University of Michigan
Department of Pediatrics
Evidence-Based Pediatrics Web Site

Question

• In pediatric patients with bacterial meningitis in developing countries, does the administration of adjunctive dexamethasone prior to parenteral antibiotics help improve outcomes (morbidity and mortality)?

Clinical Bottom Lines

1. In this, the largest placebo-controlled double-blind randomized trial of adjunctive dexamethasone therapy for bacterial meningitis, dexamethasone did not improve outcomes (morbidity and neurologic sequelae including hearing loss) in children with bacterial meningitis in Malawi, a developing nation in Africa.¹

Summary of Key Evidence

1. Prospective double-blind, placebo-controlled randomized controlled trial including 598 children 2 months to 13 years old admitted to Queen Elizabeth Central Hospital, Blantyre, Malawi with bacterial meningitis.¹
2. 307 children (51%) were randomly assigned to dexamethasone and 295 (49%) to placebo.
3. Children were randomly allocated to receive dexamethasone (0.4 mg/kg) or placebo (water) q12h X 48h. Vials were identical, labeled only with study numbers, neither the childrens' health care providers nor their parents nor outcomes assessors knew which arm the child had been assigned to.
4. Inclusion criteria: age 2 months to 13 years old and one of: admission LP with > 100 WBCs, mostly granulocytes, or positive gram stain, or culture grew bacteria. Initial enrollment was based on clinical story and initial csf hazy or cloudy.
5. Exclusion criteria: If csf later did not meet definition of meningitis as described above, child was then removed from study. Children also were excluded if they had received broad-spectrum antibiotics up to 24 hours prior to admission.
6. Outcome measures: death, full recovery, hearing loss or reduction, or various neurologic deficits.
7. Most cases of meningitis caused by four types of bacteria: S pneumoniae (238, 40%), H influenzae type b (67, 11%), N meningitidis (67, 11%), Salmonella species (29, 5%). >1/3 had received antibiotics prior to admission, mainly cotrimoxazole or penicillin, mostly orally.
8. Subgroup analysis was performed according to causative agents.
9. Dexamethasone didn't improve overall mortality (31% in both groups, RR 1.00, p = 0.093), development of neurologic sequelae (28% in both groups, RR 0.99, p = 0.97) at discharge or on 6 month followup, likelihood of discharge after full recovery, or the proportion/degree of hearing disorders in children with bacterial meningitis of all causes.
10. Children with steroids were four times more likely than placebo to develop hydrocephalus.
11. Second-line therapy was needed and imaging abnormalities on ultrasound were found more often in kids with H influenzae and Salmonella meningitis.
12. Mortality: Salmonella > S pneumoniae > H influenzae >> N meningitidis.
13. At follow-up, the two groups had similar number of kids with neurological disorders and hearing loss.
14. Mortality was much lower in patients who had received ceftriaxone v. penicillin/ chloramphenicol alone; in latter group, mortality/morbidity not affected by steroids, in ceftriaxone group, lower incidence of hearing loss (and death).
15. Death and neurologic sequelae were associated with age (inversely), malnutrition, low coma score, HIV-1 seropositivity, and by causative organism.
16. HIV-1 seropositive children were more malnourished, more likely to be left with neurologic sequelae in cases of S pneumoniae meningitis, more likely to have recurrent meningitis, and had higher case fatality rates.
17. As reported in other studies, there were no deleterious side effects recorded related to the dexamethasone.

Additional Comments

• Overall mortality in this study decreased compared to general expectations in this region (but much greater than in US) likely secondary to increased monitoring and careful management on research ward.
• Several factors influence the effect of dexamethasone on meningeal inflammation: timing of dose, severity of illness prior to diagnosis, type of antibiotic, and the concentrations and growth phase of organisms in csf.²
• One crucial difference in this study is that their first-line antibiotics were penicillin and chloramphenicol. Beta-lactam antibiotics cause rapid lysis of susceptible bacteria. Dexamethasone doesn't decrease the damage that has already developed secondary to the inflammation. Bactericidal antibiotics that work by inhibiting protein synthesis likely don't cause as rapid cell lysis as beta-lactam antibiotics such as ceftriaxone. The only benefit of dexamethasone in meningitis is to modulate the secondary meningeal inflammation following initial dose of parenteral antibiotics and the principal clinical benefit is reduction of hearing loss.²
• Another crucial difference in this study is the much higher incidence of patients in the population who were HIV-1 positive.
• Other important differences: access to healthcare, time of presentation, increased mortality, increased likelihood of having received prior antibiotics, few ICUs, lab support basic, increased incidence of anemia, and immunization status.
• Immunizations (including those against H influenzae type b) have not yet been included in Extended Programme of Immunisation.
• Among the children with H influenzae type b meningitis, per the study protocol, these children received ceftriaxone. Steroids did seem to help decreasing hearing loss in patients who had received ceftriaxone, which would be more applicable to our population. Current AAP guidelines recommend steroids in H influenzae meningitis, and possibly for S pneumoniae and meningococcal disease as well.
• Appropriate measures were taken to enroll an adequate number of patients to power the study to detect differences in the two treatment groups (presumed 20% drop-out rate and 20% mortality within 48 hours), however, study felt slightly short of goal numbers secondary to very high mortality rates (20% speculated, 31% actual).
  

Citation

1. Molyneux EM, Walsh AL, Forsyth H, et al. Dexamethasone treatment in childhood bacterial meningitis in Malawi: a randomised controlled trial. Lancet 2002; 360: 211-218.
2. McCracken GH, Jr. Rich nations, poor nations, and bacterial meningitis. Lancet 2002; 360: 183-186.

CAT Author: Joyeeta G. Dastidar, MD

CAT Appraisers: John G. Frohna, MD

Date appraised: March 1, 2006