The Clinical Effectiveness of

Epidural Bupivacaine,

Bupivacaine with Lidocaine,

and Bupivacaine with

Fentanyl for Labor Analgesia

J. Clin. Anesth. 1991;3:219-224

*Associate Professor of Anesthesiology

tClinica1 Fellow in Obstetrical Anesthesiology

$Assistant Professor of Anesthesiology

Address reprint requests to Dr. Sinatra at

the Department of Anesthesiology, T-3, Yale

University School of Medicine, Box 3333,

333 Cedar Street. New Haven, CT 06510,

USA.

Received for publication July 30, 1990; re-

vised manuscript accepted for publication

December 17, 1990.

Raymond S. Sinatra, MD, PhD,*

Robert Goldstein, MD,? Ferne B. Sevarino, MD+

Departments of Anesthesiology, Yale University School of Medicine and

Yale-New Haven Hospital, New Haven, CT.

Study Objective: To examine the efficacy of bupiuacaine alone and in combination

with lidocaine or fentanyl for epidural analgesia during labor.

Design: Randomized, single-blind study.

Setting: Labor and delivery unit at a university medical center.

Patients: Forty-five primiparas requesting epidural analgesia.

Interventions: Following epidural placement at L3-4 interspace, patients received

either bupivacaine 0.5% (Group I, n = 15), bupivacaine 0.25% with lidocaine

I % (Group 2, n = 15), or bupivacaine 0.5% with fentanyl50 pg in 10 ml of

saline (Group 3, n = 15). Patients in Groups 1 and 2 received 6 to 10 ml of

local anesthetic depending on patient height, while patients in Group 3 received 5

ml of local anesthetic plus 50 pg of fentanyl in IO ml of sa1in.e. All solutions

contained epinephrine 1:200,000.

Measurements and Main Results: Patients were assessed at regular intervals

,following administration of the epidural solution. Visual analog scale (VAS) scores

were used to measure onset of analgesia, time to complete pain relief, duration of

analgesia, and patient satisfaction with therapy. The frequency of shivering and

pruritus and the extent of sensorylmotor block also were evaluated. There were no

intragroup differences in time to complete pain relief or patient satisfaction. How-

ever, patients in Group 3 noted the most rapid onset and longest duration of pain

relief. Patients in Group 3 also experienced signzficantly less shivering and had

the lowest degree of motor block. Two patients in Group 3 experienced mild pruritus.

Conclusions: Epidurally administered fentanyl safely extended the duration of

labor analgesia while reducing bupivacaine dose requirements and magnitude of

motor block. In this setting, the combination of bupivacaine and lidocaine offered

no clinical advantage over bupivacaine alone.

0 1991 Butterworth-Heinemann Keywords: Anesthesia, obstetric; analgesia, epidural; local anesthetics,

bupivacaine, lidocaine; opioids, fentanyl.

.J. Clin. Anesth., vol. 3, May/June 1991 219

Introduction

The ideal epidurdl anesthetic for labor and delivery

should provide useful analgesia without significant

motor block or risk of maternal or fetal toxicity. In

this regard, 0.25% to 0.5% concentrations of bupi-

vacaine are commonly used, since they offer effective

and prolonged pain relief. 1,2 While bupivacaine’s pro-

longed duration of analgesia and lack of immediate

or long-term neonatal effects” offer advantages in this

setting, its slow onset and risk of maternal cardiotox-

icity, even in dilute concentrations,” represent unde-

sirable characteristics.

A number of pharmacologic adjuncts, including

bicarbonate,” opioids,6-n epinephriner and rapid-act-

ing local anesthetics, ‘“,‘1 have been added to solutions

of bupivacaine to speed analgesic onset, reduce dose

requirements, and lower risks of toxicity. Concomi-

tant administration of bupivacaine and opioid anal-

gesics appears to speed onset and improve the quality

of pain relief despite considerable reductions in local

anesthetic dose.7,1Z MaGee et al.” reported rapid onset

and reduced bupivacaine requirements when epi-

dural combinations of bupivacaine and lidocaine were

used for labor and delivery analgesia.

The following randomized, single-blind evaluation

studied the clinical effectiveness of epidural anes-

thesia observed with bupivacaine 0.5% against that

provided by combinations of bupivacaine 0.25%’ and

lidocaine 1% and bupivacaine 0.5% and fentanyl. The

evaluation measured time of onset and duration of

analgesia and determined the relative safety of each

solution administered as a single bolus dose in term

parturients in Stage I of active labor.

