Dexamethasone as adjunctive therapy in bacterial meningitis. A meta-analysis of randomized clinical trials since 1988

Abstract

Objective: To evaluate the effectiveness of dexamethasone in bacterial meningitis in the subcategories of causative organism and timing and nature of antibiotic therapy.

Data sources: MEDLINE, HEALTHLINE, and AIDSLINE were searched with the Medical Subject Headings "dexamethasone" and "meningitis" in any language. Bibliographies, conference abstracts, and the authors of identified studies were consulted.

Study selection: Randomized, concurrently controlled trials of dexamethasone therapy in childhood bacterial meningitis published from 1988 to November 1996 were selected. Of 16 studies identified, 5 were not eligible.

Data extraction: Data were extracted by means of standard outcomes in a protocol sent to all principal authors.

Data synthesis: Random-effects meta-analysis models were used to obtain summary estimates. As the incidence of severe hearing loss differed significantly by organism among control subjects, organism-specific estimates were used. In Haemophilus influenzae type b meningitis, dexamethasone reduced severe hearing loss overall (combined odds ratio [OR], 0.31; 95% confidence interval [CI], 0.14-0.69). Similar ORs were obtained after studies were stratified by the timing of administration of dexamethasone (before or with antibiotics vs later) or by type of antibiotic (cefuroxime vs other). In pneumococcal meningitis, only studies in which dexamethasone was given early suggested protection, which was significant for severe hearing loss (combined OR, 0.09; 95% CI, 0.0-0.71) and approached significance for any neurological or hearing deficit (combined OR, 0.23; 95% CI, 0.04-1.05). For all organisms combined, the pooled OR suggested protection against neurological deficits other than hearing loss but was not significant (OR, 0.59; 95% CI, 0.34-1.02). Outcomes were similar in studies that used 2 vs more than 2 days of dexamethasone therapy. Adverse effects were not significantly increased with dexamethasone except for secondary fever. The incidence of gastrointestinal tract bleeding increased with longer duration of dexamethasone treatment (0.5% in controls, 0.8% with 2 days of treatment, 3.0% with 4 days of treatment).

Conclusions: The available evidence on adjunctive dexamethasone therapy confirms benefit for H influenzae type b meningitis and, if commenced with or before parenteral antibiotics, suggests benefit for pneumococcal meningitis in childhood. Limiting dexamethasone therapy to 2 days may be optimal.