OmanDocSign in

Curriculum · Neurology

Interpretation of CSF analysis

What it is

CSF analysis will tell you one thing and will not tell you another. It will tell you whether the picture is bacterial or viral, and it does that on four parameters. It will not tell you the type of bacteria or virus; that comes from the history and from the CSF culture.

How it is diagnosed

The normal values are an opening pressure of 5-20, leukocytes under 5, a cell differential of 0, protein 15-45 and glucose 45-85. A second table sets normal against bacterial and against nothing else. There the normal protein is put at less than 0.5 g/dl, the glucose at 2.6-4.5 mmol and the CSF to plasma glucose ratio above 0.66, while the bacterial side reads WBCs above 100 to 5000 or more, neutrophil predominance, raised protein, very low glucose and a very low CSF to plasma glucose ratio.

The other four settings:

A further table sets viral against bacterial meningitis with the values written out. In viral meningitis the opening pressure is ≤250 mm H2O, the leukocyte count 50-1000, the predominance is of lymphocytes, glucose is above 45 mg/dL or 2.5 mmol/L, protein is under 200 mg/dL, and Gram stain and culture are both negative. In bacterial meningitis the opening pressure is 200-500 mm H2O, the leukocyte count 1000-5000, the predominance is of neutrophils, glucose is under 40 mg/dL or 2.2 mmol/L, protein is 100-500 mg/dL, Gram stain is positive in 60%-90% and culture is positive in 70%-85%. Two notes are set beside that bacterial side: TB and brucella = lymphocytes, and the glucose is Normal in Brucella.

The CSF is also read outside infection: in transverse myelitis it shows pleocytosis and an elevated IgG index.

What can go wrong

Each table gives its own typical leukocyte counts, and the viral and bacterial counts run into each other. The viral count is given as 25-2,000 in one and as 50-1000 in the other, and the bacterial count as above 1,000 in the first, as 1000-5000 in the second, and as above 100 to 5000 or more in a third. So the counts separate bacterial from viral by their order of size rather than by one cut point.

The Gram stain is positive in 60%-90% of bacterial meningitis and the culture in 70%-85%, so they are not positive in every case. And what CSF analysis will not tell you is the type of bacteria or virus; for that there is the history and the CSF culture.