Acta Anaesthesiol Scand 2014; ••: ••–••
Printed in Singapore. All rights reserved
Major complications of epidural anesthesia:
a prospective study of 5083 cases at a single hospital
X.-H. Kang1, F.-P. Bao1, X.-X. Xiong1, M. Li1, T.-T. Jin1, J. Shao2 and S.-M. Zhu1
1
Department of Anesthesiology, the First Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China and 2Department of
Statistics, University of Wisconsin – Madison, Madison, WI, USA
Background: We undertook a prospective study of non-
obstetric epidurals placed in surgical inpatients at a single teach-
ing hospital to evaluate the incidence of and potential risk factors
for major complications of continuous epidural anesthesia.
Methods: Demographic information, details of the epidural
procedure, and complications (from the pre-anesthetic period
through resolution) were recorded for more than 5000 surgical
inpatients who underwent continuous epidural anesthesia in
our institution between March 2009 and April 2011. The inci-
dence of and risk factors for major complications were evaluated.
Results: During the study period, 5083 patients were inter-
viewed and their details were recorded (98% capture rate). Sixty-
nine (1.36%) experienced major complications: epidural
hematoma in 1 patient (0.02%), post-operative neurologic defi-
cits in 57 patients (1.12%), post-dural puncture headache in 7
patients (0.14%), and systemic local anesthetic toxicity in 4
patients (0.08%). Only one patient had permanent sequelae: uni-
E pidural blockade is a popular technique in the
peri- and post-operative periods and during
labor. Severe complications are rare but potentially
devastating.1–5 In China, the popularity of epidural
techniques has waned in recent years, partly due to
the increasing use of peripheral nerve blocks, but
also as a consequence of growing concern about
epidural complications. A better understanding of
the pathophysiology and risk factors for complica-
tions would help guide anesthesiologists in their
use of the technique and decrease the incidence of
adverse events. Although several large studies have
evaluated the risks of epidural blocks, most have not
studied a consecutive series of patients, and have
Attributed Institution: Department of Anesthesiology, the First
Affiliated Hospital, School of Medicine, Zhejiang University,
Hangzhou, China.
Registration of Clinical Trials: http://www.chictr.org/usercenter/
project/edit.aspx?proj=2353
Registration number: ChiCTR-OCS-11001887
bs_bs_banner
© 2014 The Acta Anaesthesiologica Scandinavica Foundation.
Published by John Wiley & Sons Ltd
ACTA ANAESTHESIOLOGICA SCANDINAVICA
doi: 10.1111/aas.12360
lateral lower limb paresthesia. Identified risk factors for neuro-
logic deficits were as follows: American Society of
Anesthesiologists status II–III, siting in the lumbar region,
orthopedic and urologic surgery, multiple attempts to site an
epidural, paresthesia during insertion, a history of neuraxial
anesthesia, and use of patient-controlled epidural analgesia.
Conclusions: Serious complications were very rare; only one
patient had permanent sequelae, and a single epidural
hematoma was diagnosed. Post-operative neurologic deficits
were more common, but most complications resolved spontane-
ously within 3 months and they rarely required intervention.
Accepted for publication 18 May 2014
© 2014 The Acta Anaesthesiologica Scandinavica Foundation.
Published by John Wiley & Sons Ltd
generally included chronic pain patients, women in
labor, or combined spinal-epidural anesthesia.6–9
Consequently, the incidence of complications
reported varies enormously and may be difficult to
apply to a general surgical population, or compare
with other institutions or health-care systems. Fur-
thermore, many serious incidents resulting from
epidural anesthesia have been reported as part of
retrospective case series or individual case
reports.1,8,10–12 Only a few prospective surveys
assessing large numbers of patients have been pub-
lished, most from more than one center.13
We undertook a prospective study of the compli-
cations attributed to epidural infusions in non-
obstetric surgical inpatients in a single, large,
tertiary teaching center. We evaluated: (1) the inci-
dence and characteristics of major complications
related to continuous epidural anesthesia using
standard drugs, equipment, and techniques; and (2)
the risk factors for major complications.
1
X.-H. Kang et al.
Materials and methods
The study was registered with the Chinese Clinical
Trial Register (http://www.chictr.org/cn/about
.aspx) and was approved by the Local Ethics Com-
mittee [The Medical Ethics Committee of the First
Affiliated Hospital, School of Medicine, Zhejiang
