Neurologic Complications of Spinal Anesthesia
Terese T. Horlocker, MD
Perioperative nerve injuries have long been recognized as a
potential complication of spinal anesthesia. Fortunately, severe or
disabling neurologic complications rarely occur. Risk factors
contributing to neurologic deficit after spinal anesthesia include
spinal cord ischemia (hypothesized to be related to the use of
vasoconstrictors or prolonged hypotension), traumatic injury to
the spinal cord or nerve roots during needle or catheter
placement, pressure applied to the spinal cord due to hemorrhagic
or infectious complications, and choice of local anesthetic
solution. Practice guidelines for the safe conduct of spinal
anesthesia are influenced by knowledge of the previously
published large patient surveys as well as individual case reports
of neurologic deficits following central neural blockade.
Prevention of complications, as well as early diagnosis and
treatment are important factors in the management of regional
anesthetic risks.
Copyright 9 1998 by W.B. Saunders Company
p lerioperative nerve injuries have long been recognized as a
complication of spinal anesthesia. Fortunately, severe or
disabling neurologic complications rarely occur. Kroll et al m
examined the American Society of Anesthesiologists Closed
Claims database to determine the role of nerve damage in
malpractice claims filed against anesthesia care providers. Of
the 1,541 claims reviewed, 227 (15%) were for anesthesia-
related nerve injury, including 49 claims involving the lumbo-
sacral roots or spinal cord. Lumbosacral nerve root injuries
having identifiable etiology were associated predominantly
with a regional anesthetic technique, and were related to
paresthesias during needle or catheter placement or pain
during injection of local anesthetic.
Neurologic complications associated with spinal anesthesia
may be divided into two categories: those unrelated to the
spinal anesthetic but coincide temporally, and those that are a
direct result of the spinal anesthetic. Neurologic complications
directly related to spinal anesthesia may be caused by direct or
indirect trauma, ischemia, infection, and neurotoxicity. Postop-
erative neurologic injury due to pressure from improper
patient positioning or from tightly applied casts or surgical
dressings, as well as surgical trauma are often attributed to the
spinal anesthetic. Marinacci evaluated 542 patients with post-
operative neurologic deficits allegedly caused by spinal anesthe-
sia. 2 In only four cases were the findings actually related to the
spinal anesthetic. In the remaining 538 patients, the neurologic
deficits exhibited an apparent, but not causal, relationship to
the spinal anesthetic. Marinacci's study demonstrates the
difficulty in reporting the actual incidence, pathogenesis, and
From the Mayo Clinic, Rochester, MN.
Address reprint requests to Terese T. Horlocker, MD, 200 First Street
Southwest, Rochester, MN, 55905.
Copyright 9 1998 by W.B. Saunders Company
1084-208X/98/0204-000758.00/0
Techniques in Regional Anesthesia and Pain Management, Vol 2, No 4 (October), 1998: pp 211-218
prognosis of neurologic dysfunction that occur as a result of
spinal anesthesia.
Incidence of Neurologic Complications
Serious neurologic complications rarely occur after spinal
anesthesia. In several large studies totaling over 50,000 spinal
anesthetics, the incidence of persistent peripheral neuropathy,
including persistent paresthesia and sensory or motor dysfunc-
tion, varied from 0 to 0 . 0 8 % . 3.6 In addition, within this
combined series, there was one case of cauda equina syndrome
presenting with lower extremity weakness and impotence, one
case of paraplegia (related to a previously undiagnosed spinal
cord tumor), and three cases of meningitis. 5,6
A prospective survey in France recently evaluated the
incidence and characteristics of serious complications related
to regional anesthesia. 7 A total of 103,730 regional anesthetics
were performed over a 5-month period. The incidence of
cardiac arrest and neurologic complications was significantly
higher after spinal anesthesia than other types of regional
procedures. Among the 40,640 patients receiving spinal anes-
thesia, there were 26 cardiac arrests, six of which were fatal,
and 24 patients with neurologic complications. Overall there
were 19 cases of radiculopathy and five cases of cauda equina
syndrome following spinal anesthesia. Three patients (one
radiculopathy and two cauda equina syndrome) reported
permanent deficits. Needle trauma and local anesthetic neuro-
toxicity were identified as the etiology of most neurologic
complications. This study demonstrated the relative safety of
spinal anesthesia. However, because serious complications
were noted to occur even in the presence of experienced
anesthesiologists, continued vigilance in patients undergoing
spinal anesthesia is warranted.
Risk factors contributing to neurologic deficit after spinal
anesthesia include spinal cord ischemia (hypothesized to be
related to the use of vasoconstrictors or prolonged hypoten-
sion), traumatic injury to the spinal cord or nerve roots during
needle or catheter placement, infection, and choice of local
anesthetic solution. 7m~Practice guidelines for the safe conduct
of spinal anesthesia involve knowledge of the historical, large
patient surveys as well as individual case reports of neurologic
deficits following central neural blockade. 3,4,6 Prevention of
complications along with early diagnosis and treatment are
important factors in management of regional anesthetic risks.
Nerve Injury from Needle and Catheter
Placement
Direct needle or catheter-induced trauma rarely results in
permanent or disabling neurologic injury. A recent retrospec-
tive study of 4,767 spinal anesthetics noted the presence of a
paresthesia during needle placement in 298 (6.3%) of patients.
