Abstract
Background: Lumbar puncture (LP) for the exclusion of meningitis is included as part of routine clinical investigation of suspected sepsis in many neonatal units. A retrospective review of medical records in our unit suggested that attempts to obtain cerebrospinal fluid (CSF) were often unsuccessful and associated with complications such as apnoea, desaturation or bradycardia. The results aided diagnosis or exclusion of meningitis in only a small number of infants. This prospective study aimed to further examine the value of lumbar puncture in the routine evaluation of suspected neonatal sepsis.
Methods: Data were obtained for all lumbar punctures performed over a period of 4.5 months. Information about number of attempts to obtain CSF, procedural complications and results of microscopy and culture was collected at the time of investigation.
Results: 99 lumbar punctures were performed in 86 infants. Median (range) gestational age and weight at birth were 36(24–42) weeks and 2740(640–4510) grams respectively. The median age at the time of investigation was 2 (1–91) days. Multiple attempts were required in 54 cases with 21 of these needing ≥3 attempts. Complications were documented on 21 occasions. 20 of these were episodes of apnoea, bradycardia or desaturation and 1 infant vomited during the procedure. 4 infants required intervention with respiratory support. In 8 cases, no CSF was obtained and in a further 3, there was insufficient for microscopy. Of the remaining 62 samples, 26(42%) were too heavily blood stained to allow determination of a white cell count. Results of CSF culture were available for 79 samples. 74 yielded no growth. CSF culture was positive in 5 infants (4 bacterial, 1 viral). Treatment for meningitis was initiated in 4 infants, of whom 2 had negative CSF cultures. 3 infants with cultures positive for coagulase negative staphylococcus were not treated; the organism was presumed to be a contaminant. Positive and negative predictive values for CSF microscopy and culture in this group of infants are 40% and 97% respectively.
Conclusion: The results of this prospective study confirm that LP frequently does not contribute to the early management of infants with suspected sepsis. Complication rates for the procedure are high given the low number of positive diagnoses made. The large number of failed LPs suggests that training in the technique may be an important issue for junior doctors. However, careful clinical judgement may allow more selective use of LP as a diagnostic tool.
Article PDF
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Desurkar, A., Boyle, E. & McIntosh, N. 72 Prospective Study of Routine Lumbar Puncture in The Evaluation of Neonatal Sepsis. Pediatr Res 56, 476 (2004). https://doi.org/10.1203/00006450-200409000-00095
Issue Date:
DOI: https://doi.org/10.1203/00006450-200409000-00095