Lumbar puncture and headache

Abstract

Clinical management of meningococcal disease Coning may occur without lumbar puncture being doneEditor-Wylie et al do not give the ages of their 252 patients with meningococcal disease, but their recommendation for fewer lumbar punctures should be interpreted cautiously in children. 1 Four deaths were attributed to lumbar puncture, which was done in 198 patients.Two patients deteriorated after lumbar puncture, and postmortem examination showed brainstem herniation.Even when lumbar puncture is not performed, however, patients with meningococcal disease can deteriorate after admission and can cone secondary to raised intracranial pressure.A causal relation in these two cases is not proved.In two patients antibiotic treatment was delayed because initial examination of cerebrospinal fluid was normal.This is a criticism of clinicians, not a risk of lumbar puncture.The risk of cerebrospinal fluid initially giving false negative results and the difficulty of diagnosing meningitis clinically in young children are both recognised.If an ill child warrants a lumbar puncture then he or she should receive antibiotics for 48 hours pending the results of culture, irrespective of the initial results of microscopy.Lumbar puncture may be dangerous if patients have hypotension, coma, raised intracranial pressure, or focal signs and is unnecessary if there is a typical rash.In a series of 445 children with meningitis, the two children with meningococcal infection and cerebral herniation had contraindications to lumbar puncture. 2Lumbar puncture must be performed more readily in children than adults because of the difficulty of diagnosing meningitis clinically.In 86 children with meningococcal meningitis, only 34 had a typical haemorrhagic rash, and not always at presentation. 3 In Wylie et al's series, lumbar puncture gave a positive result initially in 177 of 198 cases and an overall positive result in 186.In children, a clinical diagnosis of meningitis is unlikely to have a similar sensitivity of 89-94%.Moreover, it is unclear how often a positive culture was obtained from another site in the 120 cases that yielded meningococci; molecular biological techniques give no information on antibiotic sensitivity.If fewer lumbar punctures are done in children with meningitis some will be missed and the start of treatment will be delayed.This increases the risk of death, 4 possibly above the 2% attributable to lumbar puncture in this series.Wylie et al have shown a high sensitivity and low mortality from lumbar puncture.Before recommending that the procedure be used less we need to be reassured that clinical diagnosis alone does not result in worse sensitivity and mortality, particularly in children.