Cerebrospinal fluid analysis: result sheets

• A result sheet for cerebrospinal fluid (CSF) asks the same four questions every time: how fast is the fluid under pressure, what does it look like, how many inflammatory cells are present, and how much protein and glucose are in it.
• Normal CSF is clear and colourless, contains very few cells, very little protein, glucose at roughly half to two-thirds of the blood level, a mildly alkaline pH, a low lactate, and an opening pressure that is comfortable rather than raised.
• Infection changes the sheet in stereotyped ways: bacterial meningitis usually brings a neutrophil-predominant, low-glucose, high-protein, high-lactate picture; viral infection usually brings a lymphocyte-predominant, near-normal-glucose sheet; tuberculosis and fungi behave like a slow-growing bacterial infection with very high protein and low glucose.
• Some non-infectious conditions shout from the sheet: subarachnoid haemorrhage leaves xanthochromia after haemoglobin breakdown; Guillain-Barre syndrome gives very high protein with almost no cells (albuminocytological dissociation); multiple sclerosis adds CSF-restricted oligoclonal bands.
• Three invented, clearly labelled sample sheets are worked through below. Every value on them is illustrative and not real patient data; the ranges, thresholds and rules used to interpret them are quoted from open sources.
• Every range on the sheet belongs to the laboratory that produced it. Read against the printed reference interval, not against any number in this page.

By the end of this page you should be able to:

1. State what opening pressure, appearance, cell count and differential, protein, glucose, lactate, Gram stain, culture and PCR each add to a CSF result sheet.
2. Read a glucose result only in paired form, with the blood glucose taken immediately before the lumbar puncture.
3. Recognise bacterial, viral, tuberculous and fungal patterns in a single look at the sheet.
4. Interpret xanthochromia, albuminocytological dissociation and CSF-restricted oligoclonal bands.
5. Spot traumatic tap and correct (or, more often, recheck) accordingly.
6. Work through three sample sheets step by step and give a one-sentence conclusion and next step for each.
7. Name the IDSA, ESCMID and NICE recommendations that define the routine CSF panel for suspected bacterial meningitis.

CSF is the fluid that bathes the brain and spinal cord. It is produced mainly in the choroid plexus, circulates through the ventricles and the subarachnoid space, and is reabsorbed at the arachnoid granulations. A laboratory examination of it is therefore a direct look at one of the most protected anatomical spaces in the body, and one of the hardest to sample. The interpretation problem is much smaller than the anxiety it generates once it is understood as pattern recognition rather than as a single abnormal line.

This page reads the numbers on the result sheet; it deliberately does not repeat how to do the procedure. The lumbar puncture and CSF topic covers indications, contraindications, the puncture itself and its complications; this page assumes the fluid has already reached the laboratory. Read both together if problems of technique or sample handling are suspected.

The disciplines that shape this panel and the ranges it uses are the same ones that shape the rest of laboratory medicine: regional bodies such as the World Health Organization (WHO) and the Uganda Clinical Guidelines, national ones such as NICE in the United Kingdom, and specialist ones such as IDSA (infectious disease) and ESCMID (clinical microbiology). NICE guideline NG240 (published 2024) is the newest open source for the routine meningitis panel; the points it fixes are set out below ^[1]. Older examples of best practice come from the same guideline bodies. Guideline editions change, and this page records only the editions it used; check the current edition before acting on any recommendation in practice.

We will build up in this order: what each test measures; reference (illustrative) ranges, with a mandatory note about laboratory dependence; causes of high and low values; the named patterns; three invented, clearly labelled sample sheets; and the pitfalls that make a sheet useless. Every number, threshold, cause and rule that is not the sample value itself carries a citation to an open source saved in sources/csf-analysis-result-sheets/.

Opening pressure. Measured when the stylet is withdrawn, before fluid is collected, with the patient relaxed and in the lateral decubitus position (or in sitting position corrected by experience, although the recumbent value is the standard). It is recorded in mmH2O or in cmH2O, and the two differ by a factor of ten (10 mmH2O = 1 cmH2O), which is the commonest reason for a pressure error. It reflects the balance between CSF production and drainage and the current state of the venous sinuses and cerebrospinal compliance, not the thickness of the meninges. Markedly elevated opening pressure is characteristic of idiopathic intracranial hypertension, venous sinus thrombosis, mass lesions with oedema, and subarachnoid haemorrhage; it is also common in cryptococcal meningitis ^[6]. A pressure at or below about 60 mmH2O deserves an explanation too: low-pressure (postural) states after leak, over-drainage and standing headache.

