I read with great interest the article by Chalupa et al. [1]. The authors have conducted a prospective hospital-based study comparing the potential biomarkers used to diagnose infection in subjects with bacterial and viral infection. The authors concluded that only procalcitonin (PCT) demonstrated superiority over the routine parameters in the differentiation of bacterial from viral infections. Based on this, the authors strongly support the use of the PCT serum level as a routine biomarker of bacterial infection. However, there a few points that need to be highlighted in order for the result to be more meaningful. It is not clear why the authors have used bacterial infection of the lungs (community-acquired bacterial pneumonia [CABP]) and urinary tract (urosepsis) to compare with the viral infections in sites other than these (i.e., of the central nervous system [CNS], liver, etc.). For the results to be meaningful and to supplement clinical decision regarding whether to start or withhold antibiotics in an individual case scenario, the authors could have compared the two infections (either bacterial or viral) at these two sites (either CABP or urosepsis) only. CABP would have been better, as urosepsis is not viral. Moreover, though the authors strongly support the use of PCT as a biomarker for bacterial infection, this issue is still debated [2, 3]. In the recent 2009 H1N1 pandemic, a higher PCT threshold was required to discriminate between bacterial and viral pneumonia, suggesting that severe viral pneumonia alone may elevate PCT [4, 5]. In another important study, 22/23 of patients with community-acquired pneumonia (CAP) with positive serology (three being atypical pneumonia and the rest were viral) were given antibiotics per their PCT algorithm [6]. Patients with atypical pneumonia (due to Mycoplasma and Chlamydia) may have low PCT levels [7]. Even patients who were bacteremic from various sources can have lower PCT levels (\0.5 ng/ml), as found in a recent study [8]. As PCT levels are documented to distinguish viral from bacterial meningitis [9, 10], the authors could have compared the PCT levels between viral or bacterial CNS infection (meningitis or encephalitis) instead of CABP or urosepsis. Another important observation that is not even included in the discussion is that, if PCT is a sensitive indicator of bacterial infection, then why there is a significant difference in the levels of PCT between patients with CABP and urosepsis (as both are presumed to be bacterial), as shown in Table 3 of the paper. As the isolated value of PCT in discriminating between bacterial and viral infection is still not clear, advanced molecular diagnostics should possibly be used to validate the ability of the serum PCT in this regard. In addition, such analysis should be able to clarify PCT responses in mixed viral and bacterial infections, which are detailed elsewhere [11]. If both pathogens (viral and bacterial) are identified, PCT levels should assist in the interpretation of the result also (e.g., colonization with one organism ersus infection with another and vice versa). In spite of all these limitations, PCT serum levels have several advantages when compared with the erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level as a biomarker for invasive bacterial disease, and I congratulate the authors for adding further data to support the use of PCT in the diagnosis of bacterial infection. R. R. Das (&) Department of Pediatrics, All India Institute of Medical Sciences (AIIMS), New Delhi 110029, India e-mail: dr_rashmipgi@yahoo.com