D-dimer can reflect severity of patients with community-acquired pneumonia and predict prognosis

Objective To investigate the utility of plasma D-dimer and fibrinogen (FIB) for the severity assessments and predicting the prognosis of patients with community-acquired pneumonia (CAP). Methods The clinical data of patients with CAP admitted to First Affiliated Hospital of Fujian Medical University were retrospectively analyzed. The patients were divided into Ⅰ-Ⅴ level groups according to pneumonia severity index (PSI), and they were divided into non-survivors and survivors according to 30-day prognosis. The data including gender, age, PSI score, platelets count (PLT), white blood cell count (WBC), D-dimer, FIB, and C-reactive protein (CRP) were compared among groups. The correlations between PSI score and D-dimer, CRP as well as FIB were analyzed by Spearman or Pearson correlation analysis. Receiver operating characteristic curve (ROC) was used to assess the prognostic value of these indicators. Results A total of 499 patients with CAP were enrolled with 298 male and 201 female, the average age was (63.4±17.8) years old, and the 30-day mortality was 6.4% (32/499). There were 77, 80, 104, 162 and 76 patients in PSI Ⅰ-Ⅴ level groups, and there were more male patients in PSI Ⅵ and Ⅴ level groups. There were no significant differences in PLT and FIB among the groups of different PSI levels, but the levels of WBC, D-dimer and CRP were significantly increased as PSI level increased from Ⅰ to Ⅴ (F1 = 3.810, χ22 = 102.361, F3 = 7.070, all P < 0.01). Compared with survivors, the non-survivors were elder (t = -4.773, P < 0.001) with lower PLT (t = 3.026, P = 0.003) and higher WBC, PSI score, D-dimer and CRP levels (t1 = -2.545, t2 = -8.421, Z3 = -6.947, t4 = -3.770, all P < 0.05). Plasma D-dimer levels in elderly patients (≥ 65 years old) were statistically higher than those in younger patients (< 65 years old; Z = -5.338, P < 0.01). It was shown by correlation analysis that PSI score was positively correlated with D-dimer and CRP (r values were 0.475 and 0.260, both P < 0.001), and no correlation was found between PSI score and FIB (r = -0.062, P = 0.170). The area under the ROC curve (AUC) for predicting 30-day death of PSI score, D-dimer and CRP was 0.858 [95% confidence interval (95%CI) = 0.802-0.914], 0.867 (95%CI = 0.812-0.922) and 0.732 (95%CI = 0.641-0.823). The combination of D-dimer and PSI score was better than any single indicator for predicting the prognosis with higher AUC up to 0.905 (95%CI = 0.867-0.944), all P < 0.001. The sensitivity and specificity for PSI in predicting 30-day death respectively were 78.1% and 82.4% with the cut-off of greater than 122, and those for D-dimer were 75.0% and 82.9% with the cut-off of greater than 2.10 mg/L, 50.0% and 84.4% for CRP with the cut-off of greater than 100.50 mg/L. Conclusions D-dimer could well reflect the severity of CAP and be a good indicator for predicting the prognosis. The combination of D-dimer and PSI might improve the accuracy in predicting prognosis.