Lumbar Puncture

Lauren M. Yorek, MD

Send Email

Recipient(s) will receive an email with a link to 'Lumbar Puncture' and will have access to the topic for 7 days.

Subject: Lumbar Puncture

(Optional message may have a maximum of 1000 characters.)



Lumbar puncture (LP) is a common diagnostic and therapeutic procedure. It is most commonly performed to obtain a sample of cerebrospinal fluid (CSF) to help establish neurological diagnoses. LP is the most accurate method for diagnosing central nervous system infection. 
CSF is produced by the choroid plexus in the brain and circulates around the brain and spinal cord within the subarachnoid space. During an LP, the spinal needle penetrates the skin, subcutaneous tissue, spinal ligaments, dura, and arachnoid before entering the subarachnoid space. Four samples of CSF are usually obtained. Usual studies include bacterial culture and Gram stain from tube 1, protein and glucose from tube 2, blood cell counts and differential cell counts from tube 3, and optional tests such as viral cultures, fungal cultures, countercurrent immunoelectrophoresis, India ink studies, or latex agglutination tests from tube 4. Common CSF findings are shown in Table 11-1. There are some CSF findings that suggest a diagnosis of bacterial meningitis. A CSF absolute white blood count (WBC) value of >500 μL, a blood glucose ratio of ≤0.4, a lactate level of ≥31.5, and the presence of bacteria on Gram stain can assist with diagnosis of bacterial meningitis. However, the absence of bacteria on Gram stain does not rule out bacterial meningitis. 
The indications for LP in neonates are not as clear as they once were. The once common practice of performing an LP in all ill newborns with suspected sepsis or respiratory distress is no longer recommended unless other findings suggest meningitis. LP is often reserved for babies who demonstrate hypothermia, hyperthermia, poor feeding 24 hours after birth, coma, or seizures. Bedside ultrasound scanning has largely replaced LP for the diagnosis of intracranial hemorrhage. Only about one half of LPs in newborns are successfully completed, and traumatic (bloody) taps are common. 
The most common complication is the post-LP (spinal) headache, which occurs in 10% to 25% of patients. The headache often persists for days. Using smaller-diameter nontraumatic needles and ensuring adequate hydration can help prevent postprocedure headache. Traditionally, it has been believed that keeping patients in the supine position after an LP would prevent headache. New data now suggest that patients should be mobilized soon after LP. When the headache is persistent, an epidural blood patch may be applied by an anesthesiologist. Traumatic (bloody) taps result from inadvertent puncture of the spinal venous plexuses and may rarely lead to spinal hematoma. Other temporary complications include shooting pains in the lower extremities and local pain in the back. 
Table 11-1.
Common Properties of Cerebrospinal Fluid
A more serious potential complication is brain herniation from elevated intracranial pressure that often is caused by a supratentorial mass. However, research has shown that severe meningitis also may cause elevated intracranial pressure and herniation. Before performing an LP, always check the optic fundi for papilledema. If increased pressure from a tumor or an intracranial bleed is suspected, an emergency noncontrast computed tomography (CT) scan should be obtained before LP to reduce the potential of herniation. Altered mentation and focal neurological findings are also indications for CT prior to performing an LP. Focal neurological findings include but are not limited to gaze palsy, facial palsy, arm or leg drift, and abnormal language function. Inadvertent aspiration of nerve roots on needle withdrawal can be prevented by replacing the stylet before needle removal. Meningitis as a result of the procedure is a theoretical complication. Epidermoid spinal cord tumors have been associated with the performance of LP in infants with unstyletted needles. 


