Special Feature: Interpreting CSF Results

By Esther Chen, MD, and Stephanie Abbuhl, MD, FACEP

Accurately measured in the supine patient, normal cerebrospinal fluid (CSF) opening pressure is typically between 150-200 mmH20.1,2 Readings may be elevated falsely in patients who are extremely flexed in the fetal position or who are sitting up, and by Valsalva maneuver. Spuriously low pressures can be seen in hyperventilated patients, if there is CSF leakage around the spinal needle, or rarely, if there is blockage of CSF by a herniating mass.1 High opening pressure may be associated with infections (bacterial, viral, and fungal meningitis; abscesses; and meningoencephalitis), tumors, intracerebral and subarachnoid hemorrhage, and pseudotumor cerebri.

Gross Examination of CSF

Normal CSF is clear and colorless. Turbid CSF results from the presence of white blood cell (WBC) counts > 200 WBC/µL or red blood cell (RBC) counts >400 RBC/µL3 and suggests a bacterial meningitis until proven otherwise. Grossly bloody CSF is apparent with >6000 RBC/µL and may be seen after a traumatic LP or in patients with subarachnoid hemorrhage.3 Xanthochromia, a pinkish or yellowish tinge to the CSF, may be apparent with RBC lysis (e.g. subarachnoid hemorrhage or delayed analysis of a traumatic LP), elevated CSF protein (>150mg/dL), and systemic hyperbilirubinemia.3

CSF Cell Count and Differential

The CSFwbc count and differential are a critical source of information, but not as simple to interpret as many charts in textbooks would imply. Cell counts must be performed promptly to prevent loss of cells from cell lysis or cell absorption to the walls of the collecting tube.1 The normal adult CSFwbc count is less than 5 cells/µL, with a lymphocytic and monocytic predominance and < 1 polymorphonuclear cell (PMN),3, 4 although PMNs > 6 have been seen in patients without meningitis, especially in traumatic LPs, where peripheral PMNs are introduced by blood contamination at the time of the procedure.5

Blood contamination during a traumatic LP requires careful analysis. CSF pleocytosis may be estimated by subtracting the contaminant WBC count (estimated using the normal peripheral WBC:RBC ratio of 1:700)3 from the actual CSFwbc count. A pre-Haemophilus influenzae vaccine study showed that this adjusted CSFwbc count accurately predicted 90% of patients with culture-positive meningitis.6 In addition, the CSFwbc count was often more than 100 times the number of WBCs attributed to the trauma. Of the patients with missed diagnoses, four were 6-week-old infants, and one was an elderly woman with overwhelming sepsis. However, because 10% of patients with meningitis still were missed, this formula must be used in conjunction with all the other available clinical and laboratory data.

Cell counts between 5 and 20 cells/µL are indeterminate and may indicate early or partially treated meningitis. A pleocytosis occasionally can be seen after generalized seizures, but this is a diagnosis of exclusion.7 Patients with bacterial meningitis tend to have a higher CSF pleocytosis (75% with >1,000 cells/µL in one study8 and a median of 1,380 cells/µL in a smaller study)9 than viral meningitis (median of 114 cells/µL)10, although there is significant overlap in these numbers. Therefore, the cell count always should be interpreted with the differential to help distinguish between those two processes.

A lymphocytic predominance typically suggests aseptic meningitis, although it does not eliminate the possibility of a bacterial etiology. Up to 14% of patients with bacterial meningitis may have more than 50% lymphocytes, especially with lower CSFwbc counts (<1,000 cells/µL) and with Listeria monocytogenes.11 Two additional studies also reported CSF lymphocytosis in 6-10% of patients with well-documented bacterial meningitis.12,13 To complicate the picture more, a lymphocytic pleocytosis also is seen in herpes simplex viral encephalitis (typically with elevated CSFprotein levels and a normal or mildly decreased CSFglucose levels), as well as in varicella-zoster, enteroviral, arthropod-borne, and other encephalitides.

