Effect of thromboprophylaxis with anticoagulant drugs on the incidence of arterial thrombotic events in medical inpatients: a systematic review

Pharmacological prophylaxis of venous thromboembolism (VTE) is recommended for medical inpatients. Since arterial thrombosis (AT) shares some risk factors with VTE, it would be reasonable to assess the efficacy of thromboprophylaxis by considering both VTE and AT as outcome events. We performed a systematic review and meta-analysis of phase III RCTs on thromboprophylaxis in medical inpatients, to evaluate the quality of reporting and the incidence of AT, and the effect of thromboprophylaxis with anticoagulants on AT incidence. Studies were identified by a combined search strategy until May 2015. Differences in outcomes among groups were expressed as pooled odds ratios (OR) and 95 % confidence intervals (CI). Statistical heterogeneity was assessed by I2 statistic. Twenty phase III RCTs, encompassing 54,742 patients, were included; of these, 3 (15 %) reported on AT as a pre-defined secondary outcome and 8 (40 %) on at least one AT outcome. Raw-unweighed incidence of fatal MI in the three RCTs is 0.37 % in patients receiving unfractionated heparin or enoxaparin, and 0.38 % in controls (OR 0.97, 95 % CI 0.62–1.52; I2 = 0 %). A non-statistically significant increase in AT is reported in patients on enoxaparin compared to control (OR 1.95, 95 % CI 0.89–4.27; I2 = 13 %). AT is underreported in RCTs on VTE prophylaxis in medical inpatients. Published data suggest that incidence of fatal MI in these patients may be clinically relevant. Insufficient data are available to draw firm conclusions on the effects of thromboprophylaxis with anticoagulants on AT incidence in this setting.