Tessa Davis. How many white cells v red cells do we allow in CSF?, Don't Forget the Bubbles, 2013. Available at:
Unless your skills are tip top, the chances are that you will have had a traumatic tap before (studies suggest up to 40% of lumbar punctures). Blood in the CSF on lumbar puncture can be a sign of a subarachnoid haemorrhage but more commonly is due to a traumatic tap (if the number of red cells in consecutive samples remains the same, it’s likely to be an SAH, but if they reduce then it’s likely due to a traumatic tap).
When the results come back with white cells and red cells, how can you work out whether or not the WCC is significant?
This DFTB post was inspired by teaching from Andrew Numa, Sydney Children’s Hospital PICU.
Red cells in the CSF means there is blood
Clear CSF has no red cells in it. So the presence of red cells indicates blood
Blood has white cells in it
This seems obvious to say but as soon as there is blood in the CSF, there will also be white cells in the CSF (unless the patient is immunosuppressed).
An average red blood cell count in a normal blood sample will be 5 x 1012 per cm3
An average white cell count in a normal blood sample will be 15 x 109 per cm3
Therefore for every white cell there will be 333 red cells.
If the patient has a raised serum white cell count then this ratio will change
Look at the patient’s FBC. If the WCC is high e.g. 30 x 109 per cm3 then this will be the ratio expected in the CSF.
If the WCC is 30, then for every white cell there will be 166 red cells.
(Similarly if the patient is anaemic then the RCC is going to be lower).
When we look at the CSF we are trying to work how many white cells are in the CSF alone
We want to work out how many white cells are coming from the CSF and NOT from the blood.
So if the patient has a serum WCC of 15, assuming a RCC of 5, and the CSF results are:
RCC 10 000/cm3
The expected number of white cells in this patient’s blood for this many red cells would be 30. But this patient has 100. So the remaining white cells are from the CSF. This would indicate meningitis.
Is this foolproof?
Calculating the expected ratios in this way is controversial and there are studies suggesting that calculating the ratio in this way isn’t actually helpful.
Osborne and Pizer (1981) did a (small) retrospective study which suggests that using this ratio can underestimate the number of white cells in the CSF and therefore under diagnose meningitis. They demonstrated that the WCC was often less than expected and hypothesised that there may be loss of white cells at the site of bleeding.
In 2006, Bonsu and Harper conducted a larger study (682 specimens) and also showed only a modest correlation between predicted and observed correlation of white cell count in blood-contaminated CSF.
However, Mazor et al (2003) studied neonates with traumatic LPs and showed that, if the WCC:RCC ratio was < 100 and an observed:predicted WCC ratio of <0.01, then the majority of patients without meningitis can be identified.
Greenberg R, Smith B, Cotton M, Moody A, Clark R, Benjamin D, Traumatic lumbar punctures in neonates: test performance of the cerebrospinal fluid white blood cell count, Pediatr Infect Dis J. 2008 December; 27(12): 1047–1051.
Bonsu BK, Harper MB, Corrections for leukocytes and percent of neutrophils do not match observations in blood-contaminated cerebrospinal fluid and have no value over uncorrected cells for diagnosis, Pediatr Infect Dis J. 2006 Mar;25(3):207.
Mazor SS, McNulty JE, Roosevelt GE, Interpretation of Traumatic Lumbar Punctures: Who Can Go Home? Pediatrics, 2003 11(3):525-528.