Patients who receive long-term oral anticoagulant (OAC) therapy often require interruption of OAC for an elective invasive procedure. Current guidelines allow bridging therapy with either unfractionated heparin (UFH) or low-molecular-weight heparin (LMWH). Apart from the risk of embolism, bleeding is an important complication in this setting and the optimal perioperative management of such patients is still under discussion. The aims of this prospective, observational, multicentre registry of patients treated by cardiologists were: 1) to evaluate current practice of perioperative management of OAC in a large outpatient cohort, 2) to document embolic and haemorrhagic events, and 3) to identify risk factors predicting adverse events. In the years 2009 and 2010, 1,000 invasive procedures (cardiac catheterisation n=533, pacemaker implantation n = 128, surgery n = 194, other n = 145) were performed in patients with OAC. Sixty- one (6.1%) of those patients did not receive bridging therapy during interruption of OAC, 937 (93.7%) patients were treated with LMWH, two patients (0.2%) received UFH. In 22 patients (2.2%) LMWHs were given in prophylactic dose, 727 patients (72.7%) were treated with halved therapeutic (i.e. weight-adapted) LMWH doses and 188 (18.8%) received full therapeutic LMWH doses. Four thromboembolic complications were observed during 30 days of follow-up (two retinal embolisms, one stroke, one myocardial infarction; 0.4%). One major bleeding (0.1%) and 35 clinically relevant bleedings (3.5%) occurred. Rehospitalisation after bleedings was necessary in 20 patients. Independent predictors for bleedings were history of mechanical heart valve replacement (MVR) (p=0.0002) and the HAS-BLED score (<0.0001), with a cut off value ≥ 3 being the most predictive variable for haemorrhage (hazard ratio 11.8, 95% confidence interval 5.6-24.9, p<0.0001). A total of 527 patients with atrial fibrillation and a CHADS₂ score ≤ 2 received halved therapeutic or full therapeutic dosages of LMWH despite a low embolic risk, whereas 49 of the patients with heart valve replacement (51%) did not receive dosages of bridging therapy as recommended in guidelines. In conclusion, in this registry of patients treated by cardiologists, 94% of patients who required interruption of OAC before invasive procedures received LMWH as a bridging therapy, of whom 73% were treated with halved therapeutic LMWH-dosages. Guideline recommendations were followed in only 31% of cases. Importantly, 69% of patients with AF were over-treated while 51% of patients with heart valve replacement were under-treated with LMWHs. A HASB-BLED score ≥ 3 was highly predictive of bleeding events.