Materials and Methods

The study was approved by the Yale University School

of Medicine Human Investigation Committee, and

informed consent was obtained from all patients prior

to epidural placement. Forty-five term primiparas in

active labor (at 3 to 6 cm cervical dilation) requesting

epidural analgesia were randomly assigned to one of

three treatment groups. Patients selected were healthy

(ASA physical status II), had uncomplicated preg-

nancies with a singleton fetus in vertex position, had

not received parenteral analgesics or sedatives, were

at least 18 but no more than 40 years of age, weighed

no more than 110 kg, and were at least 152 cm but

no more than 183 cm in height. Patients with a history

of alcohol or drug abuse or a documented sensitivity

to opioids or local anesthetics were excluded from the

study.

Following prehydration with lactated Ringer’s so-

lution ( 15 ml/kg) and routine epidural placement at

the L3-4 interspace, patients received one of three

test solutions via a 19gauge single-port epidural cath-

eter (Kendall Curity, Mansfield, MA) inserted 2.5 cm

in randomized, single-blind fashion. Patients in Group

1 (bupivacaine, n = 15) received 5 ml of 0.5% bu-

pivacaine plus epinephrine 1:200,000 followed by ad-

ditional bolus doses based on height (per standard

routine, patients 152 cm tall received an additional 1

ml, those 160 cm tall received 3 ml, and those 170 cm

and taller received 5 ml). Patients in Group 2 (bupi-

vacaine with lidocaine, n = 15) received a 5 ml mix-

ture of 0.25% bupivacaine plus l.O%, lidocaine plus

epinephrine 1:200,000, followed by additional 1 to 5

ml doses as outlined for Group 1. Patients in Group

3 (bupivacaine with fentanyl, n = 15) were given 5

ml of 0.5% bupivacaine plus epinephrine I :200,000,

followed by 50 pg of fentanyl plus 9 ml of’O.<)O/r sterile

saline. All patients remaining uncomfortable 15 min-

utes after epidural administration were given supple-

mental 1 to 2 ml boluses of their respective stud)

solution as required. Epidural catheters \vere re-

checked in all patients remaining uncomf’ortable after

30 minutes. If a catheter was found to be misplaced,

the individual was dropped from evaluation and a

second epidural was inserted. ,411 medication was ad-

ministered in divided doses with patients supine and

in left uterine displacement. Patients were monitored

continuously with electrocardiogram (IX:(;) and au-

tomated spliygmoma~ionie~er cuff.

A blinded obstetric anesthesiologist collected data

prior to epidural placement (baseline), every 5 min-

utes for the first 30 minutes following drug admin-

istration, and then every 30 minutes until epidural

reinforcement was requested. At each data collection

interval, a 10 cm VAS (anchored with 0 representing

no pain and 10 representing the worst possible pain)

was used to assess analgesic onset, duration, and qual-

ity of pain relief. A similar VAS was used to assess

patient sat’ f t’ 1s ac ion with therapy, with 0 representing

least satisfaction and 10 reflecting most satisfaction.

Onset time represented the interval between drug

administration and the first VAS score lower than

baseline. Duration was defined as the interval from

onset of analgesia until the patient requested epidural

reinforcement.

‘l’he extent of dermatomal sensory block to pin-

prick and cold (alcohol swab) and the magnitude of

motor block were assessed at each data collection in-

terval, immediately following completion of VAS

scores. Motor blockade was tested by using a modi-

fication of Bromage’s scale, ‘:s in which 0 = the ability

to maintain a leg lift for prolonged periods; 1 = the

220 J. Clin. Anesth., vol. 3, May/June 1991

ability to lift legs briefly; 2 = the ability to bend knees;

3 = the ability to wiggle toes; and 4 = no movement

in lower extremities.

Maternal side effects, including nausea, pruritus,

and shivering, were scored as 0 = absent; 1 = mild;

2 = moderate; and 3 = severe. The blinded observer

also noted whether patients were started on intrave-

nous (IV) oxytocin following epidural placement, the

frequency of fetal bradycardia [heart rate (HR) less

than 110 beats per minute (bpm)], maternal hypoten-

sion [systolic blood pressure @BP) less than 100

mmHg], the need to administer ephedrine (as per

standard routine, a decrease in SBP below 100 mmHg

mandated administration of 5 mg of ephedrine), and

total ephedrine dose requirement.