University, 79 Qingchun Road, Hangzhou City,
Zhejiang Province, China, 310003, protocol number:
(2011) Ethical Review (Scientific Research) No. (10),
approved 30 December 2011].
All anesthesiologists in the hospital participated.
All surgical inpatients undergoing epidural
anesthesia from 1 March 2009 to 30 April 2011 were
prospectively enrolled having given written consent
to participate.
The following data were recorded for each surgi-
cal inpatient undergoing epidural anesthesia while
alert: sex, age, height, weight, American Society of
Anesthesiologists (ASA) physical health status, pre-
existing neurologic conditions, insertion technique,
type and duration of surgery, block efficacy, techni-
cal complications during needle or catheter place-
ment, whether general anesthesia was also used,
and the occurrence of any major complications
during and after the operation. A checklist for pre-
existing neurologic conditions including history of
previous neuraxial anesthesia, and lumbar
intervertebral disc herniation, was documented or
diagnosed during the pre-operative interview
based on history, examination, and laboratory
studies. Insertion technique included a record of
level of needle placement, the number of attempts
to site the epidural (number of needle passes), the
use of post-operative patient-controlled epidural
analgesia (PCEA), the local anesthetic(s) adminis-
tered, and the use of epinephrine or other addi-
tives. The record of technical complications noted
at the time of siting included traumatic block place-
ment (evidence of bleeding), unplanned dural
puncture, and whether paresthesiae were elicited.
Body mass index was calculated as weight divided
by the square of the height (kg/m2). The efficacy of
the block was categorized as: (1) satisfactory
(surgery performed without additional interven-
tion); (2) unilateral anesthesia; (3) segmental or
incomplete anesthesia; or (4) no block/block
failure.
We also recorded the incidence of major compli-
cations that have previously been reported to have
arisen as a consequence of epidural anesthesia:8,13,14
(1) cardiac arrest requiring cardiac message and/or
epinephrine; (2) seizure; (3) acute respiratory failure
2
requiring tracheal intubation and/or assisted venti-
lation; (4) ‘total’ spinal anesthesia; (5) systemic local
anesthetic toxicity; (6) epidural hematoma; (7) new
or progressive post-operative neurologic deficits,
defined as the presence of motor deficits, sensory
deficits, painful paresthesiae, dysesthesias, or
hyperreflexia at the time of subsequent epidural
anesthesia; (8) post-dural puncture headache
(PDPH); (9) paraplegia; (10) cauda equina syn-
drome; (11) infectious complications such as epi-
dural abscess, meningoencephalitis, or catheter site
infections; (12) epidural catheter breakage; and (13)
death. The attending anesthesiologist responsible
for the epidural recorded all data. No attempt was
made to modify individual practice.
Post-operative follow-up was undertaken by
three anesthesiologists blinded to all details except
patient name, medical record number, surgical pro-
cedure, and date of surgery. Each patient was visited
twice on the ward on the first and second post-
operative days (24–48 h after surgery) by the same
anesthesiologist. On each occasion, a neurological
examination was performed to identify major com-
plications, and the extent of pain attributable to the
epidural itself was rated using an 11-point visual
analog scale where 0 represented ‘no pain’, 1–3 rep-
resented mild pain, 4–6 moderate pain, and 7–10
severe pain. After the initial interview, all patients
who had reported complications were contacted
again by telephone between the seventh and four-
teenth post-operative days. They were followed
until the complications had completely resolved or
the patient was unable or unwilling to be contacted
again. Duration of complications and length of
follow-up were recorded. Neurologic sequelae were
considered permanent if they lasted more than 6
months. The final declaration that follow-up was
complete was determined in each case by the two
experienced anesthesiologists who had reviewed
the initial evaluation.
Statistical analysis
Mean values are expressed as mean ± standard
deviation. Univariate analysis and multiple logistic
regression analysis were performed to identify risk
factors for complications. Univariate variables
were analyzed using the χ2 test or exact Fisher test.
Odds ratios and 95% confidence intervals (95% CI)
were obtained from stepwise logistic regression
analyses to quantify the independent risk factors. P
values of less than 0.05 were considered statistically
significant.
 Major complications of epidural anesthesia

