Lumbar puncture (LP) is one of the highest-value procedures in clinical neurology — diagnostic for meningitis, encephalitis, multiple sclerosis, neurosyphilis, idiopathic intracranial hypertension, subarachnoid hemorrhage (when imaging is non-diagnostic), and many neurodegenerative diseases via emerging CSF biomarkers. It is also one of the most commonly mis-performed neurology procedures, with avoidable post-LP headache being the dominant complication. This page covers LP technique, post-procedure care, contraindications and safety, and the interpretation of a normal CSF profile — the baseline against which every pathologic CSF is read.

🔹 Bottom Line: LP and Normal CSF

  • Normal adult opening pressure ≈ 6–25 cm H₂O in lateral decubitus, interpreted with age, body habitus, and technique. The older 6–20 range is too tight and overcalls intracranial hypertension. IIH threshold is opening pressure >25 cm H₂O with full clinical context.
  • Atraumatic (pencil-point) needles roughly halve post-LP headache vs cutting (Quincke) needles. AAN-recommended default for diagnostic LP.
  • Image first when: focal deficit, altered consciousness, new seizure, papilledema, or immunocompromise. Otherwise CT is not required before LP.
  • Normal CSF: clear, ≤5 WBC (lymphocyte/monocyte predominant), glucose 50–80 mg/dL (or ≥60% of serum), protein 15–45 mg/dL, no RBC, lactate <2.2 mmol/L.
  • Traumatic tap vs SAH: RBC dropping markedly tube 1→4 = traumatic; persistent RBC + xanthochromia (≥6 h after onset) = true SAH.
  • Suspected bacterial meningitis: don’t delay antibiotics for LP. Cultures often remain positive for the first 1–2 hours after antibiotics.

Indications

  • Diagnostic: meningitis / encephalitis; subarachnoid hemorrhage when CT is non-diagnostic and within an appropriate time window; demyelinating disease (MS, NMOSD, MOGAD); autoimmune and paraneoplastic encephalitis (antibody panels); neurosyphilis; chronic meningitides (TB, fungal, carcinomatous); idiopathic intracranial hypertension (opening pressure); CJD / prion disease (14-3-3, RT-QuIC); neurodegenerative biomarkers (AD: Aβ42/40, p-tau, t-tau); cytology for leptomeningeal carcinomatosis or CNS lymphoma.
  • Therapeutic: IIH (CSF removal); intrathecal medication (nusinersen for SMA, chemotherapy for leptomeningeal disease, baclofen via pump, antibiotics); CSF leak diagnosis and blood patch.

Contraindications and Safety

Absolute or Strong Relative

  • Increased intracranial pressure from a mass lesion with risk of brain herniation (especially with midline shift, posterior fossa mass, obstructive hydrocephalus, large posterior fossa lesion). Get imaging first when clinically suspected.
  • Coagulopathy: INR >1.4–1.5; platelets <50–80 × 10⁹/L (lab thresholds vary); active anticoagulation (DOAC, therapeutic heparin, recent thrombolysis). Correct or wait for appropriate clearance window.
  • Skin infection at the puncture site (risk of seeding the subarachnoid space).
  • Suspected spinal epidural abscess at the planned level.

When to Image First

  • Focal neurologic deficit.
  • Altered consciousness or new-onset confusion.
  • New-onset seizure within the last week.
  • Papilledema.
  • Immunocompromise (HIV, transplant, severe immunosuppression).
  • History of CNS disease that may obstruct CSF flow.
  • If imaging is needed but delays antibiotics, give empiric antibiotics before imaging in suspected bacterial meningitis.

LP Technique

Position

  • Lateral decubitus position is the only position that gives a reliable opening pressure measurement. Hips and knees flexed; back perpendicular to the bed; the line connecting the iliac crests (Tuffier’s line) passes through approximately the L4 spinous process.
  • Sitting position can be used for technically difficult LPs (obesity, scoliosis) but does not yield a valid opening pressure.

Anatomy

  • Adult spinal cord ends at the conus medullaris near L1–L2; cauda equina fills the thecal sac below.
  • Safe needle insertion levels in adults: L3–L4 or L4–L5 (occasionally L2–L3 in narrow pelvis).
  • In children, the conus is lower at birth; use L4–L5 or L5–S1 in infants.
  • Needle traverses: skin → subcutaneous fat → supraspinous ligament → interspinous ligament → ligamentum flavum (firm “pop”) → epidural space → dura → arachnoid → subarachnoid space.