Appearance. The first result on the sheet. Normal CSF is water-clear. Cloudiness suggests cells: a faint opalescence appears once the cell count rises well into the hundreds per microlitre, and obvious turbidity with higher counts. Yellow (xanthochromic) supernatant means degraded haemoglobin or very high protein. Green or pale green-yellow is sometimes seen with pus. A bloody tap looks red initially and clears as the fluid collects; true haemorrhage tends to stay uniformly bloody, and its supernatant may later be xanthochromic. Viscosity can be increased by high protein.

Red cells. Red cells in the first tube are common. With a traumatic tap they usually fall sharply between the first and last tube; with a true subarachnoid haemorrhage they remain roughly constant across all tubes and produce a yellow supernatant. Platelet aggregation or defective filters do not belong here. Expect nil to a few red cells per microlitre in health.

Cell count and differential. Counted with a haemocytometer or automated counter on fresh CSF. The differential is usually done by cytocentrifugation and staining. The normal adult count is very low, and the upper limit used by laboratories ranges from 0 to 5 cells per microlitre. Children and neonates tolerate higher counts; the newborn may have higher than the adult interval, and age-appropriate intervals always apply without infection, which is why age-appropriate intervals exist ^[1]. A neutrophil is abnormal in the absence of infection, trauma or recent convulsion; lymphocytes dominate most viral and tuberculous sheets; a mononuclear sheet occasionally appears in early viral meningitis. Malignant cells, haemophagocytes and epithelial cells are identified by morphology rather than count. ^{[3] [4]}

Protein. Total protein is measured on a clear or lightly spun supernatant. albumin is the largest contributor. Normal CSF protein is low, well under 0.5 g/L (g/L units; some reports use mg/dL, and 0.45 g/L = 45 mg/dL). Protein rises when the blood-brain barrier is leaky, when there is intrathecal inflammation, and when CSF flow is obstructed below the puncture site; it also rises simply by contamination with blood.

Glucose (always with paired serum glucose). CSF glucose is passively transported down its concentration gradient and is usually close to, but lower than, plasma glucose. The result is meaningless without a blood glucose drawn immediately before the lumbar puncture, because metabolic rate, meals and transport time all change the plasma level while the CSF level lags. NICE NG240 fixes this as a paired measurement and asks the blood to be taken immediately before the procedure ^[1]. The ratio (CSF glucose divided by blood glucose) is the more useful number; a ratio below about 0.4 is a strong pointer to a bacterial or tuberculous process ^[3].

Lactate. Lactate is a late marker of meningeal or brain cell metabolism, and it rises with anaerobic glycolysis within the CSF, particularly in bacterial and tuberculous meningitis but also after seizures, stroke and subarachnoid haemorrhage. Normal CSF lactate is low, typically under 2 mmol/L; values above 4.0 mmol/L are strongly suggestive of bacterial meningitis and values above 5.5 mmol/L of tuberculosis ^[3]. It is little changed by blood contamination and is therefore one of the most robust markers when the sheet is otherwise confusing ^[3].

Gram stain and culture. The Gram stain remains the fastest way to a diagnosis in acute bacterial meningitis, but it depends heavily on pretreatment and volume and is negative in a third or more of cases. Culture gives the organism, the sensitivities and sometimes the serogroup, and needs to be set up promptly from an adequate volume. NICE NG240 asks for both microscopy and culture and sensitivity ^[1].

Polymerase chain reaction (PCR). Multiplex PCR panels and pathogen-specific PCR identify organisms after (or despite) antibiotics, and may be positive when culture is negative. NICE NG240 includes PCR for relevant pathogens among the routine CSF investigations ^[1]. Interpret a positive PCR with the clinical picture: it may represent live infection, infection after a very short treatment course, or colonisation or shedding, depending on the organism and the assay.

Xanthochromia (spectrophotometry). Spectrophotometry for bilirubin is the reliable way to tell a real subarachnoid bleed from a blood-stained sample, because bilirubin forms from haemoglobin breakdown over hours and then persists for days. Plain visual inspection is not enough. NICE recommends that a lumbar puncture for the diagnosis of subarachnoid haemorrhage be done at least 12 hours after symptom onset, so that bilirubin has had time to form ^[2].

