Hospital-acquired neonatal bacterial meningitis: the impacts of cefotaxime usage on mortality and of amikacin usage on incidence.

All cases of bacterial meningitis in the neonatal unit at King Edward VIII Hospital, Durban for the period 1 January 1981 to 31 December 1987 were reviewed. In particular, we looked at the impact of cefotaxime on mortality rates and amikacin on the incidence of hospital-acquired Gram-negative bacillary (GNB) meningitis. Klebsiella was found to be the commonest cause of neonatal meningitis, followed by Escherichia coli and Streptococcus agalactiae. Eighty-four per cent of all cases of GNB meningitis presented more than 3 days after birth, with the vast majority being caused by gentamicin-resistant Klebsiella. A decline in the incidence of meningitis from 1.27/1000 live births in 1981 and 0.95/1000 for the period 1981-1986 to 0.22/1000 live births in 1987, with no cases of Klebsiella meningitis being seen in that year, coincided with the exclusive use of amikacin as the parenteral aminoglycoside in place of gentamicin in the unit after August 1986. The initial decline in the incidence of meningitis from 0.93/1000 in 1985 to 0.46/1000 in 1986 was attributed to the introduction in 1985 of strict hand disinfection measures to prevent cross-infection in the unit. The case mortality rate (CMR) fell from 0.65 for the period 1981-1984 to 0.42 for the period 1985-1987, and we believe this was largely a result of the introduction of cefotaxime in 1984 as first-line therapy for GNB meningitis, together with better patient care facilities.(ABSTRACT TRUNCATED AT 250 WORDS)

Meningitis caused by vancomycin-resistant Leuconostoc sp.

A rare case of purulent meningitis caused by vancomycin-resistant Leuconostoc sp. is reported. The patient was treated successfully with penicillin G, and there were no neurological sequelae. We recommend that all gram-positive cocci isolated from cerebrospinal fluid and blood be fully identified so as not to confuse Leuconostoc spp. with more commonly isolated pathogens such as pneumococci and other alpha-hemolytic streptococci.

Sensitivity and specificity of a latex agglutination test for detection of cryptococcal antigen in meningitis.

A cryptococcal latex agglutination test (Crypto-La; International Biological Laboratories, Canbury, New Jersey, USA) was evaluated for its ability to detect cryptococcal antigen in cerebrospinal fluid (CSF) specimens obtained from black patients with a clinical diagnosis of meningitis. Of the 445 Gram-stained and bacterial culture-negative CSF specimens routinely tested for cryptococcal antigen, 34 (7,6%) were positive. With the exception of 1 false-positive result, the remaining 33 specimens were obtained from 12 patients in whom the diagnosis of cryptococcal meningitis was confirmed by the isolation of Cryptococcus neoformans. The Crypto-La test gave false-positive results on 1% (4/384) of control CSF specimens tested. Nonspecific agglutination reactions were observed with 1,6% (13/829) of all CSF specimens and 22% (10) of sera tested. The ethylenediaminetetra-acetic heat-extraction method proved reliable in eliminating false-positives and nonspecific agglutination reactions in CSF and serum specimens.

Three latex agglutination tests compared with gram staining for the detection of bacteria in cerebrospinal fluid.

Three commercially available latex agglutination tests and the Gram stain were evaluated for their ability to detect Haemophilus influenzae, Streptococcus pneumoniae and Neisseria meningitidis in cerebrospinal fluid (CSF) specimens from 160 patients suspected of having meningitis. Bactigen detected 86% of all culture-positive CSF specimens, whereas Wellcogen detected 80%, Bio Merieux 59% and Gram stain 94%. Bactigen and Wellcogen detected 84% of cases of meningitis caused by H. influenzae, whereas Bio Merieux detected 53% and Gram stain 90%. Of the 27 cases of Strept. pneumoniae meningitis, all were detected by Gram stain, 96% by Bactigen, 89% by Wellcogen and 74% by Bio Merieux. For N. meningitidis, which included four isolates of group B and one group W135, detection was 80% with the Gram stain, 40% with Bactigen and 20% with Wellcogen. False-positive results were observed with Bactigen and Bio Merieux respectively on 5% of culture-positive CSF specimens for organisms other than the test organisms. With the exception of faulty batches of individual kits of Bactigen (H. influenzae B) and Wellcogen (Strept. pneumoniae) respectively, false-positives were not obtained on the 63 culture-negative CSF samples. Bactigen and Wellcogen were equally efficacious in detecting H. influenzae B. Bactigen was marginally superior to Wellcogen for the detection of Strept. pneumoniae. Bio Merieux was the least sensitive of the three latex agglutination kits tested. The Gram stain was found to be at least as sensitive or superior to latex agglutination for detecting H. influenzae B, Strept. pneumoniae and N. meningitidis in CSF specimens from patients with bacteriologically proven meningitis.