The purpose of this study was to determine if epidural fentanyl produces analgesia in laboring patients by a primary spinal or supraspinal action.


Fifty-four parturients were randomized to receive epidural 0.125% bupivacaine plus one of three treatments: epidural saline-intravenous saline, epidural fentanyl (20 microg/h)-intravenous saline, or epidural saline-intravenous fentanyl (20 microg/h). The study treatments were administered by continuous infusion, whereas epidural bupivacaine use was patient controlled.


Epidural bupivacaine use was significantly reduced by epidural (11.5+/-4.6 ml/h) but not by intravenous fentanyl (15.9+/-4.5 ml/h) compared with saline control (16+/-5.9 ml/ h). Analgesia characteristics and side effects were similar among groups.


Low-dose epidural infusions of fentanyl produce labor analgesia by a primary spinal action.

LABOR analgesia is often provided by the epidural administration of local anesthetic and opioid solutions. An opioid is combined with a local anesthetic in an attempt to reduce local anesthetic requirements and side effects, including systemic toxicity, hypotension, and motor blockade. Less leg weakness is desirable because motor block potentially prolongs second-stage labor [1]and may increase the incidence of instrument-assisted delivery [1]and patient dissatisfaction. [2]Epidural infusions that combine fentanyl and bupivacaine provide analgesia as good or better than bupivacaine alone while reducing bupivacaine requirements and motor block. [3–8] 

Opioids provide analgesia by binding to opioid receptors located spinally and supraspinally. When epidural fentanyl is administered, it either crosses the dura and binds to spinal opioid receptors, is absorbed systemically and binds to supraspinal opioid receptors, or a combination of both. Whether epidural fentanyl in patients in labor produces analgesia predominantly by a primary spinal or supraspinal action is controversial. Volunteer studies support a spinal effect from bolus epidural fentanyl administration [9,10]; similarly, clinical studies reveal more profound [11]and longer lasting [12]analgesia from bolus epidural fentanyl than analgesia from bolus intravenous fentanyl, which once again supports a spinal effect. Conversely, prospective clinical studies comparing infusions of epidural or intravenous fentanyl administered to treat postoperative pain demonstrate no differences in serum fentanyl concentrations, side effects, or analgesia, [13–21]suggesting that epidural fentanyl under these conditions produces analgesia by a systemic absorption and a supraspinal affect. Despite the prevalent use of epidural infusions containing fentanyl in laboring patients, no clinical study has compared epidural with intravenous fentanyl infusions during labor in an attempt to distinguish between spinal and supraspinal actions. The purpose of this study was to compare patient-controlled epidural bupivacaine requirements during labor with the coadministration of a continuous infusion of either epidural or intravenous fentanyl in a prospective, randomized, double-blinded trial.

After institutional review board approval and informed written consent were obtained, 54 parturients classified as American Society of Anesthesiologists physical status 1 or 2 who were in early labor (<5 cm cervical dilation) with no contraindications to epidural analgesia or study drugs were randomized by double-blinded design to receive epidural bupivacaine plus one of three treatments: epidural saline - intravenous saline, epidural fentanyl - intravenous saline, or epidural saline - intravenous fentanyl. Randomization was performed with a computer-generated Table torandom numbers.

The epidural space was identified via a loss-of-resistance technique with either saline or air using a 17-gauge, 3.5-inch Tuohy-Schliff epidural needle, and an 18-gauge, open-tip epidural catheter (B. Braun Medical, Bethlehem, PA) was inserted 4 - 6 cm into the epidural space. After insertion and negative aspiration of blood or cerebral spinal fluid, epidural catheters were tested with sequential injections of 2 ml 2% lidocaine, followed by 5 ml and 3 ml to a total of 200 mg per patient. Only patients with adequate analgesia 15 min after lidocaine administration were studied.

Epidural 0.125% bupivacaine was administered by patient-controlled epidural analgesia (PCEA) using a Bard Syringe PCA Pump (CR Bard, North Reading, MA). Settings for this were as follows: 6-ml/h basal infusion, 4-ml bolus, 10-min lock out, with a 22-ml/h limit. In addition to the PCEA, each patient was given both a continuous intravenous and epidural infusion, each at 12 ml/h (Medfusion Syringe Pump, Duluth, GA) of either saline or 1.67 [micro sign]g/ml fentanyl (20 [micro sign]g/h) as determined by the randomization. The continuous epidural infusion was connected to the epidural catheter hub with a microbore antireflux y-set (Baxter Health Care, Deerfield, IL) to minimize dead space. Similarly, intravenous infusions were connected as near to the patient as possible.

