and therapeutic recommendations,” Cox said during the presentation.

The system provided physicians with rate and rhythm control guidance, stroke prevention and risk scores, perioperative anticoagulation management, guidance on oral anticoagulation and novel oral anticoagulation, and graphics of health trends for patients with AF. The patient self-monitoring system, which was optional in this study, gave information on AF management resources, personalized alerts and recommendations, National Stroke Association guidelines and enhanced communication regarding care.

The primary efficacy outcomes were a composite of unplanned CV hospitalization or ED visit related to AF. The primary safety outcome was major bleeding during a 12-month period.

At 12 months, there were no differences between the two groups regarding the number of patients experiencing primary efficacy outcomes (HR = 1.02; 95% CI, 0.73-1.41). However, while not statistically significant, of the patients having an event, those assigned the clinical decision support system had a nearly 30% reduction in total ED visits for AF compared with those assigned usual care (incident rate ratio = 0.71; 95% CI, 0.42-1.21).

Safety outcomes also did not differ in the clinical decision support system and usual care groups. The HR for the primary safety outcome was 0.91 (95% CI, 0.32-2.6). Notably, the number of bleeding events over 12 months was low, with just 19 of these overall.

“There are several plausible reasons for why we feel that significant benefit was not seen," Cox said during the presentation. "First of all, there are real-world challenges to any health care technology assessment that pragmatic trial designs cannot control for. For instance, the [clinical decision support system] suggested approaches to care, but physicians and patients could choose to proceed otherwise. As well, there’s a steep user learning curve to new technology that takes longer to overcome in the context of a busy clinical practice.” – by Darlene Dobkowski

Reference:

Cox JL, et al. LBS.03 – Late-Breaking Clinical Trial: Harnessing technology and improving systems for global health. Presented at: American Heart Association Scientific Sessions; Nov. 10-12, 2018; Chicago.

Disclosure: The study was funded by Bayer. Cox reports he received research grant support from Bayer and honoraria from Bayer and Servier.

Editor’s Note: This article was revised Dec. 3, 2018, with updated data from Dr. Cox.