Society for Obstetric Anesthesia and Perinatology

Section Editor: Cynthia A. Wong

A Randomized Controlled Trial of the Efficacy and

Respiratory Effects of Patient-Controlled Intravenous
Remifentanil Analgesia and Patient-Controlled Epidural
Analgesia in Laboring Women
Daniel Stocki, MD,*† Idit Matot, MD,† Sharon Einav, MD,‡ Smadar Eventov-Friedman, MD,§
Yehuda Ginosar, MBBS,* and Carolyn F. Weiniger, MB ChB*║

BACKGROUND: Safe and effective alternatives are required in labor when epidural analgesia is
not appropriate. We hypothesized that patient-controlled IV remifentanil labor analgesia would
not be inferior to patient-controlled epidural labor analgesia.
METHODS: This randomized nonblinded controlled noninferiority study in healthy women with
a singleton fetus and vertex presentation was performed at 1 site. Women were randomized
to receive patient-controlled IV analgesia titrated from 20 mcg up to a maximum bolus dose of
60 mcg with a lockout interval of 1 to 2 minutes, or patient-controlled epidural analgesia 0.1%
bupivacaine with 2 mcg/mL fentanyl (initiation bolus 15 mL; maintenance bolus 10 mL, lockout
interval 20 minutes, basal infusion 5 mL/h). Crossover was permitted after 30 minutes. The pri-
mary study outcome was efficacy (assessed as hourly numerical rating scale [NRS] pain score
[11-point NRS] and maternal satisfaction [11-point NRS]); the secondary outcome was safety
(maternal apnea). Supplementary oxygen was administered continuously during the respiratory
monitoring period. During the first hour of analgesia, the heart rate, respiratory rate, pulse oxim-
etry (Spo2), and end-tidal CO2, as an indication of apnea, were compared. Apnea lasting >40
seconds was managed by light stimulation by the attending anesthesiologist.
RESULTS: Forty women were recruited to the following groups: remifentanil n = 19 (1 exclusion),
epidural n = 20. Four crossed over: 3 from the remifentanil to epidural group and 1 from the
epidural to remifentanil group. Mean (± SD) baseline NRS pain scores were similar, 8.4 ± 1.5
for remifentanil and 8.7 ± 1.2 for epidural analgesia, P = 0.52. Baseline adjusted mean NRS
reduction at 30 minutes for remifentanil was −4.5 (± 0.6) vs −7.1(± 0.6) for epidural analgesia,
P < 0.0001 for both. Pain score at 30 minutes was 3.7 ± 2.8 for remifentanil and 1.5 ± 2.2 for
epidural analgesia, P = 0.009. Remifentanil was inferior to epidural analgesia with respect to the
NRS at all time points, because the observed difference in NRS was greater than the expected
−1.5 units. Maternal satisfaction was 8.6 ± 1.4 for the remifentanil group and 9.1 ± 1.5
for epidural group, P = 0.26. Mean respiratory rate was lower in the remifentanil group, 18 ± 4
vs 21 ± 4 breaths/min in the epidural group, P = 0.03. Mean Spo2 was lower in the remifentanil
group 96.8% ± 1.4 vs 98.4 ± 1.2 for epidural group, P < 0.0001. There were 9 apnea events;
all occurred in 5 women receiving remifentanil (5/19 [26.3%], P = 0.046). Apgar scores and
neonatal respiratory outcomes were similar.
CONCLUSION: IV remifentanil is inferior to epidural analgesia for provision of labor analgesia;
however, remifentanil does provide a satisfactory level of labor analgesia. Laboring women
receiving remifentanil require suitable monitoring to detect and alert for apnea.  (Anesth &
Analg 2014;118:589–97)

From the *Department of Anesthesiology and Critical Care Medicine, Hadas-
sah Hebrew University Medical Center, Jerusalem; †Department of Anesthe-
siology and Intensive Care, Tel Aviv Medical Center, Tel Aviv; ‡Intensive Care
Unit, Shaare Zedek Medical Center; §Department of Neonatology, Hadassah
Hebrew University Medical Center, Jerusalem, Israel; and ║Department of
Anesthesia, Stanford School of Medicine, Stanford, California.
Accepted for publication July 17, 2013.
Funding: This study was supported by a research grant for Anesthesiologists
from the Hadassah Hebrew University Medical Center, Jerusalem, Israel.
Oridion® provided the capnography equipment, developed the dedicated
software, and provided the mathematician who performed data extraction.
Neither the funding body nor Oridion® had a role in study design, data inter-
pretation, writing of the manuscript, or manuscript submission for publication.
Conflict of Interest: See Disclosures at the end of the article.
Address correspondence and reprint requests to Carolyn F. Weiniger, MB
ChB, Department of Anesthesiology and Critical Care Medicine, Hadassah
Hebrew University Medical Center, Jerusalem 91120, Israel. Address e-mail
to carolynfweiniger@gmail.com.
Copyright © 2013 International Anesthesia Research Society
DOI: 10.1213/ANE.0b013e3182a7cd1b
E
pidural analgesia provides the most effective pain
relief for labor. However, epidural analgesia has risks
of failure, injury, or infection, may be contraindicated,
and requires skilled anesthesia providers. Alternative effec-
tive analgesia can be achieved using opioids.1–6
Remifentanil is a promising alternative to epidural anal-
gesia as demonstrated by a low conversion rate to neuraxial
analgesia and high maternal satisfaction.1–4 However, studies
of remifentanil use in labor describe adverse maternal respi-
ratory side effects, specifically hypoxemia detected by pulse
oximetry, despite the ultrashort duration of action of remifent-
anil.6,7 Volmanen et al.7 reported the need for frequent supple-
mental oxygen due to oxygen desaturation in laboring women
receiving remifentanil compared with epidural analgesia.
Our primary study objectives were to compare the anal-
gesia efficacy and maternal satisfaction of remifentanil labor