Automated maternal blood pressures and heart rate, tocodynomometry, and fetal heart rate were monitored throughout labor. Hypotension, defined as systolic blood pressure <100 mmHg, was treated with left uterine displacement, intravenous fluid administration, or ephedrine, as indicated. Epidural bupivacaine use, sensory levels to pinprick, motor block (0 = lifts legs, 1 = bends knee only, 2 = moves foot only, 3 = no leg movement), visual analog scores for pain (10 cm), and the presence or absence of pruritus or nausea were recorded hourly until complete cervical dilatation occurred or a decision to delivery by cesarean section was made. Patients experiencing breakthrough pain received 5-ml boluses of epidural 0.125% bupivacaine every 5 min until they were comfortable. Patients delivering within 1 h of epidural insertion were excluded from data analysis.

All statistical analyses were performed using Sigmastat (version 2.0, SPSS Inc., Chicago, IL). Unless otherwise indicated, data are presented as means +/- SD. Bupivacaine requirements were analyzed using one-way analysis of variance. Side effects were analyzed using chi-squared and Fisher's exact tests as appropriate. Visual analog scores were analyzed using one-way analysis of variance on the change from baseline. Adjustments for multiple comparisons were made when appropriate. Probability values <0.05 were considered significant. Group size was determined by power analysis ([Greek small letter beta]= 0.8) to detect a 25% reduction in epidural bupivacaine use from either epidural or intravenous fentanyl compared with the saline-only treatment group.

Fifty-four patients were randomized and 47 completed the study. Seven patients were excluded: Three delivered within 1 h of epidural catheter placement and four experienced inadequate analgesia after epidural catheter placement and lidocaine administration.

Demographic variables, parity, duration of epidural analgesia, sensory levels to pinprick, and visual analog scores for pain were similar among groups (Table 1). Similarly, side effects including hypotension, pruritus, nausea, and motor block were similar among the groups (Table 2). Overall epidural bupivacaine use was significantly reduced by epidural but not by intravenous fentanyl compared with controls (Table 3). In addition, hourly epidural bupivacaine requirements were significantly reduced by epidural fentanyl (Figure 1).

Table 1. Patient Demographics and Analgesia Characteristics 

Table 1. Patient Demographics and Analgesia Characteristics 
Table 1. Patient Demographics and Analgesia Characteristics 

Table 2. Side Effects (%) 

Table 2. Side Effects (%) 
Table 2. Side Effects (%) 

Table 3. PCEA Bupivacaine 0.125% Use 

Table 3. PCEA Bupivacaine 0.125% Use 
Table 3. PCEA Bupivacaine 0.125% Use 

Figure 1. Patient-controlled epidural analgesia (PCEA) with 0.125% bupivacaine use expressed in milliliters over time for patients given continuous infusions at 12 ml/h of epidural fentanyl (20 [micro sign]g/h), intravenous fentanyl (20 [micro sign]g/h), or saline placebo. The asterisk (*) indicates an overall significant difference between patients receiving epidural fentanyl and either intravenous fentanyl or saline placebo (P < 0.05).

Figure 1. Patient-controlled epidural analgesia (PCEA) with 0.125% bupivacaine use expressed in milliliters over time for patients given continuous infusions at 12 ml/h of epidural fentanyl (20 [micro sign]g/h), intravenous fentanyl (20 [micro sign]g/h), or saline placebo. The asterisk (*) indicates an overall significant difference between patients receiving epidural fentanyl and either intravenous fentanyl or saline placebo (P < 0.05).

Close modal

Epidural fentanyl significantly reduced the number of PCEA patient demands and PCEA doses delivered (P = 0.035). Patients administered epidural fentanyl made 8 +/- 6 (mean +/- SD) PCEA demands, and 5 +/- 4 doses were delivered throughout labor compared with 14 +/- 7 requested, 9 +/- 4 delivered doses with intravenous fentanyl, and 13 +/- 8 requested, 8 +/- 5 delivered doses placebo. There were no differences in PCEA characteristics between the intravenous fentanyl and placebo groups.

The principal finding of this study is that fentanyl infused epidurally significantly reduces epidural bupivacaine use during labor, whereas fentanyl infused intravenously does not. This reduction in bupivacaine use most likely represents the spinal effect of epidural fentanyl. If epidural fentanyl produces labor analgesia by systemic absorption and a supraspinal effect, similar reductions in bupivacaine use would be expected from fentanyl administered intravenously. In contrast, bupivacaine use was similar between intravenous fentanyl and the saline control.