Results caine (concentrations of 1–2%) and ropivacaine

Data were collected from 5185 surgical inpatients
undergoing an epidural procedure during the study
period; full records were available for 5083 (98.0%)
patients. The remaining 102 patients were excluded
as they declined to participate, or could not be suc-
cessfully interviewed.
Overview
Of the 5083 patients enrolled, 3560 (70.0%) were men
and 1523 (30.0%) were women (Table 1). The mean
age was 52.7 ± 17.0 years. A uniform 18-gauge
Tuohy epidural needle and multiport epidural cath-
eter were used in all cases. Pure or combined lido-
(concentrations of 0.375–0.75%) were used for epi-
dural anesthesia during operation. Patients were
likely to be ASA status I–II, to have had a lumbar
block, and to have undergone urologic, orthopedic,
or general surgery.
Multiple needle passes were required to site the
epidural in 491 cases (9.7%). Paresthesia during
needle or catheter placement was the most common
technical complication and occurred in 160 cases: in
12 cases, the epidural was abandoned, but in most
cases, the needle was withdrawn and the epidural
space was re-punctured. Accidental dural puncture
occurred in 24 cases during siting of the epidural; in
nine cases, the epidural was abandoned, and in 15

Table 1
Characteristics of the patient cohort.
Characteristics Number of patients Proportion
(n) (%)
Sex Male 3560 70
Female 1523 30
Age < 50 2068 40.7
≥ 50 3015 59.3
BMI < 25 3862 76.0
≥ 25 1221 24.0
ASA status I 2633 51.8
II 2233 43.9
III 217 4.3
Pre-existing neurologic condition History of neuraxial anesthesia 306 6.0
Lumbar intervertebral disc hernia 54 1.1
Type of Surgery Orthopedic 1043 20.5
Urologic 1928 37.9
General 1302 25.6
Anorectal 336 6.6
Gynecological 71 1.4
Thoracic 38 0.7
Vascular 365 7.2
Duration of surgery < 60 min 2498 49.1
≥ 60 min 2585 50.9
Puncture level Lumbar level 4090 80.5
Thoracic level 993 19.5
Multiple attempts to locate epidural space No 4592 90.3
Yes 491 9.7
Block efficacy Satisfactory 4821 94.8
Unilateral 31 0.6
Segmental or incomplete 122 2.4
No block(block failure) 109 2.2
Combined general anesthesia Yes 618 12.2
No 4465 87.8
Technical complications Traumatic bloody 65 1.3
Accidental dural puncture 24 0.5
Paresthesiae during insertion 160 3.1
Local anesthetic Lidocaine 221 4.3
Mixture of lidocaine and ropivacaine 4243 83.5
Ropivacaine 590 11.6
Post-operative PCEA Yes 873 17.2
No 4210 82.8

Proportions (%) are based on the number of patients with available data.
BMI, body mass index; ASA, American Society of Anesthesiologists; PCEA, patient-controlled epidural analgesia.

3
X.-H. Kang et al.

Table 2
Number and incidence of major complications between different surgical specialties.
Type of surgery Post-operative Epidural hematoma, Post-dural puncture Systemic local
neurologic deficits, n (%) n (%) headache, n (%) anesthetic toxicity, n (%)
Orthopedic surgery 13 (1.25%) 0 (0) 1 (0.10%) 2 (0.20%)
Urologic surgery 33 (1.71%) 0 (0) 4 (0.21%) 1 (0.05%)
General surgery 8 (0.61%) 0 (0) 1 (0.08%) 0 (0)
Anorectal surgery 1 (0.30%) 0 (0) 0 (0) 0 (0)
Gynecology 1 (1.41%) 0 (0) 0 (0) 0 (0)
Thoracic surgery 0 (0) 0 (0) 0 (0) 0 (0)
Vascular surgery 1 (0.27%) 1 (0.27%) 1 (0.27%) 1 (0.27%)
Total 57 (1.12%) 1 (0.02%) 7 (0.14%) 4 (0.08%)

Proportions (%) are based on the number of patients with available data.

