Date:  June 11th, 2014 

Guest Skeptics: Dr. Jeremy Faust is a Emergency Medicine resident Mt. Sinai Hospital, New York City. Jeremy is a self described evidence based medicine zealot. He writes a column in ACEP News on the topic of social media. Jeremy also just launched an excellent new podcast series with Lauren Westafer called FOAMCast.

Case Scenario: 24 year old woman presents to the emergency department with a 3 day history of progressively worse flu like symptoms. She has a fever 39.6C, HR 114, BP120/80 and RR 22. You notice a purpuric rash and she has clinical meningismus. Her WBC is elevated and the lactate is borderline elevated. The lumbar puncture demonstrated turbid appearing fluid and no blood.

Question: Do steroids prevent death, hearing loss or other neurologic sequelae in bacterial meningitis?

Background:  Bacterial meningitis is fatal in children 5-40% and adults 20-50%. This is even with appropriate antibiotic treatment. Concomitant inflammation of brain/meninges is commonly associated with serious neurologic sequelae. One of the worst neurologic outcomes in survivors is sensorineuroal hearing loss (SNHL). Up to 1/3 of patients will experience bilateral SNHL post meningitis. The cause of SNHL is thought to be multifactorial in bacterial meningitis (Wellman et al):Ear Anatomy

  • Extension of the infection along the 8th cranial nerve, the periotic duct and the cochlear aqueduct
  • Cochlear pathology due to serofibrinous exudate, inflammation and granulation cells
  • Septic emboli and thrombotic occlusion secondary to vasculitis
  • Pathologic formation of new bone within the otic capsule called labyrinthitis ossificans

Worse historical outcomes have previously been observed in lower income countries.

Article: Brouwer MC, McIntyre P, Prasad K, van de Beek D. Cochrane Database of Systematic Reviews 2013, Issue 6. Art. No.: CD004405. DOI: 10.1002/14651858.CD004405.pub4 PubMed ID: 23733364

  • Population: Patients with bacterial meningitis (all ages). Higher vs lower income countries defined on basis of UN Human Development Index scores (>0.7 and <0.7 respectively)
  • Intervention: Corticosteroids; most commonly dexamethasone (0.6mg/kg/day for four days)
  • Control: Usual care
  • Outcome: 
    • Mortality
    • Hearing loss
      • Severe hearing loss – bilateral loss >60dB or requiring bilateral hearing aids
      • Any hearing loss
    • Neurological sequelae
      • Focal (other than hearing), epilepsy, severe ataxia and severe memory or concentration disturbances
    • Adverse effects
      • Clinically evident GI bleed, reactive arthritis, pericarditis, herpes zoster or herpes simplex, fungal infection, recurrent fever (>38C), persistent fever (>5 days)

Authors Conclusions: “Corticosteroids significantly reduced hearing loss and neurological sequelae, but did not reduce overall mortality. Data support the use of corticosteroids in patients with bacterial meningitis in high-income countries. We found no beneficial effect in low-income countries.”

checklist-cartoonQuality Check List for Systematic Reviews:

  1. The clinical question is sensible and answerable. Agree
  2. The search for studies was detailed and exhaustive. Agree
  3. The primary studies were of high methodological quality. Agree
  4. The assessments of studies were reproducible. Agree
  5. The outcomes were clinically relevant. Agree
  6. There was low statistical heterogeneity for the primary outcome. Agree
  7. The treatment effect was large enough and precise enough to be clinically significant. Agree

Key Results:  25 studies included, N=4121 patients. 4 high quality (45% of included patients), 14 medium, 7 low.

