Full text of DOI:10.1007/BF03016829

538
Regional Anesthesia and Pain
Epidural anesthesia with lidocaine reduces propo-
fol injection pain
[L’anesthésie péridurale avec lidocaïne réduit la douleur provoquée par l’injection
de propofol]
Tatsusuke Yoshikawa MD PhD,*†‡ Zen’ichiro Wajima MD PhD,† Tetsuo Inoue MD PhD,† Akira Ogura MD PhD,§
Ryo Ogawa MD PhD§
Purpose: To determine whether epidural lidocaine reduces the
severity of propofol injection pain compared with iv lidocaine.
Méthode : L’étude clinique, prospective, randomisée et à double insu,
a été menée auprès de 120 femmes pour qui une laparotomie gyné-
cologique avait été prévue. Un cathéter péridural lombaire et un
cathéter iv, introduit dans la veine céphalique de la main la moins util-
isée, ont été employés chez toutes les patientes. Dans le groupe
témoin T, on a administré un soluté physiologique par voie péridurale,
puis iv, suivi de propofol iv. Dans le groupe P, on a donné de la lidocaïne
Methods: A prospective, randomized double-blind clinical study
was conducted in 120 female patients scheduled for elective gyne-
cological laparotomy. A lumbar epidural catheter and an iv catheter
placed in the cephalic vein of the non-dominant hand were used in
all patients. Patients of the control group (Group C) were given
epidural normal saline followed by iv normal saline then iv propo-
fol. Patients of Group E were given epidural 2% lidocaine (0.08
mL·cm–1) followed by iv normal saline and then propofol. Patients
of Group V were given epidural normal saline followed by iv 2%
lidocaine (0.05 mL·kg–1) then propofol. Pain was scored as no
pain=0, minimal pain=1, moderate pain=2, severe pain=3.
-1
à 2 % (0,08 mL·cm péridurale, puis un soluté physiologique et du
propofol iv. Dans le groupe V, on a donné un soluté physiologique
péridural suivi de lidocaïne iv à 2 % (0,05 mL·kg–1) et de propofol. La
douleur a été cotée comme suit : 0 = aucune douleur, 1 = douleur
minimale, 2 = douleur modérée, 3 = douleur sévère.
Résultats : Dans les groupes P, 1 (0–2) et V, 2 (0–2), les seuils de
douleur ont été significativement plus bas que ceux du groupe T, 2
(1–3); médiane (25e–75e percentile) (P <0,001). Les scores n’é-
taient pas différents entre les groupes P et V. La concentration plas-
matique de lidocaïne, 15 min après l’injection péridurale, était de
2,74 ± 0,54 µg·mL–1, comparé à 1,54 ± 0,31 µg·mL–1 trois minutes
après l’injection iv de lidocaïne.
Results: The pain scores, in group E; 1 (0–2) and group V; 2 (0–2),
were significantly lower than in group C; 2 (1–3); median (25th–75th
percentile) (P <0.001). There was no difference in pain score
between groups E and V. The plasma lidocaine concentration 15
min after epidural lidocaine was 2.74 ± 0.54 µg·mL–1, compared
with 1.54 ± 0.31 µg·mL–1 at three minutes after iv lidocaine.
Conclusion : L’administration péridurale et iv de lidocaïne réduit
également la sévérité de la douleur causée par l’injection de propofol
malgré de plus fortes concentrations plasmatiques de lidocaïne péridu-
rale.
Conclusion: Epidural and iv lidocaine equally reduced the severity
of propofol injection pain despite higher lidocaine plasma concen-
trations in epidurally administered lidocaine.
Objectif : Déterminer si la lidocaïne, en injection péridurale com-
parée à l’injection iv, peut réduire la douleur causée par l’administra-
tion de propofol.