Table 3
Major complications related to outcome.
Type of complication < 2 day 3–7 day 8 day–6 months > 6 months
Post-operative neurologic deficits 24 25 7 1
Epidural hematoma 1
Post-dural puncture headache 2 3 2
Systemic local anesthetic toxicity 4

cases, the epidural was sited at a lower level. No
spinal anesthesia was used. Blood was detected
during needle or catheter placement in 65 cases
(1.3%); on all occasions, a new attempt was made to
locate the epidural space and insert the catheter.
Ropivacaine was the most popular local
anesthetic. Notably, vasoconstrictors were not
added to the epidural drug infusion in a single case.
Only 873 patients (17.2%) underwent extended post-
operative PCEA; the catheter was removed immedi-
ately after surgery in the remaining patients.
Ropivacaine (concentrations of 0.125–0.18%) was the
only drug used for PCEA infusions. Analgesia was
satisfactory in 4821 cases (94.8%). Six hundred and
eighteen patients received a combined general
anesthesia technique because of surgical require-
ments or a block failure. There was no statistically
significant difference in the incidence of major com-
plications in these patients compared with those
who received epidurals alone.
Major complications
Sixty-nine patients (1.36%) experienced major com-
plications after their epidural (Tables 2 and 3),
which comprised epidural hematoma in one patient
(0.02%), post-operative neurologic deficits in 57
patients (1.12%), systemic local anesthetic toxicity in
four patients (0.08%), and PDPH in seven patients
(0.14%). The number of major complications,
excluding post-operative neurologic deficits, was
too small for meaningful statistical analysis. No
severe permanent sequelae were noted, although
one patient died 5 days after femoral head replace-
ment. The cause of death was pulmonary embolism,
and deemed not to be related to epidural anesthesia.
Post-operative neurologic deficits
Fifty-seven patients (1.12%) developed new or pro-
gressive post-operative neurologic deficits attribut-
able to epidural anesthesia. Of these, 45 reported
numbness and/or sensory deficits in the lower
extremities. Seven patients suffered from moderate
buttock and back pain radiating to the thighs, which
had recovered completely within 7 days. Five
patients reported pain and paresthesiae in the lower
limb, which was mild in three and moderate in two.
In all those who experienced major complications,
the onset of symptoms occurred from 2 h to 48 h
after surgery; 83% had transient symptoms that
recovered within 7 days. One patient reported per-
manent pain and paresthesia in the left lower limb.
The 37-year-old patient was healthy and underwent
surgery of the right knee in the supine position.
Continuous epidural anesthesia was sited at the
L2-3 interspace and paresthesia of the left thigh was
elicited during the puncture. PCEA was not used
post-operatively. Subsequent electromyography was
performed and revealed damage to the left L3 nerve

4
 Major complications of epidural anesthesia

Table 4
Potential risk factors for post-operative neurologic deficits in patients undergoing epidural anesthesia identified by univariate analysis.
Risk factor Number of patients (n) Incidence {%} P value
Sex Male 3560 1.1 0.789
Female 1523 1.2
Age < 50 2068 0.9 0.256
≥ 50 3015 1.3
BMI < 25 3862 1.1 0.683
≥ 25 1221 1.2
ASA status I 2633 0.8 0.011
II–III 2450 1.5
Type of Surgery Urologic 1928 1.7 0.016
Orthopedic 1043 1.3
General 1302 0.6
Vascular 365 0.3
Anorectal 336 0.3
Gynecological 71 1.4
Thoracic 38 0
Duration of surgery < 60 min 2498 0.9 0.182
≥ 60 min 2585 1.3
Puncture level Lumbar level 4090 1.3 0.002
Thoracic level 993 0.2
Satisfactory block Yes 4821 1.1 0.537
No 262 1.5
Lidocaine Yes 4464 1.1 0.213
No 619 1.6
Lumbar intervertebral disc hernia Yes 54 3.7 0.122
No 5029 1.1
History of neuraxial anesthesia Yes 306 4.6 < 0.001
No 4777 0.9
Multiple attempts to locate epidural space Yes 491 3.1 < 0.001
No 4592 0.9
Paresthesiae during insertion Yes 160 5.6 < 0.001
No 4923 1
Combined general anesthesia Yes 618 1.3 0.682
No 4465 1.1
Post-operative PCEA Yes 873 1.9 0.02
No 4210 1

BMI, body mass index; ASA, American Society of Anesthesiologists; PCEA, patient-controlled epidural analgesia.