Data were analyzed as follows: patient demograph-

ics and continuous variables, including VAS scores

and times, were analyzed by analysis of variance

(ANOVA) and compared with Tukey’s test of mul-

tiple comparison; differences in frequency of side ef-

fects were analyzed with chi-square tests. Data are

expressed as means 2 SD, median, or percentage. A

value of $I < 0.05 was considered to be significant.

Table 1. Patient IIemographics

Results

All patients included in the present evaluation com-

pleted the study protocol---&., there were no drop-

outs secondary to inadequate analgesia, maternal or

fetal morbidity, precipitous delivery, or emergency

cesarean section. There were no significant differ-

ences among the three treatment groups with respect

to patient demographics, cervical dilation at place-

ment of epidural analgesia, SBP, volume of lactated

Ringer’s solution infused prior to epidural placement,

or percentage of patients treated with oxytocin (Pi-

tocin) (Table 1).

Pain scores are presented in Tables 1 and 2. There

were no significant differknces among the three treat-

ment groups with respect to baseline VAS pain score

and lowest pain score achieved following administra-

tion of the study solutions. Patients in all the treatment

groups eventually attained very low pain scores (C 1);

however, individuals in Groups 1 and 2 required more

supplemental boluses and a greater amount of local

anesthetic than did those in Group 3 (Table3). Patients

in Group 1 experienced delayed analgesic onset when

Group 1 Group 2 Group 3

Age (yr)

(cm)

Wetght Height (kg)

Cervical dilation (cm)

Prehydration (ml)

Systolic blood pressure (mmHg)

Baseline pain VAS (cm)*

Percentage of patients receiving oxytocin

(4 to 6 kg/h)

iKotr: Data are means 2 SD.

*VAS = visual analog scale.

*Values are medians with range in parentheses.

27.7 5 6.7 28.7 k 5.0 27.7 k 7.0

160.2 2 5.8 164.8 ? 6.8 159.0 5 4.3

80.9 + 18.3 93.6 5 22.9 79.2 * 12.9

5.3 5 1.6 5.0 & 1.3 5.2 ? 1.2

1,423 * 252 1,515 + 288 1,390 5 277

123.7 t 12.2 125.3 2 11.7 121.7 + 9.9

9.0 (7.5 to 10.0) 8.4 (6.0 to 10.0) 8.7 (7.2 to 10.0)

38 27 35

Table 2. Pain and Satisfaction Visual Analog Scale (VAS) Scores

Group 1 Group 2 Group 3

Time to onset (min) 10.6 t 4.9t 8.2 2 3.9 5.1 +- 1.6

Time to lowest pain VAS score (min) 14.9 k 8.3 14.2 4 10.6 13.6 k 5.9

Lowest pain VAS score (cm)* 0.0 (0 to 0.3) 0.2 (0 to 0.6) 0.0 (0 to 0.3)

Time to highest satisfaction VAS score (min) 15.4 2 9.2 13.4 + 11.0 14.0 + 5.8

Highest satisfaction* VAS score (cm) 10.0 (9.8 to 10) 9.9 (9.2 to 10) 9.9 (9.6 to 10)

Duration of analgesia (min) 116.9 c 19.6 126.3 4 32.9 165.1 2 25.4$

Note: Data are means ? SD.

*Values are medians with range in parentheses.

tSignificant difference Group 1 us Group 3, p < 0.05.

$Significant difference Groups 1 and 2 US Group 3, p i 0.05.

J. Clin. Anesth., vol. 3, May/.June 1991 221

Table 3. Sensory Level and Adverse Effects

Group 1 Group 2 Group 3

Supplemental epidural bolus (ml) Total bupivacaine dose (mg) Total lidocaine dose (mg) Highest sensory block* Highest temperature block* Maximum motor blocki Number of patients experiencing hypotension

(SBP < 100 mmHg) Ephedrine dose/treated patient (mg) Shivering scale Number of patients reporting pruritus

I.8 -+ 1.3

44.6 * 10.1 -

9.4 !Z 1.2 9.2 + 1.3

1.53 5 0.0

6115

7.5 + 4.1 1.6 % 1.0

Oil5

I oi 15

9.0 i 3.9%

1.x 2 1.4

Oi l5

0.6 k I .‘L

26.” + 2.1

t-4.4 ?I- I.2

7.8 + 1.4 0.2 ” 0.4’i:

.il I .5

5.0 + 0.0

0.4 + 0.8 ‘L/l5

Note: Data are means t SD.

