Optimising the investigation of meningococcal disease

Early treatment with benzylpenicillin is important and doesn't jeopardise diagnosis

The incidence of meningococcal disease in England and Wales has remained at high levels over the past two winters, as has the proportion of cases caused by strains of serogroup C (M Ramsay, E Kaczmarski, personal communications). Clusters, also caused mainly by serogroup C strains, have increased considerably, particularly among students at schools and universities (A Rushdy, J Stuart, personal communications). While effective vaccines are awaited, current priorities are to optimise recognition, diagnosis, and management.

Administration of benzylpenicillin to suspected cases before admission to hospital reduces mortality1 and is advocated by the United Kingdom's chief medical officers. Though now used more widely, continuing failure to implement this simple measure2 may be due to a misplaced fear of obscuring the diagnosis and thereby jeopardising management.

After an injection of benzylpenicillin blood culture is rarely positive and, though cerebrospinal fluid may still yield meningococci,3 lumbar puncture is an increasingly controversial investigation in suspected meningococcal disease. Occasional deaths due to brainstem herniation, together with the recognition that negative initial results are often misleading,4 have focused attention on alternative diagnostic methods. Using a few simple alternative investigations greatly improves the chances of confirming the diagnosis.5

In about half of patients with systemic disease meningococci can be isolated from the posterior pharyngeal wall, preferably accessed through the mouth (or through the nose if the patient is unconscious or delirious); this proportion is unaffected by prior benzylpenicillin treatment.3 In as many as two thirds of patients tissue fluid aspirated from skin affected by haemorrhagic rash may show Gram negative diplococci on a stained smear or yield meningococci on culture.6 Serological diagnosis is also available, though paediatricians may be reluctant to recall and rebleed a small child recovering from the combined traumas of recent meningococcal disease and hospital admission.

Polymerase chain reaction (PCR) amplification of bacterial DNA to detect meningococci in peripheral blood or cerebrospinal fluid is now used widely in England and Wales. Performed on peripheral blood, this can be much more sensitive than blood culture,7 but specificity has yet to be evaluated in large, clinically relevant populations. New primers designed by the PHLS Meningococcal Reference Unit now permit serogrouping of meningococci in over half of cases with a positive result on PCR screening,8 providing important information for management of contacts. The most suitable specimens for PCR are cerebrospinal fluid or the residue of the first blood sample taken on admission for haematological evaluation; later blood samples are less satisfactory because of clearance of meningococcal DNA. The admitting clinician should either take two samples into EDTA tubes or ask the microbiology department to retrieve a single sample from the haematology department. Retrieval must be swift; routine samples for blood counts are normally discarded rapidly.

As the use of lumbar puncture declines, the microbiology department is less likely to be aware of newly admitted cases. Clinicians must ensure robust lines of communication between ward and laboratory. Furthermore, all cases of meningococcal disease, whether suspected or confirmed, must be reported immediately to the consultant in communicable disease control (or the consultant in public health medicine), who is responsible for identifying close contacts and organising prophylaxis. Regular joint audit of cases helps to reinforce good diagnostic, clinical, and reporting practice.

Current British guidance for managing close contacts is to offer advice and information, a chemoprophylactic antibiotic, and vaccine if appropriate. Chemoprophylaxis must be given speedily as the risk of secondary cases, though small, is highest in the days immediately after the admission of the index case to hospital. In their BMJ leading article last year Kristiansen and Knapskog proposed that close contacts aged under 15 years should be offered protective penicillin treatment in addition to chemoprophylaxis.9 However, they offered no evidence to support this additional measure, which is not national policy in Norway. We do not recommend it for Britain.

The same authors also proposed that in outbreaks with three or more linked cases chemoprophylaxis should be restricted to proved nasopharyngeal carriers. This is unwise. A single negative throat swab is unreliable in predicting freedom from meningococcal carriage; moreover, during the two to three days it takes to carry out swabbing and obtain culture results the risk is at its highest. Paradoxically, the contacts at greatest risk of disease are those who have yet to acquire the pathogenic strain. Management of clusters is complex and has been considered recently by a working group commissioned by the Public Health Laboratory Service.10

Clinical evaluation of promising conjugated serogroup C meningococcal vaccines is now well advanced, and development work at an earlier stage continues on candidate vaccines for serogroup B disease. Prospects for a reduction, and ultimately, elimination of meningococcal disease are good.