The usual goal of lumbar puncture (LP) in children is to obtain
cerebrospinal fluid (CSF) to test for markers of infection. Measuring
opening pressure is not necessary, and therefore the procedure is
straightforward. The condition of children with hypoxemia, respiratory
distress, hypotension, and tachycardia may deteriorate when they
are positioned for LP, so resuscitation and empiric administration
of IV antibiotics is needed prior to LP. In children with
thrombocytopenia or factor deficiencies, replace platelets or factor
before attempting LP.
Anticipate the procedure and its difficulties. Assemble a needle
of the correct size, the appropriate specimen containers and preprinted
labels, and ensure a quiet environment without interruptions. Explain
the procedure to the caregivers. In some institutions, written informed
consent for LP is required. Describe the process of procedural sedation
if it is needed and obtain consent.
Apply a topical anesthetic cream or spray prior to needle insertion
to reduce pain and improve the success rate of the LP.1,2 For
infants, sucking on a pacifier dipped in sucrose solution is analgesic,
calming and decreases crying. Prepare the skin using sterile technique.
Have an experienced health care provider, the “holder,” restrain
the infant or child. Wrapping the child in sheets may help limit
leg movement. Flexing the hips is more important than flexing the
neck. In addition, flexing the neck may lead to respiratory difficulty.
Whether to choose the lateral recumbent position or the sitting
position depends upon the preference of the physician. In one study
using US to measure the width of the spinous processes, the sitting
position was found to be better than the lateral decubitus position.3 Although
the sitting position may improve flexion of the hips, this position
may be more difficult for the holder to maintain.
Most LPs are performed with a 22-gauge LP needle, usually 11/2 in.
in length for infants, 21/2 in. for children
2 years to 8 years, and 31/2 in. for older
children. In obese patients, choosing an LP needle may be more difficult.
One study calculated that an LP needle length (in centimeters) of
1 + [17 × (weight in
kilograms/height in centimeters)] was most accurate.4 LP
depth was measured on abdominal CT scans to derive this formula.
Lumbar needles with a clear hub show CSF flow sooner than those with
metal or opaque hubs.
Insert the LP needle between the L4 and L5 spinous processes,
in the intervertebral space, in the midline of the back, and direct
the needle toward the umbilicus. This interspace is easily located
because it lies in line between the iliac crests. Introduce the
needle with the bevel of the needle up. Insert the needle until
the characteristic “pop” identifies introduction
into the subarachnoid space. An alternative method is to remove
the stylet from the needle5 after the needle pierces
the skin. Advance the needle, without the stylet, incrementally
until CSF flows. Occasionally, rotating the lumbar needle clockwise
or counterclockwise up to 360 degrees may help improve flow if the
bevel of the needle is sideways. When removing the lumbar needle,
replace the stylet.
Bonadio originally reported early stylet removal as the “Cincinnati method” in
1992.5 Early experience with hollow-bore needles
without a stylet associated this technique with the development
of epidermoid tumors after the procedure. However, by puncturing
the epidermis with the stylet in place and removing the stylet after
introduction through the epidermis, this complication should be
eliminated. Two separate reports found that the use of this Cincinnati
method and the administration of topical anesthetics were associated
with improved success rates.1,2
After successful entry into the subarachnoid space, collect three
tubes of CSF, each with at least 0.5 mL of fluid. Send the first
tube for cell count, which includes white blood cell and red blood
cell counts; send the second tube for protein and glucose measurement;
and send the third tube for routine bacterial culture and Gram staining
(Table 143B-1). Additional tubes may be collected
for polymerase chain reaction testing for bacteria and viruses as
needed (e.g., enterovirus, herpes simplex virus). If the child has
been pretreated with antibiotics, latex agglutination testing may
be performed for Haemophilus influenzae type
b, Streptococcus pneumoniae, group B streptococci, Escherichia
coli, and Neisseria meningitidis serogroups
A, B, C, Y, and W135. However, latex agglutination testing does
have poor sensitivity for bacterial infection.
143B-1 Normal Cerebrospinal Fluid Values in Children
| Save Table
143B-1 Normal Cerebrospinal Fluid Values in Children
|Parameter||Preterm Infant||Term Neonate||Child|
|Cell count (WBCs/mm3)||9 (0–25) WBCs (>30 cells/mm3 suggests
meningitis)||8 (0–22) WBCs (>30 cells/mm3 suggests
meningitis)||0–7 WBCs (>10 cells/mm3 in children
>1 mo suggests meningitis)|
|Polymorphonuclear neutrophils (%)||57||61||0|
|Glucose (milligrams/deciliter)||52 (24–63)||52 (34–119)||40–80|
|Protein (milligrams/deciliter)||115 (65–150)||90 (20–170)||5–40|
|Red blood cells||0||0||0|
After a failed attempt, obtain a new needle and restart the procedure.
Insertion into the L3-L4 intervertebral space or the L5-S1 intervertebral space
may be successful. US- or fluoroscopy-guided lumbar needle insertion
may be necessary when all else fails.
1. Baxter AL, Fisher RG, Burke BL, et al: Local
anesthetic and stylet styles: factors associated with resident lumbar
puncture success. Pediatrics
117(3): 876, 2006.
2. Nigrovic LE, Kuppermann N, Neuman MI: Risk factors for traumatic
or unsuccessful lumbar punctures in children. Ann Emerg
49(6): 762, 2007.
3. Molina A, Fons J: Factors associated with lumbar puncture
success. Pediatrics 118(2): 842; author ...