Early differentiation of erysipelas from deep vein thrombosis (DVT) based solely on clinical signs and symptoms is challenging. There is a lack of data regarding the usefulness of the inflammatory biomarkers procalcitonin (PCT), C-reactive protein (CRP) and white blood cell (WBC) count in the diagnosis of localized cutaneous infections. Herein, we investigated the diagnostic value of inflammatory markers in a prospective at-risk patient population. This is an observational quality control study including consecutive patients presenting with a final diagnosis of either erysipelas or DVT. The association of PCT (μg/L) and CRP (mg/L) levels and WBC counts (g/L) with the primary outcome was assessed using logistic regression models with area under the receiver-operator curve. Forty-eight patients (erysipelas, n = 31; DVT, n = 17) were included. Compared with patients with DVT, those with erysipelas had significantly higher PCT concentrations. No significant differences in CRP concentrations and WBC counts were found between the two groups. At a PCT threshold of 0.1 μg/L or more, specificity and positive predictive values (PPV) for erysipelas were 82.4% and 85.7%, respectively, and increased to 100% and 100% at a threshold of more than 0.25 μg/L. Levels of PCT also correlated with the severity of erysipelas, with a stepwise increase according to systemic inflammatory response syndrome criteria. We found a high discriminatory value of PCT for differentiation between erysipelas and DVT, in contrast to other commonly used inflammatory biomarkers. Whether the use of PCT levels for early differentiation of erysipelas from DVT reduces unnecessary antibiotic exposure needs to be assessed in an interventional trial.
Keywords: cellulitis; deep vein thrombosis; erysipelas; lower limb; procalcitonin.
© 2015 Japanese Dermatological Association.