Meningitis appears in several past paper SAQs.
- Generic bacterial meningitis is broken down by age and immune status in Question 11 from the first paper of 2015.
- Aseptic meningitis appears in Question 3.3 from the second paper of 2013
- Pneumococcal meningitis appears in Question 9 from the first paper of 2013
- Infected EVDs feature in Question 11.1 from the second paper of 2008, and Question 3.2 from the second paper of 2013 (though in the latter case the EVD is actually clean)
It is a topic which crosses the boundaries between infectious diseases questions (eg. "what organisms could be causing this?") and neurology questions ("explain this CSF picture").
One will therefore find the meningitis and encephalitis questions scattered rather arbitrarily between these two subject areas.
Meningitis is inflammation of the meninges and anything in the subarachnoid space. Encephalitis is an inflammation of the parenchyma itself, and sometimes gets referred to as "cerebritis".
A decreased level of consciousness suggests that the situation is serious; either there is increased intracranial pressure, or the sepsis is so advanced that some sort of encephalopathy has developed (however, one should point out that one of the criteria of diagnosis for septic encephalopathy is the absence of intracranial infection). Oh's Manual quotes mortality rates in excess of 90% for such bacterial meningitis. Thus, meningitis appears in the ICU only when it becomes interesting and complicated.
The main culprits are:
- Streptococcus pneumoniae (common enough to deserve its own summary chapter)
- Listeria monocytogenes
- Neisseria meningitidis
- Haemophilus influenzae
Apart from Listeria, the abovementioned bugs all spread by droplets and body fluid exposure.
In the host who has been exposed to the ICU environment and has had some intrumentation of the brain, the range of possible infectious agents is broader:
- Pseudomonas aeruginosa
- Klebsiella sp.
- Acinetobacter sp.
- Escherichia coli
One can also separate the pathogens by the age group and immune status of the host, as in Question 11 from the first paper of 2015.
|Host group||Microbial enemies||Appropriate antibiotics
(CICM model answer)
|Ceftriaxone or cefotaxime|
+ Cefepime or meropenem
|Ceftriaxone or cefotaxime
+ Benzylpenicillin for Listeria
+ vancomycin (most of the time)
|Ceftriaxone or cefotaxime
+ Benzylpenicillin for Listeria
+ vancomycin (most of the time)
Therapeutic Guidelines recommend slightly different drugs; benzylpenicillin is substituted for ampicillin and vancomycin is only offered to those with unidentified gram positive cocci in their CSF, or whenever an LP is not available. Aminoglycosides are not mentioned. The neonates get cefotaxime instead of ceftriaxone, for some reason.
eTG give their references as the 2014 NSW Health guidelines, and this 2004 article by Tunkel et al from Clinical infectious diseases. It is unclear what happened to the candidates whose antibiotic choices agreed with Therapeutic Guidelines but not with the college.
A CT before the lumbar puncture
Folk wisdom recommends against performing a lumbar puncture in a patient with increased intracranial pressure, with good reason. One does not wish to give the brainstem any reason to herniate. The 2004 published guidelines contain a often-quoted table of features which would make one want to exclude space-occupying lesions before performing an LP:
- Hypertension and bradycardia
- Focal neurological signs
- Decreased level of consciousness
- A new onset of seizures - however, this does not necessarily scream "abscess" because about 25% of meningits patients will have a seziure during their meningitis episode.
- Predisposition to infection (eg. immunocompromised state)
- History of stroke or intracranial space-occupying lesion
Oh's Manual only lists papilloedema seizures and focal neurological signs as contraindications to an LP.
Ohs Manual has a nice table on page 599 which outlines the expected CSF changes in bacterial and viral meningitis. The bacterial varierty will feature a low CSF glucose and a high CSF protein, with swarms of granulocytes in the CSF, whereas viral meningitis will have predominantly lymphocytes and not much protein.
Ventriculitis: an infected EVD
Question 11.1 from the second paper of 2008 presents the candidates with a dirty CSF sample, expecting them to conclude that the EVD is infected. The most likely organisms are usually visiting skin flora, such as Staphylococcus aureus, Staphylococcus epidermidis or Streptococcus pyogenes.
