Uses and
Technique of Pediatric Lumbar Puncture Em
\s=b\
Ward, MD, Colette A. Gushurst,
Objectives.\p=m-\Toreview diagnostic and therapeutic indi-
cations, contraindications, complications, and technique of pediatric lumbar puncture with emphasis on diagnosis of bacterial meningitis in bacteremia, cellulitis, and fever with
seizures and to discuss cerebrospinal fluid findings in partially treated infection and traumatic blood-contaminated
spinal tap.
Research Design.\p=m-\Literaturereview. Conclusions.\p=m-\Werecommend lumbar puncture for children younger than 1 year with bacteremia, children with Haemophilus influenzae type B cellulitis, and children with fever and seizures not classified as simple. Pretreatment with antibiotics rarely changes cerebrospinal fluid characteristics such that a diagnosis of bacterial meningitis would be obscured. In a traumatic spinal tap, the equation to predict cerebrospinal fluid pleocytosis based on the peripheral blood cell count is invalid. When used together, cerebrospinal fluid glucose level, Gram's staining, and observed-to\x=req-\ expected ratio of white blood cells are highly reliable in diagnosing bacterial meningitis.
(AJDC. 1992;146:1160-1165)
(LP) has been pediatrics for Lumbar puncturecentury. Quincke's13 early descriptions the used in
than a
of the procedure in late 19th century presented LP as a useful means to obtain cerebrospinal fluid (CSF) for anal¬ ysis and to remove CSF for the treatment of hydroceph¬ alus. Widespread use of LP, or spinal tap, followed in the next century, along with much debate regarding appro¬ priate application of this invasive procedure. Computed tomography and magnetic resonance imaging have be¬ come important tools in neurodiagnosis. The diagnostic use of LP has been called into question in some situations now that these new, less invasive modalities are available. However, the spinal tap remains the most commonly per¬ formed diagnostic or nonsurgical procedure in pediatrics.4 We review recent pediatrie LP literature and present summaries of indications, contraindications, complica¬ tions, techniques, and CSF findings in partially treated in¬ fection and traumatic spinal tap. We did not include indi.
more
Accepted
for publication May 18, 1992. Department of Radiology, University of New Mexico School of Medicine, Albuquerque (Dr Ward), and the Departments of Family From the
Practice (Dr Ward) and Pediatrics (Dr Gushurst), Kalamazoo Center for Medical Studies, Michigan State University, Kalamazoo. Reprint requests to Michigan State University, Kalamazoo Center for Medical Studies, 1535 Gull Rd, Suite 230, Kalamazoo, Ml 49001 (Dr
Gushurst).
MD
cations exclusive to newborns, because Halliday5 recently reported the indications for neonatal LP. DIAGNOSTIC AND THERAPEUTIC INDICATIONS
Meningitis Although LP is usually performed in cases of suspected meningitis, opinions vary on the need for spinal tap in cases of bacteremia, cellulitis, and seizures with fever to diagnose accompanying meningitis.