Importantly, four of the six patients with a persistent paresthe-
211
sia postoperatively complained of a paresthesia during needle
placement, identifying elicitation of a paresthesia as a risk
factor for a persistent paresthesia. ~ There were no permanent
neurologic complications. In four patients, the persistent
paresthesia resolved within 1 week. In the remaining two
patients, recovery was complete within 12 to 24 months. Auroy
et al 7 noted that in all 12 cases of radiculopathy in which a
paresthesia was noted during needle placement, or there was
pain upon injection of the local anesthetic, the radiculopathy
had the same distribution as the associated paresthesia.
The passage and presence of an indwelling catheter into the
subarachnoid space presents an additional source of direct
trauma. In one study, the incidence of paresthesias was 13%
with a single-dose and 30% with a continuous catheter spinal
anesthetic (CSA) technique. 12 Laboratory studies have demon-
strated demyelination and inflammation adjacent to the cath-
eter tract in both the spinal root and cord of rats following
placement of indwelling subarachnoid catheters. 13 The use of a
catheter may indirectly contribute to neurologic injury by
maldistribution and pooling of local anesthetic solution. This
is the presumed mechanism of cauda equina syndrome follow-
ing continuous spinal. 14-16
Local Anesthetic Toxicity
Neurologic complications after spinal anesthesia may be a
direct result of local anesthetic toxicity. There is both labora-
tory and clinical evidence that local anesthetic solutions are
potentially neurotoxic. 8,9,17 However, with the exception of
lidocaine, it is generally agreed that local anesthetics adminis-
tered in clinical concentrations and doses do not cause nerve
damage. 18 Prolonged exposure and/or high concentrations of
local anesthetic solutions at the spinal roots may result in
permanent neurologic deficits. Poor mixing resulting from
very slow injection rates through spinal microcatheters may
increase the risk of developing high concentrations of hyper-
baric local anesthetics in dependent areas of the spinal canal. 14
Maldistribution of local anesthetic within the cerebrospinal
fluid has been implicated in the development of cauda equina
syndrome after continuous spinal anesthesia. 14,15,~9 Transient
neurologic symptoms (or transient radicular irritation), de-
fined as pain radiating into the buttocks or legs without
sensory or motor deficits after the original anesthetic has
completely resolved, was initially described following injection
of hyperbaric 5% lidocaine. 8 However, subsequent studies have
also implicated 2% lidocaine and 0.75% bupivacaine (with and
without dextrose) as well as patient positioning. Local anes-
TABLE 1. Differential Diagnosis of Spinal Abscess, Spinal Hematoma, and Anterior Spinal Artery Syndrome
Spinal Abscess
Age of patient Any age
Previous history Infection*
Onset 1 to 3 days
Generalized symptoms Fever, malaise, back pain
Sensory involvement None or paresthesias
Motor involvement Flaccid paralysis, later spastic
Segmental reflexes Exacerbated*--Iater obtunded
Myelogram/CT scan Signs of extradural compression
Cerebrospinal fluid Increased cell count
Blood data Rise in sedimentation rate
*Infrequent findings.
Reprinted with permission from Wedel DJ, Horlocker TT: Risks of regional anesthesia-infectious,septic. Reg Anesth 21:57-61 1996.
212
thetic toxicity is discussed in greater detail elsewhere in this
issue.
Anterior Spinal Artery Syndrome
The blood supply to the spinal cord is precarious in the
watershed regions between the radicular vessels. Systemic
hypotension or localized vascular insufficiency with or with-
out a spinal anesthetic may produce spinal cord ischemia,
resulting in flaccid paralysis of the lower extremities, or
anterior spinal artery syndrome. 2~ Characteristics of anterior
spinal artery syndrome, spinal abscess, and spinal hematoma
are reported in Table 1. The use of local anesthetic solutions
containing epinephrine or phenylephrine theoretically may
produce local cord ischemia, especially in patients with micro-
vascular disease. 9,17Large studies have failed to identify the use
of vasoconstrictors as a risk factor for temporary or permanent
deficits. 3-6 Most presumed cases of vasoconstrictor-induced
neurologic deficits have been reported as single case reports,
often with several other risk factors present. 2~ However, a
recent investigation by Sakura et a121 noted the addition of
phenylephrine increased the risk of transient neurologic symp-
toms in patients undergoing tetracaine spinal anesthesia. It is
important to note that no patient in the study by Sakura et al
had sensory or motor deficits, but only radicular pain, and that
currently there is no definitive evidence that local anesthetic-
related injury is related to transient pain and dysesthesias. The
actual risk of significant neurologic ischemia in patients
administered local anesthetic solutions containing vasoconstric-
tors, therefore, remains unknown.
Spinal Hematoma
The actual incidence of neurologic dysfunction resulting from
hemorrhagic complications associated with neuraxial blockade
is unknown; however, the incidence cited in the literature is
estimated to be less than 1 in 220,000 spinal anesthetics (Fig
1). 22 Vandermeulen et a123 identified 61 cases of spinal hema-
toma associated with neuraxial anesthesia between 1904 and
1994. In 42 of the 61 patients (68%), the spinal hematomas
associated with central neural blockade occurred in patients
with evidence of hemostatic abnormality. Twenty-five of the
patients had received intravenous or subcutaneous heparin,
while an additional five patients were presumably administered
heparin, as they were undergoing a vascular surgical proce-
dure. In addition, 12 patients had evidence of coagulopathy or
were treated with antiplatelet medications, oral anticoagulants,
Anterior Spinal
Spinal Hematoma Artery Syndrome
50% over 50 years Elderly
Anticoagulants Arteriosclerosis/hypotension
Sudden Sudden
Sharp, transient back and leg pain None
Variable, late Minor, patchy
Flaccid paralysis Flaccid paralysis
Abolished Abolished
Signs of extradural compression Normal
Normal Normal
Prolonged coagulation time* Normal
HORLOCKER

Fig 1. Spinal hematoma: Lumbar myelogram (lateral view)
showing almost complete obliteration of the subarachnoid
space at the L3-4 level.
