RESUMO
OBJECTIVES: It's a retrospective study in order to determine the epidemiology of neonatal bacterial infection and to evaluate the efficiency of the antibiotic protocol in University Teaching Hospital in Dakar. MATERIAL AND METHODES: From January 1st 1997 to December 31st 1998 we have registered 7461 live births, samples of blood are taken from 2312 new-born baby and they received antibiotherapy (beta-lactamine + gentamycin) at the first day based on infections risk evaluated by anamnestic criterias. The treatment is seven to one days long, the antibiotic was adapted according to the antibiogram result. RESULTS: The neonatal infection diagnosis is confirmed in 246 cases, about 33 per 1000 live births or 10.6% of newborn babies having on antibiotherapy. Most current risk factors are premature rupture of membranes (85%) and neonatal suffering (87.8%). Isolated gerras are: Klebsiella pneumoniae (61.5%), Enterobacteria (11.5 Staphylococcus (8.7%), colibacille (6%), Streptococcus (5.5%), Enterococcus (4.1%) and Pseudomonas (2.7%). Most of these germs are resistant to antibiotics currently used in first intention (ampicillin, cefotaxim, gentamycin), in particularly 95% of Klebsiella. Most efficient antibiotics are amikacin, colistin, ceftriaxon and ciprofloxacine. Deaths occurs in 48 cases with 36 in early neonatal period, 79% of mortality rate related to infection by Klebsiella. CONCLUSION: First intention antibiotherapy must be always adapted to the bacterial ecology evolution and must be more selective by using major infections risk factors. We promote early infection diagnosis by using biologic markers which reference is represented by C Reactive Protein.