The Value of D-Dimer in Patients with Increased C-Reactive Protein Suspected of Pulmonary Embolism.

To investigate the relation between elevated levels of C-reactive protein (CRP) and D-dimer, and to study whether D-dimer levels can be interpreted in relation to elevated levels of CRP in the prediction of a pulmonary embolism in order to increase its specificity without decline in sensitivity.

Between august 2004 and april 2007 (33 months) serum levels of C-reactive protein (mmol/L) and D-dimer (mmol/L) were cross-sectionally collected and pulmonary embolisms on CT-angiograms were scored within 48 hours. The study was devided into three parts. First, characteristics of excluded persons were studied. Second, the correlation between CRP and D-dimer level was considered in those with a defined (i.e. values with < and > symbols excluded) biomarker level. Finally, the effect of CRP level on the sensitivity of D-dimer for pulmonary embolisms was examined.

CRP and D-dimer levels were positively correlated ( r = 0.37; p < 0.001), and both were increased in persons with a pulmonary embolism (CRP: p = 0.02; D-dimer: p < 0.001). 14 % of variability in D-dimer level was explained by CRP level. Median D-dimer levels were increased in the pulmonary embolism (PE) group, however, the increase in D-dimer level by CRP quartile as was found in the non-PE was not seen in de PE-group. Adding the interaction term of CRP and D-dimer to the statistical model showed some influence on the area under the curve (AUC). Nevertheless, this was not significantly different from the model with only D-dimer levels. However, when stratified for CRP quartile, ROC-curves of the predictive effect of D-dimer on pulmonary embolisms showed a decrease in AUC by increment of CRP quartile.

CRP and D-dimer are positively correlated, and both predictive of PE. The predictive value of D-dimer for PE declines by increment of CRP, although this seems to be safely for a broader range of accompanied CRP levels.

No relevant conflicts of interest to declare.