  1. Overall mortality reduction was not significant: RR 0.90 (19.9% to 17.8%, 95%CI 0.80-1.01, p=0.07)
  2. Adult mortality reduction was also not significant: RR 0.74 (95%CI 0.53-1.05, p=0.09).
  3. Reduced severe hearing loss: RR 0.67 (95%CI 0.51-0.88); not reduced when high-quality studies analyzed only.
  4. Reduced any hearing loss: RR 0.74 (95%CI 0.63-0.87)
  5. Reduced neurologic sequelae: RR 0.83 (95%CI 0.69-1.00)
  6. Increased recurrent fever: RR 1.27 (95%CI 1.09-1.47), but no other adverse events with steroids.
hearing loss

Subgroup Analyses:

    1. Some reduced mortality for S. Pneumo meningitis (RR 0.84, 95%CI 0.72-0.98), but not for H. Influenza or N. meningitidis.
    2. Reduced severe hearing loss in children with H. Influenza (RR 0.34, 95%CI 0.20-0.59) but not other non-Hemophilus species.
    3. No benefits of steroids in low income countries (mortality, severe/any hearing loss, neurologic sequelae).
    4. No benefits of steroids in worst-case scenario analyses for missing data amongst trials of high heterogeneity encountered, random effects analysis. Benefit remained significant in WSC analyses amongst trials with low heterogeneity.

Comments: Thorough Cochrane review; updated from prior 2007 review. Thorough electronic and manual searches; no mention of language restrictions.

Generally low risk of bias, although reporting bias was almost 70% amongst included trials.

Screen Shot 2014-06-11 at 6.39.33 PM

Authors analyzed data based on available-case analysis and worst-case analysis for missing data. Mild/moderate heterogeneity amongst pooled trials (0-33%). Overall mortality data showed an I2 of 21% while the any hearing loss had I2 of 24%.

A rough guide to interpreting Heterogeneity:

  •  0% to 40%: might not be important
  • 30% to 60%: may represent moderate heterogeneity
  • 50% to 90%: may represent substantial heterogeneity
  • 75% to 100%: considerable heterogeneity

All outcomes analyzed with fixed effects analytical models. We have spoken before on the SGEM about the difference between fixed and random effects model. A fixed-effect meta-analysis assumes that all the studies share the same effect size. In the random effects model we do not assume this and allow that there could be a distribution of true effect size. (Borenstein, Hedges and Rothestein).

The 2007 Cochrane Database of Systematic Review on this topic did show mortality, hearing loss and short term neurologic benefits (20 studies, 2750 patients), again favoring those from high-income countries. An evidence based summary of the 2007 review was done by Dr. S. Upadhye and published in Annals Emerg Med 2008.

The current review includes newer randomized control trials from Vietnam and Malawi, where no benefits were realized. Furthermore, a drop in mortality is noted after the introduction of H. Influenza vaccinations in higher income countries.

The randomized clinical trials included in this systematic review did not address four important issues:

  1. Minimum duration of corticosteroid therapy
  2. Type of corticosteroids (most used dexamethasome 0.4 or 0.6mg/kg/d)
  3. Maximum length of time after parenteral antibiotic therapy for starting steroids
  4. Longterm effect of corticosteroid therapy

The Bottom Line: Use dexamethasone 0.6mg/kg/day in all cases of bacterial meningitis in high income countries, it may not save lives necessarily but at least it can spare any hearing loss or other neurologic sequelae. It should be given before or with first dose of antibiotics

Case Resolution: This woman with suspected bacterial meningitis is given appropriate broad spectrum IV antibiotics. Dexamethasone 0.6mg/kg/day is given at the same time. She is sent to the ICU and you hope she does well.

Clinical Application: If practicing in a developed country you should consider giving steroids with your antibiotics in patients with bacterial meningitis.

What do I tell patients: We suspect you have bacterial meningitis. We are going to give you antibiotics. At the same time we are going to give you steroids. This has been shown to prevent swelling, decrease hearing and may help save your life.

Keener Kontest: Last weeks winner was Dr. John Haggarty from Sarasota Florida. John knew that  Carroll Spinney was the puppeteer who brought Big Bird to life on Sesame Street but also played Oscar the Grouch.

If you want to play the Keener Kontest this week then listen to the podcast for the question. Email me your answer at TheSGEM@gmail.com with “keener” in the subject line. The first person to correctly answer the question will receive a cool skeptical prize.


Upcoming conferences:

  • QueBEEM Quebec City September 29th and 30th.
  • EuroBEEM Prague December 4th and 5th

Remember to be skeptical of anything you learn,

even if you heard it on the Skeptics’ Guide to Emergency Medicine.