thrombolytics, or dextran 70 immediately before or after the
spinal or epidural anesthetic. Regional technique was also
noted. An epidural anesthetic was performed in 46 patients,
including 32 patients with an indwelling catheter. In 15 of
these 32 patients, the spinal hematoma occurred immediately
after the removal of the epidural catheter. (Nine of these
catheters were removed during therapeutic levels of hepariniza-
tion). These results suggest that catheter removal is not
entirely atraumatic, and the patient's coagulation status at the
time of catheter removal is perhaps as critical as that at the time
of catheter placement. Importantly, only 15 of 61 cases of
spinal hematoma in the series by Vandermeulen et a123 were
associated with a single dose spinal anesthetic (there were no
continuous spinal techniques implicated), again demonstrat-
ing the relative safety of this regional anesthetic technique
compared with epidural anesthesia.
Intravenous Heparin
Standard unfractionated heparin and low molecular weight
heparin (LMWH) are perhaps the most common anticoagulant
medications used for the treatment and prevention of thrombo-
embotic complications. In a study involving over 4,000 pa-
tients, including 847 patients undergoing continuous spinal
anesthesia, Rao and E1-Etr24demonstrated the safety of indwell-
ing neuraxial catheters during systemic heparinization. How-
ever, the heparin activity was closely monitored, the indwelling
NEUROLOGIC COMPLICATIONS
catheters were removed the following day at a time when
circulating heparin levels were relatively low, and patients with
a preexisting coagulation disorder were excluded. In addition,
surgery was cancelled in patients with traumatic needle
placement (and performed the following day under general
anesthesia). A subsequent study in the neurologic literature by
Ruff and Dougherty 25 reported spinal hematomas in 7 of 342
patients (2%) who underwent a diagnostic lumbar puncture
and subsequent heparinization. Traumatic needle placement,
initiation of anticoagulation within 1 hour of lumbar puncture
or concomitant aspirin therapy were identified as risk factors
in the development of spinal hematoma in anticoagulated
patients.
Mathews and Abrams 26 reported no neurologic complica-
tions in 40 cardiac surgical patients who received intrathecal
morphine through a 20 to 25-G needle 50 minutes before
complete anticoagulation with heparin. Subsequent authors
have demonstrated similarly positive outcomes with both
single dose and continuous catheter techniques in patients
undergoing cardiopulmonary bypass. However, surgery is
often cancelled or delayed if blood is present during needle or
catheter placement. 23
In summary, spinal anesthesia may be safely performed in
the patient undergoing subsequent therapeutic heparinization
provided heparinization occurs a minimum of 60 minutes after
needle placement, the heparin effect is monitored and main-
tained within acceptable levels (activated clotting time or aPTT
1.5-2.0 times baseline), and indwelling catheters are removed
at a time when heparin activity is low or completely re-
versed. 23-2~ Some authors also recommend cancellation of
surgery should bleeding occur during needle or catheter
placement. 23,24
Subcutaneous Heparin
Low-dose subcutaneous (unfractionated) heparin is adminis-
tered for thromboprophylaxis in patients undergoing major
thoracoabdominal surgery and in patients at increased risk of
hemorrhage with oral anticoagulant or low molecular weight
heparin (LMWH) therapy. A review of the literature by
Schwander and Bachmann 27 noted no spinal hematomas in
over 5,000 patients who received subcutaneous heparin in
combination with spinal or epidural anesthesia. There is a
single case of spinal hematoma associated with spinal anesthe-
sia (attempted spinal anesthesia with subsequent general
anesthesia) in the presence of low-dose heparin. 2s
Recommendations for the performance of regional anesthe-
sia in patients receiving subcutaneous unfractionated heparin
include avoidance of needle placement or catheter removal
within four hours of administration, and monitoring of antico-
agulant effect in patients with liver disease or long-term
thromboprophylaxis.23
Low Molecular Weight Heparin
Low molecular weight heparin has recently been introduced
for thromboprophylaxis following knee or hip arthroplasty.
Extensive clinical testing and use of LMWH in Europe over the
last 10 years suggested that there was not an increased risk of
spinal hematoma in patients undergoing neuraxial anesthesia
while receiving LMWH thromboprophylaxis perioperatively.23,29
However, in the time period between the release of LMWH
(enoxaparin, LOVENOX, Rhone-Poulenc Rorer Pharmaceuti-
213
cals, Collegeville, PA) for general use in the United States in
May 1993 and December 1997, over 30 cases of spinal
hematoma associated with neuraxial anesthesia administered
in the presence of perioperative LMWH prophylaxis have been
reported to the manufacturer. 3~ Nearly all of these events
occurred in patients undergoing continuous epidural anesthe-
sia and analgesia; only three were associated with a spinal
anesthetic technique. The apparent difference in incidence in
Europe compared to the United States may be a result of a
difference in dose and dosage schedule. In Europe, the
recommended dose of enoxaparin is 40 mg once daily, rather
than 30 mg every 12 hours.
