Comparison of bacterial antigen test and gram stain for detecting classic meningitis bacteria in cerebrospinal fluid.

The bacterial antigen test (BAT) screens cerebrospinal fluid (CSF) or other body fluids for antigens of classic bacterial meningitis pathogens (i.e., Streptococcus pneumoniae, Haemophilus influenzae type b [Hib], group B Streptococcus species, Neisseria meningitidis, and Escherichia coli K1) (7). The utility of the BAT has been questioned in several published reports (2, 5, 6, 8, 12, 15). Furthermore, changes in the epidemiology of bacterial meningitis related to the pneumococcal, meningococcal, and Hib vaccines have likely affected the positive predictive value of the BAT (1, 3, 9, 10, 12, 16). Consequently, many laboratories have discontinued use of the BAT. However, in the absence of clear recommendations, some laboratories may continue to offer the BAT due to conflicting evidence in the literature (5, 11, 14) or the possibility that the BAT may be valuable in diagnosing bacterial meningitis caused by culture-negative organisms (8). We retrospectively analyzed 918 CSF specimens from adults and children tested at our institution (Mayo Clinic, Rochester, MN) with the BAT (Wellcogen bacterial antigen kit; Remel, Inc., Lenexa, KS), Gram stain, and culture between January 2000 and March 2009. We further analyzed a subset of these in which at least one of the following criteria was met: (i) positive BAT result, (ii) positive Gram stain result consistent with a classical bacterial meningitis pathogen, or (iii) classical bacterial meningitis pathogen identified by culture from CSF. Fortytwo cases were identified. Results of the BAT and Gram stain were compared (Table 1). In 9 of the culture-negative CSF specimens, the culture of samples (blood or ear drainage) from other sterile body sites yielded classic meningitis organisms. For specimens from CSF, blood, or ear drainage with positive culture results, the results of the BAT and Gram stain were analyzed by organism (Table 2). For the 3 specimens in which the BAT was positive and the Gram stain was negative, the culture was also positive, rendering the BAT redundant. None of the culture-negative cases were missed by the Gram stain, again making the BAT unnecessary. To increase the sensitivity of the CSF Gram stain, a cytocentrifugation procedure was implemented in January 2007 (4, 13). All five (12%) Gram stain-negative specimens with classic bacterial meningitis pathogens present (including those in which the BAT was positive) were tested prior to the implementation of the cytocentrifugation procedure. Had these Gram stains been performed using the more sensitive cytocentrifugation protocol, the results may have been positive. Kiska et al. (8) proposed screening criteria based on CSF indices to decrease unnecessary use of the BAT. Their findings suggested that in cases of previous antibiotic treatment in which culture and Gram stain results may be negative, the BAT still may play a role in the diagnosis of bacterial meningitis. However, the findings of our large, retrospective study indicate that the BAT provides no substantial benefit beyond the Gram stain in screening for bacterial meningitis, even in cases with culture-negative results. Although the BAT may be faster than culture in the Gram stain-negative cases, routine antimicrobials are recommended until all test results are available. This moderate benefit must be contrasted with the risk of a false sense of security with a false-negative BAT, or the risk of a false-positive BAT, which would complicate care decisions.