Bacteremia.—We discuss three issues concerning bacte¬ remia and LP: (1) LP as part of a sepsis workup, (2) the risk of iatrogenic meningitis from LP, and (3) the need for re¬ peated LP. Concomitant bacterial meningitis may occur in a high percentage of children with bacteremia. In an urban teach¬ ing hospital, 5% to 6% of children who had simultaneous LPs and blood cultures had positive CSF cultures.6 In cases of "occult" bacteremia, 7% of children developed bacterial meningitis.7 Another review of outpatient therapy for bac¬ teremia without meningitis revealed that 40% of children identified as having bacteremia had simultaneously posi¬ tive CSF cultures and blood cultures.8 This was in a private office setting with good follow-up and may indicate that private practitioners have the ability to be more selective about which children undergo LP. The authors of the above studies recommended that LP be part of a sepsis workup6,8; others have suggested that LP may not be needed, in light of the risk to the patient, when meningeal signs are present.910 We agree that LP should be part of a sepsis workup, particularly in infants younger than 1 year. Questions have been raised about the possibility of pro¬ ducing meningitis during LP by introducing blood-borne organisms into the subarachnoid space.711 However, in
reports describing meningitis
complication of documented by positive culture and so it is open to interpretation whether meningitis was iatrogenic. Children with bacteremia not undergoing LP may subsequently develop meningitis.6"8 Shapiro et al7 studied risk factors for the development of bacterial meningitis in occult bacteremia and determined that the increased risk for developing meningitis after LP at the initial visit is not statistically significant. An initial CSF sample with no pleocytosis or with neg¬ ative Gram's staining or culture does not rule out early bacterial meningitis, and a second examination may be re¬ quired for the diagnosis or confirmation of aseptic menin¬ gitis.1213 A second LP is warranted if the patient's condition deteriorates or if he or she develops signs of central nersome
LP, the diagnosis of meningitis
Downloaded From: http://archpedi.jamanetwork.com/ by a University of Pennsylvania User on 05/14/2015
was
as a
not
vous
system (CNS) infection.1213 Bacteremia has been
reported as a cause of aseptic meningitis syndrome.14 The question of whether a positive second CSF sample indi¬ cates iatrogenic meningitis produced from the first LP or an evolving process that was not apparent on the initial
CNS examination remains unanswered. Cellulitis.—Meningitis may accompany or complicate cellulitis of the face or extremities.1527 In buccal cellulitis, the percentage may be as high as 8%.16 One review reported the occurrence of meningitis in head and neck cellulitis to be 26.9%.24 Some authors recommend perform¬ ing LPs in all patients with buccal cellulitis, as it is not un¬ usual for these patients to lack clinical evidence of meningeal irritation.16'2023 Others recommend LP in a facial cellulitis when there is a strong clinical suspicion of Hae¬ mophilus influenzae type infection, when there is no prob¬ able portal of entry for infection, or in a young child.21'22,25 The Table presents data on meningitis in facial cellulitis. We think the evidence is sufficiently compelling to recom¬ mend LP in infants younger than 1 year who have suspected influenzae type cellulitis anywhere on the body and particularly when there is facial cellulitis. Fever and Seizure.—The prevalence of meningitis in patients with seizure and fever is difficult to assess, because most studies have included children with simple febrile convulsions (SFCs) and complex seizures.28,29 An SFC is considered simple if it is brief and generalized and if there is no clinical evidence of intracranial infection.30 Jaffe et al30 pointed out that children with complex convulsions are more likely to have positive CSF findings. A recent survey of clinical practice in the United States suggests that routine LPs are performed in 48% of children with a first febrile seizure (FS).31 Joffe et al29 calculated that the probability that a child with seizures and fever has meningitis is only one in 1000. The prevalence based on the reviews of meningitis in cases presenting as seizure and fever ranged from 1 % to 5% .28,29 A few studies were limited to children with SFCs, and the prevalence of meningitis in those reviews ranged from 0% to 2%.30'32-33 Recommenda¬ tions as to which children with seizures and fever need LP require review in light of the population studied.34 Authors' recommendations for routine LP vary with patient age: younger than 1 year,35 younger titan 16 months,36 younger than 18 months,32,37"39 and younger than 2 years.40 Additional precautions in the evaluation of a young child with FSs are that meningitis can occur follow¬ ing initially normal CSF findings and that an extracranial source of infection does not rule out meningitis.28 Joffe et al29 found five items in the history and physical examination that discriminated significantly between chil¬ dren who had meningitis and those who did not: (1) a visit to a physician in the 48 hours before the seizure, (2) seizure on arrival at the emergency department, (3) type of seizure
(focal vs generalized), (4) suspicious findings on physical
examination, and (5) abnormal findings on neurologic ex¬
amination. When these factors were used in combination, they detected all cases of meningitis (sensitivity, 100%), and approximately 60% of children not needing an LP would have been spared this procedure. Joffe et al do not suggest that such a checklist approach be routinely used; rather, routine LP may not be warranted in all children with a first seizure and fever if these risk factors are
lacking.
