Steroids for Adult Bacterial Meningitis

March 2008

Steroids for Adult Bacterial Meningitis

Search Strategy: Starting with the Cochrane Collaboration, the top-tier EBM-resource, you search “meningitis” and obtain 22 results, the first of which is entitled “Corticosteroids for acute bacterial meningitis”. You select this Cochrane review which provides you with the additional evidence below.

Being a motivated self-learner, you attended the Society for Academic Emergency Medicine 2007 Clinical Consensus Conference in Chicago. Knowledge Translation (KT) was the topic, but after listening to experts from around the world discuss the definition of, supporting research, and science-fiction-like informatics which constitute “Knowledge Translation”, you return home unconvinced about the concept of moving from evidence to action. How does one implement KT-initiatives within an institution? Who defines the “best-evidence”? Is this a physician-led initiative or an administrative driven process? Is KT a science or is it a philosophical approach to clinical medicine?

Confused, but still motivated, you decide to initiate a novel KT project within your institution. As a four-year residency, you have access to many keen academic minds so you enlist all of them. The second year residents are tasked with identifying an institutional gap between unequivocal research evidence and current clinical practice. After extensive debate and multiple literature searches they select steroids in acute adult bacterial meningitis. To justify their selection, they perform a one-year retrospective chart review using pre-established methods to document that steroids are not being used as indicated among your patients. First year residents are next tasked with identifying and quantifying leaks within a theoretical information pipeline using Survey Monkey to identify why steroids are not being employed at your institution. Next, third year residents conduct a national survey of other institutions to identify successful models currently using steroids in meningitis. The results of this novel evidence-to-action initiative are the topic of this Journal Club.

PICO Question

Population: Adult ED patients with suspected acute bacterial meningitis

Intervention: Steroid administration in addition to usual clinical care

Comparison: Usual clinical care

Outcome: Death, post-meningitis neurological complications


First years: Corticosteroids for acute bacterial meningitis, Cochrane Database of Systematic Reviews 2007; Issue 1. Art. No. CD004405. DOI: 10.1002/14651858.CD004405.pub2 (

Second years: Dexamethasone in adults with bacterial meningitis, NEJM 2002; 347: 1549-1556. (

Third years: Trial of dexamethasone treatment for severe bacterial meningitis in adults, Intensive Care Med 1999; 25: 475-480. (

Fourth years: Dexamethasone in Vietnamese adolescents and adults with bacterial meningitis, NEJM 2007; 357: 2431-2440. (


Article 1: Corticosteroids for Acute Bacterial Meningitis: Cochrane Database of Systematic Reviews, 2007, Issue 1, Art. No.: CD004405. DOI: 10.1002/14651858. 2

Article 2: Dexamethasone in adults with bacterial meningitis NEJM 2002; 347:1549-1556

Article 3: Trial of Dexamethasone Treatment for Severe Bacterial Meningitis in Adults, Intensive Care Med 1999; 25: 475-480

Article 4: Dexamethasone in Vietnamese Adolescents and Adults with Bacterial Meningitis, NEJM 2007; 357: 2431-40

KT Presentation

Bottom Line

Here is a summary of the methods we used to illustrate and implement a Knowledge Translation Project within our residency program. Next we’ll briefly summarize our findings and then discuss the leaks within the KT-pipeline along with our proposed “plugs” to those information leaks.


  1. Identify a gap between best-evidence practice and local clinical activity via a chart review.
  2. Identify hypothetical leaks in the information pipeline between evidence and action with verification via a survey of BJH EM resident and attending physicians.
  3. Contrast local practice patterns with national standards via a separate survey using contacts at academic hospitals throughout the US and Canada provided by several PGY-III physicians.
  4. Propose a one-year plan to implement practice change locally.


KT Gap Identified:

Failure to use adjuvant steroid therapy in adult bacterial meningitis.

Summary of Supporting Evidence (from the attached critical appraisals):

In adults, dexamethasone (10mg IV q6 hours for 4-days) in acute bacterial meningitis save lives (NNT=10) and prevent long-term neurological sequelae whether administered before or after antibiotics. Except for low-income countries, steroids in children reduced hearing loss (NNT=23) regardless of the causative organism. The benefit of dexamethasone are most pronounced with S pneumonia meningitis (NNT = 5 to prevent one-death). No significant increase in adverse events was noted with dexamethasone therapy. Vietnamese adult patients previously treated with antibiotics for suspected bacterial meningitis only benefit from dexamethasone therapy if bacterial meningitis is subsequently proven by gram stain or culture-positive results. Treating other suspected meningitis patients may be harmful, probably because of the prevalence of TB meningitis.

