Background: Diagnosing venous thromboembolism (VTE) in patients with previous VTE is challenging. Only approximately 15% of these patients have a VTE recurrence ruled out without imaging, as compared with >30% in patients with a suspected first episode. Clinical decision rules (CDRs) are not well studied in this population. The aim was to assess the performance of existing CDRs (Wells and Geneva scores) in these patients.

Methods: This international prospective multicenter observational cohort study of outpatients with suspected VTE recurrence was conducted in 8 sites (November 2014-January 2019) (clinicaltrials.gov: NCT02297373). The primary outcome was the performance of CDRs. Secondary outcome included the diagnostic yield of D-dimers. Suspected recurrent VTE were adjudicated. Follow-up was 3 months.

Results: In total, 723 patients were included. Among patients with suspected isolated deep venous thrombosis (DVT), 4 of 67 patients (6.0%), 39 of 186 patients (21.0%), and 65 of 148 patients (43.9%) had recurrent VTE in the low, moderate, and high probability groups, respectively. Among patients with suspected isolated pulmonary embolism, 11 of 83 patients (13.3%), 56 of 160 patients (35.0%), and 12 to 25 patients (48.0%) had VTE recurrence in the low, moderate, and high probability groups (Wells), respectively. In patients with suspected isolated DVT and negative D-dimers, 2 of 65 patients, 3.1% (95% CI: 0.4%-10.7%) in the low and moderate probability groups had recurrent VTE at enrollment. In patients with suspected pulmonary embolism ± DVT and negative D-dimers, no recurrent VTE was diagnosed: 0 of 82, 0.0% (95% CI: 0.0%-4.4%).

Conclusion: CDRs were able to characterize patients with increasing VTE prevalence. The efficiency and safety in ruling out recurrent VTE without imaging were lower than for suspected first VTE.