root. The symptoms improved after over 10 months
and with mild persistent sequelae.
The factors analyzed for their contribution toward
the occurrence of post-operative neurologic deficits
are listed in Table 4. The incidence of neurologic
deficits was higher among ASA status II–II patients
(1.5%) and in the lumbar siting group (1.3%). We
also found that multiple attempts to identify the
epidural space increased the incidence to 3.1%, com-
pared with 0.9% for patients in whom only one pass
was necessary. In addition, the risk of neurologic
deficits was higher among patients undergoing uro-
logic surgery (1.7%) and orthopedic surgery of the
lower limb (1.4%), especially when the lithotomy
position was used. The use of post-operative PCEA
increased the incidence of numbness and sensory
deficit twofold. Our finding that post-operative neu-
rologic deficits were five times more likely in
patients with a history of previous neuraxial
anesthesia and were also more than sixfold among
patients with paresthesia during insertion.
Univariate analysis identified ASA status II–III,
siting in the lumbar region, orthopedic and urologic
surgery, multiple attempts to site an epidural, the
use of post-operative PCEA, paresthesiae during
needle or catheter placement, and a history of pre-
vious neuraxial anesthesia as risk factors for post-
operative neurologic deficits. Logistic regression
analysis found that the two factors were independ-
ent determinants: paresthesiae during needle or
catheter placement (odds ratio, 6.99; 95% CI, 3.21 to
15.24; P < 0.001) and a history of neuraxial
anesthesia (odds ratio, 4.97; 95% CI, 2.61 to 9.43;
P < 0.001).
Epidural hematoma
Epidural hematoma occurred in one patient (0.02%)
undergoing right lower limb vascular surgery. The

5
X.-H. Kang et al.
patient was a 79-year-old male with diabetes who
had undergone amputation of the left lower extrem-
ity because of arteriosclerosis obliterans 1 year
before. Pre-operative laboratory studies revealed a
normal platelet count, bleeding time, and clotting
time. In this case, the epidural was sited unevent-
fully at the L2-3 interspace, and 30 mg of
unfractionated heparin was administered intrave-
nously during the operation approximately 1.5 h
after epidural placement. The patient received 40 mg
of enoxaparin subcutaneously early post-
operatively, and the epidural catheter was acciden-
tally removed within 2 h after the administration of
enoxaparin. The patient reported severe backache,
pain, and numbness of the right foot shortly after
surgery. Magnetic resonance imaging revealed a
small epidural hematoma at L3-4. Surgical evacua-
tion was not needed; the patient was managed con-
servatively, and there were no severe sequelae.
Post-dural puncture headache
The rate of accidental dural puncture was 0.5%, and
PDPH developed in less than 30% of these cases.
Five patients reported mild PDPH and two moder-
ate PDPH. Headache resolved within 7 days in five
patients and within 8–14 days in the remaining two
(Table 3). All were managed conservatively and
none required an epidural blood patch.
Systemic local anesthetic toxicity
Four patients experienced mild systemic local
anesthetic toxicity, but none had undergone trau-
matic or bloody needle placement. Their distinctive
symptoms and signs were mild and definitely
affected the central nervous system (diplopia, dizzi-
ness, minor auditory symptoms, or metallic taste).
No patients experienced a cardiac arrest or seizure
as a consequence of local anesthetic toxicity and
there were no post-operative sequelae.
Discussion
We have established the incidence of almost all the
major complications attributed to epidural
anesthesia. In our opinion, our study design con-
tributed to better reporting. First, data were col-
lected prospectively from a single, large, tertiary
teaching hospital, allowing us to have confidence in
the consistency of data, the clinical practices used,
and the validity of the denominator. Second, we
studied only inpatients, allowing detailed inter-
views and sensory and motor examinations to be
conducted in the ward for almost every case.
6
Although permanent neurologic sequelae were rare
in our study, it could be argued that the post-
operative interview and examination influenced
patient care and might have helped to improve
outcome. Third, each patient was visited from the
first post-operative day until any complications
resolved, allowing us to detect early and transitory
symptoms. In addition, only 102 patients were
excluded and the proportion lost to follow-up was
also very low. Finally, we did not study epidurals
sited for labor. It has previously been reported that
neurologic complications after labor epidural are
commonly caused by obstetric reasons.15–17 Conse-
quently, the unspecified inclusion of obstetric
patients will likely distort any analysis.
Post-operative neurologic deficits
The etiology of a post-operative neurologic deficit
may be difficult to establish as many patients, surgi-
cal, and anesthetic risk factors may play a role.18
However, we hoped that our systematic approach
would enable us to identify subgroups of patients
with a higher or lower prevalence of risk factors.
Paresthesia during insertion and multiple attempts
to puncture significantly increase the risks of neuro-
logic deficits, leading to permanent neurologic
deficits in one case. This suggests that direct trauma
to nerve roots causes pain in the appropriate
dermatome and that this may occasionally result in
long-lasting paresthesia. We found an increased risk
for neurologic symptoms among patients undergo-
ing lumbar epidural, urologic surgery, and
orthopedic surgery of the lower limb especially in
lithotomy position. In these cases, surgical trauma,
lumbar siting, unsatisfactory patient positioning, or
incorrectly applied surgical dressings may contrib-
ute to neurologic injury.
Our results also indicate that the post-operative
PCEA was associated with the twofold increased
incidence of numbness and sensory deficit. Deposi-
tion of local anesthetics into a root sleeve over a long
time period could result in prolonged contact
between nervous tissue and the local anesthetic.
Local anesthetic alone was used for PCEA in our
study, and the higher concentration and volume of
local anesthetic used in these cases may potentially
result in neurotoxic effects. The higher volume of
the drug infused may cause nerve injury by direct
pressure or indirectly, through vascular compro-
mise. A history of previous neuraxial anesthesia, in
our region almost always epidural anesthesia, was
identified as an independent determinants of neu-
rologic deficits in our study. It may be explained by