  • Lumbar puncture tray:

  • Atraumatic (Sprotte or Pajunk) needle (which has an opening on the side at the end or the needle) or the standard (Quincke) needle (which has a standard bevel) with introducer, 22 to 26 gauge

  • Local anesthesia for injection

  • Povidone-iodine or chlorhexidine (Hibiclens) for sterilization of field

  • Four sterile collection tubes with labels one through four

  • Manometer for measurement of CSF pressure if needed

  • Dressing to apply after procedure


  • Suspected central nervous system infection

  • Suspected subarachnoid hemorrhage

  • Suspected neurosyphilis

  • Suspected Guillain-Barré syndrome

  • Support for the diagnosis of pseudotumor cerebri (i.e., increased CSF pressure without infection)

  • Serial removal of CSF

  • Support for the diagnosis of multiple sclerosis (i.e., elevated IgG level and oligoclonal banding on electrophoresis)


  • Dermatitis or cellulitis at insertion site

  • Raised intracranial pressure

  • Supratentorial mass lesions (evaluate with CT scan first)

  • Severe bleeding diathesis (relative)—increased risk of epidural hematoma

  • Lumbosacral deformity (relative)

  • Uncooperative patient

The Procedure

Step 1

Position the patient in the left lateral decubitus position, with the back near the edge of the bed or examination table and with the spine flexed and knees drawn toward the chest. Ensure the shoulders and back are perpendicular to the table. Place a pillow under the patient’s head to keep the spine as straight as possible. An alternative method is to place the patient in the sitting position, leaning on a bedside table or with two large pillows in the patient’s lap, with the spine flexed anteriorly. 
  • Pearl: Topical lidocaine (EMLA cream) has been studied in infants and found to decrease the pain response during the procedure. EMLA cream was applied in a topical dose of 1 g with an occlusion dressing placed over the site for 60 to 90 minutes before the procedure.

  • PITFALL: Avoid forced flexion of the neck during the procedure because cardiorespiratory arrest may occur if a child’s neck is excessively flexed.

  • PITFALL: Rotation of the patient beyond perpendicular can distort the appearance of the vertebral processes and make it more difficult to insert the needle in a midline position. If the patient is rotated, then insertion of the needle may be lateral and not penetrate the subarachnoid space.

Step 2

Prep the skin with povidone-iodine or chlorhexidine solution, and allow it to dry (see Appendix E). Most lumbar puncture kits include povidone-iodine swabs for sterilization of the procedure site. Set up the sterile tray, remove the tops of the sample tubes, and don a mask and sterile gloves while the povidone-iodine dries on the skin. Sterile draping typically is used for adult patients, but it can be omitted for the infant in favor of a wide prep to maximize landmark exposure and proper positioning. Inject a small amount (1 to 3 mL) of 1% lidocaine subcutaneously and into the area between the spinous processes. 
  • Pearl: Chlorhexidine may be used for patients allergic to iodine.

  • PEARL: It can be helpful to palpate the back and identify landmarks before the patient is sterilely prepped and draped. Once the patient is draped, vertebral landmarks can be more difficult to identify. It can be useful to make an indention in the skin with the end of a retractable pen that has the tip retracted. This indention will not be washed away like ink when the patient is prepped with povidone-iodine or chlorhexidine.

Step 3

The optimal needle insertion site is in the center of the spinal column, as defined by the spinous processes. The L3–4 interspace can be found where the line joining the superior iliac crests meets the spinous process of L4. Insertion is usually at the L3–4 interspace, but it may be performed one space above or below. 

Step 4

With the stylet in place, slowly insert the 22- or 20-gauge spinal needle midway between the two spinous processes. The correct angle for the needle is approximately toward the umbilicus, along the sagittal midplane of the body. If bone is encountered, withdraw the needle and change its angle. It is important to fully retract the needle because many kits include “cutting needles,” which may destroy tissue if moved within the soft tissue space. Feel for a loss of resistance, a give, or a “pop” as the needle enters the subarachnoid space, and then advance the needle 1 to 2 mm farther. The pop may not be felt in younger children. Withdraw the stylus, and check the hub for fluid. If there is no fluid, replace the stylus, and advance another fraction before repeating. 
  • Pearl: If bone is encountered, withdraw the needle and change the angle. Bone is usually encountered when the needle has been directed away from the midline. It can also be beneficial to palpate bony landmarks again and ensure that patient position is optimal. Patient movement during the procedure can alter the physician’s perception of the midline.

  • PITFALL: Make sure the bevel of the needle enters and exits the dura parallel to the long axis of the spinal column. This may lower the incidence of spinal nerve root damage and postprocedure headache.

  • PITFALL: Once the pop is felt, allow several seconds for the flow of CSF. The flow of CSF may not be immediate in some patients. This is especially true in dehydrated patients.