PMN predominance, although classically associated with bacterial meningitis, also can be found frequently in aseptic meningitis. In a study of 158 children (30 days-18 years) with acute meningitis during the peak months of enteroviral season, more than 50% PMNs were found in 56% of aseptic and 90% of bacterial causes. In addition, PMN pleocytosis was not limited to the first 24 hours of illness in patients with aseptic meningitis. The sensitivity and specificity of PMN predominance to identify a bacterial etiology was only 90% and 43%, respectively.9

CSFglucose Level

CSFglucose is derived from facilitated diffusion of plasma glucose, so normal values range from 45-80mg/dL in patients with normal fasting glucose, or approximately 60% of the serum glucose. Hypoglycorrachia (20-40mg/dL) characteristically is seen in meningitis caused by bacteria, mycobacteria, and fungi. Furthermore, low CSFglucose levels may be found in carcinomatous meningitis, granulomatous infiltration of the meninges (e.g., sarcoidosis, cysticercosis), viral infections (e.g., herpes encephalitis, mumps meningoencephalitis), and rarely, neurosyphilis.3 CSFglucose levels may be spuriously high in hyperglycemic patients and potentially mask a serious infection. In these patients, the CSF to blood glucose ratio may be more useful. One author recommends using a ratio of <0.31 to help diagnose a bacterial etiology in patients with diabetes mellitus.14 However, the sensitivity of the combination of this ratio (<0.31) with a low absolute CSFglucose level (<40mg/dL) for identifying bacterial meningitis was only 73%. It is clear that abnormally low ratios may be seen in the absence of central nervous system (CNS) disease when the serum glucose has risen rapidly. Alternatively, normal ratios can be seen in bacterial meningitis, especially after a diabetic patient has had a sudden decrease in serum glucose level, as might occur after a dose of insulin. Fluctuations in serum glucose, along with the time lag required for CSF equilibrium to occur (1-2 hours), cause a wide variation in the CSF to blood glucose ratios for simultaneously collected samples.

CSFprotein Level

The CSFprotein level alone has limited value in determining the cause of fever and headache, because it fluctuates in both infectious and noninfectious neurologic diseases. Normal values range from 20 45mg/dL, depending on the laboratory used.4 Low CSFprotein levels typically are seen in processes that either decrease protein entry or increase protein removal (e.g., CSF leaks post-LP, pseudotumor cerebri).3 Elevated CSFprotein levels may be seen in many neurologic diseases and usually reflect an abnormality in the blood brain barrier. Causes of increased CSFprotein include not only viral and bacterial meningitis (usually higher in the latter) 3, 4, 9, but also diabetes mellitus (especially when significant peripheral neuropathy is present), brain and spinal cord tumors, multiple sclerosis, CNS syphilis, myxedema, uremia, and connective tissue diseases.2

When differentiating viral from bacterial meningitis, a CSFprotein level above 100mg/dL suggests a non-viral etiology.4 Unfortunately, however, there is considerable overlap in the protein levels that can be seen in both viral and bacterial meningitis.

Microbiological Analysis of CSF

The Gram stain of CSF fluid is one of the most useful and immediately available tests used to differentiate bacterial from viral meningitis, but again, there are caveats.

In one study, organisms were detected by Gram stain in 88% of bacterial or fungal meningitis and 81% of shunt-associated meningitis, including patients on antimicrobial therapy.15 However, in other studies, the Gram stain has performed more poorly with a sensitivity of only 40-60% in patients with bacterial meningitis.16 Diagnostic accuracy increases with higher WBC counts15, number of organisms, and type of organism.4 Common bacterial causes of meningitis (e.g., Streptococcus pneumoniae, Neisseria meningitidis, Haemophilus influenzae) can be detected in more than 75% of cases.4 Lower rates are found in meningitis caused by gram-negative organisms, anaerobes, and Listeria monocytogenes.4

New CSF Markers and Tests

The most exciting new tool in CSF evaluation is PCR analysis, which has revolutionized the diagnosis of specific viral CNS infections and also can be used to detect bacterial organisms. Now more widely available, the PCR technique can detect minute quantities of DNA or RNA in as few as 2-5 hours, and does not depend upon the presence of viable organisms (especially helpful in patients who have received out-patient antibiotics or when fastidious organisms are involved). PCR analysis has a sensitivity and specificity exceeding 95% for detecting herpes simplex encephalitis and has replaced brain biopsy as the gold standard.13 In a recent study of patients with meningococcal meningitis, PCR had a sensitivity and specificity of more than 97%.17 At our institution, the sample is sent to an outside laboratory. PCR results may be available in 24-48 hours, often before culture results, which makes it extremely helpful in the subsequent management of the patient. PCR testing for common viral causes of meningitis (herpesvirus, enterovirus, varicella-zoster, and others) and for specific bacterial organisms may become routine as they are more readily available.