From the Department of Anaesthesia,* Hakujikai Memorial Hospital, Tokyo, the Department of Anaesthesia,† Chiba Hokusoh Hospital,
Chiba, the Department of Anaesthesiology,‡ Tokyo Jikeikai Medical School, Tokyo, and the Department of Anaesthesiology,§ Main
Hospital, Nippon Medical School, Tokyo, Japan.
Address correspondence to: Dr. T. Yoshikawa, 5-11-1 Shikahama, Adachi-ku, Tokyo 123-0864, Japan. Phone: +81-3-3899-1311, ext.
2116; Fax: +81-3-3855-2851; E-mail: FZN03121@nifty.ne.jp
This work was carried out at the Chiba Hokusoh Hospital, Nippon Medical School, 1715 Kamakari, Inba-mura, Inba-gun, Chiba
prefecture 270-1694, Japan.
Accepted for publication October 24, 2000.
Revision accepted March 4, 2001.

Yoshikawa et al.: PROPOFOL INJECTION PAIN
539
AIN during injection of propofol remains a
clinical problem. Our clinical experience
suggests that lumbar epidural anesthesia
patients of group E was confirmed with the pinprick
test 15 min after epidural administration of lidocaine.
–1
Normal saline (0.08 mL·cm body height) was
P
with lidocaine 2% diminishes propofol injec-
injected epidurally in patients of group V (n=37). This
was followed 12 min later by iv injection of 2% lido-
caine (0.05 mL·kg–1), and three minutes later by
propofol through the same vein as lidocaine.
tion pain, when the latter is administered after analge-
sia is established, 15 min postepidural lidocaine
administration. In the present study, we investigated
whether epidural lidocaine reduces the severity of pain
during propofol injection with iv lidocaine and mea-
sured plasma lidocaine.
Injection of lidocaine or normal saline into the
epidural space was performed using a syringe pump
(STC 525™, Terumo, Japan) at a rate of 400 mL·hr–1.
The dose of lidocaine was less than the recommended
dose for ventricular arrhythmias and possible side-
effects were explained in advance. Propofol was kept
in an incubator set at 20–23°C until just before
administration. In the operating theatre, after setting
conventional monitors and cannulating the radial
artery, propofol was administered through a three-way
stopcock directly connected to the cannula inserted
into the cephalic vein, with the iv infusion line closed,
using an electric-powered syringe pump (Graseby
Anaesthesia Pump 3500™, Graseby Medical, UK).
The rate of injection was 3 mg·kg–1min– 1and the total
dose was 2.5 mg·kg–1.
The severity of pain during injection of propofol
was rated using a four-point scale. It was considered to
be too difficult for the patient to express the injection
pain using complex terms within a short period during
induction of anesthesia. Accordingly, before the
administration of lidocaine, the patient was asked to
evaluate pain during the injection of propofol and,
when present, grade it as mild=“grade 1”, moder-
ate=“grade 2 ”, or severe=“grade 3”, as used in previ-
ous reports.1–7 Absence of pain was scored as “grade
0”. In each patient, the maximum degree of pain
before falling asleep was recorded as the pain score.
The time point when evaluating pain score was also
recorded. In patients with pain score “0”, “the time
point” was recorded as 50 sec, representing the total
period required for the injection of propofol.
Arterial blood samples were obtained at 30 sec,
one, two, three, five, seven, ten, 15 and 30 min after
epidural lidocaine and at 30 sec, one, two, three and
five minutes after iv lidocaine, to determine the plas-
ma lidocaine concentrations. Blood samples were col-
lected from the arterial cannula.
After induction of anesthesia, the lungs were venti-
lated mechanically with 30–50% oxygen in air through
an endotracheal tube, in combination with epidural
and continuous iv anesthesia using propofol. We
administered epidural analgesia postoperatively with a
combination of local anesthetics and opioids. The
relationship between the pain score and the level of
analgesia was analysed for group E. We also analysed
Methods
The study protocol was approved by the Human
Ethics Review Committee of Nippon Medical School
and a signed consent was obtained from each subject.