Patients on preoperative LMWHs can be assumed to have
altered coagulation. LMWHs are potent antithrombotic agents
with a 3 to 4 hour half-life. Approximately 50% of peak anti-Xa
activity is present 12 hours after injection. Concomitant
administration of medications affecting hemostasis, such as
antiplatelet drugs, standard heparin, or dextran represents an
additional risk of hemorrhagic complications perioperatively,
including spinal hematoma. Indeed, it appears that the clini-
cian should proceed cautiously with regional anesthesia in the
patient receiving LMWH. A single-dose spinal anesthetic may
be the safest neuraxial technique in patients receiving preopera-
tive LMWHs. Needle placement should occur at least 10 to 12
hours after the last LMWH dose. Subsequent dosing should be
delayed at least two hours after needle placement. 22,23,3~The
presence of blood during needle placement may warrant addi-
tional delay in initiation of postoperative thromboprophylaxis.
Oral Anticoagulants
Few data exist regarding the risk of spinal hematoma in
patients who are anticoagulated with warfarin. Odoom and
Sih31 performed 1,000 continuous lumbar epidural anesthetics
in vascular surgical patients who were receiving oral anticoag-
ulants preoperatively. The thrombotest (a test measuring factor
IX activity) was decreased in all patients prior to needle
placement. Heparin was also administered intraoperatively.
Epidural catheters remained in place for 48 hours postopera-
tively. There were no neurologic complications. While these
results are reassuring, the obsolescence of the thrombotest as a
measure of anticoagulation, combined with the unknown
coagulation status of the patients at the time of catheter
TABLE 2. Pharmacologic Activities of Anticoagulants and Antiplatelet Agents
Effect on
Coagulation Variables
Agent PT aPTT Time to Peak Effect
Intravenous Heparin T "rTT Minutes
Subcutaneous Hepadn T TT 40 to 50 minutes
Low Molecular -- -- 3 to 5 hours
Weight Heparin
Warfarin TTT T 4 to 6 days (Less with
loading dose)
Antiplatet Agents -- --
Aspirin -- -- Hours
Other NSAID
PT = Prothrombin time; aPTT = Activated partial thromboplastin time; ACT = Activated clotting time; T = Clinically insignificant increase; TT = Possibly
clinically significant increase; TTT = Clinically significant increase; NSAID = Nonsteroidal antiinflammatory drug.
Adapted with permission from Horlocker Tl', Wedel DJ: Regional anesthesia in the anticoagulated patient: yes or no. Sem Anesth 17:73-82, 1998.
214
removal, limit the usefulness of these results. Therefore, except
in extraordinary circumstances, spinal needle (and catheter)
placement and removal should not be performed in fully
anticoagulated patients. Theoretically, the use of spinal anesthe-
sia in patients with subtherapeutic anticoagulation (interna-
tional normalized ratio, or INR, between 1.2 and 1.5) would be
less traumatic (and consequently represent less risk) than
epidural anesthesia due to the smaller needle size and avoid-
ance of catheter placement.
Antiplatelet Medications
Orthopedic and vascular patients often receive antiplatelet
medications perioperatively for analgesia and thromboprophy-
laxis. Although Vandermeulen et a123 implicated antiplatelet
therapy in 3 of the 61 cases of spinal hematoma occurring after
spinal or epidural anesthesia, several large studies have demon-
strated the relative safety of neuraxiaI blockade in both
obstetric and surgical patients receiving these medications) 2,33
In a prospective study involving 1,000 patients, including 408
patients undergoing single dose and continuous spinal anesthe-
sia, Horlocker et a133 reported that preoperative antiplatelet
therapy did not increase the incidence of blood present at the
time of needle/catheter placement or removal, suggesting that
trauma incurred during needle or catheter placement is neither
increased nor sustained by these medications. In addition, the
incidence of blood present during needle placement was lowest
with a single dose spinal anesthetic technique. Small needle
size also decreased the likelihood of traumatic needle place-
ment. 33 The clinician should be aware of the possible increased
risk of spinal hematoma in patients receiving antiplatelet
medications who undergo subsequent heparinization. 25
In summary, the decision to perform spinal anesthesia in a
patient receiving anticoagulant or antiplatelet medications
should be made on an individual basis, weighing the small,
though definite risk of spinal hematoma with the benefits of
regional anesthesia for a specific patient. Alternative anesthetic
and analgesic techniques exist for patients considered to be at
an unacceptable risk. The patient's coagulation status should
be optimized at the time of spinal needle and catheter
placement, and the level of anticoagulation must be carefully
monitored during the period of intrathecal catheterization
(Table 2). It is important to note that patients respond with
Time to Normal
Hemostasis after
Discontinuation Comments
4 to 6 hours Monitor ACT, AP-H-, delay heparinization for
1 hour after needle placement
4 to 6 hours APTT may remain normal, anti-Xa activity
reflects degree of anticoagulation
12+ hours Anti-Xa activity not monitored. Use with cau-
tion in patients receiving epidural anal-
gesia
4 to 6 days Monitor PT in patients with indwelling cath-
eters
5 to 8 days Bleeding time not reliable predictor of
1 to 3 days platelet function
HORLOCKER
variable sensitivities to anticoagulant medications. Indwelling
catheters should not be removed in the presence of therapeutic
anticoagulation, as this appears to significantly increase the
risk of spinal hematoma. 23 In addition, communication be-
tween clinicians involved in the perioperative management of
patients receiving anticoagulants for thromboprophylaxis is
essential in order to decrease the risk of serious hemorrhagic
complications.