We believe that LP may not be necessary in all children with seizures and fever, especially those with an SFC, but
strongly urge consideration of LP in those infants younger than 1 year. Intrathecal Antibiotic Therapy.—The intrathecal route may be used to deliver antibiotics to treat unusual or re¬ sistant causes of meningitis.41"44 In newborns, intrathecal aminoglycosides have been used for the treatment of gram-negative bacillary meningitis, although efficacy over therapy has not been shown.42-44 Vancomycin hydrochloride has been used as an adjunct in the treatment of Flavobacterium meningitis and other resistant organ¬ isms.41 Candida meningitis may require intrathecal amphointravenous
tericin B.43 Intra ventricular, rather than intralumbar, anti¬ biotics may be required to achieve adequate levels in ventriculitis.42 Intralumbar antibiotics for empiric treat¬ ment of severe meningitis have also been advocated.9
Pseudotumor Cerebri Quincke1 apparently was the first person to define a syndrome with diagnostic characteristics of increased intracranial pressure and normal CSF on LP. Criteria for the diagnosis of pseudotumor cerebri are (1) symptoms and signs of increased intracranial pressure, (2) normal CSF findings (normal or low protein levels), (3) increased subarachnoid pressure on LP (>200 mm H20), (4) exclusion of a space-occupying lesion, (5) exclusion of hydrocephalus, and (6) normal mental status.45"50 Physicians have used LP as part of the treatment of pseudotumor cerebri to reduce intracranial pressure and alleviate symptoms.5054 Serial LP may not be effective in all cases.51-53 Not all authors support the use of LP to treat pseudotumor cerebri.55 However, it appears that LP, used alone or with diuretics or steroids, has provided good re¬ sults more often than not.52-54 We support the use of serial LP in the initial treatment of pseudotumor cerebri.
Malignant Neoplasms Lumbar puncture is used to obtain CSF for diagnosis and staging of CNS malignant neoplasms in childhood.5661 Lumbar puncture may be the route for chemotherapy in¬ jection into the CNS for treatment of meningeal leukemia,
meningeal carcinomatosis, non-Hodgkin's lymphoma, and lymphoblastic lymphoma.59 Lumbar puncture is also used to monitor drug levels, remission, and relapse.5759 Infusion through LP for prophylaxis may be done through the puncture site during the staging spinal tap. Myelography Lumbar puncture is the usual method of contrast injec¬ tion in spinal myelography.6265 CONTRAINDICATIONS ultrasonography and computed tomography are generally available today, it is widely held that LP should not be performed before the patient undergoes scanning if intraventricular hemorrhage or intracranial mass is suspected.5,66,67 Even when meningitis is suspected, some authors believe that LP should not be performed in a severely ill patient or one with depressed mental status.9,68"72 If increased intracranial pressure is suspected Because
or
encountered, intravenous mannitol, furosemide,
or
dexamethasone sodium phosphate therapy or other means to reduce pressure may be used before or immediately af¬ ter LP.73"76 Lumbar puncture is contraindicated when there is an infection in the region overlying the puncture site55,68,74,77
Downloaded From: http://archpedi.jamanetwork.com/ by a University of Pennsylvania User on 05/14/2015
Meningitis in Source, y
Ginsberg,16
Facial Cellulitis (Buccal, Orbital, and Periorbital)*
No. of Cases
No. With
of Cellulitis (LP Performed)
Bacteremia and Meningitis
1981
72
No. With
Meningitis Only
(66)
Comments review of Haemophilus influenzae buccal cellulitis; no
Retrospective
1
meningeal signs
158 (NG)
3
0
Feldman,21
60 (33)
0
0
0
Retrospective
165 (28)
1982
al,'9
Carter and 1983
patients;
1983
Rubinstein and Handler,' Weiss et
in 2
recommended LP in all cases of buccal cellulitis Retrospective review of orbital and periorbital cellulitis Retrospective review of periorbital and orbital cellulitis; timing of LPs unclear Retrospective review of buccal cellulitis
Baker and Bausher,22
1 with
3?