The study included 120 consecutive female patients,
ASA I or II, aged 30–60 yr, undergoing elective gyne-
cological laparotomy. None of the patients received
premedication. Patients taking sedatives or analgesic
agents and those with neurological or cardiovascular
involvement were excluded from the study.
Following injection of a small dose (<2 mL) of pro-
caine hydrochloride 1% for local anesthesia, the epidur-
al space was punctured using the loss-of-resistance
technique with normal saline at the L2–3 intervertebral
space by a paramedian approach. An epidural catheter
was inserted and its tip placed 4–5 cm cephalad.
An 18-gauge Teflon® cannula (Insyte-W™
Vialon™ E, Becton Dickinson, NJ) was inserted into
the cephalic vein of the non-dominant hand, without
the use of local anesthetic, early in the morning (about
three hours before the induction of anesthesia) in the
ward on the day of surgery and iv infusion of lactate
Ringer’s solution was started.
For arterial pressure monitoring and blood sam-
pling, a 20-gauge Teflon® cannula (Insyte-W™
Vialon™ E, Becton Dickinson) was inserted into the
radial artery on the forearm contralateral to the
venous line, following injection of a minimal dose (<1
mL) of procaine hydrochloride 1% for local anesthesia.
Patients were randomly allocated into one of three
–1
groups (C, E and V). Normal saline (0.08 mL·cm
body height) was administered epidurally in patients
of group C (control group, n=40), followed 12 min
later by iv injection of normal saline (0.05 mL·kg–1).
Three minutes later, propofol (2.5 mg·kg– 1) was
injected through the same (cephalic) vein used for
injection of normal saline.
–1
Lidocaine 2% (0.08 mL·cm ) was administered
epidurally in patients of group E (n=43). This was fol-
lowed, 12 min later, by iv injection of normal saline,
and three minutes later by injection of propofol at a
dose similar to that used for group C. Analgesia in

540
CANADIAN JOURNAL OF ANESTHESIA
TABLE Demographic data of the three groups
Group C
Group E
Group V
P value
No. of patients
Age (yr)
4 0
42.8 6.7
4 3
37
43.5 8.8 46.4 5.9 NS
Height (cm)
157.2 5.3 155.4 5.4 154.5 5.0 NS
Weight (kg)
53.6 11.2 55.4 6.4 53.7 7.6 NS
ASA grade I/II (n) 24/16
21/22
18/19
NS
Group C received epidural normal saline followed by iv normal
saline, then propofol. Group E received epidural lidocaine fol-
lowed by iv normal saline, then propofol. Group V received
epidural normal saline followed by iv lidocaine, then propofol.
Values for age, height and weight are mean SD. There were no
significant differences (NS) among the three groups.
the relationships between the pain score and plasma
lidocaine concentrations at 15 min after epidural lido-
caine, and those between pain score and plasma lido-
caine concentrations at three minutes after iv
lidocaine. Plasma lidocaine levels were measured using
an automated analyser (TDX™, ABBOTT, IL) with
fluorescence polarization immunoassay.8 The detec-
tion range of lidocaine by this method is 0.1–10
µg·mL–1, according to the specifications provided by
the manufacturer.
Data are reported as mean SD. Differences in age,
body weight, body height and plasma lidocaine con-
centrations among the three groups were examined for
statistical significance by using one-way analysis of vari-
ance (ANOVA). Differences in ASA status among
groups was analysed with the chi-square test and those
for pain score with the Kruskal-Wallis test followed by
post-hoc Dunn’s multiple comparison test (GraphPad
PRISM™ version 3.02, GraphPad Software, San Diego,
CA). The relationship between the pain score and time
of pain evaluation, that between pain score and anal-
gesic levels, and between pain score and plasma lido-
caine concentrations were analysed with Spearman’s
correction coefficient by rank. The relationship
between pain score and “time points” was analysed
excluding patients with the pain score “0” because
there were no patients in group C with pain score “0”.
