A bacteriological and epidemiological study of bacterial meningitis occurri
ng in infants under one year of age was performed from September 1981 to Ju
ne 1997 in Niamey, a city of 575,000 residents, located within the African
meningitis belt. Cases of meningitis were defined either by culture of the
cerebrospinal fluid (CSF), specific antigen agglutination, staining or cell
counts of the CSF. Over the 16 years involving both epidemic and non epide
mic periods, 1 481 infant's CSF were analysed representing 20% out of the t
otal CSF samples. The average of annual incidence rates was 511.4 cases per
100 000 infants under one year Haemophilus influenzae b represented 35.1%
of the cases, Streptococcus pneumoniae 26.3% and Neisseria meningitidis 17.
6%. The other bacteria represented 5.5% and, for 15.5% out of the analysed
CSF, the causative agent was not identified. The average annual mortality r
ate was 146.9 deaths for 100 000 infants under one year The specific case f
atality rates were 43% for H. influenzae b, 58.9% for S. pneumoniae and 17.
8% for N. meningitidis. This study showed that in Niamey, as in the rest of
the meningitis belt S. pneumoniae and H. influenzae b were the main causes
of bacterial meningitis occurring in infants under one year. However; the
specific incidence of N, meningitidis was identical for every age group bet
ween 0 and 20 years, and varied from 45 per 100,000 during non epidemic yea
r to 550 per 100,000 during epidemic year. Immunisation with conjugate vacc
ines, particularly anti-Haemophilus vaccine, appears to be the best prevent
ive measure. The systematic use of ceftriaxone in infants, during meningoco
ccal meningitis either epidemics or not, is highly recommended.