The two things one should immediately consider is
1) Get the infected EVD out
2) Start some antibiotics, which in the context of unknown sensitivities should consist of cephalosomething and vancomycin.
- No single parameter can reliably predict or exclude EVD-related infection; even CSF pleocytosis (because there is a normal granulocyte migration into the ventricles if there has been intraventricular blood)
- A "cell index" (i.e. a ratio of CSF WCC to blood WCC) is better than any sort of crude "1 to 1,500" approximation in identifying an impending infection; it may even be able to predict ventriculitis three days before the culture becomes positive.
- There is little evidence to support the practice of prophylactic catheter exchange at a predefined interval (i.e. there does not seem to be any decrease in infection risk if you change the EVD every 7 days)
- The risk of infection peaks after 2 weeks
- Risk factors for EVD infection include the following:
- EVD duration > 11 days
- Frequency of CSF sampling
- Intraventricular hemorrhage
- Surgical technique (eg. is it a subcutaneously tunneled EVD?)
"Aseptic meningitis": CSF with a negative Gram stain
There are several causes of meningitis which will not produce a positive Gram stain.
This topic enjoys a thorough exploration in this 2007 article. The table of infectious and noninfectious causes they produce is massive, and to reproduce it here would only serve to bewilder and depress the exam candidate.
- Cryptococcus neoformans
- Treponema pallidum
- Borrelia burgdorferi (Lyme disease)
Additionally, one might wish to consider non-infectious causes, such as lymphoma, vasculitis, or drug-induced meningitis (eg. due to cotrimoxazole or azathiaprine).
Investigations for aseptic meningitis
- Cryptococcal antigen
- Mycobacterial PCR
- Syphilis PCR
- HSV PCR
- VZV PCR
- India ink stain
- Fungal cultures
Dexamethasone in bacterial meningitis
The damage to neurons in bacterial meningitis is thought to be partly due to the violent host response, with much of the irreversible damage done by migrating leukocytes. This is vaguely the same justification which was given for the use of high dose steroids for septic shock in the late 1980s. In the case of meningitis, it actually sort-of seems to work.
Though it is still a debated measure, most people these days give dexamethasone to meningitis patients. This practice is supported by a large Cochrane review by Van de Beek et al (2007), which analysed the evidence from 18 studies, involving 2750 patients, and concluded that "corticosteroids significantly reduced rates of mortality, severe hearing loss and neurological sequelae".
Six years later, the 2013 review by Brouwer et al (the same one quoted by the college in their model answer to Question 11 from the first paper of 2015) was able to include 4121 participants from twenty-five trials. There were some differences in their conclusions, in comparison to the previous meta-analysis:
- The mortality benefit had shrunk into statistical non-significance.
- Deafness was reduced (RR 0.67) as were neurological sequelae of all sorts (RR 0.83).
- Mortality improved in the S.pneumoniae patients only
- In children, protection against hearing loss was only seen in the H.influenzae group.
- In third world countries, there was no benefit for anybody, except for those with tuberculosis meningitis.
- Subgroup analysis of high quality studies did not show any effect on severe hearing loss.
- There was an increased risk of recurrent fever, but no other side effects.
Still, the overall conclusion that corticosteroids should still be given to patients in high-income countries.
Adverse effects of corticosteroids in meningitis are mentioned in this 1996 article by Townsend et al. In brief, the only major concern is that corticosteroids may rapidly normalise the blood brain barrier function, decreasing the penetration of antibiotics into the CSF. This phenomenon is known purely from rabbit models. Might I add that in those models the decreased concentration of ampicillin in the rabit CSF was still well above MIC for the E.coli they injected into the rabbits. In the one known human study (Paris et al, 1994) with eleven meningitic children the CSF concentration of ceftriaxone was also unaffected. In short, the evidence to feed these concerns is far from convincing.
- Dexamethasone should be given to all adult meningitis patients in the developed world.
- Most guidelines recommend the same for children
- It should be given before the first dose of antibiotics
- It should be given for 4 days, or less if the meningitis is proven non-streptococcal (or non-Haemophilus, in the case of children)
- The daily dose should be 0.6mg/kg.