P <0.05 was regarded as significant. DeltaGraph®
4.03J for Windows™ (DeltaPoint®, Japan Polaroid
Computing) was used for illustrations.
FIGURE 1 Pain scores during injection of propofol in the three
groups. The lower and upper borders of each box show the 25th
and 75th percentiles, respectively. The thick horizontal line shows
the median, marking the 50th percentile. The vertical error bars
show the 10th percentile. Pain scores of groups E and V were sig-
nificantly lower than that in group C. There was no significant dif-
ference in pain scores between group E and group V. *P <0.001 vs
group C.
The median pain scores in groups E and V were
lower than that in group C (P <0.001). There was no
difference in the pain score between group E and
group V, as shown in Figure 1.
Plasma lidocaine concentrations in group E
increased progressively after epidural injection of the
agent and reached a peak value (2.73 0.54 µg·mL–1)
at 15 min after administration, which was significantly
higher than that at 30 min (2.12 0.56 µg·mL– 1) as
shown in Figure 2A. In group V, plasma lidocaine
concentrations at three minutes after iv injection was
1.58 0.31 µg·mL–1 (Figure 2B), which was signifi-
cantly lower than in group E (P <0.001).
The correlation coefficient (r value) of the relation-
ship between pain score and time point when evaluat-
ing pain score was 0.2 in group C, 0.5 in group E and
0.5 in group V (Figures 3A, B and C). The correlation
coefficient of the relationship between pain score and
number of dermatomes blocked in group E 15 min
after epidural lidocaine was 0.02 (Figure 4). The cor-
relation coefficient of the relationship between pain
score and plasma lidocaine concentrations was 0.07 in
group E (Figure 5A), and 0.03 in group V (Figure
Results
There were no differences in age, body weight, body
height or ASA status among the three groups (Table).
Serious symptoms or signs related to iv lidocaine were
not observed. However, 10% of patients developed
mild symptoms. These consisted of feeling of fullness in
the ear in four patients and tinnitus in three patients.

Yoshikawa et al.: PROPOFOL INJECTION PAIN
541
FIGURE 2 A, Plasma lidocaine concentrations in patients of
group E. Note the gradual increase and peak value at 15 min after
epidural administration of 2% lidocaine; B, Plasma lidocaine con-
centrations in patients of group V. Note the gradual fall after iv
administration of 2% lidocaine.
5B). There were no significant relationships between
the pain score and time point when evaluating pain
score, between the pain score and number of der-
matomes blocked, or between the pain score and plas-
ma lidocaine concentrations.
Discussion
FIGURE 3 A, Relationship between pain score and the time of
The major finding of the present study is that propo-
fol-related pain was significantly reduced after either iv
or epidural injection of lidocaine, despite higher lido-
caine plasma concentrations in patients with epidural-
ly administered lidocaine.
pain score evaluation in group C; B, Relationship between pain
score and the time of pain score evaluation in group E; C,
Relationship between pain score and the time of pain score evalua-
tion in group V. Numbers close to the solid circles represent the
number of patients for each symbol.

542
CANADIAN JOURNAL OF ANESTHESIA
FIGURE 4 Relationship between the number of analgesic der-
matomal segments evaluated 15 min after epidural injection of
lidocaine and plasma lidocaine concentrations in group E.
Among the various approaches used to reduce
propofol injection pain is the use of lidocaine. Lidocaine
is administered either mixed with propofol or prior to
injection of propofol. There is some evidence to suggest
that the analgesic effects of the former technique may
be more effective than the latter.9,10 However, mixing
of propofol emulsion with other drugs is not recom-
mended by the manufacturer,1 1 because of possible
changes in emulsion stability. Thus, the common prac-
tice of mixing lidocaine to propofol to reduce pain on
injection should be considered carefully.11 Picard and
Tramer12 used lidocaine (0.5 mg·kg–1) intravenously
with a rubber tourniquet on the forearm 30 to 120 sec
before propofol injection. They suggested that the
“tourniquet method” is the most efficacious procedure
to reduce propofol injection pain. Manger and Holak1 3
also demonstrated that application of a tourniquet
inflated at a pressure of 50 mmHg followed by iv lido-
caine (5 mL or 100 mg of lidocaine 2%) diminished
hand pain associated with subsequent propofol injec-
tion. However, even though they reported the lack of
any symptoms using this technique, it is possible that
the rapid flow of 100 mg of lidocaine into the systemic
circulation after the release of tourniquet might pro-
duce serious-side effects.