osteomyelitis
79 (73)
review of bacteremic buccal and periorbital cellulitis; H influenzae and Streptococcus
1986
pneumoniae; no meningeal signs patients; 3 CSF samples without
in 5
pleocytosis
Chartrand and
Harrison,23
72 (50)
1
(no BC)
1986
*LP indicates lumbar
puncture; NG,
not
given; CSF, cerebrospinal fluid;
and should not be performed in the presence of an untreated clotting defect or acute trauma to or preexisting compression of the spinal column (particularly the lumbar
region).68
COMPLICATIONS Acute neurologic and respiratory deterioration are the most feared complications of LP and may occur as a result of herniation or cardiopulmonary compromise. These tragic outcomes usually occur in the presence of increased intracranial pressure, typically with a mass lesion or cere¬ bral edema, or in an infant with underlying cardiopulmo¬
nary disease.7882
Non-life-threatening complications include backache, vomiting, temporary paralysis, headache, and stiff neck.83 Headache is reportedly rare in children younger than 10
years but may last for weeks.8486 Minor local soreness may also occur.85 Epidermoid tumors may result from LP, especially if performed with nonstyletted needles. This complication is not immediate, sometimes requiring years before present¬ ing symptoms occur. The tumors are thought to be caused by introduction of epidermis into the subdural space.87,88 Other complications of LP reported in infants and chil¬ dren include subdural or subarachnoid hemorrhage89; epidural hematoma; trauma to the spinal ligaments, peri¬ osteum, or intervertebral disk; entrapment of nerve root through the durai tear74; acute pyogenic spondylitis; ab¬ scess formation; other vertebral infections; and bone mar¬ row trauma.90 Paresis has resulted from puncture of the spinal cord and contrast injection into an intraspinal lesion during myelography.91 Worsening of transverse myelitis has also been reported.92 The possibility of producing meningitis by LP was discussed above. Kruesi et al85 studied side effects of LP in children and adolescents with psychiatric diagnoses. They noted no statistical difference between preference ratings that pa¬ tients gave to LP and other procedures or activities (blood drawing, electroencephalography, and attending school).
Retrospective and prospective study of buccal cellulitis; H influenzae; no meningeal signs in 3 patients
and BC, blood culture.
They concluded that psychological preparation of the pa¬ tient for the procedure may be important. TECHNIQUE When LP is clinically indicated, informed consent should be obtained and the discussion documented.93 Lo¬ cal anesthesia, usually 1% lidocaine hydrochloride infil¬ tration into the skin and subcutaneous tissues, is generally used but may be excluded in neonates.74,94 Topical anes¬ thetic cream applied to the area of puncture may reduce pain and apprehension.95 Sterile technique (gloves, iodine solution, alcohol skin preparation, and drapes) should be
used.74,94-96
The ideal needle for pediatrie LP is short, small bored, sharp, and short beveled.97-98 A smaller-bore needle makes a smaller hole in the dura, allowing less leakage through the defect.65,99103 Schreiner and Kleiman104 prospectively compared styletted LP needles, butterfly needles, and 23-gauge nonstyletted venipuncture needles and reported no statistically significant difference in traumatic spinal taps or unsuccessful spinal taps among needles. A wellfitted stylet impedes the introduction of epidermis into the subarachnoid space, therefore decreasing the risk of sub¬ sequent epidermoid tumors.87,88 After the skin has been punctured, a nonstyletted needle may be used to continue
the durai
puncture.87,88
Respiratory compromise may occur during or after LP if the neck is flexed.74'81'82-98 Hypoxemia most commonly
in newborns in the standard lateral decubitus po¬ sition with neck flexion.81-82 Thus, the lateral position should be modified with neck extension, or LP should be performed with newborns in the upright position.81-82 The sitting position may allow widening of the durai sac from the hydrostatic effect of CSF and decreased likelihood of traumatic spinal tap.97 The horizontal position (lateral de¬ cubitus) is required to measure opening pressure.74 Re¬ straint may be less difficult with the patient in the lateral decubitus position.98 Continuous monitoring of high-risk newborns during LP may be performed.94 occurs
Downloaded From: http://archpedi.jamanetwork.com/ by a University of Pennsylvania User on 05/14/2015
L3-4 and L4-5 are popular interspaces for puncture, with repeated attempt at an interspace cephalad if a traumatic spinal tap occurs on the first attempt.74'92-94-96-98 The needle should be inserted with the bevel parallel to the long axis
a
of the dura and advanced slowly.74-94-96-98 Some anesthesi¬ ologists reported that bevel orientation has no bearing on the occurrence of post-LP headache.102103 A decrease in re¬ sistance or the durai "pop" is often not felt in neonates.94-96-98 If no CSF is present on removal of the stylet, it should be replaced and the needle advanced cautiously or rotated 90°.74-94
Technique of Pediatric Lumbar Puncture Em
\s=b\
Ward, MD, Colette A. Gushurst,
Objectives.\p=m-\Toreview diagnostic and therapeutic indi-
cations, contraindications, complications, and technique of pediatric lumbar puncture with emphasis on diagnosis of bacterial meningitis in bacteremia, cellulitis, and fever with
seizures and to discuss cerebrospinal fluid findings in partially treated infection and traumatic blood-contaminated
spinal tap.
