Abstract
Objective Contamination of cerebrospinal fluid (CSF) by blood during neonatal lumbar puncture
(LP) is common and poses diagnostic difficulties. Our objectives were to determine
the number of traumatic LPs performed at the BC Children's Hospital over 9 years and
whether there was an association between traumatic LPs and demographic variables,
hospital location, or time of the procedure.
Study Design This study was a retrospective review of neonatal CSF samples from May 2006 to March
2015. The data were analyzed to establish the rate of traumatic samples and whether
there was an association between traumatic LPs and demographic variables (age, gender),
location of procedure, positive CSF culture, and/or timing of the procedure.
Results A total of 1,263 LPs were reviewed, 47.7% (n = 602) were contaminated with >400 red blood cells/high-power field. The median age
of neonates whose samples were uncontaminated was 10.580 days compared with 6.535
days in the group with contaminated samples (z = − 2.884, p = 0.004). None of the other factors studied was associated with traumatic taps. Detected
organisms included Escherichia coli (n = 12), coagulase-negative Staphylococcus (n = 7), Enterococcus faecalis (n = 3), and group B Streptococcus (n = 2).
Conclusion Nearly half of all CSF samples in the study period were contaminated. Traumatic samples
were more common in younger neonates.
Keywords
neonatal sepsis - meningitis - lumbar puncture