the presence of adhesions or scar tissue in the epi-
dural space, which could increase local pressure in
the epidural space during drug infusion or make the
epidural more technically difficult to site, increasing
the risk of nerve trauma.
Notably, we found that only seven patients expe-
rienced moderate post-operative low back or
buttock pain radiating to the thighs, all of which
had resolved completely within 7 days. These symp-
toms are suggestive of transient neurologic symp-
toms (TNS), a term coined to describe typical pains
experienced within 24 h of spinal anesthesia.19 Few
studies have examined the incidence of TNS
after epidural anesthesia.20,21 Our findings suggest
that it can occur after epidural anesthesia, but
while the precise etiology of TNS is not known,18,22
further study is required to establish whether the
diagnostic criteria for TNS also apply to epidural
anesthesia.
Epidural hematoma
The risk of neuraxial hemorrhagic events may
increase in patients receiving anticoagulants, and
the time interval between epidural manipulation
and anticoagulant drug administration is of para-
mount importance.23 For thromboprophylaxis, it is
advisable that low-molecular-weight heparin
(LMWH) be used as soon as possible. Our hospitals
have adopted the routine that the first enoxaparin
dose (of 40 mg) is typically given subcutaneously
within 2–6 h following vascular surgery.24 This high-
lights the risk of neuraxial hemorrhagic events in
patients, even when the insertion of the epidural is
considered ‘atraumatic’.
In our epidural hematoma case, the use of LMWH
in the immediate perioperative period and the acci-
dental removal of the epidural catheter within 2 h of
LMWH administration were believed to be contrib-
uting factors. Siting an indwelling epidural catheter
rather than performing a single-shot injection
appears to further increase the risk.25 It is recom-
mended that LMWH administration should be
delayed no less than 6 h following placement of an
epidural. Furthermore, the removal of the epidural
catheter should be delayed by no less than 10 h after
a dose of 40 mg or less of enoxaparin or 5000 U or
less of dalteparin, but no less than 24 h after a dose
higher than 40 mg/5000 U.26
Although the incidence of complications caused
by epidural hematomas is low, the consequences
can be disastrous and irreversible if not promptly
recognized and treated (usually not much later than
10–12 h from the start of symptoms26). In our study,
Major complications of epidural anesthesia
the single epidural hematoma was likely detected
quickly as a result of the detailed neurological
examinations we performed on all patients. A thor-
ough initial assessment and risk–benefit analysis
regarding the use of epidural anesthesia, careful
catheter placement and removal, and close monitor-
ing of post-operative neurologic signs can mitigate
the risk of an epidural hematoma.
Post-dural puncture headache
It is important to diagnose PDPH correctly: we
defined PDPH as a headache that developed within
7 days of dural puncture that worsened within
15 min of standing and diminished or disappeared
within 30 min of lying down.27,28 The incidences of
post-operative orthostatic and non-orthostatic head-
ache were similar in our study, and eight patients
were erroneously given the diagnosis of PDPH by
their surgeon. This finding is in broad agreement
with other studies, which found that almost half
post-operative headaches were tension headaches
or migraine.29 This underlines the importance of
considering other potential diagnoses in patients
with a headache who have also recently had a dural
puncture.
Systemic local anesthetic toxicity
Notably, no systemic toxicity from local anesthetics
occurred in the 65 patients in whom blood was
detected during puncture. We attribute this to the
care taken in making a new attempt to site the epi-
