Venograms were used to determine the presence of venous obstruction. Interrelations between the incidence of venous obstruction and patient-or devicerelated parameters were identified using Fisher's exact test and univariate logistic regression. Multivariate logistic regression was used to identify independent predictors of venous obstruction. Results: 456 patients met the inclusion criteria (330 males, 126 females, 67.8 ± 12.9 years). 100 patients underwent first implantation, and 356 patients underwent device revision (mean time since implantation 82.5 ± 75.3 months). Venous obstruction was present in 11.0 % and 30.1 % before implantation and revision, respectively. Only presence of ventricular escape rhythm was significantly related to venous occlusion (p < 0.001) prior to first implantation. Prior to revision, significant predictors were male sex (p = 0.01), time since implantation (p < 0.0001), presence of escape rhythm (p = 0.02), compromised coagulation (p = 0.02), phenprocoumon (p = 0.005), and peripheral arterial disease (p = 0.01). Conclusion: Although several risk factors could be identified, reliable prediction of venous obstruction was not possible. Therefore, we advocate performing venography in all patients prior to device revision or upgrade to avoid complications. In cases of first device implantation, the risks associated with venography should be weighed against the surprisingly high rate of deep upper vein obstruction.This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.