External validation of the DAYS score for suspected deep vein thrombosis

Abstract

Background

Diagnosing deep vein thrombosis (DVT) involves clinical assessment, D-dimer testing, and imaging. The DAYS score, a novel 2-item prediction tool combined with D-dimer, demonstrated promising performance but required external validation.

Objectives

This study aimed to validate 2 DVT prediction scores: the DAYS score and a newly developed DVT score.

Methods

Data were collected from a prospective Norwegian DVT management study (2015-2018; NCT02486445). The DAYS score includes 2 items (DVT most likely diagnosis and calf swelling), while the new score comprises tenderness along the deep veins and previous venous thromboembolism. DVT was considered excluded if no items were present and D-dimer was < 1.0 μg/mL or if ≥1 item was present and D-dimer was < 0.5 μg/mL. The 2-tier Wells score served as the reference. Safety was defined as the number of missed DVT cases divided by the total number of patients classified as having DVT excluded and was set at 2%.

Results

Among 1312 patients (median age, 64 years; IQR, 52-73 years; 55% women), 261 (20.0%) had confirmed DVT. The DAYS score excluded DVT in 455 patients (34.6%), of whom 11 were diagnosed with DVT (failure rate, 2.4 %; 95% CI, 1.2-4.2). The new score excluded DVT in 519 patients (39.6%) and missed 7 cases with confirmed DVT (failure rate, 1.3%; 95% CI, 0.5-2.8). The Wells score excluded DVT in 271 patients (20.6%), missing only 2 cases with confirmed DVT.

Conclusion

While both the DAYS score and the new score demonstrated low failure rates, they exceeded the predefined safety threshold.