March 2014 • Volume 118 • Number 3 www.anesthesia-analgesia.org 589

Effects of IV Remifentanil Vs Epidural for Labor
analgesia with standard treatment (epidural analgesia); the
secondary outcome was the incidence of apnea.
METHODS
This single center, randomized nonblinded controlled non-
inferiority design trial was performed in a Jerusalem tertiary
hospital labor and delivery suite with 5600 deliveries per
year, an epidural analgesia rate of 59% and cesarean deliv-
ery rate of 21%. After receiving IRB and Ministry of Health
approval (057708HMO), eligible women were approached
for recruitment between February 2010 and August 2010 in
the labor ward when the investigator was available. Written
informed consent was obtained before enrollment. The trial
was registered at www.clinicaltrials.gov (NCT00801047).
Eligible women included healthy, ASA physical status
class I or II, age 18 to 40 years, body weight <110 kg, gesta-
tional age >36 completed weeks, with singleton pregnancy
and vertex presentation. Women with the following criteria
were excluded: contraindication to epidural analgesia, opi-
oid administration in the previous 2 hours, previous uterine
surgery, preeclampsia, inability to understand the consent
form, nasal obstruction for any reason and medical indica-
tion for epidural analgesia (e.g., cardiac disease, suspected
difficult airway), or nonreassuring fetal heart rate (FHR)
tracing. Induction of labor (oxytocin or prostaglandin E2
suppository) and augmentation of established labor (oxyto-
cin) before enrolment were not exclusion criteria.
Women were approached on arrival in the delivery suite
or as soon as possible thereafter and gave verbal consent;
however, they were not enrolled until they were in the
active phase of labor, as determined by the attending obste-
trician. Women were informed before study enrolment that
conversion to the other treatment would be possible at any
time during labor beginning 30 minutes after analgesia ini-
tiation with the study technique. Written informed consent
was signed at the time of request for analgesia. Randomized
treatment allocation was revealed when the woman was in
active labor with cervical dilatation ≥2 cm and requesting
analgesia, as determined by the obstetrician in attendance.
Randomization and group allocation were determined:
cards were divided into groups of 8 cards. Each group
contained 4 allocation cards for remifentanil and 4 alloca-
tion cards for epidural analgesia (ratio 1:1), and 8 opaque
envelopes numbered in groups from 1–8, 9–16, etc. were
assigned to each group of cards. The cards were placed
face down, manually shuffled, randomly selected, and then
inserted into the numbered, opaque envelopes by a per-
son not involved in the study. These envelopes were then
sealed. Treatment assignment was revealed by breaking the
seal of an envelope in consecutive order from number 1.
Laboring women randomized to the remifentanil group
received remifentanil patient-controlled IV analgesia
(PCIA); the bolus dose was titrated to effect from 20 mcg up
to a maximum of 60 mcg as required; the lockout interval
was initially set at 2 minutes, without a background infu-
sion. The PCIA bolus/lockout interval was titrated to an end
point of either patient comfort, or a maximal bolus dose of
60 mcg/minimal lockout interval of 1 minute by the recruit-
ing anesthesiologist at any time during labor. The PCIA
pump tubing was “piggybacked” into the distal most port
590   
www.anesthesia-analgesia.org
of the mainline IV fluid tubing. The mainline tubing con-
tained an antireflux valve designed to prevent remifentanil
inadvertently backing up in the IV line during administra-
tion. The recruiting anesthesiologist (a resident performing
a mandatory research project) remained by the woman’s
bedside until the end of treatment with remifentanil.
For women randomized to receive epidural analge-
sia, a 17-gauge Tuohy needle was inserted by the midline
approach using loss of resistance to air at intervertebral
space L3-4 or L2-3. An incremental initial loading dose
of 15 mL of 0.1% bupivacaine with 50 mcg fentanyl was
administered followed by patient-controlled epidural
analgesia infusion of 0.1% bupivacaine with 2 mcg/mL
fentanyl: basal infusion of 5 mL/h, patient-controlled
bolus 10 mL, and lockout interval 20 minutes. Additional
epidural bolus doses (either 0.1% bupivacaine 10 mL dur-
ing the first stage of labor or 1% lidocaine 8 mL during the
second stage of labor) were administered by the anesthe-
siologist to treat breakthrough pain. If epidural analgesia
failed, the epidural catheter was reinserted. After epidural
analgesia administration, the recruiting anesthesiologist
remained by the woman’s bedside for the first hour and
then remained in the labor ward, dedicated to her care,
until delivery.
Maternal safety aspects: Respiratory monitoring was
performed using a Capnostream® capnograph (Oridion®,
Jerusalem, Israel) with an oral-nasal cannula, sampling
from both the nose and mouth (Oridion®). Laboring
women receiving remifentanil were monitored through-
out labor, whereas women receiving epidural analgesia
were monitored for a minimum of 1 hour after induction
of analgesia. This difference was due to attrition with the
respiratory monitor throughout labor in the epidural group
as women wished to remove the respiratory monitor and
nasal prongs. However, women in the remifentanil group
were encouraged to remain monitored for safety reasons.
All women received continuous supplementary oxygen
(2 liters/min) through an oral-nasal cannula throughout the
respiratory monitoring period.
The respiratory monitor recorded continuous waveform
end- tidal carbon dioxide (etco2), respiratory rate, pulse
oximetry (Spo2) and heart rate. Carbon dioxide (CO2) was
monitored by continuously drawing gas through the sam-
pling line to an infrared source and detector within a remote
monitor. This device avoids significant entrainment of room
air by sampling 50 mL/min; after measurement, the sample
is defused into the environment. The small sampling vol-
ume reduces errors associated with traditional sidestream
capnography. Dilution of sampled gas and blockage of the
sampling line by water are potential sources of inaccurate
measurements when large volumes are sampled.8
All alarms were audible, but the beat-to-beat sound on
the respiratory monitor was turned off so as not to alert
the laboring women to changes. The settings of the Spo2
alarm and respiratory rate alarm acted, for the purpose
of the study, as an alert for potential impending apnea.
Spo2 <94%, respiratory rate below 8 breaths per minute,
etco2 ± 15 mm Hg from baseline or apnea longer than
20 seconds, activated an audible alarm. Throughout the
respiratory monitoring period, the recruiting anesthesi-
ologist was present in the labor room, maintaining visual
anesthesia & analgesia
contact with both the woman’s face and respiratory moni-
tor. Apnea triggered a staged intervention starting with
a verbal reminder to breathe after 40 seconds, unless
Spo2 decreased below 90%, in which case the response
was initiated earlier. If there was no response, a light tap
was given on the woman’s arm or shoulder in combina-
tion with a verbal command to take a deep breath. The
anesthesia provider in the room manually documented
potential causes of artifacts (e.g., uterine contractions, dis-
connections, removal of nasal prongs, eating, or drinking).
In addition, the presence of the anesthesiologist fulfilled a
United States Food and Drug Administration recommen-
dation for a trained person to be present for monitored
anesthesia care.
FHR was monitored to ensure neonatal safety using
cardiotocography.9 The local IRB required that remifent-
anil therapy be withdrawn in favor of alternative analgesia
(epidural, nitrous oxide) if the FHR tracing became nonreas-
suring (decreased variability, bradycardia, or late decelera-
tions). The IRB also required the presence of a neonatologist
at each birth after remifentanil treatment, in case of neonatal
cardiorespiratory depression.
Data Acquisition and Retrieval
Each woman was asked about antenatal class atten-
dance and her plans for labor analgesia. Demographic
and medical data were obtained from personal inter-
views before analgesia initiation and throughout labor.
Physiological data were recorded from ongoing assess-
ment of respiratory and fetal monitors. Maternal pain
was assessed using an 11-point verbal numerical rating
scale (NRS) of 0 to 10, where 0 is no pain and 10 is the