Step 5

After fluid is obtained, obstruct the passage of fluid with the stylet or your thumb. Place the stopcock and manometer onto the hub of the needle. As the CSF rises in the manometer, observe the color of the fluid and the opening pressure (Table 11-1). 
  • Pearl: The CSF opening pressure normal value is 6 to 14 mm Hg.

  • PITFALL: Have the patient relax his or her legs to prevent falsely elevating the opening pressure.

  • PITFALL: Accurate pressure measurements can only be made in the lateral decubitus position.

Step 6

Turn the stopcock to allow 2 to 3 mL of the CSF in children or 4 to 5 mL in adults to flow into each test tube. If desired, measure the closing pressure, but this has little value and removes additional CSF. Replace the stylus, and withdraw the needle. Wash off the povidone-iodine, and cover the puncture site with a sterile dressing. A small bandage is included in most kits to cover the procedure site. 
  • PEARL: Allow the fluid in the manometer tube to flow into the tubes first to decrease the amount of CSF removed.

  • PITFALL: If the tubes are not prelabeled, make sure to place the tubes in order so that you can easily identify and label each tube after the procedure.


  • Implantation of epidermoid tumors, which occur when epidermoid tissue is implanted into the spinal canal during a lumbar puncture: This occurs with the use of unstyletted needles or with needles that have poorly fitting stylets. These tumors cause pain to the back and lower extremities years after spinal puncture.

  • Aspiration of a nerve root into the spinal space, which occurs when the needle is withdrawn without the stylet in place.

  • Postprocedure headache:

    • Headache occurs in 5% to 40% of all lumbar punctures.

    • Headaches can start up to 48 hours after the procedure and usually last 1 to 2 days (sometimes up to 14 days).

    • Headache is associated with sitting upright and ceases with lying down.

    • They are thought to be caused by leakage of fluid through the dural puncture site.

    • Incidence of headache may be higher with “cutting needles” and larger-diameter needles.

    • Epidural blood patch may be performed to alleviate a persistent headache. This procedure is usually performed by an anesthesiologist.

  • Infection of CSF:

    • Infection on CSF may occur if local tissue infection is present over the LP site. Overlying cellulitis is a contraindication to LP.

    • It has also been suggested that infection may be caused by introduction of bacteria from infected blood (sepsis or bacteremia) during an LP. This has not been proven, and reported cases most likely represent CSF infection that was not yet detectable at the time of LP.

  • Postprocedure herniation:

    • This occurs in 2% to 3% of patients after LP.

    • Onset of symptoms is noted within 12 hours of procedure and manifested by loss of consciousness.

    • Many of these patients have a normal CSF opening pressure.

    • Most patients improve within 48 hours of symptom onset.

    • Risk of this complication can be decreased with the use of small-caliber spinal needles and ICP-lowering agents when needed.

  • Backache and radicular symptoms: Minor backache occurs in up to 90% of patients because of local trauma of the spinal needle.

Pediatric Considerations

Topical lidocaine (EMLA cream) has been studied in infants and found to decrease the pain response during the procedure. EMLA cream was applied in a topical dose of 1 g with an occlusion dressing placed over the site for 60 to 90 minutes before the procedure. 

Postprocedure Instructions

Once the procedure has been successfully performed, ensure that all four collection tubes are sealed and labeled in order of CSF collection. CSF can then be sent to the laboratory for pertinent testing. Always ask the laboratory to keep the CSF available in case further studies become necessary. 

Evaluation of Suspected Meningitis

  • Use Gram stain, culture, glucose, protein, and lactate dehydrogenase (LDH).

  • Other tests that may be performed are CSF counterimmunoelectrophoresis (CIE), CSF latex agglutination (LA), and coagulation immunoelectrophoresis.

  • Commercial kits are available to detect the many common organisms that cause meningitis.

  • All of these tests have a low sensitivity for bacterial meningitis; however, they have a much higher specificity. False-negative antigen tests may be seen with elevated rheumatoid factor and complement levels.

  • Polymerase chain reactions hold promise for future rapid diagnosis of infection and may be available in some hospital settings.