Patients should be closely monitored in the perioperative
period for early signs of cord compression such as severe back
pain, progression of numbness or weakness, and bowel and
bladder dysfunction. If spinal hematoma is suspected, radio-
graphic confirmation must be immediately sought. Delay may
lead to irreversible cord ischemia. The treatment of choice is
decompressive laminectomy. Recovery is unlikely if surgery is
postponed more than 8 hours. 23
Infectious Complications
Bacterial infection of the central neuraxis may present as
meningitis or cord compression secondary to abscess forma-
tion (Fig 2). The infectious source can be exogenous due to
contaminated equipment or medication, or endogenous second-
ary to a bacterial source in the patient seeding to the remote
site of needle or catheter placement. In addition, indwelling
catheters may be colonized from a superficial (skin) site and
Fig 2. Infected spinal hematoma: MRI of cervical spine
demonstrating spinal abscess/hematoma.
NEUROLOGIC COMPLICATIONS
subsequently serve as a wick for spread of infection from the
skin to the epidural or intrathecal space.
Although individual cases have been reported in the litera-
ture, serious central neural infections such as arachnoiditis,
meningitis, and abscess following spinal anesthesia are rare. In
a combined series of more than 65,000 spinal anesthetics, there
were only three cases of meningitis. 2~ Few data suggest that
spinal anesthesia during bacteremia is a risk factor for infection
of the central neuraxis. Although the authors of the previous
studies did not report how many patients were febrile during
administration of the spinal anesthetic, a significant number of
the patients included in these studies underwent obstetric or
urologic procedures, and it is likely that some patients had
bacteremia after (and perhaps during) needle or catheter
placement. Despite the apparent low risk of central nervous
system infection following regional anesthesia, anesthesiolo-
gists have long considered sepsis to be a relative contraindica-
tion to the administration of spinal anesthesia. This impression
is based largely on anecdotal reports and conflicting laboratory
and clinical investigations.
Dural puncture has long been considered a risk factor in the
pathogenesis of meningitis. Exactly how bacteria cross from
the blood stream into the spinal fluid is unknown. The
presumed mechanisms include introduction of blood into the
intrathecal space during needle placement and disruption of
the protection provided by the blood-brain barrier. However,
lumbar puncture is often performed in patients with fever or
infection of unknown origin. If dural puncture during bacter-
emia results in meningitis, definite clinical data should exist. In
fact, clinical studies are few, and often antiquated.
Initial laboratory and clinical investigations were performed
over 80 years ago. Weed et a134 demonstrated that lumbar or
cisternal puncture performed during septicemia (produced by
intravenous injection of a gram-negative bacillus) invariably
resulted in a fatal meningitis. Wegeforth and Latham ~5reported
their clinical observations on 93 patients suspected of having
meningitis who received a diagnostic lumbar puncture. Blood
cultures were taken simultaneously. The diagnosis was con-
firmed in 38 patients. The remaining 55 patients had normal
cerebrospinal fluid (CSF). However, 6 of these 55 patients were
bacteremic at the time of lumbar puncture. Five of the six
bacteremic patients subsequently developed meningitis. It was
implied, but not stated, that patients with both sterile blood
and CSF cultures did not develop meningitis. Unfortunately,
these lumbar punctures were performed during two epidemics
of meningitis which occurred at a military instillation, and it is
possible that some (or all) of these patients may have devel-
oped meningitis without lumbar puncture. These two histori-
cal studies provided support for the claim that lumbar punc-
ture during bacteremia was a risk factor for meningitis.
Carp and Bailey36 investigated the association between
meningitis and dural puncture in bacteremic rats. Twelve of
forty rats subjected to cisternal puncture with a 26-G needle
during an E. coli bacteremia subsequently developed meningi-
tis. In addition, bacteremic animals not undergoing dural
puncture, as well as animals undergoing dural puncture in the
absence of bacteremia did not develop meningitis. Meningitis
occurred only in animals with a blood culture result of more
than 50 colony-forming units/mL at the time of dural punc-
ture. Treatment of a group of bacteremic rats with a single dose
of gentamycin immediately prior to cisternal puncture elimi-
215
nated the risk of meningitis, as none of these animals devel-
oped infection.
This study demonstrates that dural puncture in the presence
of bacteremia is associated with the development of meningitis
in rats, and that antibiotic treatment before dural puncture
reduces this risk. Unfortunately, this study did not include a
group of animals that were treated with antibiotics after dural
puncture. Because many surgeons defer antibiotic therapy
until after cultures are obtained, the actual clinical scenario
remains unstudied. There are several other limitations to this
study. While Escherichia coli is a common cause of bacteremia,
it is an uncommon cause of meningitis. In addition, the
investigators knew the sensiti~ify to the bacteria injected,
allowing for appropriate antibiotic coverage. They also per-
formed a cisternal puncture (rather than lumbar puncture) and
used a 26-G needle, producing a relatively large dural defect in
the rat compared to humans. Finally, no local anesthetic was
injected. Local anesthetic solutions are typically bacteriostatic,
which may reduce the risk of meningitis in normal clinical
settings.
In summary, numerous clinical and laboratory studies have
suggested an association between dural puncture during bacter-
emia and meningitis (Table 3). The data are not equivocal,
however. The clinical studies are limited to pediatric patients
who are historically at high-risk for meningitis. Many of the
original animal studies used bacterial counts that were far in
excess of those noted in humans in early sepsis, making central
nervous system (CNS) contamination more likely. Despite
these conflicting results, it is generally recommended that
except in the most extraordinary circumstances, central neural
block should not be performed in patients with untreated
systemic infection. However, patients with evidence of sys-
temic infection may safely undergo spinal anesthesia, provided
appropriate antibiotic therapy is initiated prior to dural punc-
ture, and the patient has demonstrated a response to therapy,
such as a decrease in fever. Spinal anesthesia may be safely
performed in patients at risk for low-grade transient bacter-
emia after dural puncture.