Our findings are in contrast to those of studies comparing postoperative analgesia from intravenous and epidural infusions of fentanyl [13–21]and may represent differences in study populations and study design. These postoperative pain studies used infusions of plain epidural or intravenous fentanyl, whereas our patients received a combination of fentanyl and bupivacaine. Because opioids and local anesthetics interact synergistically, [22–23]lower doses of fentanyl are required to produce analgesia when combined with bupivacaine than when administered alone. Typically, patients require 50 - 150 [micro sign]g/h fentanyl to relieve postoperative pain, whereas labor analgesia is enhanced by 15 - 25 [micro sign]g/h fentanyl when combined with epidural bupivacaine. [3–8]The higher doses of fentanyl infused to treat postoperative pain may mask the spinal effects of epidural fentanyl. In addition, pain during the early stages of labor is distinct from pain during the later stages of labor and postoperative pain, because the former is transmitted primarily by visceral afferents and the later types of pain are transmitted primarily by somatic afferents. [24–26]Opioids, including fentanyl, are much more effective in the treatment of early labor pain. [27–29] 

In many studies, epidural fentanyl is administered through a lumbar epidural catheter, even though analgesia is sought in a thoracic dermatome. [13,16,17,20]Evidence suggests that spinal analgesia from bolus fentanyl administered epidurally is segmental in nature. Analgesia restricted to the dermatomes of the lower extremity was observed in volunteers when bolus fentanyl was administered through lumbar epidural catheters. [9,10]Furthermore, pharmacokinetic studies in patients with chronic pain [30]given bolus lumbar epidural fentanyl demonstrate cerebrospinal fluid concentrations of fentanyl 10 times greater in the lumbar cerebrospinal fluid than in cervical cerebrospinal fluid. Although the extent of segmental analgesia from low-dose lumbar epidural fentanyl infusions is unknown, it is likely to include the T10 - L1 segments, [24–26]which lie near the site of lumbar epidural catheter placement at which afferents relaying labor pain enter the spinal cord.

Although other studies use patient-controlled analgesia, [14,16,18]our study differed in that while patients controlled the epidural administration of bupivacaine, fentanyl infusions were constant. This study design allows for a better delineation of the analgesic properties of a fixed dose of fentanyl, especially when comparing one route of administration with another. Had fentanyl been combined with epidural bupivacaine, epidural fentanyl doses would have varied with PCEA use, whereas intravenous fentanyl doses would remain constant and unaffected by PCEA use, making comparisons between groups difficult. Furthermore, higher doses of epidural fentanyl potentially favor systemic reabsorption and supraspinal analgesia, as observed with the treatment of postoperative pain.

In this study, epidural fentanyl reduced epidural bupivacaine requirements by 28% compared with either intravenous fentanyl or a saline placebo. This finding corresponds closely to a recent report estimating a 31% reduction in epidural bupivacaine use with the coadministration of epidural 2 [micro sign]g/ml fentanyl. [8] 

Several limitations of this study must be noted. Only one dose of epidural fentanyl was administered to labor patients receiving PCEA bupivacaine. Thus our results may not be applicable to varying doses of epidural fentanyl, fentanyl administered alone, or to different patient populations. In addition, fentanyl was administered for an average of 4 h in our study. It is possible that the spinal effects of fentanyl diminish over time and that, as more fentanyl is administered, supraspinal analgesia may predominate. And, finally, although epidural bupivacaine use was reduced with epidural fentanyl, no significant reductions in side effects, including motor block, were associated with the decreased bupivacaine use. This probably reflects study design because we had insufficient power to test for differences in side effects. The initial power analysis was designed only to test for predicted reductions in bupivacaine use. Nevertheless, patients given epidural fentanyl did appear to have less motor block than did patients in the other two groups (Table 2), although the difference was not significant. A larger study would be required to test for differences in motor block and side effects from the administration of epidural fentanyl.

In conclusion, low-dose epidural fentanyl infusions reduce epidural bupivacaine use and enhance labor analgesia primarily by a spinal, rather than a supraspinal, effect.