dural and the particular attention paid to these
patients. Accidental injection of inappropriate solu-
tions can be very harmful, and the low incidence in
our cohort may reflect the low doses, slow and frac-
tionated injection, and use of a test dose that are our
routine clinical practices.
Our study had some limitations. First, it was not
double blind, because the anesthesiologists per-
forming epidural anesthesia recorded data about
patient characteristics and epidural insertion. Nev-
ertheless, the investigators who contacted patients
and collected post-operative data were unaware of
all aspects of anesthesia care, so it is unlikely that
investigator bias influenced our findings. Second,
electromyography studies were only performed for
patients with longer term neurologic deficits.
In conclusion, we observed major complications
as a consequence of epidural anesthesia in surgical
inpatients in 1.36% of cases. Serious complications
were very rare; only one patient had permanent
sequelae, and a single epidural hematoma was diag-
nosed. However, post-operative neurologic deficits
7
X.-H. Kang et al.
were more common and require vigilance from cli-
nicians. ASA status II–III, siting in the lumbar
region, orthopedic and urologic surgery, multiple
attempts to site an epidural, paresthesiae during
insertion, a history of neuraxial anesthesia, and the
use of post-operative PCEA were identified as risk
factors for neurologic deficit. Furthermore, logistic
regression analysis identified paresthesiae during
insertion (odds ratio, 6.99; 95% CI, 3.21 to 15.24) and
a history of neuraxial anesthesia (odds ratio, 4.97;
95% CI, 2.61 to 9.43) as independent risk contribu-
tions. Most of the complications resolved spontane-
ously within 3 months and all were managed
conservatively. We suggest that the prevention and
management of the complications of epidural
anesthesia should begin at the pre-operative visit
with a careful evaluation of the patient’s medical
history. Appropriate judgment of the risks, post-
operative interview, and examination may help to
decrease the incidence of major complications of
epidural techniques.
Conflicts of interest: The authors have no conflicts
of interest.
Funding: Departmental funding only.
References
1. Christie IW, McCabe S. Major complications of epidural
analgesia after surgery: results of a six-year survey. Anaes-
thesia 2007; 62: 335–41.
2. Lee LA, Posner KL, Domino KB, Caplan RA, Cheney FW.
Injuries associated with regional anesthesia in the 1980s and
1990s. Anesthesiology 2004; 101: 143–52.
3. Moen V, Dahlgren N, Irestedt L. Severe neurological com-
plications after central neuraxial blockades in Sweden 1990–
1999. Anesthesiology 2004; 101: 950–9.
4. Volk T, Wolf A, Van Aken H, Buerkle H, Wiebalck A,
Steinfeldt T. Incidence of spinal haematoma after epidural
puncture: analysis from the German network for safety
in regional anaesthesia. Eur J Anaesthesiol 2012; 29: 170–
6.
5. De Sèze MP, Sztark F, Janvier G, Joseph PA. Severe and
long-lasting complications of the nerve root and spinal cord
after central neuraxial blockade. Anesth Analg 2007; 104:
975–9.
6. Moen V, Irestedt L, Dahlgren N. Major complications of
central neuraxial block: the Third National Audit Project:
some comments and questions. Br J Anaesth 2009; 103:
130–1.
7. Grewal S, Hocking G, Wildsmith JAW. Epidural abscesses.
Br J Anaesth 2006; 96: 292–302.
8. Cook TM, Counsell D, Wildsmith JA, Royal College of
Anaesthetists Third National Audit Project. Major complica-
tions of central neuraxial block: report on the Third National
Audit Project of the Royal College of Anaesthetists. Br J
Anaesth 2009; 102: 179–90.
9. Koryachkin VA, Glushchenko VA, Strashnov VI. Regional
anesthesia: combined spinal epidural anesthesia. Anesteziol
Reanimatol 2007; 5: 72–4.
8
10. PitkaNen MT, Aromaa U, Cozanitis DA, Forster JG. Serious
