Despite the development of vaccines which have substantially reduced the acquisition of some strains of meningitis (McIntyre et al, 2012), bacterial meningitis maims or kills approximately 20% of people that contract the disease (van de Beek et al, 2012), and is, therefore, a medical emergency in which the patient requires immediate medical intervention. There are approximately 3 200 cases of bacterial meningitis and septicaemia in the UK and Ireland annually, of which meningococcal infection is the most common (Meningitis Research Foundation, 2014a). As the patient's deterioration in health can often be rapid and unpredictable (Weisfelt et al, 2008), this has serious implications for ambulance clinicians in recognising and treating community-acquired meningococcal meningitis.
Diagnosis
Because bacterial meningitis is life-threatening, early diagnosis and treatment is vital (Brouwer et al, 2012)—the necessity of which is underlined in a study that demonstrated an association between increased three-month mortality in patients and delayed antibiotic treatment of three hours or more post-hospital admission (Auburtin et al, 2006). Research has suggested that there is also an incremental effect of delay in treatment on the rate of mortality: the greater the delay, the greater the mortality rate (Proulx et al, 2005).
However, despite widespread acknowledgement of the necessity for early diagnosis and treatment, there are apparent ambiguities in the Association of Ambulance Chief Executives (AACE) 2013 algorithm for patients with suspected meningococcal meningitis and septicaemia. The algorithm states that in the presence of a non-blanching rash, antibiotics should be administered either on scene or en route if there is a possibility of a delayed transfer (Association of Ambulance Chief Executives, 2013). However, in the absence of a non-blanching rash, the algorithm dictates that the patient should be transferred to hospital under time-critical conditions, but without the administration of benzylpenicillin (Association of Ambulance Chief Executives, 2013), principally so dexamethasone can be administered within four hours of the first antibiotic dose (Meningitis Research Foundation, 2011). However, in the book version of the guideline, there is asterisked small print that comments: ‘If bacterial meningitis is suspected and urgent transfer is not possible, administer antibiotics even in the absence of a nonblanching rash’ (Association of Ambulance Chief Executives, 2013: 127). For the same algorithm in the pocket-sized version of the guidelines, this comment is omitted altogether. This appears to offer conflicting advice, suggesting on one hand a non-blanching rash is the sole confirmatory factor on which a diagnosis should be based, and yet conversely suggesting that in some circumstances, treatment can be initiated simply on a high index of suspicion.
The highly publicised petechial rash presented as the hallmark of bacterial meningitis, including within the AACE guideline, is not necessarily a reliable indicator. Whereas the Meningitis Research Foundation (2011) asserts that a non-blanching rash is present in most cases, various studies have shown only 11%–26% of patients with confirmed bacterial meningitis actually present with a rash (Fitch and van de Beek, 2007), and the AACE (2013) guideline itself refers to the presence of a non-blanching rash in only 40% of paediatric cases. Compounding this is the fact that a petechial rash may also be present in less serious conditions, such as viral meningitis (Schut et al, 2008), and the recognition that of the children admitted to primary care with petechial rashes, the majority will not have meningococcal disease (Meningitis Research Foundation, 2011). It could be argued, therefore, that the guidelines for the pre-hospital assessment and treatment of bacterial meningitis are overreliant on the presence of a non-blanching rash as a means of diagnosis, and this is supported in a study by van de Beek et al (2004), who identify that the absence of a rash has a significant association with an unfavourable patient outcome.
Although the National Institute for Health and Care Excellence (NICE) provide an extensive list of specific signs and symptoms of bacterial meningitis, and, additionally, common and less common non-specific symptoms, the fact that they also acknowledge that specific symptoms may not manifest at all in some patients (National Institute for Health and Care Excellence, 2013)) highlights the challenge the clinician faces when attempting to diagnose the disease, particularly in the absence of a non-blanching rash.