FIGURE 5 A, Relationship between pain score and plasma lido-
caine concentrations 15 min after epidural injection of lidocaine in
group E; B, Relationship between pain score and plasma lidocaine
concentrations three minutes after iv injection of lidocaine in
group V.
use of the tourniquet method were associated with the
least pain. Several studies have examined the cause of
propofol pain and the mechanisms of pain alleviation
by various therapeutic techniques. Scott et al.9 sug-
gested that propofol injection pain results from activa-
tion of the kinin cascade system and that lidocaine acts
as a stabilizer of this system, preventing the release of
pain-related mediators. When using the “tourniquet
method”, lidocaine held in the vein for a certain peri-
od of time results in anesthesia of the vein.1 3
Therefore, the efficacy of iv lidocaine should be high-
er with a tourniquet.
Picard and Tramer1 2 reviewed the results of using
15 different pharmaceutical agents by analysing 12
different physiological variables, in order to analyse
the most appropriate techniques to reduce the inci-
dence and severity of propofol injection pain. They
concluded that injection of lidocaine/propofol and
Propofol is presented in emulsion form due to its
low water solubility. The emulsion droplets are less than
1 µm in diameter.11 Nathanson et al.5 demonstrated

Yoshikawa et al.: PROPOFOL INJECTION PAIN
543
immediate changes in the stability of propofol emulsion
when 10–20 mg of lidocaine solution is added to 200
mg propofol. However, the changes in droplet size of
propofol emulsion following the addition of <20 mg
lidocaine to 200 mg propofol (final lidocaine concen-
tration, 0.95%), are unlikely to be of clinical impor-
tance. An increase in the droplet size to >5 µm poses
the risk of pulmonary fat embolism.14,15 In this regard,
Ho and colleagues1 6 reported that the optimal concen-
tration of lidocaine that reduced pain on injection of a
propofol-lidocaine mixture was 0.1%. This concentra-
tion is within the acceptable range suggested by
Nathanson et al.5 The mechanism explaining the effica-
cy of this technique we used remains unclear.1 1
School, Tokyo, Japan, for his continued encouragement
throughout the study. We also thank Dr. Nobuo
Nishimura, President of the Gerontological Medicine
Institute, Hakujikai Memorial Hospital, for his invalu-
able help in reviewing the manuscript.
References
1 Helbo-Hansen S, Westergaard V, Korgh BL, Svendsen
HP. The reduction of pain on injection of propofol:
the effect of addition of lignocaine. Acta Anaesthesiol
Scand 1988; 32: 502–4.
2 Valtonen M, Iisalo E, Kanto J, Rosenberg P. Propofol as
an induction agent in children: pain on injection and
pharmacokinetics. Acta Anaesthesiol Scand 1989; 33:
152–5.
Although iv lidocaine without a tourniquet prior to
propofol injection, our results show that the analgesic
effect of epidural lidocaine was not significantly different
from that of iv bolus injection of lidocaine and the pain
score after propofol injection was reduced to a similar
level in both methods. The analgesic effects of epidural
lidocaine on pain during injection of propofol might be
explained by the following mechanisms: (a) a direct
effect of epidural anesthesia; (b) the systemic effect of
lidocaine following absorption from the epidural space.