All patients with an established local or systemic infection
should be considered at risk for developing infection of the
TABLE 3. Meningitis After Dural Puncture
Number
Author, of
Year Patients Population Microorganism(s)
Wegeforth, 93 Military personnel N. meningitidis
1919 S. pneumonia
Pray, 1941 416 Pediatric with bacteremia S. pneumonia
Eng, 1981 1,089 Adults with bactermia Atypical and typical bac-
teria
Teele, 1981 271 Pediatric with bacteremia S. pneumonia
N. meningitis
H. influenza
Smith, 1986 11 Preterm with neonatal
sepsis
*Significant association (P < .001).
Spontaneous meningitis = concurrent bacteremia and meningitis (without a preceeding lumbar puncture).
Lumbar puncture-induced meningitis = positive blood culture with sterile CSF on initial exam; subsequent positive CSF culture (same organism present
in blood).
Reprinted with permission from Wedel DJ, Horlocker TT: Risks of regional anesthesia-infectious, septic. Reg Anesth 21:57-61,1996.
216
CNS. A delay in diagnosis and treatment of even a few hours
significantly worsens neurologic outcome. Bacterial meningitis
is a medical emergency. Mortality is approximately 30%, even
with antibiotic therapy. Meningitis presents most often with
fever, severe headache, altered level of consciousness, and
meningismus. The diagnosis is confirmed with a lumbar
puncture. Lumbar puncture should not be performed if spinal
abscess is suspected, as contamination of the intrathecal space
may result. CSF examination in the patient with meningitis
reveals leukocytosis, a glucose level of less than 30 mg/dL, and
a protein level greater than 150 mg/dL.
The clinical course of epidural abscess progresses from
spinal ache and root pain, to weakness (including bowel and
bladder symptoms) and eventually paralysis. The initial back
pain and radicular symptoms may remain stable for hours to
weeks. However, the onset of weakness often progresses to
complete paralysis within 24 hours. Although the diagnosis
was historically made with myelogram, radiologic examination
using computed tomography (CT) scan or magnetic resonance
imaging (MRI) is currently recommended. A combination of
antibiotics and surgical drainage remains the treatment of
choice. As with spinal hematoma, neurologic recovery is
dependent on the duration of the deficit and the severity of
neurologic impairment before treatment.
Preexisting Neurologic Disorders
Patients with preexisting neurologic disease present a unique
challenge to the anesthesiologist. The cause of postoperative
deficits is difficult to evaluate because neural injury may occur
as a result of surgical trauma, tourniquet pressure, prolonged
labor, improper patient positioning, or anesthetic technique.
Progressive neurologic diseases such as multiple sclerosis may
coincidentally worsen perioperatively, independent of the
anesthetic method. The most conservative medico-legal ap-
proach is to avoid regional anesthesia in these patients.
However, high-risk patients, including those with significant
cardiopulmonary disease, may benefit medically from regional
anesthesia and analgesia. The decision to proceed with a
regional anesthetic in these patients should be made on a
Patients Patients With
With Lumbar
Spontaneous Puncture-induced
Meningitis Meningitis Comments
38 of 93 (41%) 5 of 93, including.5 of 6 LPs performed during
bacteremic patients meningitis epidemics
86 of 386 (22%) 8 of 30 (27%) 80% of patients with men-
ingitis <2 yrs of age
30 of 919 (3.3%) 3 of 170 (1.8%) Atypical organisms respon-
sible for lumbar puncture
induced meningitis
2 of 31 (8.7%) 7 of 46 (15%)* All cases of meningitis
occurred in children <1
yr of age. Antibiotic
therapy reduced risk
0% 0%
HORLOCKER
case-by-case basis. Meticulous regional anesthetic technique
should be observed to minimize further neurologic injury.
Disorders of the Central Nervous System
Patients with preexisting neurologic disorders of the CNS,
such as multiple sclerosis or amyotrophic lateral sclerosis
(ALS), lumbar radiculopathy, and ancient poliomyelitis present
potential management dilemmas for anesthesiologists. The
presence of preexisting deficits, signifying chronic neural
compromise, theoretically places these patients at increased
risk for further neurologic injury. It is difficult to define the
actual risk of neurologic complications in patients with preex-
isting neurologic disorders who receive regional anesthesia; no
controlled studies have been performed, and accounts of
complications have appeared in the literature as individual case
reports. The decision to use regional anesthesia in these
patients is determined on a case-by-case basis and involves
understanding the pathophysiology of neurologic disorders,
the mechanisms of neural injury associated with regional
anesthesia, and the overall incidence of neurologic complica-
tions after regional techniques. Although laboratory studies
have identified multiple risk factors for the development of
neurologic injury after regional anesthesia, clinical studies are
lacking. Even less information is available for the variables
affecting neurologic damage in patients with preexisting neuro-
logic disease. Several neurologic conditions warrant a more
comprehensive discussion.
Multiple Sclerosis
Multiple sclerosis is a degenerative disease of the CNS,
characterized by multiple sites of demyelination in the brain
and spinal cord. The peripheral nerves are not involved. The
course of the disease consists of exacerbations and remissions
of symptoms, and the unpredictability in the patient's changing
neurologic status must be appreciated when selecting an
anesthetic technique. Stress, surgery, and fatigue have been
implicated in the exacerbation of multiple sclerosis. Epidural
and, more often, spinal anesthesia have been associated with
relapse of multiple sclerosis, although the evidence is not
strong. 37 The mechanism by which spinal anesthesia may
exacerbate multiple sclerosis is unknown, but it may be direct
local anesthetic toxicity. Epidural anesthesia has been recom-
mended over spinal anesthesia because the concentration of
local anesthetic in the white matter of the spinal cord is
one-fourth the level after epidural administration. 38 A dilute
solution of local anesthetic with spinal or epidural anesthesia is
also advised. Because multiple sclerosis is a disorder of the
CNS, peripheral nerve blocks do not affect neurologic function
and are considered appropriate anesthetic techniques.
