Effect of Clinical Decision Support Systems on the Antithrombotic Treatment for Atrial Fibrillation Patients: A Systematic Review and Meta-analysis (Preprint)

BACKGROUND Atrial fibrillation (AF) is one of the most common arrhythmogenic diseases with high risk of disability and mortality, thereby greatly reducing the quality of life. Thromboembolic prophylaxis plays an essential role in AF therapy. Clinical decision support systems (CDSS) is available for management of AF patients with regard to antithrombotic treatment. OBJECTIVE To systematically review the association between clinical decision support systems (CDSS) and the antithrombotic treatment for the management of atrial fibrillation (AF) patients. METHODS We searched the electronic databases PubMed, MEDLINE., Embase, The Cochrane Library, and Biosis Preview for published randomized controlled trials (RCTs) on the relationship between CDSS and the management of AF patients from inception to April 2021. Two researchers screened these studies independently, extracted data, assessed the risk of bias and evaluated the CDSS features. The primary outcome was the proportion of antithrombotic treatment prescriptions in agreement with recommendations in the guidelines, and the secondary outcome was stroke morbidity and the incidence of adverse events. Meta-analysis was done using Revman5.4.1 and Stata16.1. RESULTS We included six RCTs, involving 20,562 subjects (11,334 in the intervention group and 9,228 in the control group). The 14.265 subjects had a primary outcome (7,930 in the intervention group and 6,335 in the control group). The proportion of antithrombotic treatment prescriptions in agreement with recommendations in the guidelines in the intervention group was slightly higher than that in the control group (RR=1.03, 95% CI: 1.01–1.05, P<.001). Stroke morbidity was not significantly different (RR=1.07, 95% CI: 0.94–1.22, P=.33), but adverse events were lower in the intervention group than that in control group (RR=0.79, 95% CI :0.64–0.98, P=.03). We detected no publication bias for the primary outcome in the meta-analysis (P=.89 for the Egger test and P=.81 for Begg’s test). CONCLUSIONS The use of CDSS improved physicians’ compliance with AF guidelines for antithrombotic treatment and decreased adverse events, but did not lower the stroke morbidity.