Blood culture-proven neonatal sepsis and resistance trends in a tertiary NICU: A retrospective study

INTRODUCTION

Neonatal sepsis remains a major contributor to neonatal mortality in India, accounting for nearly one-fifth of all neonatal deaths. Its incidence is estimated between 2.7 and 17/1,000 live births in community-based studies and can be as high as 30/1,000 in hospital settings.^[1-3] Early-onset sepsis (EOS) is typically associated with vertical transmission from the mother, whereas late-onset sepsis (LOS) is more often linked to nosocomial or community-acquired infections.

The growing burden of multidrug-resistant (MDR) pathogens in neonatal intensive care units (NICUs) poses a significant challenge for effective management. Although empirical antibiotic therapy is routinely initiated in suspected cases, the inappropriate and frequent use of broad-spectrum agents has accelerated antimicrobial resistance.^[4,5] Among the limited therapeutic options, carbapenems continue to demonstrate relatively high efficacy; however, increasing dependence on these agents may further drive resistance.^[6,7] In this context, developing and adhering to unit-specific antibiograms is essential to guide rational empirical therapy, optimise clinical outcomes and curb resistance trends.

MATERIAL AND METHODS

RESULTS

During the 3-month study period, a total of 2,796 neonates were delivered, of whom 1,013 required admission to the NICU. Among these, 542 were clinically suspected of sepsis, and 73 (13.4%) were confirmed to have blood culture-positive sepsis [Figure 1].

Close

Figure 1:

Study flow chart. NICU: Neonatal intensive care unit

Export to PPT

The majority of affected neonates were preterm (83.6%) and female (61.6%). EOS accounted for 57.5% of cases, while LOS was observed in 42.5% [Table 1].

Klebsiella species emerged as the predominant pathogen, isolated in over half of the cases (50.7%), followed by Acinetobacter (23.3%) and Staphylococcus aureus (16.4%). Less common isolates included coagulase-negative staphylococci (5.5%) and Enterococcus species (4.1%) [Table 2].

Antibiotic susceptibility testing revealed that carbapenems remained highly effective, with meropenem and imipenem showing sensitivity rates of 96.4% and 96.2%, respectively. Among Gram-positive isolates, linezolid demonstrated the highest efficacy (94.7%). In contrast, aminoglycosides such as gentamicin showed only moderate sensitivity (50%), while amikacin (29.7%) and ciprofloxacin (44%) had poor activity. Cephalosporins performed worst, with sensitivity rates<3%, highlighting widespread resistance [Table 3].

Clinical outcomes demonstrated that 59 neonates (80.8%) survived and were discharged, whereas 14 (19.2%) succumbed to sepsis, yielding a case fatality rate of nearly one in five among culture-proven cases.

DISCUSSION

The present study demonstrates that Gram-negative organisms were the predominant cause of culture-proven neonatal sepsis, accounting for nearly three-fourths of all isolates (74%). Klebsiella spp. (50.7%) and Acinetobacter spp. (23.3%) emerged as the leading pathogens. These findings are consistent with reports from NICUs across South Asia, where Gram-negative organisms – particularly Klebsiella and Acinetobacter – have been identified as the major contributors to neonatal sepsis and mortality.^[1,2] Their predominance reflects both vertical transmission from colonised mothers and horizontal transmission within resource-constrained hospital environments.^[3]

Carbapenems showed excellent efficacy in our study, with sensitivity rates exceeding 96% for both meropenem and imipenem. This reaffirms their role as the cornerstone of empiric and targeted therapy for neonatal sepsis in high-burden NICU settings.^[4] However, the reliance on carbapenems raises concern, as inappropriate or prolonged use can accelerate the emergence of carbapenem-resistant Enterobacteriaceae (CRE), a phenomenon increasingly reported in low- and middle-income countries.^[5,6] Conversely, the strikingly poor performance of cephalosporins (cephalosporins strikingly poor performance of cephalosporins-fluoroquinolones mirror global resistance patterns, highlighting the diminished utility of these once-standard agents.^[7] For Gram-positive pathogens, linezolid maintained high sensitivity (94.7%), corroborating international observations that it remains a robust therapeutic option for MDR Gram-positive sepsis.^[8]

The observed case fatality rate of 19.2% underscores the severity of neonatal sepsis and the persistent challenge of reducing mortality in this vulnerable population. Prematurity and infections due to MDR organisms were key contributors to adverse outcomes, in line with global evidence that immature immune responses, coupled with limited therapeutic options, significantly worsen prognosis.^[9,10] The distribution of pathogens by timing of sepsis also merits attention: Coagulase-negative staphylococci (CONS) were more common in LOS, reflecting nosocomial acquisition,^[11] whereas Enterococcus spp. in EOS likely indicates vertical transmission from maternal genital or gastrointestinal flora.^[12]

These findings carry important clinical and public health implications. First, there is a pressing need to integrate rapid diagnostic tools, such as molecular assays or point-of-care culture techniques, to enable the timely initiation of targeted therapy.^[13] Second, strengthening antimicrobial stewardship programs within NICUs is crucial to rationalise antibiotic use, prevent overreliance on carbapenems and slow the march toward pan-resistance.^[14] Third, enhanced infection prevention and control strategies – including strict hand hygiene, aseptic delivery practices and judicious use of invasive devices – are essential, particularly for preterm neonates who face the highest risk.^[15] Ultimately, the development of unit-specific antibiograms and periodic resistance surveillance can guide empiric therapy more effectively, thereby improving clinical outcomes.^[16]

In conclusion, our study reinforces the shifting microbial landscape of neonatal sepsis toward Gram-negative dominance, highlights the precarious dependence on carbapenems and emphasises the urgent need for stewardship and preventive strategies. Addressing these challenges is vital for reducing the burden of sepsis-related mortality in neonatal populations, especially in resource-limited settings.

CONCLUSION

Neonatal sepsis in our setting was predominantly caused by Gram-negative organisms, with carbapenems remaining the most effective therapeutic option. High mortality was observed among preterm infants and those with MDR infections. These findings underscore the need for rapid diagnostics, antimicrobial stewardship and strengthened infection prevention strategies to improve outcomes in this vulnerable population.