Research Design.\p=m-\Literaturereview. Conclusions.\p=m-\Werecommend lumbar puncture for children younger than 1 year with bacteremia, children with Haemophilus influenzae type B cellulitis, and children with fever and seizures not classified as simple. Pretreatment with antibiotics rarely changes cerebrospinal fluid characteristics such that a diagnosis of bacterial meningitis would be obscured. In a traumatic spinal tap, the equation to predict cerebrospinal fluid pleocytosis based on the peripheral blood cell count is invalid. When used together, cerebrospinal fluid glucose level, Gram's staining, and observed-to\x=req-\ expected ratio of white blood cells are highly reliable in diagnosing bacterial meningitis.
(AJDC. 1992;146:1160-1165)
(LP) has been pediatrics for Lumbar puncturecentury. Quincke's13 early descriptions the used in
than a
of the procedure in late 19th century presented LP as a useful means to obtain cerebrospinal fluid (CSF) for anal¬ ysis and to remove CSF for the treatment of hydroceph¬ alus. Widespread use of LP, or spinal tap, followed in the next century, along with much debate regarding appro¬ priate application of this invasive procedure. Computed tomography and magnetic resonance imaging have be¬ come important tools in neurodiagnosis. The diagnostic use of LP has been called into question in some situations now that these new, less invasive modalities are available. However, the spinal tap remains the most commonly per¬ formed diagnostic or nonsurgical procedure in pediatrics.4 We review recent pediatrie LP literature and present summaries of indications, contraindications, complica¬ tions, techniques, and CSF findings in partially treated in¬ fection and traumatic spinal tap. We did not include indi.
more
Accepted
for publication May 18, 1992. Department of Radiology, University of New Mexico School of Medicine, Albuquerque (Dr Ward), and the Departments of Family From the
Practice (Dr Ward) and Pediatrics (Dr Gushurst), Kalamazoo Center for Medical Studies, Michigan State University, Kalamazoo. Reprint requests to Michigan State University, Kalamazoo Center for Medical Studies, 1535 Gull Rd, Suite 230, Kalamazoo, Ml 49001 (Dr
Gushurst).
MD
cations exclusive to newborns, because Halliday5 recently reported the indications for neonatal LP. DIAGNOSTIC AND THERAPEUTIC INDICATIONS
Meningitis Although LP is usually performed in cases of suspected meningitis, opinions vary on the need for spinal tap in cases of bacteremia, cellulitis, and seizures with fever to diagnose accompanying meningitis.
