Comparison of ampicillin plus gentamicin vs. penicillin plus gentamicin in empiric treatment of neonates at risk of early onset sepsis

Aim: We aimed to compare the clinical efficacy of ampicillin (AMP) vs. penicillin (PEN) both combined with gentamicin in the empirical treatment of neonates at risk of early onset neonatal sepsis (EOS). Methods: We performed an open label cluster randomized equivalence study in both Estonian neonatal intensive care units, including neonates with suspected EOS, aged less than 72 h. Primary end-point was clinical failure rate, expressed by need for change of antibiotic regimen within 72 h and/or 7-day all cause mortality. Bowel colonization was followed with biweekly perineal swab cultures. Results: Incidence of proven EOS was 4.9%. Among neonates receiving AMP (n = 142) or PEN (n = 141) change of antibiotic regimen within 72 h (10/142 vs. 10/141; OR 1.02; 95% CI 0.40-2.59), 7-day mortality (11/142 vs. 14/141; OR 0.76; 95% CI 0.33-1.75) and over-all treatment failure (20/142 vs. 20/141; OR 1.01; 95% CI 0.52-1.97) occurred at similar rates. The only differences in gut colonization were lower number of patients colonised with enterococci, S. aureus and AMP resistant Acinetobacter spp. in AMP and lower number of those with S. haemolyticus and S. hominis in PEN arm. Conclusions: AMP and PEN combined with gentamicin have similar effectiveness in the empiric treatment of suspected neonatal EOS.