Diabetes Mellitus
A substantial proportion of diabetic patients report clinical
symptoms of a peripheral neuropathy. However, a subclinical
peripheral neuropathy may be present before the onset of pain,
paresthesia, or sensory loss and may remain undetected
without electrophysiologic testing for slowing of nerve conduc-
tion velocity. The presence of underlying nerve dysfunction
suggests that patients with diabetes may have a decreased
requirement for local anesthetic. The diabetes-associated micro-
NEUROLOGIC COMPLICATIONS
angiopathy of nerve blood vessels decreases the rate of
administration, resulting in prolonged exposure to local anes-
thetic solutions. The combination of these two mechanisms
may cause nerve injury with an otherwise safe dose of local
anesthetic in diabetic patients.
In a study examining the effect of local anesthetics on nerve
conduction block and injury in diabetic rats, Kalichman and
Calcutt 39 reported that the local anesthetic requirement is
decreased and the risk of local anesthetic-induced nerve injury
is increased in diabetes. These findings support the suggestions
that diabetic patients may require less local anesthetic to
produce anesthesia and that a reduction in dose may be
necessary to prevent neural injury by doses considered safe in
nondiabetic patients.
In summary, major complications after regional anesthetic
techniques are rare, but can be devastating to the patient and
the anesthesiologist. Prevention and management begin during
the preoperative visit with a careful evaluation of the patient's
medical history and appropriate preoperative discussion of the
risks and benefits of the available anesthetic techniques.
Patients with preexisting neurologic disorders such as multiple
sclerosis, polio, or ALS may develop new neurologic deficits
perioperatively. It is often difficult to differentiate between
surgical or anesthetic causes, z The medicolegal issue, however,
remains, and if a regional anesthetic is indicated (or re-
quested), the patient's preoperative neurologic examination
should be formally documented and the patient must be made
aware of the possible progression of the underlying disease
process. Although stable preexisting neurologic conditions
such as an inactive lumbosacral radiculopathy or hemiparesis
associated with an ancient cerebrovascular accident are not
contraindications to spinal anesthesia, the underlying etiology
of such neurologic deficits requires careful evaluation.
Patients with preoperative neurologic deficits may undergo
further nerve damage more readily from needle or catheter
placement, local anesthetic systemic toxicity, and vasopressor-
induced neural ischemia. Care should be taken to minimize
needle trauma and intraneuronal injection. 4~ Dilute or less
potent local anesthetic solutions should be used when feasible
to decrease the risk of local anesthetic toxicity.39
The use of epinephrine-containing solutions in patients with
preexisting neurologic deficits is controversial. The potential
risk of vasoconstrictor-induced nerve ischemia must be weighed
against the advantages of improved quality and duration of
block. Because epinephrine and phenylephrine also prolong
the block and therefore neural exposure to local anesthetics,
the appropriate concentration and dose of local anesthetic
solutions must be thoughtfully considered. Patients with
microvascular disease in combination with an underlying
neuropathy, such as those with diabetes, may be most sensitive
to the effects of vasoconstrictors.
Efforts should also be made to decrease neural injury in the
operating room through careful patient positioning. Postopera-
tively, these patients must be followed closely to detect
potentially treatable sources of neurologic injury, including
constrictive dressings, improperly applied casts, and increased
pressure on neurologically vulnerable sites. New neurologic
deficits should be evaluated promptly by a neurologist to
document formally the patient's evolving neurologic status,
arrange further testing, and provide long-term follow-up.