Consequently, the reliability of the other signs nd symptoms listed by NICE should also be addressed. A study by Schut et al (2008) found that only 44% of patients presented with the traditional triad of fever, altered mental state, and neck stiffness, and the collection of symptoms were more prevalent in pneumococcal than meningococcal meningitis. However, the same study recognised that if headache as a symptom was added to the triad, 95% of patients displayed at least two of these four signs and symptoms (Schut et al, 2008), with fever having the highest incidence, being evident in 95% of patients (Fitch and van de Beek, 2007). Such is prevalence of these symptoms that Attia et al (1999) believe that the clinician can have 100% confidence in excluding meningitis from their diagnosis in the absence of all of these symptoms. However, Weisfelt et al (2008) state that of the above symptoms, a reduced GCS is one of the primary determinants of poor patient survival, with Lu et al (2002) specifying that the first dose of antibiotic should be administered prior to the patient's GCS deteriorating to below ten.
NICE also lists Kernig's sign and Brudzinski's sign among its more specific signs and symptoms (National Institute for Health and Care Excellence, 2013). Kernig's Sign refers to the occurrence of flexion contracture in the legs or occasionally also in the arms. By having the patient in a supine position with hips and knees in flexion, the patient's knee is then slowly extended. Severe stiffness and pain (spasms) involving the hamstring muscles as a result of stretching the inflamed sciatic nerve roots at angles less than 135˚ constitutes a positive sign (Verghese and Gallemore, 1987). Conversely, Brudzinski's Sign involves a health professional attempting to raise the head of a patient who is in a supine position towards their chest. By placing one hand on the back of the patient's head and the other on the patient's chest to prevent them from rising, in cases of meningitis, flexion of hips and knees during the procedure will occur. This is because passive flexion of the neck stretches the nerve roots through the meninges, both of which are inflamed, causing pain and involuntary movement of lower extremities (Saberi and Sayed, 1999).
However, Fitch and van de Beek (2007) noted that although there was a rate of 95% specificity to bacterial meningitis in patients that tested positive to both Kernig's and Brudzinksi's Sign, the signs overall only demonstrated a 5% sensitivity, and therefore, it was not possible to rule out meningitis in the absence of these traditionally ascribed meningitis signs.
It is clear that without reliable and specific determining characteristics, confidently diagnosing bacterial meningitis—particularly in a pre-hospital environment—is highly challenging. Indeed, the symptoms that manifest during the prodromalphase are so generalised and common to many regular viral illnesses that on their first visit to a GP, nearly 50% of children with meningococcal disease are sent home and consequently are more likely to die (Meningitis Research Foundation, 2011). Even more specific symptoms such as stiffness or rigidity could be attributed to disorders such as tetanus and drug-induced reactions (Schut et al, 2008), and differential diagnoses for disorders that can mimic symptoms of meningitis also include subarachnoid haemorrhage, brain abscesses, lupus, and other types of meningitis (Schut et al, 2008).
It might aid ambulance clinicians, therefore, if the AACE removed the presentation of a non-blanching rash as the sole pre-requisite for treatment, and instead implemented, for example, a numerical scale similar to the Oostenbrink Clinical Decision Rule for Predicting Bacterial Meningitis Risk in Children with Meningeal Signs (Oostenbrink et al, 2001), with the presence of certain symptoms given a corresponding numerical value, and the more specific signs and symptoms weighted more heavily. If the patient scores over a certain threshold of points, the clinician can conclude they have a sufficiently high index of suspicion of the probability of bacterial meningitis to justify beginning treatment. It would thereby eliminate the ambiguities contained in the AACE guideline, and would also ensure a thorough examination of the febrile patient.
Treatment
While diagnosis is evidently deeply challenging, the treatment is also similarly controversial. As established above, benzylpenicillin is the antibiotic administered by UK ambulance clinicians to combat meningococcal meningitis. Over 20 years ago, the pre-hospital use of benzylpenicillin was supported and promoted in concurrently-released studies by Cartwright et al (1992) and Strang and Pugh (1992), who both noticed an association between the pre-hospital administration of antibiotics and a significant reduction of up to 40% in the mortality rate of those cases. However, a later Danish study conversely noted a rise in the mortality rate in those that had been given antibiotics prior to hospital admission, and queried whether this could be attributed to the antibiotic-induced release of endotoxins precipitating shock (Sørensen et al, 1998)—an effect that was considered but excluded by Strang and Pugh (1992). Later research supports the findings of the Danish study, but is still unable to establish the biological mechanism for this adverse reaction, although it does acknowledge the possibility that benzylpenicillin may be disproportionately administered to patients with more severe or advanced symptoms (and so a clearer diagnosis), and therefore for whom the prognosis is already poorer (Harnden et al, 2006). There is, however, very little research on the subject, as is evident from the chronology of the reports, and all recognise that further research and trials are required to firmly establish a causal link between pre-hospital antibiotics and mortality rate.
