Abstract
No data exist on the usefulness of the delta neutrophil index (DNI) to discriminate pulmonary tuberculosis (PTB) from community-acquired pneumonia (CAP). We performed a retrospective cohort study involving patients with PTB (n = 62) and CAP (n = 215), and compared their initial DNI levels. The median DNI values were 0% (interquartile ranges [IQR] 0-0.2%) and 1.6% (IQR 0.7-2.9%) in PTB and CAP, respectively, which was significantly lower in PTB patients (P < 0.001). Sixty-nine percent of patients with PTB had DNI value of 0%; however, only 15% of patients with CAP had 0% DNI. The discriminatory power of the DNI for diagnosing PTB was high with 89% sensitivity and 67% specificity at a DNI cut-off ≤ 1.0% (area under the curve, 0.852). The diagnostic sensitivity and negative predictive value (NPV) for PTB were 89% (55/62) and 95% (145/152) at the DNI cut-off ≤ 1.0%, respectively, and in multivariate analyses after adjusting for other factors (smoking, no fever, upper lobe involvement), DNI ≤ 1.0% remained significant (odds ratio, 15.265; P < 0.001). We demonstrated that the DNI was lower in PTB compared with CAP, and an initially elevated DNI (>1.0%) may be useful to rule out the possibility of PTB due to its high NPV.