SBP = systolic blood pressure.

*Thoracic dermatomal level.

‘rModification of Bromage’s scale.

$Significant difference (;roups I and 2 US Group 3, /J -C 0.05.

PSignificant difference Group 2 11.‘ Group ?J, p < 0.05.

compared with individuals in Group 3. However, there

were no significant intergroup differences in time re-

quired to achieve lowest VAS score (Table 2).

There were no intergroup differences in satisfaction

or time necessary to achieve the highest satisfaction score

(TubG 2). Duration of analgesia was significantly longer

in Group 3 than in Groups 1 and 2.

Patients in Group 3 had a significantly lower degree

of motor block than those in Groups 1 and 2 (Tablr

3). There were no intergroup differences in sensory

block to pinprick or cold temperature (Table 3).

Following administration of epidural study solu-

tions, more episodes of maternal hypotension oc-

curred in Group 2 than in the other two groups (Tublp

3). Patients in Croup 2 required significantly more

IV ephedrine to maintain SBP above 100 mmHg, while

those in Group 3 required the least (Group 2 7)s Group

3). No episodes of fetal bradycardia or arrhythmia OI

need for emergency cesarean delivery occurred.

With regard to other adverse events, patients in Group

3 had a significantly lower frequency and severity of

shivering than individuals in Groups 1 and 2 (Table 3).

There were no episodes of increased maternal sedation

and no intergroup differences in sedation scores. Mild

abdominal pruritus, which did not require treatment,

was noted in two patients in Group 3.

Discussion

The goal of epidural analgesia for labor is to provide

rapid and reliable pain relief with minimal maternal

or fetal side effects. In this regard, the epidural so-

222 J. Clin. Anesth., vol. 3, May/June 1991

lutions used in the present investigation offered uni-

formly excellent analgesia and high patient satisfaction.

This single-bolus-dose study did not evaluate neonatal

outcome or mode of delivery, since analgesic tech-

niques following termination of the evaluation period

were varied and individualized to patient status. While

it is recognized that concentrations and total milli-

grams of local anesthetic administered are higher than

amounts reported to provide useful labor analge-

sia,‘,‘.‘-’ dosages represent standard of care at this in- c stitution in otherwise unmedicated primiparas.

Although there were no intergroup differences in time

until peak analgesia, defined as the lowest VAS pain

score, or in patient satisfaction with therapy, individ-

uals treated with bupivacaine and fentanyl experi-

enced rapid onset, the longest duration of pain relief,

the lowest degree of shivering, and the least motor

block. The mixture of bupivacaine and lidocaine of-

fered little clinical advantage in this setting.

A number of clinical studies have documented the

effectiveness and safety of epidurally administered

fentanyl.“,’ This highly lipid-soluble k-agonist rapidly

crosses dura, activates spinal opioid receptors in dor-

sal horn, and selectively suppresses c-fiber mediated

nociceptive input. I5 While not effective by itself,‘“,” it

appears to potentiate the activity of epidurally ad-

ministered local anesthetics and improves both the

quality and duration of labor analgesia. Onset of an-

algesia was particularly rapid in patients receiving fen-

tanyl. This finding may reflect rapid systemic uptake

and initial supraspinal effects in addition to spinal

cord-mediated activity .’ Supraspinal analgesia is of

short duration and provides negligible benefit after

20 minutes,” whereas more profound spinal modu-

lation of pain peaks at 10 to 15 minutes and is main-

tained for 2 or more hours.6,7J2 In this regard, patients

receiving bupivacaine with fentanyl experienced re-

ductions in pain at the first observation interval (5

minutes); however, analgesia was incomplete, and time

to lowest pain score was no more rapid than that ob-

served in the two nonopioid treatment groups.

