Usefulness of admission hematologic parameters as diagnostic tools in acute pulmonary embolism

Abstract

The purpose of this study was to determine the role of red cell distribution width (RDW), neutrophil-lymphocyte ratio (NLR), and platelet-lymphocyte ratio (PLR) in the diagnostic phase of acute pulmonary embolism (PE). We screened 248 consecutive patients who were admitted to the emergency service with PE foremost in the differential diagnosis. Based on spiral computed chest tomography, the patients were divided into two groups. There were 112 confirmed cases of acute PE and 138 patients without PE. Blood samples were obtained within 2 hours of presentation and before starting any medication. There were no significant differences between the PE and the non-PE groups with respect to sex, age, frequency of disease, serum creatinine, sodium, and potassium (p > 0.05 for all). NLR, RDW, and PLR were higher in patients with PE than those without PE. High-sensitivity C-reactive protein, D-dimer, and troponin levels were also higher in patients with PE. RDW values were positively correlated with troponin levels (r = 0.147, p = 0.021). There were no correlations between RDW and NLR, PLR, or D-dimer. NLR had a highly positive correlation with PLR (r = 0.488, p < 0.001). In multivariate logistic regression analysis, troponin I, D-dimer, high-sensitivity C-reactive protein, and RDW were found to be independent predictors of PE [odds ratio (95% confidence interval) respectively: 5.208 (2.534-10.704), 1.242 (1.094-1.409), 1.005 (1.000-1.010), 1.175 (1.052-1.312)]. In receiver operating characteristic analysis of the patients in the study, RDW >18.9 predicted acute PE with a sensitivity of 20.7% and a specificity of 93.4%. In conclusion, RDW can be considered useful as a diagnostic measure for patients with suspected acute PE.

Keywords: Neutrophil lymphocyte ratio; Platelet lymphocyte ratio; Pulmonary embolism; Red cell distribution width; Troponin I.

Figure 1

Receiver operating characteristic curve analysis of study patients for predicting pulmonary embolism. The area under the curve is 0.559. Criterion is over 18.9 with 20.7% sensitivity and 93.4% specificity. Standard error = 0.0369, 95% confidence interval = 0.495–0.622, z statistic = 1.602, p = 0.1091.