Abstract
Background: atrial fibrillation (AF) patients with obesity and high thromboembolic risk need oral anticoagulant therapy. Limited data are available on direct oral anticoagulants (DOACs) in this population, and a point-of-care method has been validated to support rapid clinical decisions and to identify on-off plasma concentration thresholds. Methods: This is a monocentric, cross-sectional diagnostic accuracy study on obese AF outpatients referred to Policlinico Umberto I of Rome. Urinary Dipsticks were assessed with separate pads for factor Xa (FXA-i) and thrombin inhibitor (THR-i) and compared to the reference standard of trough and peak plasma concentrations with chromogenic assays/dTT and prespecified plasma thresholds for each DOAC. Study endpoints were the sensitivity, specificity, positive and negative predictive values (PPV and NPV) of DOACs Dipstick compared to plasma concentrations. Sub-analyses according to obesity severity and type of DOAC were performed. Results: 320 paired plasma and urine samples were available from 160 enrolled patients (mean age 73.2 ± 9.1 years). Compared to trough plasma concentrations, DOACs Dipstick showed a sensitivity of 99.24% (mean, 95% confidence interval, CI 95.82-99.98), specificity of 6.89% (0.85-22.76), PPV 82.80% (81.32-84.18), NPV 66.67% (15.79-95.52). On the other hand, compared to peak plasma concentrations, DOACs Dipstick showed a sensitivity of 97.8% (93.7-99.5), specificity of 0% (0.0-15.4), and PPV of 85.9% (85.6-86.2). Urinary Dipstick showed a sensitivity of 99.10% (95.4-100.0), specificity of 4.70% (0.60-16.20) and a PPV and NPV of 74.50% (73.2-75.8) and 66.70 (15.7-95.6), compared to plasma thresholds > 30 ng/mL of FXA-I and THR-I. Sub-analyses showed similar results between FXA-i and THR-i. Conclusions: The urine point-of-care has high sensitivity, acceptable PPV, but low specificity and NPV in AF obese patients and may be useful only in selected clinical scenarios.