The duration of pain relief in the bupivacaine with

fentanyl group was longer than that recently reported

by Lirzin et al. I8 Patients in that study, however, were

evaluated at a later stage of labor. It is also recognized

that duration of epidural fentanyl analgesia may be

increased when administered in larger volumes of so-

lution.‘” Increasing the volume of epidural injectate

is believed to provide greater dispersion in the epi-

dural space and to allow a larger number of fentanyl

molecules to reach opiate receptors concentrated in

the thoracic/lumbar spinal cord. The relatively large

volume of epidurally administered fentanyl solution

appeared to dilute the activity of bupivacaine and in-

fluenced its localization in that gross motor function

was virtually unaffected, while sensory blockade to

temperature tended to spread to higher dermatomal

levels. As in previous evaluations,‘,‘* this higher der-

matomal block was imperceptible to the patient, did

not influence sympathetic tone, and was not associ-

ated with a higher frequency of hypotension or need

for vasopressors. It is unclear whether dilution and

greater epidural dispersion would have influenced

the onset, duration, and effectiveness of analgesia

provided by bupivacaine and bupivacaine with lido-

Caine. It is conceivable that greater dispersion would

have reduced the magnitude of motor blockade.

Solutions of lidocaine are commonly used for ep-

idural analgesia in obstetrics because they provide

rapid onset of pain relief with minimal effects on uter-

ine tone”’ and neonatal outcome.21 In contrast to the

findings by MaGee et al, I1 the combination of bupi-

vacaine and lidocaine in the present study offered no

clinical advantage over bupivacaine alone in terms of

onset, duration of analgesia, or degree of motor block

and shivering. Duration of analgesia in this group was

significantly shorter than that noted with bupivacaine

and fentanyl. These findings reinforce the contention

that combinations of local anesthetics with different

pharmacodynamics offer no clinical advantage over

therapeutic doses of the individual agents used.*‘”

*Howell P, Davies W, Wrigley M, Tan P, Morgan B: Are two better

than one? An assessment of a mixture of bupivacaine and ligno-

Caine with adrenaline 1:200,000 for elective epidural caesarean

section. Proceedings of the Society qf Obstetric Anesthesia and Perinatology,

Madison, WI, May 1990;22:E-16.

Perhaps the only advantage to patients in this group

was a bupivacaine-sparing effect in which equivalent

quality of analgesia and overall patient satisfaction

were achieved with significant reductions in total bu-

pivacaine dosage. This finding mirrored a similar re-

duction observed in the bupivacaine with fentanyl

group. It remains unclear whether reductions in bu-

pivacaine dosage improve patient safety by lowering

risks of cardiotoxicity.lz

In the present study, patients receiving bupivacaine

and lidocaine tended to require a greater amount of

ephedrine to maintain SBP above 100 mmHg. How-

ever, transient episodes of hypotension and IV doses

of vasopressor did not influence fetal HR or varia-

bility. Few other maternal adverse events were noted

in any of the study groups. Patients receiving bupi-

vacaine with fentanyl had a lower frequency of shiv-

ering and a higher frequency of mild pruritus than

did individuals in the two other treatment groups.

Epidural opioids previously have been shown to re-

duce maternal shivering associated with lumbar epi-

dural analgesia.“R Whether this effect represents a

supraspinal mode of action or direct inhibition of spinal

reflexes remains unclear. Pruritus is a well-docu-

mented side effect associated with epidural opioids.

As in previous evaluations of epidural fentanyl,lZ,i’i

this effect was of mild severity and isolated to the

abdomen. The frequency of fentanyl-induced pruri-

tus may be directly related to the volume of solution

administered rather than the total dose.”

Finally, while patients treated with bupivacaine and

fentanyl noted prolonged analgesia, an absence of

shivering, and an intact ability to move and adjust

body position, such desirable characteristics did not

improve overall patient satisfaction. Thus, while pa-

tient satisfaction with therapy may reflect the com-

posite of useful analgesia with minimal adverse effects,

patients experiencing complete relief of labor pain

appear to overlook any associated inconveniences.

In conclusion, solutions of bupivacaine alone and

in combination with a second local anesthetic or opioid

provided useful epidural analgesia in laboring par-

turients. The combination of bupivacaine and fen-

tanyl offered the longest duration of analgesia and

the lowest frequency of shivering and motor block.

The combination of bupivacaine and lidocaine of-

fered little clinical advantage over bupivacaine alone.

References

1. Li DF, Rees GAD, Rosen M: Continuous extradural

infusion of 0.0625% or 0.125% bupivacaine for pain

.J. Clin. Anesth., vol. 3, May/-June 1991 223

relief in primagravid labor. BrJ Anaesth 1985;57:264-

70.

2. Thorburn J, Moir DD: Extradural analgesia: the influ-

ence of volume and concentration of bupivacaine on

the mode of delivery, analgesic efficacy, and motor block.