Bacteremia.—We discuss three issues concerning bacte¬ remia and LP: (1) LP as part of a sepsis workup, (2) the risk of iatrogenic meningitis from LP, and (3) the need for re¬ peated LP. Concomitant bacterial meningitis may occur in a high percentage of children with bacteremia. In an urban teach¬ ing hospital, 5% to 6% of children who had simultaneous LPs and blood cultures had positive CSF cultures.6 In cases of "occult" bacteremia, 7% of children developed bacterial meningitis.7 Another review of outpatient therapy for bac¬ teremia without meningitis revealed that 40% of children identified as having bacteremia had simultaneously posi¬ tive CSF cultures and blood cultures.8 This was in a private office setting with good follow-up and may indicate that private practitioners have the ability to be more selective about which children undergo LP. The authors of the above studies recommended that LP be part of a sepsis workup6,8; others have suggested that LP may not be needed, in light of the risk to the patient, when meningeal signs are present.910 We agree that LP should be part of a sepsis workup, particularly in infants younger than 1 year. Questions have been raised about the possibility of pro¬ ducing meningitis during LP by introducing blood-borne organisms into the subarachnoid space.711 However, in
reports describing meningitis
complication of documented by positive culture and so it is open to interpretation whether meningitis was iatrogenic. Children with bacteremia not undergoing LP may subsequently develop meningitis.6"8 Shapiro et al7 studied risk factors for the development of bacterial meningitis in occult bacteremia and determined that the increased risk for developing meningitis after LP at the initial visit is not statistically significant. An initial CSF sample with no pleocytosis or with neg¬ ative Gram's staining or culture does not rule out early bacterial meningitis, and a second examination may be re¬ quired for the diagnosis or confirmation of aseptic menin¬ gitis.1213 A second LP is warranted if the patient's condition deteriorates or if he or she develops signs of central nersome
LP, the diagnosis of meningitis
Downloaded From: http://archpedi.jamanetwork.com/ by a University of Pennsylvania User on 05/14/2015
was
as a
not
vous
system (CNS) infection.1213 Bacteremia has been
reported as a cause of aseptic meningitis syndrome.14 The question of whether a positive second CSF sample indi¬ cates iatrogenic meningitis produced from the first LP or an evolving process that was not apparent on the initial
CNS examination remains unanswered. Cellulitis.—Meningitis may accompany or complicate cellulitis of the face or extremities.1527 In buccal cellulitis, the percentage may be as high as 8%.16 One review reported the occurrence of meningitis in head and neck cellulitis to be 26.9%.24 Some authors recommend perform¬ ing LPs in all patients with buccal cellulitis, as it is not un¬ usual for these patients to lack clinical evidence of meningeal irritation.16'2023 Others recommend LP in a facial cellulitis when there is a strong clinical suspicion of Hae¬ mophilus influenzae type infection, when there is no prob¬ able portal of entry for infection, or in a young child.21'22,25 The Table presents data on meningitis in facial cellulitis. We think the evidence is sufficiently compelling to recom¬ mend LP in infants younger than 1 year who have suspected influenzae type cellulitis anywhere on the body and particularly when there is facial cellulitis. Fever and Seizure.—The prevalence of meningitis in patients with seizure and fever is difficult to assess, because most studies have included children with simple febrile convulsions (SFCs) and complex seizures.28,29 An SFC is considered simple if it is brief and generalized and if there is no clinical evidence of intracranial infection.30 Jaffe et al30 pointed out that children with complex convulsions are more likely to have positive CSF findings. A recent survey of clinical practice in the United States suggests that routine LPs are performed in 48% of children with a first febrile seizure (FS).31 Joffe et al29 calculated that the probability that a child with seizures and fever has meningitis is only one in 1000. The prevalence based on the reviews of meningitis in cases presenting as seizure and fever ranged from 1 % to 5% .28,29 A few studies were limited to children with SFCs, and the prevalence of meningitis in those reviews ranged from 0% to 2%.30'32-33 Recommenda¬ tions as to which children with seizures and fever need LP require review in light of the population studied.34 Authors' recommendations for routine LP vary with patient age: younger than 1 year,35 younger titan 16 months,36 younger than 18 months,32,37"39 and younger than 2 years.40 Additional precautions in the evaluation of a young child with FSs are that meningitis can occur follow¬ ing initially normal CSF findings and that an extracranial source of infection does not rule out meningitis.28 Joffe et al29 found five items in the history and physical examination that discriminated significantly between chil¬ dren who had meningitis and those who did not: (1) a visit to a physician in the 48 hours before the seizure, (2) seizure on arrival at the emergency department, (3) type of seizure
(focal vs generalized), (4) suspicious findings on physical
examination, and (5) abnormal findings on neurologic ex¬
amination. When these factors were used in combination, they detected all cases of meningitis (sensitivity, 100%), and approximately 60% of children not needing an LP would have been spared this procedure. Joffe et al do not suggest that such a checklist approach be routinely used; rather, routine LP may not be warranted in all children with a first seizure and fever if these risk factors are
lacking.