Local Practice Pattern:

Based upon a twelve-month IRB-approved retrospective chart review conducted by Dr. Kopp, Asaro, and Carpenter for 2006-2007, 65 patients were identified with an EM-diagnosis of meningitis or encephalitis of which six were excluded for repeat visit (2) or lack of LP (4). Of the remaining, 25% (15/59) had positive-cultures: S. pneumonia (2), S. dysgalactiae (1), Coag-negative Staph (1), Cryptococcus (2), various viruses (9). In total, 13.6% (8/59) had received steroids, half before or concurrent with antibiotics. Both S. pneumonia meningitis patients received steroids as did six aseptic meningitis patients.

BJH Survey:

Hypothetical leaks in the information pipeline (illustration of theoretical model attached) based upon BJH survey with 86% (76/88) response rate with 59% reporting that they do not routinely use steroids in adult meningitis. When analyzing the responses of only those not using steroids to determine why steroids were not being employed for this indication the following “leaks” were discovered:

  • Awareness – 28.9% were unaware of any research supporting steroids in adult meningitis; 61% were uncertain of the correct dose of steroids for adult meningitis adjunctive therapy; 39% were uncertain of the timing of steroids relative to antibiotics.
  • Acceptance – 66.7% suggested physician peers doubted the efficacy of steroids in meningitis; 50% had never been taught that steroids played a role in the management of adult bacterial meningitis.
  • Applicable – 33% placed less importance on using steroids in adult bacterial meningitis because no formal QI measures exist.
  • Able – 33% felt ED overcrowding precludes timely recognition of adult bacterial meningitis; 48% felt that the atypical clinical manifestations of adult bacterial meningitis precluded timely administration of steroids;
  • Act Upon – 87% noted a lack of pre-existing order sets within HMED which limits recall during busy clinical shifts; 76% obtained a head CT before LP on all suspected meningitis patients.

The leading findings which would increase opponents likelihood of using steroids in adult bacterial meningitis were a positive CSF gram stain (76%), an “abnormal CSF WBC” which was undefined (76%), any combination of altered mental status, stiff neck, photophobia, headache, petechial rash or fever (70%), abnormal CSF glucose (57%).

International Survey:

54% (36/54) response rate from across the United States and McMaster University in Canada. Overall, 52.8% “almost always” give steroids to suspected adult bacterial meningitis. Among those who do give steroids nationally, 74% believed that the supporting literature was indisputable. The leading local motivators respondents identified were Journal Club (78%), didactics (39%), and Quality Improvement reminders (17%). Barriers outside institutions now using adjuvant steroids in meningitis had discovered included forgetfulness in the hectic ED (56%), difficulty distinguishing viral from bacterial meningitis before LP (50%), and delays in diagnosing CNS infection (39%). Ways to Implement Change at BJH: Based upon these findings, the following solutions were proposed to address each of the hypothetical leaks:

  • Awareness – Journal Club discussion and links to critical appraisal of the best-evidence on the topic..
  • Acceptance – the bulk of the evidence and opinion favors use of steroids in bacterial meningitis including attached summaries by EM Abstracts, Annals EM Evidence-Based Reviews, a recent NEJM letter, and our own peer-review of the literature summarized in the attached critical appraisals. With caveats in Third World children or adults with high-likelihood to have TB, HIV, or HSV meningoencephalitis, steroids ought to be strongly considered when treating suspected bacterial meningitis. This likely represents a Standard of Care for Emergency Medicine.
  • Applicable – recognizing that we had 65 cases of suspected meningitis in one-year, steroids-in-meningitis is not a “big-ticket item” like multisystem trauma, acute coronary syndrome, or stroke which we see in far greater volume. Nonetheless, individual efforts to optimize the outcome of these rare patients is every physicians’ duty.
  • Able – To standardize this process, 10 mg dexamethasone IV every 6-hours for 4-days should be considered before, with, or as soon as possible after antibiotics in suspected adult bacterial meningitis in the following scenarios:
  • Act Upon – The above dexamethasone dose and caveats should be added to the HMED CSF order set to facilitate recall.

Knowledge Translation Project Lessons Learned

  1. Disease prevalence will be important in selecting future labor-intensive KT projects. Although adjuvant steroids in definite meningitis is indicated based upon the best-available evidence, meningitis is a rare entity in the post-vaccination era and will struggle to maintain the attention of more voluminous EM-problems.
  2. Class champions are imperative to moving a multi-component, interdependent project like this forward.
  3. Survey wording cannot be over-emphasized. Although reviewed by multiple co-investigators before sending out, wording of some questions remained esoteric and unclear limiting our ability to draw conclusions based upon responses.