worst pain imaginable. NRS pain scores were recorded
before analgesia, 30 minutes after analgesia was admin-
istered, then hourly up to a total of 7 hours, and within 1
hour after delivery.
The respiratory monitor data were downloaded onto a
USB memory stick. Lowest Spo2 was noted during all apnea
events. The number of apnea events was counted for each
woman (respiratory rate of zero for at least 20 seconds).
Apnea events for the remifentanil group were recorded
throughout the entire labor as capnography was used for
the labor duration in these women, and the Spo2 signals
were examined during all apnea events. The relationship
between administered remifentanil dose and apnea was
assessed. For apnea events, the interval examined for each
apnea event included the start of the apnea episode until
40 seconds after the end of the apnea episode. Due to the
respiratory monitoring protocol whereby the anesthesi-
ologist intervened after 40 seconds of apnea, all apnea
events >60-second duration were assumed to be artifacts
and removed. The incidence and number of hypoxemic
events (Spo2 <94%) were counted for each woman. The
mean Spo2 per woman was used to compare between the
2 study groups. Numbers of alarm triggers were recorded
for apnea, hypoxemia, and respiratory rate. Sedation scores
were recorded hourly after initiating analgesia (4-point
scale, 0 = spontaneous eye opening to 3 = unrousable).10
Maternal satisfaction with analgesia was assessed using an
March 2014 • Volume 118 • Number 3 www.anesthesia-analgesia.org 591
NRS score (0–10) 30 minutes after analgesia administration
and after delivery.
Maternal requests for additional analgesia were manu-
ally recorded in the research form hourly throughout labor.
The investigator inquired whether opioid side effects (i.e.,
pruritus and/or nausea and vomiting) were present or
absent. Oral temperature was measured both at the onset of
analgesia and within 1 hour of delivery.
Cervical dilation was assessed by the midwife, and all
changes were recorded until delivery. The time from request-
ing analgesia until 10-cm cervical dilation was also noted by
the study anesthesiologist. Use of oxytocin to augment the
labor was documented, as was mode of delivery. Decisions
to perform operative delivery were made by the attending
obstetrician based on standard practice. The duration of the
first and second stages of labor were abstracted from the
electronic medical record as recorded by the midwife. The
first stage was defined as time from the first notation of a
cervical dilatation >2 cm with regular contractions until the
cervix was noted to be dilated to 10 cm.
After delivery, face-to-face follow-up was performed on
the first postpartum day for both mother and child by one of
the investigators (DS). Umbilical artery pH and base excess
were abstracted from the medical record. Apgar scores at 1
and 5 minutes were determined by the midwife or pediatri-
cian (neither were involved in the study). Neonatal inter-
ventions required in the immediate postnatal period (e.g.,
manual ventilation with bag/mask or supplemental oxy-
gen) were noted. Umbilical artery pH and base excess were
measured. Heart rate, respiratory rate, Spo2, and tempera-
ture were recorded by the neonatologist during the initial
newborn assessment in the postnatal ward, approximately
30 to 60 minutes after delivery. Any exceptional event was
recorded by the neonatologist.
Study End Points
The primary study objective was to demonstrate noninfe-
riority of remifentanil labor analgesia compared with epi-
dural analgesia in laboring women, measured by hourly
assessment of NRS for pain throughout the duration of
labor and maternal satisfaction. The secondary outcome
was the incidence of apnea.
Statistical Methods
Sample Size
This study was designed to determine whether remifentanil
analgesia in labor is noninferior to epidural analgesia with
respect to the subject-reported pain using NRS. This efficacy
end point was chosen rather than a potentially rarer safety
end point. The sample size required to test this hypothe-
sis with 90% power at a 1-sided 2.5% level of significance,
assuming a standard deviation (SD) of 13 mm and a nonin-
feriority margin of −15 mm, was estimated as 17 women per
study group. The sample size was based on previous find-
ings of 10 mm SD, where 1 NRS unit was assumed equiv-
alent to 10 mm on the 100-mm visual analog scale.11 That
study reported a comparison of analgesic efficacy of a high
versus low epidural dose for labor analgesia. For the current
study, we elected to use the more conservative SD of 13 mm,
(1.3 units), to increase the sample size. The noninferiority
Effects of IV Remifentanil Vs Epidural for Labor
margin of 15 mm (or 1.5 units) was selected since this was
not considered to be a clinically significant difference in
reported pain scores. A noninferiority test was used because
epidural analgesia has been shown to be more effective than
remifentanil analgesia and was not expected to be worse.
Allowing for a 15% dropout rate, 20 subjects were random-
ized to each study arm. We assumed that the remifentanil
analgesic effect was not inferior to the epidural analgesia
effect if the lower limit of the 95% confidence interval (CI) of
the difference between the epidural and remifentanil mean
NRS scores was >−1.5 units, that is, −1.5 was not contained
within the 95% CI of the difference. A lower NRS score is
desirable; thus, a negative difference means that the epi-
dural score is lower (better) than the remifentanil score.
Data Analysis
Data were analyzed using the intention-to-treat principle.
Demographic data and baseline clinical characteristics
were tabulated and compared between the 2 study groups.
Continuous data are summarized by a mean and SD and
compared with 2 sample t tests (respiratory rate, saturation,
etco2, cervical dilatation, maternal satisfaction, tempera-
ture, duration of labor, birth weight, umbilical artery pH
and base excess, neonatal assessments). Wilcoxon Mann-
Whitney U was used for nonparametric tests for nonnor-
mally distributed data after assessment for normality using
Q-Q plots (number of respiratory, hypoxemia, apnea alarm
Figure 1. Trial enrollment profile. The exclusion was due to obstetrician concern regarding potential effects of remifentanil patient-controlled
IV analgesia on cardiotocography monitoring, which demonstrated fetal heart rate nonvariability, after enrollment but before analgesia.
592   
www.anesthesia-analgesia.org
occurrences per woman). Categorical data are presented as
counts and percentages and compared with the χ2 test or
Fisher exact test (number of women with: apnea alarm trig-
gered, requests for additional analgesia, nausea, pruritus,
sedation, oxytocin augmentation, instrumental or cesarean
delivery, neonatal resuscitation). Repeated measures analy-
sis of variance was performed (SAS® PROC MIXED proce-
dure) to model the NRS scores changes over time, in which
the NRS score differences from baseline were modeled as
a function of the baseline NRS, time, study group, and
the study group * time interaction. LSMeans (model esti-
mated means) for planned comparisons of the differences
in NRS scores between the groups per time point were cal-
culated from the interaction term and are presented with
95% confidence intervals (all Bonferroni adjusted). Several
covariance structures were assessed in the model-building
process; the compound symmetry structure had the low-
est AIC (Akaike information criterion) value and was used
in the final model. These 95% CIs were used to test the
primary efficacy end point. A P value of 0.05 or less was
considered statistically significant. Nominal P values are
presented.
Data retrieved from the capnograph were analyzed by
a mathematician using Matlab version 7.9.0, Natick, MA:
The MathWorks Inc., 2003. Due to the 1-hour length of the
epidural data strip, group allocation was implicitly obvi-
ous; hence, the mathematician was not blinded to group
anesthesia & analgesia
Table 1.  Population Characteristics
Remifentanil, n = 19 Epidural, n = 20 P
Maternal age (y)b 31 ± 5 30 ± 6 0.49
Weight (kg)b 72 ± 10 73 ± 13 0.89
Height (cm) 166 ± 4 162 ± 6 0.02
Parity (n)a 1 (1–1) 1 (0–1) 0.27
Nulliparas (n, %)c 4 (21%) 6 (30%) 0.72
Planned to take epidural analgesia before request for analgesia (n, %)b 15 (79%) 17 (85%) 0.70
Attended antenatal course (n, %)b 14 (73%) 12 (60%) 0.50
a
Mann-Whitney U test median (IQR).
b
t test mean ± SD; χ2.
c
Fisher exact test.

