To the Editors: Murphy et al1 have postulated that the combination of 2 serial, normal, immature-to-total-neutrophil (I:T) ratios and a negative blood culture at 24 hours virtually rules out infection in a neonate. Of the 3154 neonates in their 1999 to 2008 cohort, 23 (0.73%) were diagnosed with true sepsis, 119 presumed sepsis (blood culture negative, I:T ratio abnormal). The difference between 1473 noninfected neonates and 1539 culture-negative and normal I:T-ratio neonates is not clear. Also, based on the choice of laboratory test that was used (I:T ratio) in combination with blood culture, the diagnostic standard needs to be reanalyzed. We agree with the authors that serial estimation of I:T ratio would be better than a single value, as is well documented for other conventional markers of neonatal sepsis such as C-reactive protein.2 I:T ratio is a time-consuming and labor-intensive test. Wide interreader differences of band neutrophil identification and I:T ratio are documented between evaluators of the same blood smear, thus diminishing the significance of the test.3 It would be worthwhile to know whether the 2 I:T ratios in the given study were determined by the same observer or by different people with variable hematologic experience. Also, were the 2 I:T ratios determined in isolation or as a component of several markers? In the latter event, was I:T ratio the best performer? In this study, 2 normal white blood cell count screenings and a negative blood culture at 24 hours had a very low positive predictive value (PPV; 8.8%). This low PPV is in accordance with previous literature,4 although we agree with the authors that their study was about determining which infants were likely not infected and thus more dependent on negative predictive value. But an early marker of neonatal sepsis should have not only high negative predictive value but also moderate PPV.5 The low specificity (51%) and PPV partially defeat the purpose of screening because 1473 of 1615 (91.2%) noninfected neonates were also exposed to unwarranted antibiotics. We evaluated expression of CD64 on neutrophils as an early indicator of neonatal sepsis and to help guide antibiotic treatment in comparison with conventional markers, including I:T ratio. It was a prospective case-control study on 90 neonates; 60 sick neonates and 30 healthy controls. We found alterations in I:T ratio to be significant in differentiating between sick neonates and controls (P = 0.037). However, I:T ratio could not distinguish culture-positive (n = 24) from culture-negative (n = 36) neonates (P = 0.189). I:T ratio remained normal in 29% of culture-positive neonates and was elevated in 52.7% of culture-negative neonates. This manifested as low positive likelihood ratio of 1.342 and negative likelihood ratio of 1.619. Of the 32 neonates with early-onset sepsis, 11 were evaluated within the first 24 hours of life. Klebsiellea species was isolated from 1 of 11 neonates (9%). I:T ratio was normal in 7 of 11 sick neonates (63.6%). According to Murphy et al, a second I:T ratio should have been done in all of them. This need was partially circumvented by using a combination of conventional markers, the results of which were available concurrently. Of the 7 of 11 sick neonates, conventional sepsis screen was positive in 4 neonates. Ideally, only the remaining 3 neonates would be model candidates for a repeat evaluation, not all 7. We agree with Murphy et al that a raised I:T ratio, despite a negative culture result, as occurred in 3 of our 11 sick neonates, should be considered as a presumptive evidence of infection and antibiotics be administered to them. Thus, either a combination or serial evaluation of a marker is likely to be useful than a single value. Sushant Soni, MB BS Neelam Wadhwa, MD Department of Pathology University College of Medical Sciences University of Delhi, Shahdra Rajive Kumar, MD Department of Laboratory Oncology Institute of Rotary Cancer Hospital AIIMS, Ansari Nagar M.M.A. Faridi, MD Department of Pediatrics University of Delhi, Shahdra Delhi, India