Needle Choice (Most Important Modifiable Risk Factor)

  • Atraumatic (pencil-point) needles (Sprotte, Whitacre): substantially lower post-LP headache rate than cutting (Quincke) needles. Preferred whenever available.
  • Smallest practical gauge: 22G or 24G atraumatic preferred for diagnostic LP; reserve 20G when CSF flow is poor or large volume needed quickly.
  • If a cutting (Quincke) needle is used: orient the bevel parallel to the long axis of the spine (i.e., bevel facing the patient’s flank) to separate rather than cut dural fibers.

🔹 Clinical Relevance: Atraumatic Needles Halve Post-LP Headache

Head-to-head trials and meta-analyses show that pencil-point (Sprotte, Whitacre) needles roughly halve the rate of post-LP headache versus traditional cutting (Quincke) needles, and AAN guidance now recommends atraumatic needles as the default for diagnostic LP. Cutting needles still dominate in many inpatient and ED procedure trays, often because operators trained on Quincke needles and atraumatic needles feel different — there’s no clear “dural pop” (the pencil tip separates rather than cuts dural fibers) and CSF flow can be slower.

  • Stock atraumatic needles in every neurology and ED procedure tray.
  • Teach the introducer technique: a short introducer through skin/SQ tissue, then the atraumatic needle through the introducer.
  • Reserve cutting needles for situations where CSF flow with an atraumatic needle is genuinely inadequate.

Procedure Sequence

  1. Position the patient; identify Tuffier’s line; mark the intended interspace.
  2. Sterile prep; drape; sterile gloves.
  3. Local anesthetic (1–2% lidocaine) to skin and deeper tissues along the planned trajectory.
  4. Insert the spinal needle through the marked interspace, angled slightly cephalad (toward the umbilicus), with stylet in place.
  5. Advance through the ligaments; a subtle change in resistance often signals the ligamentum flavum and dural penetration. Withdraw the stylet to check for CSF flow at intervals.
  6. Once CSF flows: attach the manometer (in lateral decubitus only) and measure opening pressure.
  7. Collect 4 sequential tubes (typical adult: ~2–4 mL each), labeled 1–4. Tube 1 usually for cell count / chemistry, tube 4 for cell count comparison (to distinguish traumatic tap from true blood), tubes 2–3 for microbiology, cytology, antibody panels, or saved CSF.
  8. Replace the stylet before withdrawing the needle (evidence suggests this reduces post-LP headache).
  9. Apply a sterile dressing.

Tube-Number Convention

Tube Typical use
1 Cell count and differential; glucose; protein
2 Microbiology (Gram stain, bacterial culture, viral PCR, fungal, AFB)
3 Special studies (OCB / IgG index, antibody panels, cytology, flow cytometry, 14-3-3, RT-QuIC, AD biomarkers, lactate)
4 Repeat cell count (to compare to tube 1 — distinguishing traumatic tap from true RBC)

Lab-specific tube assignments vary; always confirm with the receiving lab.

Opening Pressure

  • Measured in lateral decubitus only; patient relaxed and legs partially extended (sustained Valsalva or knee-to-chest position falsely elevates the reading).
  • Normal adult: roughly 6–25 cm H₂O in lateral decubitus, interpreted with age, body habitus, and technique. The historical “6–20” range is too tight and risks overcalling intracranial hypertension on borderline LPs.
  • Elevated >25 cm H₂O: supports IIH (no mass, normal CSF, normal-to-large ventricles, papilledema), meningitis, hemorrhage, CVT, mass effect — always interpret with the full clinical context.
  • Low <6 cm H₂O: spontaneous intracranial hypotension / CSF leak; over-shunting; severe dehydration.
  • Pediatric thresholds are different (lower in young children; ≥28 cm H₂O typically used as elevated in children >1 year).

Normal Adult CSF

Parameter Normal range Notes
Opening pressure ~6–25 cm H₂O Lateral decubitus, relaxed; modern reference range. IIH threshold >25 cm H₂O with clinical context.
Appearance Clear, colorless Cloudy = WBC, RBC, protein, or microbes; xanthochromic = bilirubin/oxyhemoglobin after SAH
WBC ≤5 cells/µL Predominantly lymphocytes/monocytes; neutrophils essentially absent in normal CSF
RBC 0 (none acceptable in atraumatic tap) Persistent RBC across tubes 1→4 suggests true bleed
Glucose 50–80 mg/dL, OR ≥60% of serum glucose Always interpret with simultaneous serum glucose
Protein 15–45 mg/dL Higher in lumbar than cisternal CSF; rises with age
Lactate <2.2 mmol/L Elevated in bacterial meningitis; not elevated in viral
IgG index ≤0.66 Elevated in MS, other inflammatory CNS disease
Oligoclonal bands (OCB) None unique to CSF ≥2 unique CSF bands = pathologic; MS, neurosyphilis, SSPE, chronic CNS infection