Oligoclonal bands (OCB). Oligoclonal bands are immunoglobulin bands seen in isoelectric focusing of CSF but not of paired serum. Two or more CSF-restricted bands demonstrate intrathecal immunoglobulin synthesis. Their classic associations are multiple sclerosis and a long list of chronic infections and inflammatory disorders of the nervous system. The 2024 revision of the McDonald criteria also accepts the kappa free light chain index (a value of 6.1 or more as the proposed cut-off) as an alternative marker ^[5]. A negative OCB result does not exclude multiple sclerosis in a patient who otherwise meets the criteria, and a positive result does not, by itself, make the diagnosis; clinical and imaging evidence must also be present.

Other special tests. Depending on the indication: adenosine deaminase (ADA) for tuberculous meningitis, cryptococcal antigen (CrAg) or India ink for fungal disease, flow cytometry for malignant cells, cytology for malignant infiltrate, virology by PCR for enteroviruses and herpesviruses, and, when demyelination is suspected, the IgG index and kappa free light chain index. The named reference ranges for these come from their method, not from this page.

The intervals below are one series of adult examples drawn from sources opened in writing this page. They are shown in SI units with conventional units in brackets. They are not your laboratory's ranges. These tests are measured in different ways, in different physical phases of the fluid, and against different reference populations; the intervals change with age, sex, pregnancy, prior illness, the laboratory method and the collection tube. The reporting laboratory's printed range always wins. Neonates, children, older adults, pregnant patients and post-neurosurgical patients have their own intervals and, often, their own rules for a traumatic tap.

A plain note repeating the rule, because it is the one thing readers get wrong on an exam or at the bedside: the laboratory's own reference interval, with the method and population it used, is the only interval that may be called normal for this patient. If a range in a textbook, a menu, or this page disagrees with the laboratory report, the report's range wins ^[4]. Paediatric and pregnancy intervals for this fluid are very different from adult ones, and neonates in particular tolerate far higher cell counts without infection ^[1].

The thresholds used later in interpretation (not the raw intervals) are:

• Glucose under 2.2 mmol/L or a CSF:serum ratio under 40 to 50% is read as a decreased CSF glucose ^[3].
• Lactate above 5.5 mmol/L is a strong marker of tuberculous meningitis (and high bacterial lactate is also expected) ^[3].
• Protein above 1 g/L, frequently 1 to 3 g/L is typical of tuberculous meningitis ^[3].
• Lymphocytic pleocytosis of 10 to 500 cells per microlitre, with more than half lymphocytes, is compatible with tuberculosis ^[3].
• A raised opening pressure above 200 mmH2O was present in most patients in a published cryptococcal meningitis cohort ^[6].

The most useful causes to carry with you:

• High protein without many cells is the classic albuminocytological dissociation of Guillain-Barre syndrome, where protein is raised (commonly 1 to 10 g/L) while the white cell count stays normal, under 5 cells per microlitre, although a count under 50 cells per microlitre remains compatible with the diagnosis (Brighton criterion) ^{[4] [7]}. Nerve conduction studies confirm it; not the sheet alone.
• Normal or nearly normal protein with low glucose is so unusual that it forces re-sampling: it occurs in hypoglycaemia, in Mollaret (recurrent) meningitis, or in assay error.
• A traumatic tap adds whole blood, so it raises white cells proportional to the blood haematocrit and inflates the protein and glucose in the same proportion. The exact correction is unreliable (it depends on the haemoglobin and reabsorption), so recheck on a fresh sample and read the biodependent markers (lactate, PCR, xanthochromia) rather than trusting a corrected number.
• Cell count and glucose fall with antibiotic pre-treatment. NICE NG240 explicitly warns that the diagnostic reliability of every CSF parameter drops after antibiotics, which is why an early safe lumbar puncture is strongly preferred ^[1].
• Very high protein (well above 1 g/L) in meningitis should make you think of tuberculosis (1 to 3 g/L), cryptococcosis, fungal disease, malignant seeding, established bacterial meningitis or subarachnoid block, and (when chronic) syphilis.
• Neonates and very young infants require different thresholds, because normal white cell counts and protein are higher than adult intervals in the first weeks of life (NICE notes that normal thresholds may be higher in babies under 3 months) and glucose ranges differ ^[1].