Br J Anaesth 1981;53:933-9.

3. Scanlon JW, Ostheimer GW, Lurie AO: Neurobehav-

ioral responses and drug concentrations in newborns

after epidural anesthesia with bupivacaine. Anesthesiol-

ogy 1976;45:400-5.

4. Marx GF: Cardiotoxicity of local anesthetics-the plot

thickens [Editorial]. Anesthesiology 1984;60:3-5.

5. Brown DJ, Morison DH, Covino BG: Comparison of

carbonated bupivacaine and bupivacaine hydrochlo-

ride for extradural anesthesia. Br JAnaesth 1980;52:419-

21.

6. Justins DM, Francis D, Houlton PG, Reynolds F: A

controlled trial of extradural fentanyl in labor. Br J

Anaesth 1982;54:405-14.

7. Vella LM, Willats DG, Knott C, Lintin DJ, Reynolds F:

Epidural fentanyl in labor. Anaesthesia 1985;40:741-7.

8. Naulty JS, Malinow A, Hunt CO, et al: Epidural

butorphanol-bupivacaine for analgesia during labor and

delivery [Abstract]. Anesthesiology 1986;65:A369.

9. Abboud TK, Sheikh-01-Eslam A, Yanagi T, Costandi

J, Zakarian M: Safety and efficacy of epinephrine added

to bupivacaine for lumbar epidural analgesia [Abstract].

Anesthesiology 1987;61:A406.

10. Cohen SE, Thurlow A: Comparison of chloroprocaine-

bupivacaine mixture with chloroprocaine and bupiva-

Caine used individually for obstetric analgesia. Anesthe-

siology 1979;51:288-91.

11. MaGee DA, Sweet PT, Holland AJ: Epidural anaes-

thesia with mixtures of bupivacaine and lidocaine. Can

Anaesth Sot J 1983;30: 174-8.

12. Cohen SE, Tan S, Albright GA, Halpern J: Epidural

fentanylibupivacaine mixtures for obstetric analgesia.

Anesthesiology 1987;67:403-7.

13. Bromage PR: Mechanism of action. In: Bromage PR,

ed. Epidurul Analgesia. Philadelphia: WB Saunders,

1978: 144.

14. Chestnut DH, Vandewalker GW, Owen CL, Bates JN,

Choi WW: The influence of continuous epidural bu-

pivacaine analgesia on the second stage of labor and

method of delivery in nulliparous women. Anesthrsiolo<q

1987;66:774-80.

15. Cousins MJ, Mather LE: Intrathecal and epidural

administration of opioids. Anesthe.riology 1984;63 :276-

310.

16. Carrie LES, O’Sullivan GM, Seegobin R: Epidural fen-

tanyl in labor. Anuesthesia 1981;36:965-9.

17. Justins DM, Knott C, Luthman J, Reynolds F: Epidural

versus intramuscular fentanyl: analgesia and pharma-

cokinetics in labor. Anaesthesia 1983;38:937-42.

18. Lirzin JD, Jacquinot P, Jorrot J, Dailland P, ConseilleI

C: Effect of diluting fentanyl on epidural bupivacaine

during labor analgesia. Reg Anesth 1989; 14:279-81.

19. Birnbach DJ, Johnson MD, Arcario T, Datta S, Naulq

JS, Ostheimer GW: Effect of diluent volume on anal-

gesia produced by epidural fentanyl. Anusth Analg

1989;68:808-9.

20. Craft JB, Epstein BS, Coakley CS: Effect of’ lidocaine

with epinephrine versus lidocaine (plain) on induced

labor. Anesth Analg 1972;51:243-6.

2 1. Abboud I‘K, David S, Shakuntala N, Costandi J, Yanagi

‘I‘, Sze-Ua Y: Maternal, fetal and neonatal effects of

lidocaine with and without epinephrine for epidural

anesthesia in obstetrics. Anesth Analg 1984;63:973-9.

22. de Jong RH, Bonin JD: Mixtures of local anesthetics

are no more toxic than the parent drugs. Anesthesiology

1981;54: 177-9.

23. Sevarino FB, Johnson MD, Lema MJ, Datta S, Ostheimer

GW, Naulty JS: The effect of epidural sufentanil on

shivering and body temperature in the parturient. A,tvsth

.Analg 1989:68:530-3.

224 J. Clin. Anesth., vol. 3, May/June 1991