Furthermore, the effects of epidural lidocaine are likely
to represent the central sedative effects of spinal deaf-
ferentation and involvement of the spinal cord in sys-
temic analgesia. In our study, the level of analgesia did
not exceed the level of T5, there were no signs of sym-
pathetic nerve block in the upper extremities, and there
was no significant relationship between the pain score
and the level of analgesia. In group E, since lidocaine
acted epidurally as well as systemically, the analgesic
effect on pain during propofol injection should have
been greater in these patients compared to those of
group V. However, analysis of the pain scores showed no
difference between the two groups.
In conclusion, the present study shows that lidocaine
epidural anesthesia reduces pain during propofol injec-
tion. We also confirmed, in another group of patients,
that pre-treatment with iv lidocaine reduces propofol-
related pain. Both methods are useful to lessen propo-
fol injection pain when the latter is used for the
induction or maintenance of anesthesia. Our results also
indicate that it is unnecessary to use iv lidocaine to pre-
vent pain on injection of propofol when patients are
scheduled to receive epidural lidocaine anesthesia.
Further studies are necessary to investigate the pain
reducing mechanism of epidural lidocaine.
3 King SY, Davis FM, Wells JE, Murchison DJ, Pryor PJ.
Lidocaine for the prevention of pain due to injection of
propofol. Anesth Analg 1992; 74: 246–9.
4 Lee TW, Loewenthal AE, Strachan JA, Todd BD. Pain
during injection of propofol. The effect of prior admin-
istration of thiopentone. Anaesthesia 1994; 49: 817–8.
5 Nathanson MH, Gajraj NM, Russell JA. Prevention of
pain on injection of propofol: a comparison of lido-
caine with alfentanil. Anesth Analg 1996; 82: 469–71.
6 Yokota S, Komatsu T, Komura Y, et al. Pretreatment
with topical 60% lidocaine tape reduces pain on injec-
tion of propofol. Anesth Analg 1997; 85: 672–4.
7 Iwama H, Nakane M, Ohmori S, et al. Nafamostat
mesilate, a kallikrein inhibitor, prevents pain on injec-
tion with propofol. Br J Anaesth 1998; 81: 963–4.
8 Caplan YH, Levine B. Application of the Abott TDx
lidocaine, phenytoin, and phenobarbital assays to post-
mortem Blood Specimens. J Anal Toxicol 1988; 12:
265–7.
9 Scott RPF, Saunders DA, Norman J. Propofol: clinical
strategies for preventing the pain of injection.
Anaesthesia 1988; 43: 492–4.
10 Brooker J, Hull CJ, Stafford M. Effect lignocaine on
pain caused by propofol injection (Letter). Anaesthesia
1985; 40: 91–2.
11 Lilley EMM, Isert PR, Carasso ML, Kennedy RA. The
effect of the addition of lignocaine on propofol emul-
sion stability. Anaesthesia 1996; 51: 815–8.
12 Picard P, Tramer MR. Prevention of pain on injection
with propofol: a quantitative systemic review. Anesth
Analg 2000; 90: 963–9.
13 Mangar D, Holak EJ. Tourniquet at 50 mm Hg fol-
lowed by intravenous lidocaine diminishes hand pain
associated with propofol injection. Anesth Analg 1992;
74: 250–2.
14 Driscoll DF, Bhargava HN, Zaaim RH, Babayan VK,
Bistrain BR. Physicochemical stability of total nutrient
admixtures. Am J Health-Syst Pharm 1995; 15: 623–4.
Acknowledgements
We are grateful to Professor Dr. Yoshikiyo Amaki,
Department of Anaesthesiology, Tokyo Jikeikai Medical

544
CANADIAN JOURNAL OF ANESTHESIA
15 Lewis JS. Justification for use of 1.2 micron end-line fil-
ters on total nutrient admixtures. Hospital Pharmacy
1993; 28: 656–8.
16 Ho C-M, Tsou M-Y, Sun M-S, Chu C-C, Lee T-Y. The
optimal effective concentration of lidocaine to reduce
pain on injection of propofol. J Clin Anesth 1999; 11:
296–300.