[Resistance to antimicrobials and combination therapy assessment in neonatal sepsis]

To estimate antibiotic resistance among bacterial strains isolated from patients with neonatal sepsis from January 1994 through December 1998, and to assess the predictive value of the checkerboard method for selecting treatment with combination antibiotic therapy in seriously ill patients.

The study of strain susceptibility was carried out by microdilution in broth, and the checkerboard method in broth trays was used to assess the efficacy of antibiotic combination therapy.

Fifty per cent of the strains of coagulase-negative Staphylococcus spp. and 37% of S. aureus strains, which are the main pathogens involved in neonatal sepsis, were resistant to methycillin. Of these strains, 94.5% were resistant to several antibiotics. Four strains of coagulase-negative Staphylococcus spp. and one strain of Enterococcus faecium were resistant to vancomycin (minimal inhibitory concentration = 128 micro g/mL). We noted a high probability of clinical efficacy (P < 0.01) with antibiotic combinations having a synergistic effect in vitro, whereas antagonic reactions were seen in close association with a highly significant probability of clinical therapeutic failure.

The main cause of neonatal sepsis in the Neonatal Intensive Care Unit at "América Arias" Hospital in Havana, Cuba, were strains of Staphylococcus resistant to methycillin and other antibiotics. In seriously ill patients (according to policy established at this institution), the first line of treatment is a combination of an aminoglycoside antibiotic and a beta-lactamic antibiotic. This study showed that the clinical efficacy of combination antibiotic therapy is not determined by the resistance pattern of the strain and that good clinical results can be attained, in all likelihood, when a synergistic effect can be shown in vitro.