Dear Sir,D-dimers are specific fibrin degradationproducts. Elevated plasma levels indicatean activation of the coagulation system [1].Modern latex agglutination assays are thuswidely performed in clinical practice to eval-uate suspected acute venous thromboembo-lism (VTE) and disseminated intravascularcoagulation (DIC), to monitor thrombolytictherapies, or to assess the postoperative riskfor VTE. With a reported sensitivity of up to98–100% and a high negative predictive val-ue, a normal D-dimer assay is useful for rul-ing out acute VTE. However, this laboratorytest is known to be of modest specificity forthe diagnosis of acute VTE in different pa-tient groups [2, 3]. In hemodialysis patientsthe diagnostic value of D-dimers is uncer-tain. End-stage renal disease and especiallyhemodialysis have been shown to inducechanges of coagulation and fibrinolytic fac-tors consistent with a procoagulatory state[4]. Therefore, we analysed the prevalence ofelevated D-dimer levels and the impact ofcomorbidity causing a procoagulatory state.The D-dimer plasma levels from 54 he-modialysis patients without clinical signs ofacute VTE were examined. All patients un-derwent chronic hemodialysis therapy at thesame centre. The blood specimen was takenbefore starting a single dialysis sessionand before administration of anticoagulantdrugs. We used a quantitative assay based onagglutination of latex microparticles coatedwith monoclonal antibodies specific forD-dimers, the STA-Liatest D-dimer test(Roche Diagnostica, Switzerland) with a cut-off level of 0.5 mg/l [6, 7]. Plasma D-dimerlevels were determined for the whole cohort.For further evaluation patients were dividedin two subgroups, according to the presence(group A) or absence (group B) of procoagu-latory risk factors. Malignancies, infectiousdiseases, severe congestive heart failure andatrial fibrillation, coagulopathies and surgi-cal interventions within the last month wereconsidered to predispose for elevated D-dimer levels, according to results of previousstudies and general clinical experience [1,5].The mean age of all examined patients(18 female/36 male) was 61.2 years (27.9–79.2), while it was 66.8 years (41.9–79.2) inthe subgroup with (n = 20) and 57.9 years(27.9–79.2) in the subgroup without predis-posing comorbidities (n = 34). 59% of allpatients had plasma D-dimer levels of0.5 mg/l or above. In the two subgroups, 16 of20 specimens were found to be superior to0.5 mg/l in the presence, and 16 of 34 in theabsence of predisposing factors, yielding aprevalence for elevated D-dimers of 80 and47%, respectively. This difference betweenthe two subgroups was considered statistical-ly significant with a p value of 0.023 by Fish-er’s exact test. With mean D-dimer levels of1.21 mg/l (95% CI 0.77–1.66 for group A)and 0.67 mg/l (95% CI 0.40–0.94 for groupB), respectively, the resulting mean differ-ence of 0.54 mg/l was also significant (two-tailed t test, p = 0.026).The present study confirms previous in-vestigations indicating a high prevalence ofprocoagulatory changes in hemodialysis pa-tients. Overall we observed an elevated plas-ma D-dimer rate of 59% about 44 h after thelast hemodialysis. 37% of the patients hadone or more procoagulatory risk factors, andelevated D-dimers were found in 80% ofthese cases. The diagnostic value of a D-dimer assay, e.g. for suspected acute VTE, iscertainly low in this special population. Asexpected, the presence of procoagulatory fac-tors was associated with a higher mean age.In the absence of predisposing factors(group B) the mean D-dimer level and theprevalence of elevated values (0.5 mg/l ormore) were both significantly lower com-pared with those of the predisposed patients(group A). With a prevalence of 47% suppos-ingly false positive D-dimer assays with re-gard to acute VTE, this subgroup of hemodi-alysis patients was surprisingly comparableto other, nonhemodialysis patient groups.Taking the reported high sensitivity of thisrapidly available, relatively inexpensive,diagnostic test into consideration, it may be