Chestnut DH, Vandewalker GE, Owen CL: The influence of continuous epidural bupivacaine analgesia on the second stage of labor and method of delivery in nulliparous women. Anesthesiology 1987; 66:774-80
Russell R: Assessment of motor blockade during epidural analgesia in labour. Int J Obstet Anesth 1992; 1:230-4
Elliot RD: Continuous infusion epidural analgesia for obstetrics: Bupivacaine versus bupivacaine-fentanyl mixture. Can J Anaesth 1991; 38:303-10
Cohen S, Tan S, Albright G, Halpren J: Epidural fentanyl-bupivacaine mixtures for obstetric analgesia. Anesthesiology 1987; 67:403-7
Vella L, Willatts D, Knott C, Lintin DJ, Justins DM, Reynolds F: Epidural fentanyl in labour: An evaluation of the systemic contribution to analgesia. Anaesthesia 1985; 40:741-7
Cellano D, Capogna G: Epidural fentanyl plus bupivacaine 0.125 percent for labour: Analgesic effects. Can J Anaesth 1988; 35:375-8
Chestnut DH, Owen CL, Bates JN, Ostman LG, Choi WW, Geiger MW: Continuous infusion epidural analgesia during labor: A randomized, double-blind comparison of 0.0625% bupivacaine/0.0002% fentanyl versus 0.125% bupivacaine. Anesthesiology 1988; 68:754-9
Lyons G, Columb M, Hawthorne L, Dresner M: Extradural pain relief in labour: Bupivacaine sparing by extradural fentanyl is dose dependent. Br J Anaesth 1997; 78:493-7
Liu SS, Gerancher JC, Bainton BG, Kopacz DJ, Carpenter RL: The effect of electrical stimulation at different frequencies an perception and pain in human volunteers: Epidural versus intravenous administration of fentanyl. Anesth Analg 1996; 82:98-102
Coda BA, Brown MC, Schaffer R, Donaldson G, Jacobson R, Hautman B, Shen DD: Pharmacology of epidural fentanyl, alfentanil, and sufentanil in volunteers. Anesthesiology 1994; 81:1149-61
Justins DM, Francis D, Houlton PG, Reynolds F: A controlled trial of extradural fentanyl in labour. Br J Anaesth 1982; 54:409-13
Reynolds F, O'Sullivan G: Epidural fentanyl and perineal pain in labour. Anaesthesia 1989; 44:341-4
Ellis DJ, Millar WL, Reisner LS: A randomized double-blind comparison of epidural versus intravenous fentanyl infusion for analgesia after cesarean section. Anesthesiology 1990; 72:981-6
Welchew EA, Breen DP: Patient-controlled on-demand epidural fentanyl. Anaesthesia 1991; 46:438-41
Salomaki TE, Laitinen JO, Nuutinen LS: A randomized double-blind comparison of epidural versus intravenous fentanyl infusion for analgesia after thoracotomy. Anesthesiology 1991; 75:790-5
Grant RP, Dolman JF, Harper JA, White SA, Parsons DG, Evans KG, Merrick PC: Patient-controlled epidural fentanyl compared with patient-controlled intravenous fentanyl for post-thoracotomy pain. Can J Anesth 1992; 39:214-9
Sandler AN, Stringer D, Panos L, Badner N, Friedlander M, Koren G, Katz J, Klein J: A randomized, double-blind comparison of lumbar epidural and intravenous fentanyl infusions for post thoracotomy pain relief. Anesthesiology 1992; 77:626-4
Glass PS, Estok P, Ginsberg B, Goldberg JS, Sladen RN: Use of patient-controlled analgesia to compare the efficacy of epidural to intravenous fentanyl administration. Anesth Analg 1992; 74:345-51
Guinard JP, Mavrocordatos P, Chiolero R, Carpenter RL: A randomized comparison of intravenous versus lumbar and thoracic epidural fentanyl for analgesia after thoracotomy. Anesthesiology 1992; 77:1108-15
Baxter AD, Laganiere S, Samson B, Stewart J, Hull K, Goernert L: A comparison of lumbar epidural and intravenous fentanyl infusions for post-thoracotomy analgesia. Can J Anaesth 1994; 41:184-91
Harukuni I, Yamaguchi H, Sato S, Naito H: The comparison of epidural fentanyl, epidural lidocaine, and intravenous fentanyl in patients undergoing gastrectomy. Anesth Analg 1995; 81:1169-74
Maves TJ, Gebhart GF: Antinociceptive synergy between intrathecal morphine and lidocaine during visceral and somatic nociception in the rat. Anesthesiology 1992; 76:91-9
Penning JP, Yakish TL: Interaction of intrathecal morphine with bupivacaine and lidocaine in the rat. Anesthesiology 1992; 77:1186-200
Chestnut D: Obstetrical Anesthesia, Principle and Practice, First Edition. St. Louis, Mosby, 1994, pp 316-7
Bonica JJ: The nature of pain of parturition. Clin Obstet Gynaecol 1975; 2:511
Gould SF: Obstetrics, Normal and Problem Pregnancies. Edited by SG Gabbe, JR Niebyl, JL Simpson. New York, Churchill Livingstone, 1986, pp 3-38
Scott PV, Bowen FE, Cartwright P, Rao BC, Deeley D: Intrathecal morphine as sole analgesic during labour. BMJ 1980; 281:351-3
Baraka A, Noueihid R, Hajj S: Intrathecal injection of morphine for obstetric analgesia. Anesthesiology 1981; 54:136-40
Honet J, Arkoosh V, Norris M, Huffnagle HJ, Silverman NS, Leighton BL: Comparison among intrathecal fentanyl, meperidine, and sufentanil for labor analgesia. Anesth Analg 1992; 75:734-9
Gourley GK, Murphy TM, Plummer JI, Kowalski SR, Cherry DA: Pharmacokinetics of fentanyl in lumbar and cervical CSF following lumbar epidural and intravenous administration. Pain 1989; 38:252-9