We believe that LP may not be necessary in all children with seizures and fever, especially those with an SFC, but
strongly urge consideration of LP in those infants younger than 1 year. Intrathecal Antibiotic Therapy.—The intrathecal route may be used to deliver antibiotics to treat unusual or re¬ sistant causes of meningitis.41"44 In newborns, intrathecal aminoglycosides have been used for the treatment of gram-negative bacillary meningitis, although efficacy over therapy has not been shown.42-44 Vancomycin hydrochloride has been used as an adjunct in the treatment of Flavobacterium meningitis and other resistant organ¬ isms.41 Candida meningitis may require intrathecal amphointravenous
tericin B.43 Intra ventricular, rather than intralumbar, anti¬ biotics may be required to achieve adequate levels in ventriculitis.42 Intralumbar antibiotics for empiric treat¬ ment of severe meningitis have also been advocated.9
Pseudotumor Cerebri Quincke1 apparently was the first person to define a syndrome with diagnostic characteristics of increased intracranial pressure and normal CSF on LP. Criteria for the diagnosis of pseudotumor cerebri are (1) symptoms and signs of increased intracranial pressure, (2) normal CSF findings (normal or low protein levels), (3) increased subarachnoid pressure on LP (>200 mm H20), (4) exclusion of a space-occupying lesion, (5) exclusion of hydrocephalus, and (6) normal mental status.45"50 Physicians have used LP as part of the treatment of pseudotumor cerebri to reduce intracranial pressure and alleviate symptoms.5054 Serial LP may not be effective in all cases.51-53 Not all authors support the use of LP to treat pseudotumor cerebri.55 However, it appears that LP, used alone or with diuretics or steroids, has provided good re¬ sults more often than not.52-54 We support the use of serial LP in the initial treatment of pseudotumor cerebri.
Malignant Neoplasms Lumbar puncture is used to obtain CSF for diagnosis and staging of CNS malignant neoplasms in childhood.5661 Lumbar puncture may be the route for chemotherapy in¬ jection into the CNS for treatment of meningeal leukemia,
meningeal carcinomatosis, non-Hodgkin's lymphoma, and lymphoblastic lymphoma.59 Lumbar puncture is also used to monitor drug levels, remission, and relapse.5759 Infusion through LP for prophylaxis may be done through the puncture site during the staging spinal tap. Myelography Lumbar puncture is the usual method of contrast injec¬ tion in spinal myelography.6265 CONTRAINDICATIONS ultrasonography and computed tomography are generally available today, it is widely held that LP should not be performed before the patient undergoes scanning if intraventricular hemorrhage or intracranial mass is suspected.5,66,67 Even when meningitis is suspected, some authors believe that LP should not be performed in a severely ill patient or one with depressed mental status.9,68"72 If increased intracranial pressure is suspected Because
or
encountered, intravenous mannitol, furosemide,
or
dexamethasone sodium phosphate therapy or other means to reduce pressure may be used before or immediately af¬ ter LP.73"76 Lumbar puncture is contraindicated when there is an infection in the region overlying the puncture site55,68,74,77
Downloaded From: http://archpedi.jamanetwork.com/ by a University of Pennsylvania User on 05/14/2015
Meningitis in Source, y
Ginsberg,16
Facial Cellulitis (Buccal, Orbital, and Periorbital)*
No. of Cases
No. With
of Cellulitis (LP Performed)
Bacteremia and Meningitis
1981
72
No. With
Meningitis Only
(66)
Comments review of Haemophilus influenzae buccal cellulitis; no
Retrospective
1
meningeal signs
158 (NG)
3
0
Feldman,21
60 (33)
0
0
0
Retrospective
165 (28)
1982
al,'9
Carter and 1983
patients;
1983
Rubinstein and Handler,' Weiss et
in 2
recommended LP in all cases of buccal cellulitis Retrospective review of orbital and periorbital cellulitis Retrospective review of periorbital and orbital cellulitis; timing of LPs unclear Retrospective review of buccal cellulitis
Baker and Bausher,22
1 with
3?