Furthermore, there should be an awareness of a world-wide increase in penicillin-resistant strains of meningitis, particularly as some penicillinresistant strains of meningococcal disease have also been identified ( Joint Formulary Committee, 2011). A Spanish study reported a rise in such cases from 9.1% in 1986 to 71.4% in 1997 (Latorre et al, 2000), and there is concern that such strains will become similarly prevalent in the UK (Meningitis Research Foundation, 2014b). A thirdgeneration cephalosporin such as cefotaxime is now often the preferred treatment for those with penicillin-resistant strains of meningococcal meningitis (van de Beek et al, 2012), and can also be administered to those in whom benzylpenicillin is contraindicated due to severe allergy ( Joint Formulary Committee, 2011). However, at £4.31 for a 1 g vial, cefotaxime is over twice the price of a 1.2 g vial of benzylpenicillin ( Joint Formulary Committee, 2011), although it could also be argued that the relative infrequency of its use means that there would not necessarily be a greatly increased expenditure overall. However, as with benzylpenicillin, there should also be research into whether the pre-hospital administration of cefotaxime would be advantageous to the patient.
As there appears to be an association between the severity of subarachnoid inflammation and the outcome for the patient (van de Beek et al, 2012), recent studies have focused on the adjunctive use of corticosteroids such as dexamethasone to potentially suppress the inflammation (Chaudhuri, 2004). Dexamethsone is already a drug approved for paramedic use by the AACE, although it is currently only indicated for croup (Association of Ambulance Chief Executives, 2013). As the first dose of dexamethasone should be administered concurrently with the first dose of antibiotics (Brouwer et al, 2010), this has implications for treatment commenced prior to hospital admission. While some studies suggest that dexamethasone should be withheld from patients with comorbidities such as diabetes or alcoholism (Chaudhuri, 2004), there have been favourable responses for its use in trials conducted in developed countries (van de Beek et al, 2012), and research has found no detrimental long-term outcome in those with bacterial meningitis who received dexamethasone treatment (Weisfelt et al, 2006). As a result, dexamethasone is recommended for use with benzylpenicillin when the clinician prefers to begin treatment prior to hospital transfer (Chaudhuri, 2004). Furthermore, dexamethasone is suitable for ambulance use as the solution has a three year shelf life and does not require refrigeration (Chaudhuri, 2004), and is also inexpensive, costing 83p for a 1 ml ampoule (Joint Formulary Committee, 2011).
Similarly, studies are being conducted into the adjunctive use of oral paracetamol for patients with bacterial meningitis. Pelkonen et al (2011) found that a 48 hour regimen of paracetamol combined with a cefotaxime infusion was associated with three days of lowered mortality. Again, this is an area that requires further research and trials, but it is another treatment available to ambulance clinicians should a favourable causal connection be proven.
Conclusions
It appears that of the variety of life-threatening conditions which ambulance clinicians can be expected to diagnose and treat, bacterial meningitis has one of the most ambiguous protocols, and most under-researched supporting evidence with regard to pre-hospital treatment. This could possibly be because its relative infrequency precipitates a lack of urgency in research, but with such significant consequences for patients, administering the appropriate treatment at the appropriate time is vital. The combination of a meningitis diagnostics scale and further investigation into the biological mechanisms of pre-hospital antibiotics and adjunctive treatments should begin to determine a more concrete protocol for addressing bacterial meningitis. Additionally, with the increasing development of penicillin-resistant strains of bacterial meningitis, the ambulance service perhaps should also consider pre-emptive research into the implementation of nonpenicillin antibiotics as a first line of treatment.
Conflict of interest: none declared