BACKGROUND: Patients with atrial fibrillation (AF) may be susceptible to gastrointestinal (GI) effects of medications used for thromboprophylaxis. OBJECTIVE: To determine the incidence of thromboprophylaxis discontinuation due to GI symptoms and GI events. METHODS: A systematic search of MEDLINE (1950-November 2011) and Cochrane Central (through third quarter of 2011) databases was conducted for randomized trials of adults with AF published in English full-text and reporting on discontinuation due to GI symptoms or the incidence of various GI symptoms. RESULTS: Of studies comparing oral direct thrombin inhibitors (DTIs) to vitamin K antagonists (VKAs), 67%, 100%, and 100% found significantly higher incidences of discontinuation due to GI symptoms, upper GI symptoms, and nondistinct GI symptoms with DTIs, respectively. When comparing aspirin to control therapy, 50% and 33% of trials found significantly higher incidences of discontinuation due to GI symptoms and nondistinct GI symptoms with aspirin, respectively. Aspirin significantly increased the incidence of upper GI symptoms compared to VKAs in 1 trial, but VKAs did not increase the incidence of discontinuation due to GI symptoms and nondistinct GI symptoms versus control. In the only trial comparing a factor Xa inhibitor to VKAs, no significant difference in diarrhea was seen, and when a factor Xa inhibitor was compared to aspirin, there was no significant difference in nondistinct GI symptoms. The incidence of discontinuation due to GI symptoms (4 trials) ranged from 0.0% to 0.4% for control, 0.0% to 2.4% for aspirin, 0.0% to 0.63% for VKA, and 1.2% to 4.7% for oral DTI recipients. The incidence of upper GI symptoms (2 trials) was 1.2% for aspirin, 0.0-5.8% for VKA, 4.2% for VKA plus aspirin, and 11.3-11.8% for DTI recipients. The incidence of nondistinct GI symptoms (7 trials) was 0.0-12.4% for control, 0.2-17.9% for aspirin, 0.0-14.1% for VKA, 23% for DTI, and 2.42% for factor Xa inhibitor recipients. Five trials evaluated diarrhea incidence, which ranged from 0.0% to 12.4% for VKA and 3.2% to 12.2% for oral DTI, and was 5.31% for factor Xa inhibitor recipients. One trial evaluated nausea, with incidences of 1.5% and 2.7% for VKA and DTI recipients, respectively. CONCLUSIONS: GI adverse events are common in AF trial populations. DTIs were associated with more discontinuations due to GI symptoms and aspirin appeared to increase the incidence of GI symptoms compared to control.
BACKGROUND:Patients with atrial fibrillation (AF) may be susceptible to gastrointestinal (GI) effects of medications used for thromboprophylaxis. OBJECTIVE: To determine the incidence of thromboprophylaxis discontinuation due to GI symptoms and GI events. METHODS: A systematic search of MEDLINE (1950-November 2011) and Cochrane Central (through third quarter of 2011) databases was conducted for randomized trials of adults with AF published in English full-text and reporting on discontinuation due to GI symptoms or the incidence of various GI symptoms. RESULTS: Of studies comparing oral direct thrombin inhibitors (DTIs) to vitamin K antagonists (VKAs), 67%, 100%, and 100% found significantly higher incidences of discontinuation due to GI symptoms, upper GI symptoms, and nondistinct GI symptoms with DTIs, respectively. When comparing aspirin to control therapy, 50% and 33% of trials found significantly higher incidences of discontinuation due to GI symptoms and nondistinct GI symptoms with aspirin, respectively. Aspirin significantly increased the incidence of upper GI symptoms compared to VKAs in 1 trial, but VKAs did not increase the incidence of discontinuation due to GI symptoms and nondistinct GI symptoms versus control. In the only trial comparing a factor Xa inhibitor to VKAs, no significant difference in diarrhea was seen, and when a factor Xa inhibitor was compared to aspirin, there was no significant difference in nondistinct GI symptoms. The incidence of discontinuation due to GI symptoms (4 trials) ranged from 0.0% to 0.4% for control, 0.0% to 2.4% for aspirin, 0.0% to 0.63% for VKA, and 1.2% to 4.7% for oral DTI recipients. The incidence of upper GI symptoms (2 trials) was 1.2% for aspirin, 0.0-5.8% for VKA, 4.2% for VKA plus aspirin, and 11.3-11.8% for DTI recipients. The incidence of nondistinct GI symptoms (7 trials) was 0.0-12.4% for control, 0.2-17.9% for aspirin, 0.0-14.1% for VKA, 23% for DTI, and 2.42% for factor Xa inhibitor recipients. Five trials evaluated diarrhea incidence, which ranged from 0.0% to 12.4% for VKA and 3.2% to 12.2% for oral DTI, and was 5.31% for factor Xa inhibitor recipients. One trial evaluated nausea, with incidences of 1.5% and 2.7% for VKA and DTI recipients, respectively. CONCLUSIONS: GI adverse events are common in AF trial populations. DTIs were associated with more discontinuations due to GI symptoms and aspirin appeared to increase the incidence of GI symptoms compared to control.