BACKGROUND: Previous research suggests uncertainty whether or not there is any additional benefit in adding antiplatelet therapy (APT) to anticoagulation therapy (ACT) in patients with high-risk atrial fibrillation (AF) in terms of reduction in vascular events, including stroke. The existing guidelines acknowledge an increased risk of bleeding associated with such a strategy; however, there is no consensus on the treatment pathway. OBJECTIVES: To determine, by undertaking a systematic review, if the addition of APT to ACT is beneficial compared with ACT alone in patients with AF who are considered to be at high risk of thromboembolic events (TEs). DATA SOURCES: Data sources included bibliographic databases {the Cochrane Library [Cochrane Central Register of Controlled Trials (CENTRAL)], MEDLINE, MEDLINE In-Process and Other Non-Indexed Citations, EMBASE, ClinicalTrials.gov, National Institute for Health Research (NIHR) Clinical Research Network Portfolio, Current Controlled Trials (CCT) and the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP)}, reference lists from identified systematic reviews and relevant studies, and contact with clinical experts. Searches were from inception to September 2010 and did not use language restrictions or study design filters. REVIEW METHODS: Studies of any design were included to evaluate clinical effectiveness, including randomised controlled trials (RCTs), non-randomised comparisons, cohort studies, case series or registries, longitudinal studies, systematic reviews and meta-analyses, and conference abstracts published after 2008. Inclusion criteria consisted of a population with AF, at high-risk of TEs, aged ≥ 18 years, on combined ACT and APT compared with others on ACT alone or ACT plus placebo. Inclusion decisions, assessment of study quality and data extraction were undertaken using methods to minimise bias. RESULTS: Fifty-three publications were included, reporting five RCTs (11 publications), 18 non-randomised comparisons (24 publications) and 18 publications that reported reviews, which added no further data. There was variation in the population, types and doses of ACT and APT, definitions of outcomes, and length of follow-up between the studies. There was a paucity of directly randomised high-quality RCTs, whereas non-randomised comparisons were found to have significant confounding factors. No studies looked at the effect of ACT plus APT compared with ACT alone on vascular events in patients with AF following acute coronary syndrome (ACS) or percutaneous coronary intervention. In most studies, significant differences in event rates were not seen between the patients on combined therapy compared with those on ACT alone for outcomes such as stroke (including haemorrhagic and ischaemic strokes), rates of transient ischaemic attacks, composite end points of stroke and systemic embolism (SE), SE alone, acute myocardial infarction, mortality (vascular or all cause) or bleeding events. There was conflicting evidence regarding rates of major adverse events consisting of composite end points, although event rates were generally low. LIMITATIONS: An attempt was made to identify all of the available evidence around the subject despite the dearth of directly randomised studies using a robust review methodology. There was a paucity of directly randomised evidence to undertake a meta-analysis for the merits of one technology over another. The selection criteria were kept necessarily broad with regard to the population, intervention and comparator in order to capture all relevant studies. CONCLUSIONS: This systematic review suggests that there is still insufficient evidence to advocate a clear benefit of the addition of APT to ACT compared with ACT alone in reducing the risk of vascular events in a population of patients at high risk of TEs resulting from AF. It is recommended that a definitive prospective RCT needs to be undertaken in a population at high risk of atherosclerotic coronary artery and other vascular events in addition to being at high risk of AF-mediated TEs. From the UK context, at the time of writing, any future trial should compare adjusted-dose warfarin [international normalised ratio (INR) 2.0-3.0] plus aspirin (75-325 mg) with adjusted-dose warfarin (INR 2.0-3.0). However, given the emergence of newer anticoagulation agents (dabigatran, rivaroxaban and apixaban) this prioritisation may need to be revisited in the future to reflect current best clinical practice. FUNDING: The National Institute for Health Research Health Technology Assessment programme.