Clinical history and gastrointestinal bleeding in patients taking oral anticoagulants

Background: Common risk factors for gastrointestinal bleeding (GIB) are advanced age and the use of antiplatelet or anticoagulants drugs for the prevention of cardiovascular diseases. Methods: In this prospective real-world observational study, oral anticoagulated patients were recruited and followed between June 2013 and December 2019. The primary end-point was to evaluate a possible relationship between bleeding events and patients' clinical history of gastrointestinal disease prior to the start of the therapy. The secondary end-points were time of GIB appearance and the percentage of idiopathic or provoked events, i.e., bleeding due to a gastrointestinal disease. In case of GIB event all the patients were studied by means of endoscopic procedures. Cox regression was used to calculate the relative hazard ratios (HRs) of GIB for each considered clinical variable. Results: 734 patients on both VKAs or DOACs were studied. Overall, 46 hemorrhagic events were recorded: 6 were major bleeding (0.42/100 patient-years) while 43 were clinically relevant non major bleeding (2.8/100 patient-years). The Cox regression analysis did not show any relationships among GIB and the variables considered. Conclusion: The patients' clinical history is neither a predictor for GIB bleeding nor a guide to the choice of the oral anticoagulant to be administered. Routinely applying bleeding risk screening, such as occult blood in the stool, should be added to the periodic laboratory checks for early recognition of patients at higher risk of GIB.