When to do a lumbar puncture in a neonate.

Correspondence When to do a lumbar puncture in a neonate Sir, We enjoyed Dr Halliday's annotation on neonatal lumbar puncture but felt that his discussion of the role of lumbar puncture in suspected sepsis left an unfortunate question mark hanging over this procedure.' It is important we feel to distinguish between early onset and late onset sepsis. In early onset sepsis the usual route of infection is by ascending infection that causes neonatal pneumonia, often clinically and radiologically indistinguishable from hyaline membrane disease, and septicaemia. Frequently babies with group B streptococcal pneumonia are septicaemic at birth,2 and as meningitis occurs secondary to high level bacteraemia it is scarcely surprising that about 30% of babies with early onset group B streptococcal sepsis have meningitis.3 ' Dr Halliday quotes Eldadah and colleagues' two year study,5 which showed only five cases of group B streptococcal pneumonia and septicaemia, none of whom had meningitis, as a reason for not doing early lumbar punctures. Over a five year period (1984-9) with 5000 to 6000 live births per year we have seen seven cases of early onset meningitis (<48 hours old): four were due to group B streptococci, and one each to Listeria monocytogenes, Escherichia coli, and Streptococcus mitis. There were no clinical features to distinguish them from septicaemia without meningitis. In two of the seven cases blood cultures were negative. A positive lumbar puncture caused us to revise our antibiotic regime in two cases. The situation with suspected late onset sepsis is superficially similar but the organisms are different. Over the same five year period, we have seen four cases of Gram negative bacillary meningitis (one pseudomonas, one klebsiella, one achromobacter, and one with both klebsiella and E coli). In all cases the blood cultures were positive, but a positive lumbar puncture altered our antibiotic regime. It is sometimes argued that if the same antibiotic regime is used to treat late onset septicaemia alone as is used to treat meningitis, then one can rely on blood cultures without performing a lumbar puncture. However, about 10% of late onset meningitis occurs with negative blood cultures.6 If, as is increasingly the case, antibiotics are stopped after two or three days in the face of negative blood cultures, the occasional case of meningitis will be missed and will progress. Although rare, such an outcome would be disastrous. In addition, we have seen two unsuspected cases of Candida albicans …