Other diagnostic tests include India ink stain and cryptococcal antigen detection (serum and CSF) for cryptococcal meningitis. Patients with Lyme disease can present with aseptic meningitis and should have detectable serum and CSF antibodies. Latex agglutination and counterimmune electrophoresis for bacterial antigens can be extremely helpful, especially in partially treated cases. Other new markers and diagnostic tests to detect bacterial meningitis are currently under investigation, such as CSF procalcitonin,18 CSF C-reactive protein, and specific immunologic tests for bacterial or mycobacterial antigens and antibodies.1

Dr. Chen is Assistant Professor, University of Pennsylvania Health System, Philadelphia. Dr. Abbuhl, Medical Director, Department of Emergency Medicine, The Hospital of the University of Pennsylvania; Associate Professor of Emergency Medicine, University of Pennsylvania School of Medicine, Philadelphia, is on the Editorial Board of Emergency Medicine Alert.


1. Greenlee JE. Approach to diagnosis of meningitis. Cerebrospinal fluid evaluation. Infect Dis Clin North Am 1990;4:583-598.

2. Weiner HL, Rae-Grant A. Neurology. Philadelphia: Lippincott Williams & Wilkins; 2004.

3. Fishman R. Cerebrospinal fluid in diseases of the nervous system. Second ed. Philadelphia: Saunders; 1992.

4. Graves M. Cerebrospinal fluid infections. Boston: Kluwer; 1989.

5. Hayward RA, et al. Are polymorphonuclear leukocytes an abnormal finding in cerebrospinal fluid? Results from 225 normal cerebrospinal fluid specimens. Arch Intern Med 1988;148:1623-1624.

6. Mayefsky JH, et al. Determination of leukocytosis in traumatic spinal tap specimens. Am J Med 1987; 82: 1175-1181.

7. Roos KL. Pearls and pitfalls in the diagnosis and management of central nervous system infectious diseases. Semin Neurol 1998;18:185-196.

8. Arevalo CE, et al. Cerebrospinal fluid cell counts and chemistries in bacterial meningitis. South Med J 1989;82:1122-1127.

9. Negrini B, et al. Cerebrospinal fluid findings in aseptic versus bacterial meningitis. Pediatrics 2000;105:316-319.

10. Elmore JG, et al. Acute meningitis with a negative Gram’s stain: Clinical and management outcomes in 171 episodes. Am J Med 1996;100:78-84.

11. Powers WJ. Cerebrospinal fluid lymphocytosis in acute bacterial meningitis. Am J Med 1985;79:216-220.

12. Geiseler PJ, et al. Community-acquired purulent meningitis: A review of 1,316 cases during the antibiotic era, 1954-1976. Rev Infect Dis 1980;2:725-745.

13. DeBiasi RL, et al. Use of PCR for the diagnosis of herpesvirus infections of the central nervous system. J Clin Virol 2002;25:S5-11.

14. Powers WJ. Cerebrospinal fluid to serum glucose ratios in diabetes mellitus and bacterial meningitis. Am J Med 1981;71:217-220.

15. Dunbar SA, et al. Microscopic examination and broth culture of cerebrospinal fluid in diagnosis of meningitis. J Clin Microbiol 1998;36:1617-1620.

16. Gutsin DL, et al. Acute bacterial and viral meningitis in adults: Time-sensitive strategies for optimizing patient outcomes. Emerg Med Rep 1998;19:94-105.

17. Richardson DC, et al. Evaluation of a rapid PCR assay for diagnosis of meningococcal meningitis. J Clin Microbiol 2003;41:3851-3853.

18. Viallon A, et al. High sensitivity and specificity of serum procalcitonin levels in adults with bacterial meningitis. Clin Infect Dis 1999;28:1313-1316.