International normalized ratio (INR) and aPTT values associated with dabigatran plasma levels≤20 ng/ml – Development and validation of a predictive model including 164 patients.

Introduction Prior to surgery or in emergency situations, there is a need to rule out potential anticoagulant effects in patients with a possible unknown intake of oral anticoagulants. Whereas calibrated anti-Xa-based assays to measure the anticoagulant effect of anti-Xa-inhibitors are broadly available, this is not the case for dabigatran. The purpose of this study was to develop and validate a predictive model to ensure dabigatran levels≤20 ng/ml using INR and aPTT values. Method We conducted a two-center cohort study including 164 patients (male n=81; median [min-max] age 69 yrs. [14-96]) consecutively ascertained between 01/2017 and 09/2025 in whom dabigatran plasma levels were analyzed along with INR, aPTT, renal and liver function. A receiver operating characteristic (ROC) curve analysis was performed in the derivation cohort (D: enrolled between 01/2017-06/2021) and was validated in an independent cohort (V: ascertainment 07/2021-09/2025). APTT values were measured with HemosIL®​ SynthASil (Werfen) and dabigatran with the HemosIL Direct Thrombin Inhibitor Assay (Werfen), both on the ACL TOP 750. Cut-off values used in the model were≤1.3 for INR and≤40 sec. for aPTT values. In addition, within the entire cohort logistic regression adjusted for age and gender was performed to discriminate between dabigatran plasma levels below/above the chosen cut-off. Results Whereas patients in the derivation cohort were younger compared to subjects in the validation cohort (62 vs. 72 yrs; p=0.05), no significant differences were found when comparing median [min-max] values for dabigatran concentrations (D: 60 ng/ml [10-840] vs. V: 70 ng/ml [10-840]; p=0.42), aPTT levels (D: 38 sec. [27-87]; V: 42 sec. [25-120]; p=0.1) and INR values (D: 1.1 [0.9-2.3]; V: 1.2 [0.93-4.5]; p=0.6). When using the defined cut-off in the derivation cohort specificity to rule out dabigatran concentrations above 20 ng/ml calculated via area under the ROC curve (AUC) was 100 for INR, 92.77 for aPTT and 97.22 when using the combined model INR≤1.3 and aPTT≤40 sec. Similar results were found in the validation cohort: Specificity in the INR-model was 100, 97.8 for aPTT values≤40 sec., and 97.73 when including both parameters, i.e. INR and aPTT. Of note, sensitivity was low in both cohorts. Logistic regression adjusted for age and gender performed in 164 patients enrolled revealed an Odds of 11.6 to rule out dabigatran plasma concentrations>20ng/ml in the patients investigated. Conclusion Our prediction model demonstrates that INR in combination with aPTT values are useful to rule out potential anticoagulant effects in patients receiving dabigatran medication in the absence of specific anti-IIa assays. When using the cut-off values≤1.3 for INR and≤40 sec. for aPTT dabigatran plasma concentrations>20 ng/ml can be safely ruled out. This model is valid for the specific aPTT reagent used and has to be re-evaluated with different aPTT reagents. Publication History Article published online: 17 February 2026 © 2026. Thieme. All rights reserved. Georg Thieme Verlag KG Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany