Background The Perioperative Anticoagulation Use for Surgery Evaluation study prospectively evaluated a prespecified periprocedural interruption strategy of direct oral anticoagulants (DOACs) among patients with atrial fibrillation. Coagulation testing is widely available and frequently requested prior to invasive procedures. Coagulation assays display poor sensitivity to clinically relevant DOAC concentrations. Objectives Determine the utility of routinely available coagulation testing at predicting a DOAC concentration of <30 ng/ml among patients in the preprocedural setting. Methods We calculated the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and positive and negative likelihood ratio (LR+ and LR−) of a normal coagulation assay result for identifying patients with a preprocedural DOAC level < 30 ng/ml. Results We identified weak or very weak correlations between coagulation assay results and DOAC levels in the preprocedural setting, except for a moderate correlation between the thrombin time (TT) and dabigatran concentrations (ρ = 0.68; p < .001). The prothrombin time (PT) and activated partial thromboplastin time (APTT) demonstrated modest sensitivity (78.9% to 88.2%) and PPVs (76.4% to 93.1%) but poor specificity (13.2% to 53.3%) and NPVs (16.3% to 30.2%) across all three DOACs. A normal TT was associated with 100% specificity and PPV values for a dabigatran level < 30 ng/ml. A normal APTT among patients on dabigatran was associated with an LR+ of 1.671 (95% confidence interval [CI] 1.297, 2.154) and an LR− of 0.395 (95% CI 0.207, 0.751) for levels <30 ng/ml. Conclusions The PT and APTT perform poorly at safely identifying patients with negligible DOAC levels in the preprocedural setting.