The common named combinations and what each suggests:

• Normal spectrum: pressure normal, clear fluid, white cells under 5 per microlitre, protein 0.18 to 0.45 g/L, glucose normal with ratio above 0.5, lactate under 2 mmol/L, no xanthochromia, no CSF-restricted OCBs. The bacterial panel is negative and the PCR is negative. Conclusion: no meningeal infection; look at the specimen quality and the clinical picture before settling on "normal" if symptoms are pointing that way.
• Bacterial meningitis (overt): CSF opacity or turbidity; white cell count raised, usually neutrophil-predominant; CSF glucose reduced (often below 2.2 mmol/L, or a CSF:serum ratio below 0.4); protein raised, often above 1 g/L; lactate raised, often above 4 mmol/L; Gram stain positive in roughly two thirds with adequate volume and before antibiotics. If the aspiration was done after antibiotics started, all of these fall toward normal; a second (they are not available) sample or the clinical picture must be relied on ^{[1] [3]}.
• Viral (aseptic) meningitis: clear or only slightly opalescent; white cell count raised modestly, mononuclear or mixed, usually in low hundreds, and never the millions seen in bacteria; CSF glucose preserved at roughly half to two-thirds of blood; protein normal or mildly high; lactate in or near the normal range. Enterovirus, herpes simplex 2, varicella-zoster and HIV are common causes. CSF is the only way to separate viral from bacterial in the first hours. ^[3]
• Tuberculous meningitis: clear or opalescent fluid; lymphocytic pleocytosis of 10 to 500 cells per microlitre, sometimes neutrophils in the first few days; glucose low, under 2.2 mmol/L or ratio under 40 to 50%; protein high, 1 to 3 g/L; lactate high, above 4 to 5.5 mmol/L; opening pressure often raised ^[3]. Gram stain and AFB smear are usually negative; culture is positive slowly (weeks); Xpert Ultra PCR is the faster diagnostic step per WHO, but its yield depends on volume and handling ^[3]. Chest signs or extrapulmonary TB support the diagnosis, and MRI of the basal meninges may be abnormal. A raised ADA and high IFN-gamma support, but do not prove, the diagnosis.
• Fungal / cryptococcal meningitis: often subacute lymphocytic; glucose reduced or near normal; protein raised; opening pressure very high in a proportion, including values above 200 mmH2O in one published cohort ^[6]. India ink or CrAg (cryptococcal antigen) in CSF and blood are the useful tests. Characteristic in advanced HIV infection, but also occurs in transplant and immunocompetent patients, especially with Cryptococcus gattii.
• Partial (or antibiotic-pretreated) bacterial meningitis: the sheet may be almost normal or a mix, with neutrophils in small numbers, a mildly raised protein and normal glucose. Whenever antibiotics have been given before sampling, the sheet is less reliable and audited against the blood CRP, procalcitonin and the kinetics of the illness, not against normal ranges ^[1].
• Guillain-Barre syndrome: albuminocytological dissociation. Opening pressure normal. Glucose, cell count and lactate normal; protein high, often over 1 g/L and commonly 1 to 10 g/L, with a white cell count under 5 per microlitre (under 50 per microlitre remains compatible with the diagnosis) ^{[4] [7]}. The CSF is usually clear. The statement "high protein with a normal cell count" is one of the most useful single-line diagnoses in neurology ^[4]. ^[1]
• Multiple sclerosis. Clear fluid; cell count 0 to a few per microlitre, a mild lymphocytic pleocytosis, protein in or close to the normal range, normal glucose, lactate probably normal; CSF-restricted oligoclonal bands or kappa FLC index of 6.1 or more are the characteristic abnormality ^[5]. If demyelination is the question, the IgG index is often mildly raised. The CSF is not the diagnostic test for MS on its own; it supports an MRI diagnosis made by imaging criteria.
• Subarachnoid haemorrhage. Opening pressure often high. CSF uniformly bloody in all tubes; spectrophotometry for bilirubin (xanthochromia) positive when the sample is taken at least 12 hours after symptom onset; supernatant yellow. A negative troponin-to-brain paragraph: red cells in the CSF do not diagnose SAH unless the characteristics above are present. NICE directs an LP after a negative CT when symptoms are more than 6 hours old, and waits at least 12 hours before performing it ^[2].
• Traumatic tap without any underlying disease: first-tube red cells with a low but positive white cell count that falls away over three tubes; protein and glucose altered only in proportion to the added blood; no xanthochromia and no bilirubin peak on spectrophotometry; and no bacterial signal on stain or PCR.
• Idiopathic intracranial hypertension (benign intracranial hypertension): high opening pressure, sometimes very high, with normal CSF chemistry and cells. The diagnosis is by exclusion, imaging and eye examination; treatment is not a CSF-pathology treatment.
• Autoimmune encephalitis / parainflammatory disease: CSF may be entirely normal, or show a mild lymphocytic pleocytosis, variably raised protein and normal glucose; oligoclonal bands can be present.

