Abstract
BACKGROUND: Patients with chronic obstructive pulmonary disease (COPD) have a high incidence of cardiovascular disease including thromboembolisms. Fibrin degradation products, like D-dimer, have been associated with death from all causes in healthy individuals and COPD patients. We aimed to determine the (i) association between D-dimer levels and all-cause mortality and time being alive and out of a hospital, (ii) possible modifying effect of anticoagulant treatment,, and (iii) distribution of D-dimer in patients with moderate to severe COPD. METHODS: Results of routinely measured stable phase D-dimer samples from COPD-outpatients at Copenhagen University Hospital - Herlev and Gentofte, COPD-outpatient clinic were collected using the Danish registries. These were used to examine whether COPD-patients with a D-dimer level in the upper quartile, had a higher risk of death from all causes within 365 days. RESULTS: In the unadjusted Cox proportional hazards regression we found an association between high D-dimer and all-cause mortality: Hazard ratio (HR): 2.3 (95% Confidence Interval (CI) 1.1-4.7). In the fully adjusted regression, the HR was 1.8 (CI 0.8-3.9). We did not find any interaction between D-dimer and anticoagulant or antiplatelet therapy. For the secondary outcome, proportion of days alive and out of hospital in 365 days (pDAOH), the unadjusted multiple linear regression had an association between high D-dimer level and pDAOH: -2.7% points (pp) (CI -3.9 pp - -1.5 pp), which was attenuated to -1,7pp (-2.9pp - -0.4pp) in the fully adjusted regression. CONCLUSIONS: In patients with moderate to severe COPD, patients with a high level of D-dimer were more likely to die; however, the signal was not strong in the adjusted analyses and our results do not support unselected risk stratification with D-dimer in COPD-outpatients.