Traumatic Tap vs True Blood (SAH)

  • Traumatic tap: RBC count decreases substantially from tube 1 to tube 4 (often by >50–75%); CSF supernatant is colorless after centrifugation.
  • True subarachnoid blood: RBC count stable across tubes; supernatant is xanthochromic (yellow-pink) due to bilirubin breakdown of hemoglobin. Xanthochromia appears 6–12 hours after bleed and can persist 2–4 weeks.
  • Spectrophotometry is more sensitive than visual inspection for xanthochromia and is used in some centers for delayed SAH workup.
  • Correction formulas (e.g., subtracting WBC and protein proportional to RBC contamination) help interpret cell count and protein when blood is present.

Post-LP Headache

Mechanism

  • Persistent CSF leak through the dural puncture site lowers CSF volume / pressure.
  • Brain sags caudally in the upright position, traction on pain-sensitive structures (dural sinuses, bridging veins, trigeminal afferents).
  • Compensatory vasodilation may contribute.

Clinical Features

  • Bilateral, often frontal or occipital headache.
  • Positional — worse on standing, better on lying flat (key diagnostic feature).
  • Onset within 48 hours; usually resolves within 1 week.
  • Associated: neck stiffness, nausea, tinnitus, hearing changes, mild photophobia.
  • Rarely: subdural hematoma (especially in older adults or after prolonged untreated leak).

Risk Factors

  • Cutting (Quincke) needle (largest modifiable risk).
  • Larger needle gauge.
  • Multiple attempts.
  • Younger age (peak 20s–40s).
  • Female sex.
  • Low BMI.
  • History of prior post-LP headache.

Prevention

  • Atraumatic (pencil-point) needles: the most effective single intervention; meta-analyses consistently show ~50% reduction in post-LP headache rate.
  • Smallest practical gauge.
  • Bevel orientation parallel to spine if cutting needle is used.
  • Replace stylet before withdrawing needle.
  • Bed rest after LP does not prevent post-LP headache (multiple RCTs); routine post-procedure bed rest is no longer recommended.
  • Hydration does not have proven preventive benefit either, though it is reasonable.

Treatment

  • Conservative: lying flat, oral hydration, simple analgesics (acetaminophen, NSAIDs).
  • Caffeine: 300–500 mg orally or IV; helps a subset; evidence modest.
  • Theophylline: alternative methylxanthine.
  • Epidural blood patch: 15–20 mL of autologous blood injected into the epidural space at the LP site; gold standard for persistent or severe post-LP headache; success rate ~70% with first patch, higher with repeat.

Other LP Complications

  • Cerebral herniation: rare but catastrophic. Risk in unrecognized mass lesion or severe ICP elevation. Image-first selectively (see indications above).
  • Epidural / subdural hematoma: especially in coagulopathy or anticoagulation; usually self-limiting but can cause cauda equina syndrome if large.
  • Infection: meningitis if procedure unsterile or LP performed through infected skin. Rare with proper technique.
  • Nerve root irritation: transient lancinating leg pain during needle placement (very common, generally benign — reposition needle slightly).
  • Cauda equina syndrome: rare; from epidural hematoma or rarely from needle injury.
  • Persistent CSF leak: can require surgical repair if recurrent epidural blood patches fail.

LP in Special Situations

Anticoagulation / Antiplatelet

  • Warfarin: reverse to INR ≤1.4; PCC + vitamin K if urgent.
  • DOACs: hold for ≥24–48 hours depending on agent and renal function; idarucizumab (dabigatran) or andexanet alfa (Xa inhibitors) if urgent.
  • Therapeutic heparin: stop and verify aPTT normal; LMWH at least 12–24 hours after last dose.
  • Aspirin alone: generally acceptable without hold; do not delay urgent LP.
  • Clopidogrel: hold 5–7 days for elective; weigh urgency against bleeding risk.
  • Recent thrombolysis: avoid LP for at least 24 hours after tPA.

Coagulopathy

  • Platelets <50 × 10⁹/L: transfuse before LP if non-urgent.
  • INR >1.4: correct with vitamin K, PCC, or FFP.
  • Severe thrombocytopenia from ITP or TTP: individualized risk-benefit.