Every sheet below is an illustrative sample invented for teaching. None of it is real patient data, a real laboratory report, or any patient of any kind. The sample numbers are not claims; the ranges, thresholds and rules used to interpret them are. Sample results are never evidence.

Sample sheet 1 - normal. Conceived as a healthy 32-year-old person with a normal neurological examination; not a real patient.

Interpretation:

1. Opening pressure 120 mmH2O is within the range usually cited for adults and, with a normal chemistry and cell count, this is a normal-pressure, non-inflammatory picture. Local criteria for raised pressure still apply.
2. The fluid is clear, colourless, with no red cells; there is no sign of blood contamination or bleed ^[4].
3. White cells 2 per microlitre and a fully mononuclear differential sit inside the example adult interval of 0 to 5 cells per microlitre. The sheet should be read as non-inflammatory ^[4].
4. Protein 0.35 g/L sits inside the example interval of 0.18 to 0.45 g/L ^[4].
5. Glucose 3.4 mmol/L sits inside the example interval of 2.5 to 3.5 mmol/L, and the paired serum glucose is 5.6 mmol/L, giving a ratio of 0.61, above the reference cut-off ^{[3] [4]}.
6. Lactate 1.6 mmol/L is within the interval reported for this fluid by the laboratory used; the stain, culture and molecular panel are all negative.
7. Xanthochromia is absent and there are no CSF-restricted oligoclonal bands, which excludes haemorrhage and intrathecal synthesis on this sample.

Conclusion: this is a normal cerebrospinal fluid result sheet in an apparently normal adult. No further CSF investigation is warranted from these results alone, and retesting is not indicated unless new symptoms develop ^[1].

Sample sheet 2 - bacterial meningitis. Conceived as a 19-year-old residential student with fever, headache and photophobia; not a real patient.

Interpretation:

1. Opening pressure 240 mmH2O is raised by the usual adult criteria and fits acute meningitis; combined with a turbid appearance this is not a normal tap, although the collection itself was never traumatic and the result is not an artefact ^[6].
2. Red cells 15 per microlitre are few enough to accept as a slightly bloodstained sample rather than as a true haemorrhage; the absence of xanthochromia supports this ^[4]. ^[2]
3. White cells 2,650 per microlitre with over 90% neutrophils is a strongly neutrophilic picture: far above the adult reference interval and the clear-cut signature of bacterial meningitis rather than viral ^[1].
4. Protein 2.10 g/L is far above 1 g/L and well above the example interval of 0.18 to 0.45 g/L, consistent with a ruptured meningeal barrier in bacterial meningitis ^[4].
5. Glucose 0.9 mmol/L is under the 2.2 mmol/L threshold and the CSF:serum ratio of about 0.15 is far below 0.4, both matching a heavily metabolising bacterial process ^[3].
6. Lactate 5.8 mmol/L lies well above the 4.0 mmol/L marker of bacterial meningitis; because lactate is little affected by blood contamination, the red cells do not invalidate it ^[3].
7. Gram stain shows gram-negative diplococci and the PCR is positive for relevant pathogens (meningococcus consistent); the culture may take longer to confirm the organism and sensitivities ^[1].
8. Xanthochromia absent excludes haemorrhage, and there are no CSF-restricted oligoclonal bands.

Conclusion: this sheet is classic bacterial (likely meningococcal) meningitis: raised pressure and neutrophilic pleocytosis, low glucose, raised protein and lactate, and a positive PCR. The next step is immediate intravenous antibiotic and adjunctive steroid therapy according to local and IDSA/ESCMID/NICE guidance, isolation precautions, notification and an urgent senior review; antibiotics must not be delayed to wait for culture ^[1].