Table 2.  Repeated Measures ANOVA Was Performed (SAS® PROC MIXED Procedure) to Model the NRS
Scores Over Time
Model estimated mean differences, 95% CI, (Bonferroni adjusted) P, (Bonferroni adjusted)
(epidural-remifentanil)
Baseline 0.28 −2.74 to 3.31 0.73
30 min −2.21 −5.24 to 0.81 0.0081
1h −1.95 −5.23 to 1.33
2h −3.51 −7.31 to 0.28 0.0008
3h −2.35 −6.62 to 1.91
4h −2.81 −7.87 to 2.25
5h −0.13 −5.78 to 5.53
6h −3.55 −10.83 to 3.74 0.08
Post delivery −1.96 −5.09 to 1.18
Numerical rating scale (NRS) score was modeled as a function of time, group, and the group * time interaction. LSMeans (model estimated means) for the NRS
scores in each group per time point presented in Figure 2 and the differences between the groups at each time point were estimated from the model (from the
interaction term) and are presented with respective Bonferroni adjusted 95% confidence intervals, and level of significance. The null hypothesis of noninferiority
is tested at each time point via the CIs, if −1.5 is not contained within the interval the remifentanil group is deemed noninferior to the epidural group. This did
not occur at any time point.

allocation. However, she was blinded to our study hypoth- remifentanil was significantly less effective than epidural
esis that remifentanil was noninferior to epidural analgesia analgesia, including at the postpartum assessment.
and that remifentanil may have respiratory effects different One woman allocated to the epidural analgesia group
from those seen with epidural analgesia. For the analysis, refused to use the respiratory monitor; therefore, these
the mean ± SD for each recorded signal (etco2, respiratory data are reported for 19 women per group. Apnea occurred
rate, and Spo2) over the first hour after analgesia initiation a total of 9 times in 5 women during the first hour in the
were used for each woman as a summary statistic. The data remifentanil group and did not occur in any woman in the
were compared between groups using 2 sample t tests. The
incidence of occurrence of an alarm trigger for each woman
was compared using Fisher exact test. Median, range [IQR],
number of alarm triggers per woman are reported. The cor-
relation between apnea and the total remifentanil dose were
analyzed by Spearman Rank Correlation.

RESULTS
During the study period, 144 women were approached,
40 women consented to enroll, and 1 was subsequently
excluded Figure 1. Demographic characteristics are reported
in Table 1.
Remifentanil was found to be inferior to epidural anal-
gesia with respect to the NRS scores at all time points; the
observed difference in NRS was greater than the expected
−1.5 unit difference in NRS scores. NRS pain scores are pre-
sented in Table 2. There was no significant difference between
baseline NRS pain scores in the 2 groups. Scores were sig-
nificantly lower at 30 minutes in both groups (adjusted mean
[± SD] change for remifentanil −4.7 ± 0.6, epidural −7.2 ±
0.6, P < 0.0001). Both remifentanil and epidural analgesia
resulted in a significant decrease from baseline NRS scores
over time (Fig. 2). Assessing the change in NRS scores over 6
hours via model of repeated measures analysis of variance,
Figure 2. The figure depicts the model estimated mean (LSmeans)
(with error bars) changes from baseline for both study groups at
each time point. Since the figure depicts model estimated means
the SE of the estimate is presented. The change from baseline in
numerical rating scale (NRS) score was modeled using repeated
measures analysis of variance model. NRS score differences from
baseline were modeled as a function of the baseline NRS time,
group, and the group * time interaction. The baseline values were
similar in both groups (remifentanil 8.4 ± 1.5, epidural 8.7 ± 1.2,
P = 0.52). The NRS score is reduced in both groups compared with
baseline. The number of data points used at each time is shown in
the raw data, Table 2, and the attrition is manifested in the figure by
the longer error bars.

March 2014 • Volume 118 • Number 3 www.anesthesia-analgesia.org 593

Effects of IV Remifentanil Vs Epidural for Labor

Table 3.  Respiratory Data During First Hour After Analgesia Initiation

Remifentanil, n = 19 Epidural, n = 19a 95% CI of difference P
Respiratory rate (bpm)d 18.2 ± 4.1 21.1 ± 3.9 −2.9 (−5.6 to 0.2) 0.03
Saturation (%)d 96.8 ± 1.4 98.4 ± 1.2 −1.6 (−2.49 to 0.76) <0.001
etco2 (mm Hg)d 34.2 ± 1.8 32.9 ± 2.4 1.3 (−0.14 to 2.6) 0.08
Respiratory rate alarm triggered (<8 bpm) (n, (%))e 10 (52.6%) 11 (57.9%) −0.28 to 0.36 1.00
Number of respiratory rate alarms triggered per 1;0–34[0–3] 1;0–43[0–3] 0.65b
woman (<8 bpm)c
Hypoxemia alarm (Sao2 <94%) triggered (n, (%))e 13 (68.4%) 3 (15.8%) 0.17 to 0.74 0.003
Number of hypoxemia alarms triggered per woman 4;0–18[0–9] 0;0–23[0] 0.002b
(Sao2 <94%)c
Apnea alarm triggered (>20 s of zero respiratory 5 (26.3%) 0 −0.0038 to 0.51 0.046
rate) (n, (%))e
Number of apnea alarms triggered per woman (> 0;0–2[0–0] 0 0.018b
20 s of zero respiratory rate)c
Data for apnea alarm triggers have unequal variance.
n = number.
a
One patient refused to use the capnograph throughout the first hour, so 19 patients are analyzed.
b
Levene’s test showed that data for alarms for hypoxemia and respiratory rate triggers have equal variance, despite the zero median.
c
Mann-Whitney U test median;range [IQR]).
d
Independent sample t test, mean ± SD.
e
Fisher exact test.

Figure 3. Twenty-seven apnea events were detected (no breath for at
least 20 seconds) in the remifentanil patient-controlled IV analgesia
(PCIA) group. The graph represents the lowest Spo2 recorded during
each apnea event. Spo2 remained above 94% throughout 16/27
(59.3%) apnea events and above 92% throughout 18/27 (66.7%)
apnea events.
Figure 4. This figure presents an example of an etco2 wave graph
and saturation graph during an apnea event. The interval in which
the etco2 is zero represents respiratory rate of zero. The apnea event
lasted over 30 seconds without a decrease in Spo2. The Spo2 does
not alert the staff to a respiratory problem, whereas the respiratory
rate monitoring would prompt a response.

epidural group (Table  3). The hypoxemia alarm data and
other respiratory variables are reported in Table 3.
In the remifentanil group, 27 apnea events occurred in 9
women over the duration of monitoring; 14 events occurred in
the first 2 hours of remifentanil administration. Three women
had 1 apnea event, 1 had 2 events, 2 had 3 events, 2 had 4 events,
and 1 woman had 8 apnea events (all after the first 2 hours of
administration). There was no correlation between total remi-
fentanil dose per kilogram body weight (independent variable)
and the number of apnea episodes divided by recording length
in minutes (dependent variable), R = 0.13, P = 0.30.
The Spo2 signals were examined during all 27 apnea
events in remifentanil women (Fig. 3). Spo2 remained above
94% throughout 16/27 (59.3%) apnea events recorded in
the remifentanil group and above 92% throughout 18/27
(66.7%) apnea events. Only in 6/27 (22.2%) events did the
Spo2 decrease below 94% before apnea occurred. Figure  4
shows a typical apnea event lasting over 30 seconds with
time-matched capnography and Spo2 graphs.
Most women in both groups were rated as awake on the
4-point sedation scale, except 6/19 (32%) in the remifent-
anil group who were easily arousable and 1 in the epidural
group (5%) (Table 4).10 Delivery modes were similar in both
groups (Table 5).
Two neonates in the remifentanil group and none in
the epidural group required an intervention after delivery
(Table  6). One neonate received stimulation and supple-
mental oxygen, after the mother received a total labor dose
of 3890 mcg remifentanil over 281 minutes (4.7 hours). The
other neonate required stimulation and manual ventilation

594   
www.anesthesia-analgesia.org anesthesia & analgesia
Table 4.  Maternal Data
Remifentanil, n = 19 Epidural, n = 20 P
Cervical dilatation at onset of analgesia (cm)d 4.3 ± 1.6 4.3 ± 1.5 0.98
Crossover to other treatment group (n, %)e 3 (16%) 1 (5%) 0.34
Maternal satisfaction at 30 min after initiating analgesia 8.4 ± 2.3 8.4 ± 2.5 0.93
(NRS 0–10)d
Request for additional analgesia 1 h after initiating 0/15 (0%)a 3/18 (17%)a 0.10
analgesia (n, %)
Maternal satisfaction postpartum (NRS 0–10)d 8.6 ± 1.4 9.1 ± 1.5 0.26
Nausea 1 h after initiating analgesia (n) 3/15 (20%)a 1/18 (6%)a 0.31
Total remifentanil administered (mcg) 1725 ± 1392d NA NA
1600(350–1600)c
Total bupivacaine administered (mg) NA 27.1 ± 30.2d NA
34.4(20.9–34.4)c
Nausea postpartum (n, %)f 3/19 (16%) 0/17 (0%)b 0.23
Pruritus 1 hour after initiating analgesia (n)f 6/15 (40%)a 14/18 (78%)a 0.04
Pruritus postpartum (n, %)f 0/19 (0%) 4/18 (22%)b 0.046
Sedation (awake) at 1 h after initiating analgesia (n, %) 13/19 (68%) 19/20 (95%) 0.09
Temperature 30 min post analgesia initiation (°C)d 36.6 ± 0.4 36.6 ± 0.3 0.99
Temperature within 1 h postpartum (°C)d 36.9 ± 0.6 36.8 ± 0.5 0.49
NRS = Numerical Rating Scale; NA = not available.
a
Four women in the remifentanil group and 2 women in the epidural group delivered within 1 hr of initiating labor analgesia.
b
Women who had a cesarean delivery were not included in the numbers for postpartum nausea and pruritus as their surgical anesthesia may contribute to the findings.
c
median(range).
d
mean ± SD.
e
all crossovers were due to failure of allocated analgesia, and there were no crossovers due to fetal heart rate abnormality.
Fisher exact test.
f 

Table 5.  Obstetric End Points

Remifentanil, n = 19 Epidural, n = 20 P
Duration of 1st stage of labor (min)b 329 ± 215 404 ± 259 0.34
Duration of 2nd stage of labor (min)b 35 ± 41 69 ± 81 0.11
Augmentation with oxytocin (n, %)c 6 (32%) 10 (50%) 0.24
Vacuum delivery (n, %)c 2 (11%) 1 (5%) 1.00
Cesarean deliveryb (n, %)c 0 (0%) 4 (20%) 0.11
a
Indications for cesarean delivery were 1 case each of arrest of descent, poor progress, failed vacuum delivery, and fetal bradycardia.
b
t test mean ± SD.
c
Fisher exact test.

Table 6.  Neonatal Data

Remifentanil, n = 19 Epidural, n = 20 P
Birth weight (g)a 3416 ± 484 3360 ± 428 0.71
Apgar 1 minb 9(9-9) 9(9-9) 0.93
Apgar 5 minb 10(10-10) 10(10-10) 0.89
Umbilical artery pHa 7.28 ± 0.06 7.31 ± 0.07 0.32
Umbilical artery base excessa −3.6 ± 2.2 −3.5 ± 1.6 0.89
Respiratory rate at 1st assessment (bpm)a 49 ± 6 48 ± 7 0.51
Heart rate at 1st assessmenta 153 ± 13 161 ± 8 0.10
Oxygen saturation (Sao2) at 1st assessment (%)a 95.2 ± 1.2 94.7 ± 1.5 0.33
Temperature at 1st assessment (°C)a 36.8 ± 0.3 36.9 ± 0.3 0.49
Resuscitation at birth (n, %)c 2 (11%) 0 (0%) 0.23
a
t test mean ± SD.
b
Mann Whitney U test median (IQR).
χ.
c 2

using bag and mask, with recovery within 30 seconds after
maternal administration of 3415 mcg remifentanil throughout
labor over 174 minutes (2.9 hours). Both newborns quickly
recovered without further interventions; 1-minute Apgar
scores were 7 and 6, and 5-minute Apgar scores were 8 and
10, respectively. There were no reported adverse events in the
neonatal ward in either group.
DISCUSSION
This study aimed to show that IV remifentanil as a labor
analgesic is not inferior to epidural analgesia. Although
both are effective at reducing NRS pain scores, remifentanil
is inferior to epidural with regard to the magnitude of the
pain score reduction. Pain scores were higher at all time
points than the expected −1.5 unit difference in NRS scores.
The current study demonstrates respiratory morbidity
with remifentanil which may be an unacceptable risk. The
data retrieved from the capnography monitor report apnea
without desaturation (Sao2 <94%), suggesting that pulse
oximetry is inadequate as an early alert to apnea.
Epidural analgesia has been shown to reduce NRS scores
to a greater extent than remifentanil; however, women

March 2014 • Volume 118 • Number 3 www.anesthesia-analgesia.org 595

Effects of IV Remifentanil Vs Epidural for Labor
receiving remifentanil expressed their satisfaction with
being able to “control” their analgesia and adjust it accord-
ing to their need.7,12 Involvement in decision making may
override other factors influencing satisfaction, including
pain.13 The introduction of routinely available remifentanil
PCIA as an alternative to epidural analgesia in 1 labor unit
led to a decrease in the epidural rate.2 Remifentanil PCIA
has been proposed as a viable routine alternative to epi-
dural analgesia in the labor ward.14
Maternal respiratory side effects of remifentanil have been
reported previously.5,6,15 Hypoxemia has been reported also
with systemic meperidine analgesia. Comparison of meperi-
dine and nitrous oxide to remifentanil show the latter provided
better analgesia with fewer adverse effects.1,16,17 However, 1 life-
threatening respiratory complication of remifentanil in labor
analgesia has been reported in the literature; other cases may
not be reported because of publishing bias.15 Administration
of supplemental oxygen in our study may have delayed the
occurrence of desaturation; however, it did not prevent apnea.
Oxygen supplementation during remifentanil administration
is recommended in laboring women by some investigators.6,7,18
Remifentanil resulted in higher mean etco2 than epidural anal-
gesia, although the difference was not significant. Since both
groups received oxygen during the entire monitoring period, it
is not reasonable to assume that supplemental oxygen was an
additional cause for hypercarbia in women receiving remifen-
tanil. Both Tveit et al.6 and Volmanen et al.7 used pulse oxim-
etry and found desaturation at effective doses of remifentanil
in labor but did not report whether apnea occurred. The Food
and Drug Administration requires that remifentanil be admin-
istered only in a monitored anesthesia care unit with continu-
ous presence of personnel trained in airway management, as
was done in the current study.19
Hypoxemic and apneic events in the remifentanil group
during labor were seen both during and after the first 2
hours and occurred throughout labor over a wide range of
doses. Thus, a “safe” dose or duration of administration of
remifentanil cannot be recommended based on the results
presented in this study. This study is one of the few studies
which offered remifentanil from time of request for analgesia
up until delivery. The bolus dose of remifentanil in the cur-
rent study was adjustable up to 60 mcg. Increasing the remi-
fentanil dose during labor can avoid regression to baseline
pain levels that may be observed when a fixed remifentanil
dose is used.4,6,18,20 Opioid-induced hyperalgesia or tolerance
may play a role in increased requirements. Both bolus and
continuous infusion have been described, although continu-
ous infusion may provide superior analgesia.5,21 This study
supports previous findings in which no correlation between
dose (adjusted for body weight) and hypoxemia was found.6
The current study supports previous findings that
maternal remifentanil may be safe for neonates, although
the study sample size is small. Remifentanil does cross the
placenta but is rapidly metabolized by blood and tissue
esterases in the fetus.22,23 Remifentanil PCIA may currently
be the optimal alternative among opioid analgesics to mini-
mize adverse neonatal side effects.24
Our study design had several limitations. Postevent
analysis of data obtained from the respiratory monitor was
complicated by some artifacts, which were not noted by the
596   
www.anesthesia-analgesia.org
anesthesiologist monitoring the patient, caused, for exam-
ple, by hand squeezing interfering with the Spo2 probe dur-
ing contractions. Connecting the Spo2 probe elsewhere was
considered inconvenient for the woman. The complexity of
analyzing apnea events was also confounded by women
temporarily removing the nasal prongs to scratch her nose,
drinking, eating, or breath holding during a contraction.
Nonbiased identification of artifacts during labor may have
been possible with an additional observer throughout labor,
although this was considered an unacceptable invasion of pri-
vacy in the study institution. The data from the current study
population suggest the different rate of apnea occurrence in
the 2 groups was unlikely to be by chance (P < 5%). However,
the study sample size was not based upon apnea as an out-
come. This difference in apnea occurrence would be expected
to be greater were a larger sample to be used. The accurate
measurement of etco2 requires a closed circuit. However, the
significant entrainment of room air was avoided by using a
capnograph providing a breath sampling rate of 50 mL/min
to the device. Neither laboring women nor the anesthesiolo-
gist were blinded to the study group assignment. This would
have required placing an epidural catheter in all women.
Most likely, both the woman and the treating team would
have guessed treatment assignment. Furthermore, a poten-
tial benefit to encourage enrollment was the fact that epidural
catheter insertion was unnecessary. To avoid bias, study par-
ticipants, the obstetricians/nurse-midwives, and the math-
ematician performing respiratory analyses were not aware
of the primary study end points. Laboring women were told
that the study was performed to assess the feasibility of using
remifentanil in the labor ward. (Interestingly, many women
who received epidural analgesia were not delighted with their
group allocation.) The epidural group underwent respiratory
monitoring for only 1 hour rather than during the entire labor.
Prestudy testing suggested that there would be high attrition
with respiratory monitoring in the epidural group.
Apnea alone may not be an unacceptable risk of remifent-
anil analgesia if it is easily resolved. However, the natural his-
tory of these apnea events was not evaluated in the current
study. We only evaluated the occurrence of apnea using bolus
doses of remifentanil. It is possible that a continuous infu-
sion, or a combination of bolus doses and a continuous infu-
sion, may have a different impact on the occurrence of apnea.
In conclusion, remifentanil administration during labor
requires appropriate monitoring to detect and alert for
maternal apnea. Capnography may be used to identify
respiratory depression, and simple interventions were suf-
ficient to restore adequate ventilation when apnea occurred
in our population. Although remifentanil analgesia is infe-
rior to epidural analgesia, it may provide a satisfactory
alternative when epidural analgesia is not desired or per-
mitted. Future studies should consider administration of
remifentanil in the labor ward with particular reference to
respiratory monitoring and manpower requirements. E
DISCLOSURES
Name: Daniel Stocki, MD.
Contribution: This author is the principal investigator, helped
study design, recruitment, and performance, collect the data,
analyze the data, and prepare the manuscript.
anesthesia & analgesia
Attestation: This author approved the final manuscript.
Conflicts of Interest: The author has no conflicts of interest to
declare.
Name: Idit Matot, MD.
Contribution: This author helped study design, performance,
analyze the data, and prepare the manuscript.
Attestation: This author approved the final manuscript.
Conflicts of Interest: The author has no conflicts of interest to
declare.
Name: Sharon Einav, MD.
Contribution: This author helped study design, analyze the
data, and prepare the manuscript.
Attestation: This author approved of the final manuscript.
Conflicts of Interest: Oridion has provided travel support for
conference presentation to Shaare Zedek Medical Center, for
which Dr. Einav has been a beneficiary.
Name: Smadar Eventov-Friedman, MD.
Contribution: This author helped study design, collect the
data, analyze the data, and prepare the manuscript.
Attestation: This author approved of the final manuscript.
Conflicts of Interest: The author has no conflicts of interest to
declare.
Name: Yehuda Ginosar, MBBS.
Contribution: This author has helped study design, analyze
the data, and prepare the manuscript.
Attestation: This author approved of the final manuscript.
Conflicts of Interest: The author has no conflicts of interest to
declare.
Name: Carolyn F.Weiniger, MB ChB.
Contribution: This author is the principal investigator, helped
study design and recruitment, collect the data, analyze the data,
prepare the manuscript, and is archival author.
Attestation: This author approved of the final manuscript.
Conflicts of Interest: Oridion has provided travel support
for conference presentation to Hadassah Hebrew University
Medical Center, for which Dr. Weiniger has been a beneficiary
This manuscript was handled by: Cynthia A. Wong, MD.
ACKNOWLEDGMENTS
The authors are grateful to the invaluable assistance and
insight from Professor Felicity Reynolds with regard to study
design and manuscript presentation. Michal Ronen, PhD, and
Keren DavidPur, BSc, from Oridion® analyzed the capnogra-
phy data and Lisa Deutsch, PhD, assisted with other statistical
analyses.
REFERENCES
1. Leong WL, Sng BL, Sia AT. A comparison between remifentanil
and meperidine for labor analgesia: A systemic review. Anesth
Analg 2011;113:818–25
2. Hill D. Remifentanil in obstetrics. Curr Opin Anaesthesiol
2008;21:270–4
3. D’Onofrio P, Novelli AM, Mecacci F, Scarselli G. The efficacy
and safety of continuous intravenous administration of remi-
fentanil for birth pain relief: an open study of 205 parturients.
Anesth Analg 2009;109:1922–4
4. Douma MR, Verwey RA, Kam-Endtz CE, van der Linden PD,
Stienstra R. Obstetric analgesia: a comparison of patient-con-
trolled meperidine, remifentanil, and fentanyl in labour. Br J
Anaesth 2010;104:209–15
March 2014 • Volume 118 • Number 3 www.anesthesia-analgesia.org 597
5. Volmanen P, Palomäki O, Ahonen J. Alternatives to neuraxial
analgesia for labor. Curr Opin Anaesthesiol 2011;24:235–41
6. Tveit TO, Seiler S, Halvorsen A, Rosland JH. Labour analgesia:
a randomised, controlled trial comparing intravenous remifen-
tanil and epidural analgesia with ropivacaine and fentanyl. Eur
J Anaesthesiol 2012;29:129–36
7. Volmanen P, Sarvela J, Akural EI, Raudaskoski T, Korttila K,
Alahuhta S. Intravenous remifentanil vs. epidural levobu-
pivacaine with fentanyl for pain relief in early labour: a ran-
domised, controlled, double-blinded study. Acta Anaesthesiol
Scand 2008;52:249–55
8. Kasuya Y, Akça O, Sessler DI, Ozaki M, Komatsu R. Accuracy
of postoperative end-tidal Pco2 measurements with main-
stream and sidestream capnography in non-obese patients and
in obese patients with and without obstructive sleep apnea.
Anesthesiology 2009;111:609–15
9. Thurlow JA, Laxton CH, Dick A, Waterhouse P, Sherman L,
Goodman NW. Remifentanil by patient-controlled analgesia
compared with intramuscular meperidine for pain relief in
labour. Br J Anaesth 2002;88:374–8
10. Demiraran Y, Kocaman B, Akman RY. A comparison of the post-
operative analgesic efficacy of single-dose epidural tramadol
versus morphine in children. Br J Anaesth 2005;95:510–3
11. Atiénzar MC, Palanca JM, Torres F, Borràs R, Gil S, Esteve I.
A randomized comparison of levobupivacaine, bupivacaine
and ropivacaine with fentanyl, for labor analgesia. Int J Obstet
Anesth 2008;17:106–11
12. Evron S, Ezri T, Protianov M, Muzikant G, Sadan O, Herman
A, Szmuk P. The effects of remifentanil or acetaminophen with
epidural ropivacaine on body temperature during labor. J
Anesth 2008;22:105–11
13. Hodnett ED. Pain and women’s satisfaction with the experi-
ence of childbirth: a systematic review. Am J Obstet Gynecol
2002;186:S160–72
14. Hill D. Remifentanil patient-controlled analgesia should
be routinely available for use in labour. Int J Obstet Anesth
2008;17:336–9
15. Waring J, Mahboobi SK, Tyagaraj K, Eddi D. Use of remifent-
anil for labor analgesia: the good and the bad. Anesth Analg
2007;104:1616–7
16. Blair JM, Dobson GT, Hill DA, McCracken GR, Fee JP. Patient
controlled analgesia for labour: a comparison of remifentanil
with pethidine. Anaesthesia 2005;60:22–7
17. Volmanen P, Akural E, Raudaskoski T, Ohtonen P, Alahuhta S.
Comparison of remifentanil and nitrous oxide in labour analge-
sia. Acta Anaesthesiol Scand 2005;49:453–8
18. Douma MR, Middeldorp JM, Verwey RA, Dahan A, Stienstra R.
A randomised comparison of intravenous remifentanil patient-
controlled analgesia with epidural ropivacaine/sufentanil dur-
ing labour. Int J Obstet Anesth 2011;20:118–23
19. Landow L, Kahn RC, Wright C. FDA’s role in anesthetic drug
development. Anesthesiology 1999;90:882–9
20. Balki M, Kasodekar S, Dhumne S, Bernstein P, Carvalho JC.
Remifentanil patient-controlled analgesia for labour: optimiz-
ing drug delivery regimens. Can J Anaesth 2007;54:626–33
21. Blair JM, Hill DA, Fee JP. Patient-controlled analgesia for
labour using remifentanil: a feasibility study. Br J Anaesth
2001;87:415–20
22. Kan RE, Hughes SC, Rosen MA, Kessin C, Preston PG, Lobo
EP. Intravenous remifentanil: placental transfer, maternal and
neonatal effects. Anesthesiology 1998;88:1467–74
23. Ross AK, Davis PJ, Dear Gd GL, Ginsberg B, McGowan FX,
Stiller RD, Henson LG, Huffman C, Muir KT. Pharmacokinetics
of remifentanil in anesthetized pediatric patients undergo-
ing elective surgery or diagnostic procedures. Anesth Analg
2001;93:1393–401, table of contents
24. Reynolds F. The effects of maternal labour analgesia on the
fetus. Best Pract Res Clin Obstet Gynaecol 2010;24:289–302