Suspected Bacterial Meningitis

  • Do not delay antibiotics waiting for LP. Empiric antibiotics + dexamethasone before LP if there will be any delay.
  • Blood cultures before antibiotics if possible.
  • Image-first only if focal deficit, altered consciousness, new seizure, papilledema, or immunocompromised — and even then, antibiotics should not wait.

Suspected SAH

  • Non-contrast CT within 6 hours: ~100% sensitive — LP may not be needed.
  • CT >6 hours after onset: LP with spectrophotometric or visual xanthochromia assessment is still the standard for delayed presentations.

Idiopathic Intracranial Hypertension (IIH)

  • Document opening pressure carefully in lateral decubitus.
  • Diagnostic threshold: opening pressure >25 cm H₂O in adults (≥28 cm H₂O in children) with normal CSF composition and normal imaging (small or normal ventricles).
  • Therapeutic CSF removal can transiently improve symptoms; not a substitute for definitive management (acetazolamide, weight loss, surgical CSF diversion).

Pediatric LP

  • Insertion at L4–L5 or L5–S1 (conus is more caudal in infants).
  • Smaller needles (22–25G).
  • Topical anesthesia (EMLA) helpful.
  • Side position with infant flexed; avoid hyperflexion of neck (compromises airway).
  • Sedation in young children may improve safety and success.

Reporting a Normal CSF Profile

A normal CSF report should specify:

  • Indication for LP.
  • Technique (patient position, needle type and gauge, attempts, level).
  • Opening pressure (lateral decubitus only).
  • Appearance of each tube.
  • Cell counts (WBC, RBC) for tubes 1 and 4 if both sent.
  • Differential of WBC.
  • Glucose (with simultaneous serum glucose).
  • Protein.
  • Specialized studies sent and results when available.
  • Interpretation in clinical context.

Pitfalls and Pearls

  • Image-first when: focal deficit, altered consciousness, new seizure, papilledema, immunocompromised. Otherwise, image-first is not required.
  • Antibiotics before LP if there is any delay in suspected bacterial meningitis. Cultures may still be positive within 1–2 hours.
  • Lateral decubitus for opening pressure: sitting position falsely elevates it.
  • Atraumatic needle: ~50% reduction in post-LP headache; should be the default.
  • Smallest practical gauge; bevel parallel to spine if Quincke is used.
  • Replace the stylet before withdrawing the needle.
  • Bed rest does not prevent post-LP headache; routine bed rest is no longer recommended.
  • Tubes 1 vs 4: distinguishes traumatic tap (RBC drops) from true SAH (RBC stable).
  • Xanthochromia: appears 6–12 hours after SAH; spectrophotometry more sensitive than eye.
  • Normal adult opening pressure: 6–20 cm H₂O.
  • Normal CSF WBC: ≤5/µL, lymphocyte/monocyte predominance.
  • CSF glucose: always interpret with simultaneous serum glucose; ratio ≥0.6 normal.
  • Normal CSF protein: 15–45 mg/dL; rises with age.
  • Epidural blood patch: gold standard for persistent post-LP headache; ~70% success first try.
  • SAH workup: non-contrast CT within 6 hours nearly always sufficient; LP reserved for delayed presentations.
  • Anticoagulation hold periods: vary by agent — confirm before LP.
  • Document tube assignments: lab requirements vary.

References

  1. Engelborghs S, Niemantsverdriet E, Struyfs H, et al. Consensus guidelines for lumbar puncture in patients with neurological diseases. Alzheimers Dement (Amst). 2017;8:111-126.
  2. Nath S, Koziarz A, Badhiwala JH, et al. Atraumatic versus conventional lumbar puncture needles: a systematic review and meta-analysis. Lancet. 2018;391(10126):1197-1204.
  3. Mowery NT, Bechtel L, McMurry T, et al. Spinal needles and post-dural puncture headache. Curr Pain Headache Rep. 2024;28(4):223-232.
  4. Doherty CM, Forbes RB. Diagnostic lumbar puncture. Ulster Med J. 2014;83(2):93-102.
  5. Wright BL, Lai JT, Sinclair AJ. Cerebrospinal fluid and lumbar puncture: a practical review. J Neurol. 2012;259(8):1530-1545.
  6. Williams J, Lye DC, Umapathi T. Diagnostic lumbar puncture: minimizing complications. Intern Med J. 2008;38(7):587-591.
  7. Whiteley W, Al-Shahi R, Warlow CP, et al. CSF opening pressure: reference interval and the effect of body mass index. Neurology. 2006;67(9):1690–1691.