Sample sheet 3 - tuberculous meningitis. Conceived as a 44-year-old adult with subacute headache, weight loss and mild fever; not a real patient.

Interpretation:

1. Opening pressure 220 mmH2O is raised and the fluid is opalescent rather than turbid; together with the subacute time course this is more suggestive of a chronic meningeal process than of pyogenic meningitis ^{[3] [6]}.
2. Red cells 8 per microlitre are few and look like minor blood admixture rather than true haemorrhage, supported by negative xanthochromia ^[4].
3. White cells 185 per microlitre with 82% lymphocytes is a lymphocytic pleocytosis: well above the 10 to 500 cells per microlitre compatible with tuberculosis, and a very different direction of change from the neutrophilic picture of bacterial meningitis ^[3].
4. High protein at 1.85 g/L (often above 1 g/L, frequently 1 to 3 g/L in tuberculosis) points to a thick basal exudate and fits the picture ^[3].
5. Glucose 1.5 mmol/L is below 2.2 mmol/L and the CSF:serum ratio is 0.25, far below 0.4, both pointing to heavy mycobacterial metabolism ^[3].
6. Lactate 6.2 mmol/L is above the 5.5 mmol/L marker quoted for tuberculous meningitis. The number is compatible, not diagnostic: cryptococcal and bacterial meningitis can also raise lactate above 4 mmol/L ^[3].
7. ADA is mildly raised at 7.8 U/L, consistent with higher CSF ADA activity in tuberculosis than in viral or cryptococcal meningitis, although no universal cut-off exists and the laboratory's own cut-off should be used ^[3].
8. Gram stain and AFB smear are negative; Xpert Ultra PCR is positive, which is the single most useful rapid test named in the WHO guidance and is consistent with definite disease ^[3].
9. xanthochromia absent and no CSF-restricted oligoclonal bands complete the excluded differentials.

Conclusion: this sheet is compatible with definite tuberculous meningitis - a lymphocytic CSF with very low glucose and high protein and lactate, plus a positive Xpert Ultra. The adult next step is anti-tuberculous therapy with adjunctive corticosteroids plus appropriate management of raised intracranial pressure and partner tracing, following WHO and national guidance, with close monitoring for hydrocephalus and stroke. Sample values are illustrative; clinician-directed treatment decisions rest on the clinical picture and local epidemiology.

• Antibiotics before the tap. NICE NG240 warns explicitly that earlier antibiotics reduce the diagnostic reliability of every CSF parameter ^[1]. A normal-ish sheet after intravenous ceftriaxone does not exclude bacterial meningitis. Repeat sampling and heavier reliance on blood PCR or other microbiological methods may be needed.
• No paired glucose. A CSF glucose on its own is uninterpretable, and a person with systemic hyperglycaemia can look to have a high CSF glucose while their ratio is already falling. Take the blood immediately before the procedure ^[1].
• Traumatic tap. The first tube is the one most distorted. If tuberculosis, cryptococcosis or malignant infiltration is the suspicion, send a later tube for cytology and reserve DNA for a tube clear of blood. A genuinely blood-stained tap ruins cell differentials and reasonably reduces the protein reading. Rely on lactate (little changed by blood), PCR (which still detects organisms), and the clinical picture instead.
• Looking for xanthochromia too early. Xanthochromia only becomes reliable once enough time has passed: NICE waits at least 12 hours after symptom onset before performing a diagnostic lumbar puncture ^[2]. An LP done within the first few hours may be negative and must be repeated if clinical suspicion remains ^[2].
• Using the wrong units for pressure. The pressure is written in mmH2O, and a misplaced decimal or a confusion with cmH2O changes it tenfold. Record position, the height reading, and conversion in every clinical note.
• Calling a normal number "normal". A white cell count of 4 in an immunosuppressed patient with overwhelming cryptococcal meningitis is not reassuring if the protein, glucose and lactate disagree. Cell counts can be surprisingly low in fungal, tuberculous and partially treated bacterial disease. The sheet is a set of linked clues, not a single line.
• Assuming culture is sufficient. Culture can take days and may be negative, particularly after antibiotics. PCR, Gram stain and paired blood cultures add independent information. NICE NG240 lists all three kinds of investigation for this reason ^[1].
• Storing or transporting mishap. Delay between collection and analysis degrades glucose, strips red cells and alters cell counts. NICE NG240 asks that cell counts, total protein and glucose be available within four hours of lumbar puncture, and reminds the clinician to store the remaining CSF for later tests ^[1].
• Interpreting post-tap leaks as abnormalities. Re-sampling a leaky, recently punctured patient can produce haemorrhagic and inflammatory artefacts for days. A stable re-sits level per week and experience is needed.
• Confusing blood contamination with anaemia. A slightly reddish supernatant is a clue to recent red blood cells but requires centrifugation to settle; a protein that rises out of proportion suggests true vasculature rather than gross red-cell mass. Send both tubes to the laboratory and let the haematologist sort it out.
• Forgetting geography and age. Cryptococcal meningitis and neurotuberculosis are far commoner in some parts of the world and in certain immunocompromised groups; viral meningitis is common in children and joggers; meningococcal disease clusters in young adults in dormitories. A sheet is never read without the patient.
• Assuming a normal CSF excludes central nervous system infection in an immunosuppressed patient, an early infection, a partially treated patient, or a deep brain abscess with a reactive meningeal sheet. Repeat and reassess.

