Should GPs give penicillin for suspected meningococcal septicaemia?
I found Dr Rosie Hague's article ‘Meningococcal disease warrants prompt referral' (Practitioner 2008;252:17-20) very helpful but was confused by her advice that GPs should give parenteral antibiotics as soon as meningococcal disease is suspected.
I read a report last year of an increase in mortality rate in children with clinically diagnosed bacterial meningitis who were given parenteral benzylpenicillin prior to hospital admission by GPs.
If a GP is in close proximity to the referring hospital and the casualty department has been informed of the case, should the GP still give the drug or is it better for the patient to be treated in hospital?
Dr Irene Cotter, GP
Dr Rosie Hague replies:
The advice to administer parenteral antibiotics as soon as meningococcal disease is suspected is based on the observation that those patients who present with meningococcal septicaemia (fever and petechiae/ purpura) can progress very rapidly from mild symptoms to severe sepsis syndrome and death. This process cannot be influenced until antibiotics, along with supportive measures such as oxygen and intravenous fluids, are commenced.
There are studies indicating that delay in administration of antibiotics in those presenting to the emergency department is associated with a worse prognosis.1 Clearly randomised controlled trials are not possible in this situation, but the assumption has been that giving penicillin in primary care will ‘buy more time' and potentially save lives.
I assume that the report to which you are referring is the paper by Hahne et al, published in the BMJ in 2006.2 This was a systematic review of the effectiveness of antibiotics given before admission in reducing mortality from meningococcal disease. In my opinion, this is not the easiest paper to read, and the individual papers cited produced conflicting results.
However, the main message is that while the results suggested that those with severe disease who were given parenteral antibiotics in primary care had a worse outcome, this finding is likely to be confounded by the fact that those treated had the most severe disease, and so were less likely to do well.
A helpful accompanying editorial in the same issue3 rightly emphasises that the most important role for GPs is to recognise the disease and to make sure that the patient gets to hospital as soon as possible. The benefit of parenteral penicillin in the absence of provision of other supportive measures remains unproven.
The D level recommendation in the SIGN guideline4 reflects the current Health Protection Agency guidelines that, after considering this evidence, parenteral antibiotics should still be given as soon as meningococcal disease is suspected.5
It is important to recognise that the condition discussed here is that of meningococcal septicaemia, not bacterial meningitis. In the most fulminant cases where any delay in administration of antibiotics may be crucial, the organism has not even got as far as the meninges, and the signs prompting treatment are therefore fever and petechiae.
There is no evidence supporting the administration of antibiotics to those presenting with signs of meningitis in the absence of a rash. In this case, the causative organism may not be N. meningitidis. There is evidence that in other forms of bacterial meningitis, dexamethasone should be given before, or at the same time, as the first dose of antibiotic to reduce the risk of deafness. If parenteral antibiotics have already been given, this opportunity will have been lost.
The most recent published guideline on this subject by an EFNS Task Force6 recommends that anyone suspected of having meningitis should be admitted to hospital within 90 minutes of the diagnosis being suspected,
and antibiotics commenced after a lumbar puncture (if there are no contraindications) within 60 minutes of hospital admission, and no longer than three hours after first contact with the health service.
This may only be achievable in remote and rural areas if antibiotics are given in primary care, but otherwise the ideal pathway is rapid transit to hospital, where appropriate investigation and all the elements of treatment can be commenced.