217
 References
1. Kroll DA, Caplan RA, Posner K, et al: Nerve injury associated with
anesthesia. Anesthesiology 73:202-207, 1990
2. MarinacciAA: Neurological aspects of complications of spinal anesthe-
sia with medico-legal implications. Bull Los Angeles Neurol Soc
25:170-192, 1960
3. Phillips OC, Ebner H, Nelson AT, Black MH: Neurologic complications
following spinal anesthesia with lidocaine: a prospective review of
10,440 cases. Anesthesiology 30:284-289, 1969
4. Dripps RD, Vandam LD: Long-term follow-up of patients who received
10,098 spinal anesthetics: failure to discover major neurological
sequelae. JAMA 156:1486-1491, 1954
5. Sadove MS, Levin MJ, Rant-Sejdinaj I: Neurological complications of
spinal anaesthesia. Can Anaes Soc J 8:405-416, 1961
6. Moore DC, Bridenbaugh LD: Spinal (subarachnoid) block: a review of
11,574 cases. JAMA 195:123-128, 1966
7. Auroy Y, Narchi P, Messiah A, et al: Serious complications related to
regional anesthesia. Anesthesiology 87:479-486, 1997
8. Schneider M, Ettlin T, Kaufmann M, et al: Transient neurologic toxicity
after hyperbaric subarachnoid anesthesia with 5% lidocaine. Anesth
Analg 76:1154-1157, 1993
9. Myers RR, Heckinan HM: Effects of local anesthesia on nerve blood
flow: studies using lidocaine with and without epinephrine. Anesthesi-
ology 71:757-762, 1989
10. Ready LB, Plumer MH, Haschke RH, et al: Neurotoxicity of intrathecal
local anesthetics in rabbits. Anesthesiology 63:364-370, 1985
11. Horlocker TT, McGregor DG, Matsushige DK, et al: A retrospective
review of 4,767 consecutive spinal anesthetics: central nervous
system complications. Anesth Analg 84:578-584, 1997
12. Dripps RD: A comparison of the malleable needle and catheter
techniques for continuous spinal anesthesia. NY State J Med 50:1595-
1599, 1950
13. Myers RR, Sommer C: Methodology for spinal neurotoxicity studies.
Reg Anesth 18:439-447, 1993
14. Ross BK, Coda B, Heath CH: Local anesthetic distribution in a spinal
model: a possible mechanism of neurologic injury after continuous
spinal anesthesia. Reg Anesth 17:69-77, 1992
15. Rigler ML, Drasner K, Krejcie TC, et al: Cauda equina syndrome after
continuous spinal anesthesia. Anesth Analg 72:275-281, 1991
16. Lambert DH, Hurley RJ: Cauda equina syndrome and continuous
spinal anesthesia. Anesth Analg 72:817-819, 1991
17. Kalichman MW, Calcutt NA: Local anesthetic-induced conduction
block and nerve fiber injury in streptozotocin-diabetic rats. Anesthesi-
ology 77:941-947, 1992
18. Selander D: Neurotoxicity of local anesthetics: animal data. Reg
Anesth 18:461-468, 1993
19. Drasner K: Models for local anesthetic toxicity from continuous spinal
anesthesia. Reg Anesth 18:434-438, 1993
20. Kane RE: Neurologic deficits following epidural or spinal anesthesia.
Anesth Analg 60:150-161,1981
21. Sakura S, Sumi M, Sakaguchi Y, et al: The addition of phenylephrine
contributes to the developmentof transient neurologic symptoms after
218
spinal anesthesia with 0.5% tetracaine. Anesthesiology 87:771-778,
1997
22. Tryba M: Epidural regional anesthesia and low molecular heparin: Pro
(German). An&sth Intensivmed Notfallmed Schmerzther 28:179-181,
1993
23. Vandermeulen EP, Van Aken H, Vermylen J: Anticoagulants and
spinal-epidu ral anesthesia. Anesth Analg 79:1165-1177, 1994
24. Rao TLK, EI-Etr AA: Anticoagulation following placement of epidural
and subarachnoid catheters: An evaluation of neurologic sequelae.
Anesthesiology 55:618-620, 1981
25. Ruff RL, Dougherty JH: Complications of lumbar puncture followed by
anticoagulation. Stroke 12:879-881,1981
26. Mathews ET, Abrams LD: Intrathecal morphine in open heart surgery
(letter). Lancet 2:543, 1980
27. Schwander D, Bachmann F: Heparin and spinal or epidural anesthe-
sia: decision analysis. [review] Ann Fr Anesth Reanim 10:284-296,
1991
28. Dupeyrat A, Dequire PM, Merouani A, et al: Subarachnoid hematoma
and spinal anesthesia. Ann Fr Anesth Reanim 9:560-562, 1990
29. Bergqvist D, Lindblad B, Matzsch T: Low molecular weight heparin for
thromboprophylaxis and epidural/spinal anaesthesia: Is there a risk?
Acta Anaesthesiol Scand 36:605-609, 1992
38. Horlocker TT, Heit JA: Low molecular weight heparin: Biochemistry,
pharmacology, perioperative prophylaxis regimens, and guidelines for
regional anesthetic management. Anesth Analg 85:874-885, 1997
31. Odoom JA, Sih IL: Epidural analgesia and anticoagulant therapy.
Anaesthesia 38:254-259, 1983
32. CLASP (Collaborative Low-Dose Aspirin Study in Pregnancy): CLASP:
a randomized trial of low-dose aspirin for the prevention and treatment
of pre-eclampsia among 9,364 pregnant women. Lancet 343:619-
629, 1994
33. Horlocker "17, Wedel DJ, Offord KP, et al: Preoperative antiplatelet
therapy does not increase the risk of spinal hematoma associated
with regional anesthesia. Anesth Analg 80:303-309, 1995
34. Weed LH, Wegeforth P, Ayer JB, et al: The production of meningitis by
release of cerebrospinal fluid during an experimental septicemia.
JAMA 72:190-193, 1919
35. Wegeforth P, Latham JR: Lumbar puncture as a factor in the causation
of meningitis. Am J Med Sci 158:183-202, 1919
36. Carp H, Bailey S: The association between meningitis and dural
puncture in bacteremic rats. Anesthesiology 76:739-742, 1992
37. Crawford J, James FM, Noite H, et al: Regional analgesia for patients
with chronic neurologic disease and similar conditions. Anaesthesia
36:821, 1981
38. Warren TM, Datta S, Ostheimer GW: Lumbar epidural anesthesia in a
patient with multiple sclerosis. Anesth Analg 61:1022-1023, 1982
39. Kalichman MW, Calcutt NA: Local anesthetic-induced conduction
block and nerve fiber injury in streptozotocin-diabetic rats. Anesthesi-
ology 77:941-947, 1992
40. Selander D, Dhuner KG, Lundborg G: Peripheral nerve injury due to
injection needles used for regional anesthesia. Acta Anaesthesiol
Scand 21:182-188, 1977
41. Selander D, Sjostrang J: Longitudinal spread of intraneurally injected
local anesthetics. Acta Anaesthesiol Scand 22:622-634, 1978
HORLOCKER