• CSF examination is interpreted as one sheet, not as any single value, and is read in the context of the history, examination and systemic blood tests.
• Normal adult CSF is clear, nearly acellular (under about 5 white cells per microlitre, all mononuclear), with protein of 0.18 to 0.45 g/L and glucose at roughly half to two-thirds of plasma glucose ^{[1] [3] [4]}.
• The lumbar puncture and CSF reference page covers technique and safety; this one covers interpretation, and the two complement rather than replace each other.
• NICE NG240 (published 2024) lists the routine minimum panel for suspected bacterial meningitis: white and red cell count with differential, total protein, glucose with paired serum glucose, Gram stain, culture and sensitivity, and PCR for relevant pathogens; the clinically important tests must be returned within four hours ^[1].
• CSF glucose counts only when drawn in parallel with blood glucose taken immediately before the procedure ^[1].
• Bacterial meningitis usually shows a neutrophil-predominant pleocytosis, reduced glucose (under 2.2 mmol/L or ratio under 0.4), raised protein and raised lactate; lactate above 4 mmol/L is strongly suggestive of bacterial disease ^{[1] [3]}.
• Viral meningitis leaves glucose normal and protein near normal, with a lymphocytic or mixed pleocytosis and lactate in the normal range.
• Tuberculous meningitis behaves subacutely: 10 to 500 lymphocytes, glucose under 2.2 mmol/L or ratio under 40 to 50%, protein 1 to 3 g/L, lactate above 4 to 5.5 mmol/L, with a positive Xpert Ultra as the gold standard rapid test ^[3].
• Cryptococcal meningitis is a slow lymphocytic, low-glucose picture with very high pressure in many patients; diagnosis is by CrAg and culture ^[6].
• Guillain-Barre syndrome gives albuminocytological dissociation: high protein with a normal or near-normal cell count and normal pressure ^[4].
• Multiple sclerosis gives a normal small-volume CSF with CSF-restricted oligoclonal bands or an elevated kappa free light chain index of 6.1 or more ^[5].
• Xanthochromia on spectrophotometry, with the sample taken at least 12 hours after onset, is the NICE-recommended way to confirm a subarachnoid haemorrhage ^[2].
• Raised opening pressure with a normal sheet points to idiopathic intracranial hypertension; negative xanthochromia plus red-cell-contaminated blood points to a traumatic tap.
• Prior antibiotics make every number less reliable; a normal sheet after treatment does not exclude infection ^[1].
• Ranges differ by laboratory, age, sex and method; the laboratory's own range always wins ^[4].
• All of the guidance cited was published between 2024 and 2026 and should be re-checked before clinical use; IDSA, ESCMID and NICE all publish variants of the meningitis pathway and may differ on antibiotic and steroid details.

Related topics, as plain text until they are linked in this build: lumbar puncture and cerebrospinal fluid interpretation (the indications, contraindications and technique), idiopathic intracranial hypertension, bacterial meningitis, viral meningitis, tuberculous meningitis, cryptococcal meningitis, Guillain-Barre syndrome, multiple sclerosis, subarachnoid haemorrhage, abnormal liver function tests (a result-sheet exercise in the same style), serum electrolytes (another result-sheet exercise), reading a laboratory report, acquired intracranial infection, hydrocephalus, brain abscess.