Subarachnoid Hemorrhage

  • CSF should be examined for xanthochromia, which is produced by the lysis of red blood cells (RBCs) in the CSF. RBC lysis begins to occur approximately 2 hours after exposure to CSF. The CSF is centrifuged and then examined for clarity. Collection of CSF within 12 hours of symptom onset of suspected subarachnoid hemorrhage may reveal false-negative results as a result of this phenomenon.

Coding Information and Supply Sources

Spinal tray sets may be obtained through Arrow Medical Products, Ltd., 2400 Bernville Road, Reading, PA 19605. Phone: 1-800-523-8446. Web site:
Many kits and supplies from various companies, including Baxter and American Hospital Supply (, can be obtained from Cardinal Health, Inc., 7000 Cardinal Place, Dublin, OH 43017 (phone: 1-800-234-8701; and from Owens and Minor, 4800 Cox Road, Glen Allen, VA 23060-6292 (phone: 804-747-9794; fax: 804-270-7281). 


Chordas C. Post-dural puncture headache and other complications after lumbar puncture. J Pediatr Oncol Nurs.  2001;18:244–259. [View Abstract]
Errando CL, Peiro CM. Postdural puncture upper back pain as an atypical presentation of postdural puncture symptoms. Anesthesiology .  2002;96:1019–1020. [View Abstract]
Flaatten H, Thorsen T, Askeland B Puncture technique and postural postdural puncture headache: a randomised, double-blind study comparing transverse and parallel puncture. Acta Anaesthesiol Scand .  1998;42:1209–1214. [View Abstract]
Grande PO, Myhre EB, Nordstrom CH Treatment of intracranial hypertension and aspects on lumbar dural puncture in severe bacterial meningitis. Acta Anaesthesiol Scand .  2002;46:264–270. [View Abstract]
Hasbun R, Abrahams J, Jekel J Computed tomography of the head before lumbar puncture in adults with suspected meningitis. N Engl J Med.  2001;345:1727–1733. [View Abstract]
Holdgate A, Cuthbert K. Perils and pitfalls of lumbar puncture in the emergency department. Emerg Med.  2001;13:351–358. [View Abstract]
Levine DN, Rapalino O. The pathophysiology of lumbar puncture headache. J Neurol Sci.  2001; 192:1–8. [View Abstract]
Marton KI, Gean AD. The spinal tap: a new look at an old test. Arch Intern Med.  1986;104:840–848. [View Abstract]
Nigrovic LE, Kupperman N, Macias CG Clinical prediction rule for identifying children with cerebrospinal fluid pleocytosis at very low risk for bacterial meningitis. JAMA.  2007;297(1):52–60. [View Abstract]
Roberts J, Hedges J. Clinical Procedures in Emergency Medicine . 4th ed. Philadelphia: Saunders, An Imprint of Elsevier;  2004:1197–1219.
Straus S, Thorpe K, Holroyd-Leduc J. How do I perform a lumbar puncture and analyze the results to diagnose bacterial meningitis? JAMA .  2006;296:2012–2022. [View Abstract]
Thoennissen J, Herkner H, Lang W Does bed rest after cervical or lumbar puncture prevent headache? A systematic review and meta-analysis. Can Med Assoc J .  2001;165:1311–1316. [View Abstract]
Thomas SR, Jamieson DR, Muir KW. Randomised controlled trial of atraumatic versus standard needles for diagnostic lumbar puncture. BMJ .  2000;321:986–990. [View Abstract]
Van Creval H, Hijdra A, de Gans J. Lumbar puncture and the risk of herniation: When should we first perform a CT? J Neurol .  2002;249(2):129–137.
Van de Beek D, de Gans J, Tunkel AR Community-acquired bacterial meningitis in adults. N Engl J Med .  2006;354:44–53.
Weise KL, Nihata MC. EMLA for painful procedures in infants. J Pediatr Health Care.  2005;19(1):42–49. [View Abstract]
2008 MAG Mutual Healthcare Solutions, Inc.’s Physicians’ Fee and Coding Guide. Duluth, Georgia. MAG Mutual Healthcare Solutions, Inc. 2007.