Editor,—Pollard et alpresented a comprehensive personal view on the emergency management of meningococcal disease.1 I wish however to take issue with one point concerning lumbar puncture. Lumbar puncture should certainly be deferred in certain instances but should not be avoided as could be interpreted from the article. All children with suspected meningitis should, in my opinion, have a lumbar puncture at some stage in their illness. The reasons for lumbar puncture include:

•

the presence or absence of meningitis should influence the choice and, perhaps, duration of antibiotic treatment

•

the presence or absence of meningitis should influence fluid management once the initial shock is treated

•

accurate anatomical diagnosis of meningitis is important for epidemiological purposes

•

the presence or absence of meningitis is very relevant to neurodevelopmental prognosis and possible hearing impairment.

 I increasingly meet paediatric trainees who seem to accept that a clinical and polymerase chain reaction based diagnosis of meningitis is sufficient. I would prefer if Pollardet al replaced (in the figure) the capitalised order DO NOT LUMBAR PUNCTURE (sic) with the instructions DEFER LUMBAR PUNCTURE and discuss its performance later …