Background and aims: Blood culture (BC) remains gold standard for the evaluation and diagnosis of neonatal sepsis. Time when BC becomes positive and the type of microorganism isolated are crucial in deciding the antimicrobial management. Likely pathogenicity of organisms growing in BC could potentially be predicted based on the "time to positivity" (TTP). We aimed to estimate the predictive value of isolating a likely pathogenic organism depending on TTP; evaluate the aetiological trend and neonatal mortality rate due to culture-proven neonatal sepsis for over a decade and verify whether the application of a "36 hour rule" to discontinue empiric antibiotics in well newborn infants with negative BC would be safe.Methods: Retrospective review of BC results over a 14-year period from a regional neonatal unit in Ireland. Laboratory data were independently extracted in relation to BC results from the laboratory information management system (LIMS-iLAB). Neonatal mortality data were collected from multiple sources. Statistical analysis included logistic regression, chi-square, and Mann-Whitney U-test.Results: Over a 14-year period 11,432 neonatal BC specimens were incubated of which 605 (5.3%) turned positive. Overall, the commonest organism grown was coagulase-negative Staphylococcus (CoNS), 416 (68.8%). Main pathogenic organisms were Staphylococcus aureus 23 (3.8%), Enterococcus spp. 22 (3.6%), E.coli 21 (3.5%), group B Streptococcus (GBS) 18 (3.0%), and Klebsiella species 9 (1.5%). Gram-negative organisms had the shortest TTP, with Klebsiella spp. having a median TTP of 10 h and E. coli 11 h. For Gram-positives, GBS had a median TTP of 12 h, Enterococcus species 14 h, with S. aureus growing at a median time of 15 h. All of the Klebsiella spp. and other Coliforms were detected within 24 h, with, 95.2% of E.coli, 94.4% of GBS, 95.5% of Enterococci, and 95.7% of S. aureus, flagging positive in 24 h. Using logistic regression the omnibus test of the coefficients in the resulting model was significant (p < .001). Our observed coefficient (β) for TTP was 0.144; shorter the TTP higher was the likelihood of isolating a pathogenic organism, with an odds ratio (OR) of 1.155. We also report a relatively low blood culture proven sepsis-specific neonatal mortality rate of 0.403/1000 live births and in all such instances observed TTP was less than 24 h.Conclusion: Duration of this study exceeds that of most of the neonatal blood culture TTP analysis published to-date. A shorter TTP is an important adjunct to suggest the growth of a pathogenic organism while managing suspected neonatal sepsis. TTP if < 24 h per se would not necessarily confirm the growth of a highly pathogenic organism; however, if a positive growth is likely to happen for a significant neonatal pathogen, in more than 98% the TTP would be within 24 h. This offers the clinician more of negative predictive value than a positive one; when there is no growth in BC. Our observation on TTP reiterate the National Institute of Health and Care Excellence (NICE) guideline of discontinuation of empiric antibiotics after 36 h in and clinically well and BC negative newborn infants.
Keywords: Antibiotics; ELBW; NICU; VLBW; blood culture; neonatal sepsis; time to positivity (TTP).