complications associated with spinal and epidural anaesthe-
sia in Finland from 2000 to 2009. Acta Anaesthesiol Scand
2013; 57: 553–64.
11. Aldrete JA. Neurologic deficits and arachnoiditis following
neuroaxial anesthesia. Acta Anaesthesiol Scand 2003; 47:
3–12.
12. Lala PS, Langar V, Rai A, Singh R. A rare complication of
epidural anaesthesia a case report with brief review of lit-
erature. Indian J Anaesth 2011; 55: 629–30.
13. Auroy Y, Benhamou D, Bargues L, Ecoffey C, Falissard B,
Mercier FJ, Bouaziz H, Samii K. Major complications
of regional anesthesia in France: the SOS Regional
Anesthesia hotline service. Anesthesiology 2002; 97: 1274–
80.
14. Finucane BT. Complications of regional anesthesia, 2nd edn.
New York: Springer Science + Business Media, 2007.
15. Moen VA, Irestedt LB. Neurological complications following
central neuraxial blockades in obstetrics. Curr Opin
Anaesthesiol 2008; 21: 275–80.
16. Wong CA. Nerve injuries after neuraxial anaesthesia and
their medicolegal implications. Best Pract Res Clin Obstet
Gynaecol 2010; 24: 367–81.
17. Birnbach DJ, Hernandez M, van Zundert AAJ. Neurologic
complications of neuraxial analgesia for labor. Curr Opin
Anaesthesiol 2005; 18: 513–7.
18. Horlocker TTMD, Wedel DJMD. Complications of spinal
and epidural anesthesia. Reg Anesth Pain Med 2000; 25:
83–98.
19. Freedman JM, Li DK, Drasner K, Jaskela MC, Larsen B, Wi S.
Transient neurologic symptoms after spinal anesthesia: an
epidemiologic study of 1,863 patients. Anesthesiology 1998;
89: 633–41.
20. Shifman E, Floka S, Got I. Transient neurological symptoms
of epidural analgesia in labor. Acta Anaesthesiol Scand 2009;
53(Suppl. 119): 62.
21. Mathur SK, Chakraborty A, Jain S, Jain A. Thoracic epidural
catheterization leading to delayed transient neurological
symptoms with normal imaging findings. Acta Anaesthesiol
Scand 2008; 52: 862–3.
22. Zaric D, Christiansen C, Pace NL, Punjasawadwong Y. Tran-
sient neurologic symptoms after spinal anesthesia with lido-
caine versus other local anesthetics: a systematic review of
randomized, controlled trials. Anesth Analg 2005; 100:
1811–6.
23. Horlocker TT, Wedel DJ, Rowlingson JC, Enneking FK,
Kopp SL, Benzon HT, Brown DL, Heit JA, Mulroy MF,
Rosenquist RW, Tryba M, Yuan CS. Regional anesthesia in
the patient receiving antithrombotic or thrombolytic therapy
American Society of Regional Anesthesia and Pain Medicine
Evidence-Based Guidelines (third edition). Reg Anesth Pain
Med 2010; 35: 64–101.
24. Samama CM, Albaladejo P, Benhamou D, Bertin-
Maghit M, Bruder N, Doublet JD, Laversin S, Leclerc S,
Marret E, Mismetti P, Samain E, Steib A. Venous
thromboembolism prevention in surgery and obstetrics:
clinical practice guidelines. Eur J Anaesthesiol 2006; 23:
95–116.
25. Horlocker TT, Wedel DJ, Benzon H, Brown DL, Enneking
KF, Heit JA, et al. Regional anesthesia in the anticoagulated
patient: defining the risks. Reg Anesth Pain Med 2004; 29 (2
Suppl. 1): 1–12.
26. Breivik H, Bang U, Jalonen J, Vigfússon G, Alahuhta S,
Lagerkranser M. Nordic guidelines for neuraxial blocks in
disturbed haemostasis from the Scandinavian Society of
Anaesthesiology and Intensive Care Medicine. Acta
Anaesthesiol Scand 2010; 54: 16–41.
 Major complications of epidural anesthesia

27. Bezov D, Lipton RB, Ashina S. Post-dural puncture head- Address:

ache: part I diagnosis, epidemiology, etiology, and patho- Sheng-mei Zhu
physiology. Headache 2010; 50: 1144–52. Department of Anesthesiology, the First Affiliated Hospital,
28. Bezov D, Ashina S, Lipton R. Post-dural puncture headache: School of Medicine, Zhejiang University
part II – prevention, management, and prognosis. Headache 79 Qingchun Road
2010; 50: 1482–98. Hangzhou
29. Stella CL, Jodicke CD, How HY, Harkness UF, Sibai BM. 310003 Zhejiang Province
Postpartum headache: is your work-up complete? Am J China
Obstet Gynecol 2007; 196: 318.e1–7. e-mail: smzhu20088@163.com