osteomyelitis
79 (73)
review of bacteremic buccal and periorbital cellulitis; H influenzae and Streptococcus
1986
pneumoniae; no meningeal signs patients; 3 CSF samples without
in 5
pleocytosis
Chartrand and
Harrison,23
72 (50)
1
(no BC)
1986
*LP indicates lumbar
puncture; NG,
not
given; CSF, cerebrospinal fluid;
and should not be performed in the presence of an untreated clotting defect or acute trauma to or preexisting compression of the spinal column (particularly the lumbar
region).68
COMPLICATIONS Acute neurologic and respiratory deterioration are the most feared complications of LP and may occur as a result of herniation or cardiopulmonary compromise. These tragic outcomes usually occur in the presence of increased intracranial pressure, typically with a mass lesion or cere¬ bral edema, or in an infant with underlying cardiopulmo¬
nary disease.7882
Non-life-threatening complications include backache, vomiting, temporary paralysis, headache, and stiff neck.83 Headache is reportedly rare in children younger than 10
years but may last for weeks.8486 Minor local soreness may also occur.85 Epidermoid tumors may result from LP, especially if performed with nonstyletted needles. This complication is not immediate, sometimes requiring years before present¬ ing symptoms occur. The tumors are thought to be caused by introduction of epidermis into the subdural space.87,88 Other complications of LP reported in infants and chil¬ dren include subdural or subarachnoid hemorrhage89; epidural hematoma; trauma to the spinal ligaments, peri¬ osteum, or intervertebral disk; entrapment of nerve root through the durai tear74; acute pyogenic spondylitis; ab¬ scess formation; other vertebral infections; and bone mar¬ row trauma.90 Paresis has resulted from puncture of the spinal cord and contrast injection into an intraspinal lesion during myelography.91 Worsening of transverse myelitis has also been reported.92 The possibility of producing meningitis by LP was discussed above. Kruesi et al85 studied side effects of LP in children and adolescents with psychiatric diagnoses. They noted no statistical difference between preference ratings that pa¬ tients gave to LP and other procedures or activities (blood drawing, electroencephalography, and attending school).
Retrospective and prospective study of buccal cellulitis; H influenzae; no meningeal signs in 3 patients
and BC, blood culture.
They concluded that psychological preparation of the pa¬ tient for the procedure may be important. TECHNIQUE When LP is clinically indicated, informed consent should be obtained and the discussion documented.93 Lo¬ cal anesthesia, usually 1% lidocaine hydrochloride infil¬ tration into the skin and subcutaneous tissues, is generally used but may be excluded in neonates.74,94 Topical anes¬ thetic cream applied to the area of puncture may reduce pain and apprehension.95 Sterile technique (gloves, iodine solution, alcohol skin preparation, and drapes) should be
used.74,94-96
The ideal needle for pediatrie LP is short, small bored, sharp, and short beveled.97-98 A smaller-bore needle makes a smaller hole in the dura, allowing less leakage through the defect.65,99103 Schreiner and Kleiman104 prospectively compared styletted LP needles, butterfly needles, and 23-gauge nonstyletted venipuncture needles and reported no statistically significant difference in traumatic spinal taps or unsuccessful spinal taps among needles. A wellfitted stylet impedes the introduction of epidermis into the subarachnoid space, therefore decreasing the risk of sub¬ sequent epidermoid tumors.87,88 After the skin has been punctured, a nonstyletted needle may be used to continue
the durai
puncture.87,88
Respiratory compromise may occur during or after LP if the neck is flexed.74'81'82-98 Hypoxemia most commonly
in newborns in the standard lateral decubitus po¬ sition with neck flexion.81-82 Thus, the lateral position should be modified with neck extension, or LP should be performed with newborns in the upright position.81-82 The sitting position may allow widening of the durai sac from the hydrostatic effect of CSF and decreased likelihood of traumatic spinal tap.97 The horizontal position (lateral de¬ cubitus) is required to measure opening pressure.74 Re¬ straint may be less difficult with the patient in the lateral decubitus position.98 Continuous monitoring of high-risk newborns during LP may be performed.94 occurs
Downloaded From: http://archpedi.jamanetwork.com/ by a University of Pennsylvania User on 05/14/2015
L3-4 and L4-5 are popular interspaces for puncture, with repeated attempt at an interspace cephalad if a traumatic spinal tap occurs on the first attempt.74'92-94-96-98 The needle should be inserted with the bevel parallel to the long axis
a
of the dura and advanced slowly.74-94-96-98 Some anesthesi¬ ologists reported that bevel orientation has no bearing on the occurrence of post-LP headache.102103 A decrease in re¬ sistance or the durai "pop" is often not felt in neonates.94-96-98 If no CSF is present on removal of the stylet, it should be replaced and the needle advanced cautiously or rotated 90°.74-94
