Limited diagnostic value of Wells-score and D-dimer testing in hospitalized dermatologic patients with symptoms of deep vein thrombosis.

BACKGROUND: D-dimer analysis and clinical probability scoring (Wells-score) show a high sensitivity and negative predictive value for the exclusion of deep vein thrombosis (DVT).

OBJECTIVE: To identify the diagnostic performance of D-dimer testing and Wells-score in hospitalized patients with dermatologic conditions.

METHODS: In this retrospective cohort study, 109 examinations in 102 patients were performed by Wells-score, Tina-quant D-dimer testing and whole-leg duplex ultrasonography or phlebography.

RESULTS: DVT was confirmed in 14 patients. The Wells-score alone allowed no discrimination of DVT and non-DVT patients. D-dimer testing identified all cases of DVT (100% sensitivity). Only 16 patients showed D-dimers within normal limits and none was diagnosed with DVT (100% negative predictive value). A high rate of false-positive D-dimer results (72%) led to a low specificity (17%). The number needed-to-test to exclude one DVT was 6.8. Based on multivariate statistical analysis, increased D-dimer levels were significantly associated with the dermatologic main diagnosis (p = 0.008), age (p = 0.001) and with the presence of DVT (p = 0.011). The highest D-dimer values were found in non-DVT patients with metastasized or systemic malignancies (median 2.48 mg/L) or inflammatory skin conditions (e.g., generalized psoriasis, median 2.22 mg/L).

CONCLUSIONS: Wells-score and D-Dimer testing were of limited diagnostic value because of many false-positive results. Required imaging procedures were reduced by only 16 cases (15%). Therefore, we suggest directly investigating hospitalized dermatologic patients with suspected DVT and skin diseases associated with high D-Dimer levels, by whole-leg compression ultrasonography.