Are Infant CSF Reference Values Accurate?

William T. Basco, Jr., MD


March 14, 2011

Normative Cerebrospinal Fluid Profiles in Febrile Infants

Byington CL, Kendrick J, Sheng X
J Pediatr. 2011;158:130-134

Study Summary

Cerebrospinal fluid (CSF) white blood cell (WBC) counts are often assessed as part of the initial workup for infants with fever. Newer techniques allow for more complete viral testing of CSF and sera to determine whether previously suggested normative levels (obtained from studies with either small numbers of subjects or incomplete ascertainment of viral infection) were accurate.

The study, completed at a single medical center in the southwest United States, was a retrospective analysis of CSF results in febrile infants. Of 1779 infants (0-90 days of age) who were initially enrolled, 823 infants fulfilled criteria for being presumptively uninfected (negative urine, blood, and CSF bacterial cultures and negative test results for enterovirus via polymerase chain reaction) and 743 infants had atraumatic lumbar punctures (defined as red blood cell [RBC] count < 1000/mm3). Of infants without infection or traumatic lumbar punctures, 677 (91%) had CSF WBC < 14.5/mm3, and these infants comprised the final group from which normative CSF profiles were derived.

The mean CSF WBC count was < 10/mm3 during the first, second, and third months of life (6.1/mm3 in month 1; 3.1/mm3 in month 2; 3.0 in month 3). When the first month was broken down by weeks, mean WBC counts in the CSF were, respectively, 6.3, 6.8, 6.9, and 5.0, during weeks 1-4. Even in uninfected infants with traumatic lumbar punctures (CSF RBCs of < 10,000), the mean (6.5/mm3) and median (5.0/mm3) CSF WBC counts were < 10/mm3. Finally, the range of CSF WBC counts in children who were not infected was 0-18/mm3 in children during first month of life, 0-8.5/mm3 during the second month of life, and 0-8.5/mm3 during the third month of life. Although an overall correlation between CSF RBC counts and CSF WBC counts was found, the correlation explained only a small amount of the variability in WBC counts, suggesting that one should not correct WBC in CSF for RBC numbers.


Although Byington and colleagues did not set out to validate traditional cutoff levels for WBC counts in CSF in the evaluation of febrile infants (eg, ≤ 10 WBC per high-powered field are "low risk" for serious bacterial infection[1,2]) the data do provide additional face validity for that approach. It is worth emphasizing that the range of CSF WBCs in the uninfected infants was 0-20/mm3. The study is notable for having a large number of very young infants who had extensive testing for bacterial or viral infection. These data may very well represent the most comprehensively defined noninfected cohort yet reported in the pediatric literature.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: