Anda di halaman 1dari 11

Original Research

Maternal Pulse Pressure and the Risk of


Postepidural Complications
A Randomized Controlled Trial
Justin R. Lappen, MD, Stephen A. Myers, DO, Norman Bolen, MD, Brian M. Mercer, MD,
and Edward K. S. Chien, MD, MBA

OBJECTIVE: To test the hypothesis that increasing the occurrence of the primary outcome from 27% in the
intravenous fluid bolus volume at the time of epidural 500-mL group to 13.5% in the 1,500-mL group
placement in women with narrow pulse pressure would (two-sided a50.05).
reduce postepidural fetal heart rate (FHR) abnormalities, RESULTS: From October 2015 to November 2016, 276
hypotension, and resuscitative obstetric interventions. women were allocated to receive a 500-mL (n5139) or
METHODS: We performed a single-center randomized 1,500-mL (n5137) fluid bolus. One hundred thirty-eight
controlled trial. Eligible participants were normotensive women were evaluated in the reference group. Demo-
with a nonanomalous singleton gestation at or after graphic, obstetric, and labor characteristics were similar
35 weeks and with a narrow pulse pressure (less than between groups. The 1,500-mL group had significantly
45 mm Hg) on admission. Enrolled patients remained fewer postepidural FHR abnormalities (38.0% compared
eligible for randomization at epidural request if they with 51.8%, relative risk 0.73, 95% CI 0.56–0.96, P5.02).
were within 6 hours of admission and the FHR remained Maternal systolic hypotension (10.2% compared with
category 1. Patients were allocated to a 500-mL (insti- 34.5%, relative risk 0.30, 95% CI 0.17–0.51, P,.001) and
tutional standard) or 1,500-mL intravenous fluid bolus composite postepidural interventions (18.3% compared
at epidural placement. A reference group with admis- with 44.2%, relative risk 0.42, 95% CI 0.28–0.62, P,.001)
sion pulse pressure 50 mm Hg or greater was also were also less frequent in the 1,500-mL group. Fetal heart
evaluated. The primary outcome was a category 2 or 3 rate abnormalities remained significantly less frequent in
FHR pattern within 60 minutes after the epidural test the reference group than among women with a narrow
dose. Evaluated secondary outcomes included maternal pulse pressure on admission for delivery.
hypotension and composite resuscitative interventions CONCLUSION: A 1,500-mL intravenous fluid bolus in
to correct FHR abnormalities or hypotension. We women with a narrow pulse pressure decreases the risk of
calculated that 276 women (138/group) would provide postepidural FHR abnormalities (number needed to
80% power to detect a relative 50% reduction in the treat57), results in less frequent postepidural hypoten-
sion, and reduces the need for resuscitative interventions.
From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Admission pulse pressure may be used to individualize
Gynecology, and the Department of Anesthesiology, MetroHealth Medical Center, intrapartum fluid management at the time of initiation of
Cleveland, Ohio.
neuraxial labor analgesia.
Presented at the Society for Obstetric Anesthesia and Perinatology 49th Annual
Meeting, May 10–14, 2017, Bellevue, Washington. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov,
Each author has indicated that he has met the journal’s requirements for author-
NCT02565485.
ship. (Obstet Gynecol 2017;130:1366–76)
Corresponding author: Justin R. Lappen, MD, Department of Reproductive DOI: 10.1097/AOG.0000000000002326
Biology, Case Western Reserve University School of Medicine, Division of
Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University
Hospitals Cleveland Medical Center, MAC 5034, 11100 Euclid Ave, Cleveland,
OH 44106; email: Justin.Lappen@UHhospitals.org.
I n contemporary obstetric practice, neuraxial anal-
gesia is the most commonly used method of pain
management for delivery, with approximately 60–
Financial Disclosure
The authors did not report any potential conflicts of interest. 70% of U.S. women receiving epidural anesthesia.1,2
Neuraxial analgesia induces a sympathetic blockade,
© 2017 by The American College of Obstetricians and Gynecologists. Published
by Wolters Kluwer Health, Inc. All rights reserved. which results in decreased maternal systemic vascular
ISSN: 0029-7844/17 resistance and increased venous capacitance. These

1366 VOL. 130, NO. 6, DECEMBER 2017 OBSTETRICS & GYNECOLOGY

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
alterations in the maternal hemodynamic profile affect
Box 1. Exclusion Criteria
maternal blood pressure (BP) and uteroplacental
blood flow and contribute to maternal hypotension ! Hypertensive disorder (chronic hypertension, gesta-
and abnormal fetal heart rate (FHR) patterns, which tional hypertension, preeclampsia, or eclampsia)19
! Gestational or pregestational diabetes mellitus
occur with reported frequencies of 14% and 30%,3–5 ! Cardiomyopathy or maternal congenital heart disease
respectively, after initiation of regional anesthesia. ! Pulmonary edema
These abrupt hemodynamic changes generally ! Renal insufficiency (serum creatinine 1.0 mg/dL or
occur in the setting of a constant maternal intravas- greater at any point during pregnancy)
cular volume. Given that the uteroplacental circula- ! Substance abuse (not including tobacco)
! Medical contraindication to neuraxial anesthesia
tion has limited capacity for autoregulation,6,7 uterine such as thrombocytopenia (platelet count less than
perfusion is sensitive to changes in both maternal BP 80,000/microliter)
and venous return.8 This physiology represents the ! Maternal age younger than 18 y
basis for the administration of an intravenous (IV) ! Inability to consent in English
bolus of crystalloid solution at the time of epidural ! Admission systolic blood pressure less than 90 mm Hg
! Multiple gestation
placement. However, the evidence supporting IV ! Suspected intrauterine growth restriction
fluid loading is conflicting,9–12 and a Cochrane review ! Congenital or chromosomal fetal anomalies
concluded that IV fluid preload does not decrease the ! Category 2 or 3 FHR pattern20 before randomization
risk of maternal hypotension or FHR abnormalities in
women receiving low-dose epidural labor analgesia.13
Although the administration of IV fluid preload is at MetroHealth Medical Center (September 1, 2015).
a common clinical practice, the American Society of Written informed consent was obtained from all trial
Anesthesiologists Task Force on Obstetric Anesthesia participants before study entry. All authors take
has stated that “administration of a fixed volume of responsibility for the completeness of the reporting
intravenous fluid is not required before neuraxial and fidelity of the report to the study protocol.
analgesia is initiated.”14 Eligible participants were healthy, normotensive
To date, published randomized trials regarding women with a viable, nonanomalous singleton gesta-
IV fluid loading at epidural placement have used an tion at or after 35 weeks who presented for delivery
empiric approach without consideration of clinical (spontaneous labor or induction of labor) and desired
factors, such as maternal volume status, that may neuraxial labor analgesia. Trial exclusion criteria are
modify the effect of this intervention. Observational listed in Box 1.
and experimental studies have demonstrated that On admission for labor management and before
pulse pressure (systolic–diastolic BP [mm Hg]) is an the initiation of intrapartum maintenance IV fluid,
indicator of central volume status with a narrow pulse potentially eligible women underwent BP evaluation to
pressure reflecting hypovolemia.15–18 However, it re- determine admission pulse pressure. Blood pressure
mains unknown whether pulse pressure can be used measurements were obtained for all study patients
for prospective identification and tailored prophylaxis using automated oscillometric BP monitors with the
to prevent postepidural FHR abnormalities and patient in a semirecumbent position with left lateral
maternal hypotension. We designed this study to displacement. Measurements were obtained between
evaluate the hypothesis that providing a larger vol- contractions and at least 1 minute from completion of
ume of IV fluid at the time of intrapartum epidural a contraction as assessed using a tocodynamometer. A
placement for women with a narrow pulse pressure pulse pressure value below 45 mm Hg (2 SDs below
(less than 45 mm Hg) would decrease the frequency of the mean at 36 weeks of gestation based on published
postprocedure FHR abnormalities. Secondarily, we longitudinal data)21 on both of the first two consecutive
sought to assess whether providing a larger volume of clinical BP measurements was required for trial enroll-
IV coload to women with a narrow pulse pressure ment. Although additional vital signs were collected
would decrease the frequency of maternal hypoten- during intrapartum management, only the admission
sion and resuscitative obstetric interventions. pulse pressure values were used for study eligibility.
Consenting patients remained eligible for random-
MATERIALS AND METHODS ization if they were within 6 hours from admission
This was a randomized controlled trial performed at at the time of epidural request, had received only
MetroHealth Medical Center in Cleveland, Ohio, maintenance IV fluid (125 mL/h of lactated Ringer’s,
between October 2015 and November 2016. This the routine intrapartum infusion rate for women at our
study was approved by the institutional review board institution), and the FHR pattern remained category 1

VOL. 130, NO. 6, DECEMBER 2017 Lappen et al Pulse Pressure and Postepidural Complications 1367

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
from admission to epidural request. Patients were ran- Patient vital sign data including temperature, BP,
domly assigned in a one-to-one ratio to receive either heart rate, respiratory rate, and oxygen saturation
a 500-mL (institutional routine) or 1,500-mL IV bolus were collected on all patients in a standardized
of lactated Ringer’s at the time of epidural placement. manner at prespecified time intervals (routine institu-
A computer-generated, block randomization plan was tional clinical practice). A complete set of vital sign
created with a block size of 10. Given the effect of data was obtained at the time of epidural request
maternal morbid obesity on volume of distribution (prebolus), immediately after the IV fluid bolus
and the risk of postepidural FHR abnormalities,22 the (postbolus), and at the time of the administration of
randomization scheme was stratified by body mass the epidural test dose. Postbolus vital sign data were
index (BMI, calculated as weight (kg)/[height (m)]2; less obtained only if the IV fluid bolus was completed
than 40, 40 or greater). The allocation for each study before the epidural test dose. After the epidural test
participant was located in sequentially numbered, dose, heart rate and BP were obtained at 5-minute
sealed opaque envelopes in a locked research office intervals for the first 30 minutes and then 10-minute
on the labor and delivery unit until allocation was as- intervals for an additional 30 minutes. Subsequently,
signed, which occurred immediately before the study vital sign data were collected at frequencies deter-
intervention. A single member of the research team mined by clinical care practices of our labor and
(J.R.L.) generated the random allocation sequence, delivery unit. After the epidural test dose, BP meas-
which was destroyed after the sealed envelopes were urements were obtained using an automated cuff with
generated. Study patients were enrolled and subse- patients in a semirecumbent position with left lateral
quently allocated by members of the study team displacement. Continuous maternal pulse oximetry
(except E.K.S.C.) and by resident and attending physi- and continuous external (or internal) fetal monitoring
cians on labor and delivery. In this pragmatic trial, were routinely performed for all participants. Fetal
obstetric, anesthesia, and nursing providers were not monitoring was not typically performed during epi-
masked to study group assignment after allocation. dural placement.
The study protocol specified that the IV fluid bolus The management of postepidural FHR abnormal-
be infused over 30 minutes and initiated with patient ities and maternal hypotension, including the admin-
epidural request and with preparation for or initiation istration of vasopressors, additional IV fluid, and
of epidural placement by the anesthesiology team supplemental oxygen; intentional position changes;
(ie, coload). Women assigned to the 500-mL group fetal scalp electrode placement; and obstetric inter-
received an IV fluid bolus on a programmed infusion ventions such as operative vaginal and cesarean
pump (rate 999 mL/h), whereas women in the 1,500- delivery, was performed according to the clinical
mL group received a bolus off pump with a pressure judgment of the obstetric and anesthesiology teams.
bag. The duration of the IV fluid bolus and the date Artificial rupture of membranes within 60 minutes of
and time of initiation of analgesia were documented in epidural placement was permitted only if medically
the medical record by nursing and anesthesia staff. indicated (eg, placement of fetal scalp electrodes for
Before study initiation, the technique for epidu- FHR abnormalities). On completion of the epidural
ral placement and dosing was standardized on our placement and 60 minutes of postprocedure monitor-
labor and delivery unit. After epidural catheter ing, participants were considered to have completed
placement and administration of a test dose (3 mL the trial and were managed according to their health
of 1.5% lidocaine with 5 micrograms/mL epineph- care provider’s clinical judgment.
rine to rule out the presence of an intravascular or The primary outcome for this trial was the
intrathecal catheter), epidural analgesia was initiated occurrence of a category 2 or category 3 FHR pattern
with a bolus dose of 5 mL of 0.125% bupivacaine within the 60 minutes after the epidural test dose.20 A
with 10 micrograms/mL of fentanyl and 5 micro- single member of the research team (E.K.S.C.) who
grams/mL epinephrine. The level of sensory block was masked to patient study group assignment inter-
and patient pain scores were assessed by anesthesi- preted all FHR patterns. The postepidural FHR trac-
ology teams. Pain scores were rated on an 11-point ing was analyzed within four 15-minute epochs and
scale in which 0 represented no pain and 10 the worst was categorized as category 1, 2, or 3. Uterine con-
pain imaginable. If a patient did not achieve a pain traction frequency was also quantified within each
score of 4 or less at 15 minutes after the initial epoch.
epidural bolus, additional bolus doses of 3 mL of the Prespecified secondary outcomes included mater-
same solution were administered until adequate nal hypotension (more than a 20% decrease in systolic
analgesia was achieved. or diastolic BP from the admission value), composite

1368 Lappen et al Pulse Pressure and Postepidural Complications OBSTETRICS & GYNECOLOGY

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
resuscitative interventions (administration of vaso- institutional delivery data consecutively beginning
pressors, additional IV fluid, supplemental oxygen, on the day trial enrollment was initiated. To facilitate
intentional maternal position changes, or emergent matching by BMI in a group receiving usual care
operative vaginal or cesarean delivery) to correct (and not a randomly allocated intervention), refer-
FHR abnormalities or maternal hypotension, and ence group patients were not enrolled prospectively
adverse events (pulmonary edema) within 60 minutes or consented as trial participants.
after epidural placement. An analysis of vital sign A previous retrospective study by Miller et al18
trends and obstetric and neonatal outcomes according reported that new-onset FHR abnormalities occur
to study group was also planned. One member of the after epidural placement in 27% of women with an
research team (J.R.L.) abstracted information from admission pulse pressure less than 45 mm Hg and in
maternal and neonatal charts, including demographic 6% of women with an admission pulse pressure of
and clinical characteristics, medical and obstetric 45 mm Hg or greater. We calculated that 276 women
history, and maternal and neonatal outcomes. Data (138/group) would provide 80% power to detect a rel-
were entered in REDCap, a Health Insurance Porta- ative 50% reduction in the occurrence of the primary
bility and Accountability Act-certified, secure web- outcome from 27% in the 500-mL group to 13.5% in
based data storage platform for research studies.23 the 1,500-mL group (two-sided alpha .05). For a one-
To further characterize the effect of pulse pres- to-one-to-one ratio between study groups, 138 women
sure on the primary and secondary study outcomes, were evaluated in the reference group.
we evaluated a reference group of women with Analyses were performed according to the
a normal admission pulse pressure (50 mm Hg or intention-to-treat principle using Stata 13.1. Continu-
greater on the first two consecutive measurements) ous variables were compared using Wilcoxon rank-
matched according to BMI who otherwise met all sum and analysis of variance where appropriate.
study inclusion and exclusion criteria. Women in the Categorical variables were compared using x2, Fisher
reference group received routine obstetric and anes- exact, and Kruskal-Wallis tests where appropriate.
thesia management and received a 500-mL IV bolus Number needed to treat calculations were performed
of lactated Ringer’s at the time of epidural place- for primary and secondary study outcomes. Among
ment. Reference group patients were identified from three-group analyses in which a statistically significant

Fig. 1. Trial screening, enrollment, and randomization. FHR, fetal heart rate; IV, intravenous.
Lappen. Pulse Pressure and Postepidural Complications. Obstet Gynecol 2017.

VOL. 130, NO. 6, DECEMBER 2017 Lappen et al Pulse Pressure and Postepidural Complications 1369

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
Table 1. Baseline Characteristics of Study Participants

Study Intervention Groups Reference Group


(Pulse Pressure Less Than (Pulse Pressure 50 mm
45 mm Hg) Hg or Greater)
500-mL Bolus 1,500-mL Bolus 500-mL Bolus
(n5139) (n5137) (n5138) P*
Admission
Age (y) 26.065.1 26.965.3 25.565.3 .08
Race .92
Caucasian 65 (46.8) 62 (45.3) 58 (42.0)
African American 56 (40.3) 52 (38.0) 57 (41.3)
Asian 3 (2.2) 3 (2.2) 2 (1.5)
Other 15 (10.8) 20 (14.6) 21 (15.2)
Gestational age (wk) 39.1 (1.1) 39.1 (1.3) 39.0 (1.2) .38
Nulliparous 45 (32.4) 43 (31.4) 67 (48.6) .006†
Prior cesarean delivery 11 (7.9) 8 (5.8) 9 (6.5) .83
BMI (kg/m2) 31.165.5 31.566.0 32.166.6 .39
40 or greater 9 (6.5) 9 (6.6) 16 (11.6) .70
Tobacco abuse 24 (17.3) 29 (21.2) 21 (15.2) .22
Admission diagnosis‡ .007§
Induction of labor 51 (36.7) 69 (50.4) 46 (33.3)
Spontaneous labor 88 (63.3) 68 (49.6) 92 (66.7)
Admission cervical examination (cm) 2.9 (1.3) 3.1 (1.5) 3.2 (1.7) .24
Admission vital signs║
Systolic blood pressure (mm Hg) 114.5 (7.5) 114.4 (7.4) 126.8 (7.3) ,.001†
Diastolic blood pressure (mm Hg) 76.1 (7.2) 75.8 (6.7) 70.5 (7.2) ,.001†
Pulse pressure (mm Hg) 38.4 (3.9) 38.6 (4.1) 56.3 (4.4) ,.001†
Mean arterial pressure (mm Hg) 88.9 (7.0) 88.7 (6.7) 89.3 (6.9) .75
Heart rate (bpm) 91.4 (13.5) 91.0 (13.0) 90.4 (12.3) .82
Respiratory rate (bpm) 18.3 (1.4) 18.3 (1.4) 18.4 (1.6) .78
Temperature (˚C) 36.7 (0.2) 36.8 (0.2) 36.8 (0.2) .27
Oxygen saturation (%) 99.0 (1.1) 99.0 (1.1) 99.1 (1.1) .98
Admission laboratory data
Hemoglobin (g/dL) 11.461.4 11.4 (1.4 11.3 (1.3 .72
Platelet count (thousands/microliter) 218.0659.5 216.1657.3 226.7664.1 .31
Epidural request or placement
Admission to epidural request (min) 150 (78–241) 192 (90–287) 138 (58–332) .28
IV fluid before epidural request (mL) 313 (163–502) 402 (188–615) 286 (121–692) .12
Narcotics before epidural request 12 (8.6) 9 (6.6) 18 (13.0) .08
Oxytocin before epidural request 83 (59.7) 87 (63.5) 60 (43.5) .002†
Oxytocin infusion rate at request (milliunits/min) 8.567.3 9.4 (7.4 8.3 (6.7 .28
Cervical dilation at epidural request (cm) 3.461.3 3.661.3 3.561.2 .60
Duration of IV fluid bolus (min)¶ 29.861.8 32.268.0 30.160.9 ,.001§
Start of IV bolus to epidural test dose (min)¶ 10 (5–16) 14 (8–21) 11 (6–17) ,.001§
Total volume epidural solution (mL)# 5.361.9 5.261.8 5.161.3 .63
Additional 3-mL epidural bolus# .22
1 12 (8.6) 11 (8.0) 11 (8.0)
2 1 (0.7) 2 (1.5) 1 (0.7)
Postepidural (first 60 min)
Oxytocin after epidural placement 84 (60.4) 91 (66.4) 70 (50.7) .01†
Dose (milliunits/min) 8.667.0 9.967.4 8.366.7 .28
Rupture of membranes (spontaneous or artificial) 35 (25.2) 37 (27.0) 26 (18.8) .13
BMI, body mass index; bpm, beats per minute; IV, intravenous.
Data are mean6SD, n (%), or median (interquartile range) unless otherwise specified.
Study intervention groups compared using x2 or Wilcoxon rank-sum where appropriate. No statistically significant differences in baseline
characteristics were present except for those noted in the ‡ and ¶ footnotes.
* Three-group comparison using Kruskal-Wallis or analysis of variance where appropriate. If three-group analysis was statistically
significant, pairwise comparison between all groups was performed with indicating reference group significantly different than both
study intervention groups (†), or indicating reference group significantly different than 1,500-mL group (§).

Indicating reference group significantly different than both study intervention groups.

Indicating significant difference between study intervention groups with P5.02.
§
Indicating reference group significantly different than 1,500-mL group.

Systolic and diastolic blood pressure, pulse pressure, and mean arterial pressure are the average of the two consecutive measurements
required for trial enrollment. Other vital sign values are from the first set obtained on presentation.

Indicating significant difference between study intervention groups with P,.01.
#
Solution of 0.125% bupivacaine with 10 micrograms/mL of fentanyl and 5 mg/mL epinephrine.

1370 Lappen et al Pulse Pressure and Postepidural Complications OBSTETRICS & GYNECOLOGY

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
result was obtained, post hoc pairwise comparisons erence group was similar to the study intervention
were performed with Dunn adjustment for multiple groups regarding baseline, intrapartum, and anesthe-
comparisons. Maternal vital sign trends were analyzed sia characteristics except that these women were more
using two-way analysis of variance with Tukey correc- likely to be nulliparous, were more frequently admit-
tion for multiple comparisons. A P value of ,.05 was ted in spontaneous labor, and were less likely to
considered statistically significant for all analyses. receive oxytocin before or after epidural placement.
An independent data safety monitor was identi- Complete data were available for all women in
fied to review all cases of pulmonary edema related to the trial and no patients were lost to follow-up. The
the study intervention. Given the rarity of pulmonary frequency of the primary outcome, a category 2 or 3
edema in healthy obstetric patients and the exclusion FHR pattern, was significantly lower in the 1,500-mL
of patients with pre-existing cardiac disease or pre- group than in the 500-mL group (38.0% compared
eclampsia, a 5% rate of pulmonary edema at 50% of with 51.8%, relative risk [RR] 0.73, 95% CI 0.56–0.96,
study enrollment was specified a priori as a trigger to P5.02). The number needed to treat to prevent one
stop the study. No interim analyses were planned or primary outcome event is seven (95% CI 4–48).
performed. Postepidural systolic hypotension (10.2% com-
pared with 34.5%, RR 0.30, 95% CI 0.17–0.51,
RESULTS P,.001) and diastolic hypotension (55.5% compared
Between October 2015 and November 2016, 338 with 75.5%, RR 0.73, 95% CI 0.62–0.88, P,.001) were
women meeting study criteria were enrolled and 276 less frequent in the 1,500-mL group than the 500-mL
women underwent randomization, 139–500 mL and group (Table 2). Additionally, the composite rate of
137–1,500 mL of lactated Ringer’s at the time of epi- interventions to treat an abnormal FHR pattern or
dural placement (Fig. 1). The most common reason maternal hypotension was lower in the 1,500-mL
for exclusion from randomization after enrollment group than the 500-mL group (18.3% compared with
was a duration of more than 6 hours from admission 44.2%, RR 0.42, 95% CI 0.28–0.62, P,.001). The
to epidural request. Table 1 displays baseline charac- number needed to treat to prevent one additional
teristics of the study cohort and intrapartum and anes- intervention for an abnormal FHR pattern or mater-
thesia characteristics at the time of epidural request nal hypotension was four (95% CI 3–6). Women allo-
and after epidural placement. The two study interven- cated to 1,500 mL were significantly less likely to
tion groups were similar at baseline except that receive one or more additional IV fluid boluses
women in the 1,500-mL group were more likely to (7.3% compared with 27.3%) or undergo intentional
have been admitted for induction of labor than those position changes (14.6% compared with 29.5%) in the
in the 500-mL group (50.4% compared with 36.7%, 60 minutes after epidural test dose (P#.003 for each).
P5.02). Labor and anesthesia characteristics at epidu- There were no statistically significant differences in
ral request and after epidural placement were also supplemental oxygen use, vasopressor administration,
similar between the study intervention groups with and emergent operative vaginal or cesarean delivery
the exception that the time to complete the IV fluid among groups.
bolus was significantly shorter in the 500-mL group In comparison with the two study intervention
(29.861.8 compared with 32.268.0 minutes, P,.001) groups, the reference group demonstrated a signifi-
and the time from the start of the IV bolus to the cantly lower rate of category 2 or 3 FHR patterns in
epidural test dose was longer in the 1,500-mL group the 60 minutes after initiation of epidural analgesia
(14 [8–21] compared with 10 [5–16] minutes, P5.005). (17.4% compared with 38.0% compared with 51.8% in
Baseline characteristics of the reference group are the reference, 1,500-mL, and 500-mL groups, respec-
also depicted in Table 1. As anticipated, mean admis- tively, P,.001) (Table 2). The frequency of maternal
sion BP and pulse pressure were significantly different hypotension (systolic or diastolic) was also lower in
between the reference group and the study interven- the reference group compared with both study inter-
tion groups. Higher systolic BP (126.867.3 compared vention groups (40.6% compared with 55.5% com-
with 114.467.4 compared with 114.567.5 mm Hg, pared with 75.5% in the reference, 1,500-mL, and
P,.001), lower diastolic BP (70.567.2 compared with 500-mL groups, respectively, P,.01). The frequency
75.866.7 compared with 76.167.2 mm Hg, P,.001), of systolic hypotension was significantly lower in the
and a wider pulse pressure (56.364.4 compared with reference group compared with the 500-mL group;
38.664.1 compared with 38.463.9 mm Hg, P,.001) however, it was lowest in the 1,500-mL study inter-
were present in the reference group compared with vention group (24.6% compared with 10.2% com-
the 1,500-mL and 500-mL groups. Otherwise, the ref- pared with 34.5% in the reference, 1,500-mL, and

VOL. 130, NO. 6, DECEMBER 2017 Lappen et al Pulse Pressure and Postepidural Complications 1371

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
Table 2. Primary and Secondary Outcomes

Study Intervention Groups Reference Group


(Pulse Pressure Less Than (Pulse Pressure 50 mm
45 mm Hg) Hg or Greater)
500-mL Bolus 1,500-mL Bolus RR 500-mL Bolus
Outcome (n5139) (n5137) (95% CI) P* (n5138) P†

Primary outcome
Category 2 or 3 FHR pattern 72 (51.8) 52 (38.0) 0.73 (0.56–0.96) .02
Secondary outcomes
Category 2 or 3 FHR pattern 72 (51.8) 52 (38.0) 24 (17.4) ,.001‡
Hypotension within 60 min of 105 (75.5) 76 (55.5) 0.73 (0.62–0.88) ,.001 56 (40.6) .01‡
epidural§
Systolic hypotension 48 (34.5) 14 (10.2) 0.30 (0.17–0.51) ,.001 34 (24.6) .005‡
Diastolic hypotension 101 (72.7) 76 (55.5) 0.76 (0.64–0.92) .003 47 (34.1) ,.001‡
Composite interventions for FHR 61 (44.2) 25 (18.3) 0.42 (0.28–0.62) ,.001 21 (15.2) .55
or hypotension
Intentional position change 41 (29.5) 20 (14.6) 0.49 (0.31–0.80) .003 11 (8.0) .14
Supplemental oxygen 26 (18.7) 16 (11.7) 0.62 (0.35–1.11) .10 8 (5.8) .13
Additional IV fluid bolus 38 (27.3) 10 (7.3) 0.27 (0.14–0.51) ,.001 15 (10.9) .42
Vasopressor administration 22 (15.8) 12 (8.8) 0.55 (0.29–1.07) .07 14 (10.1) .72
Transfer to operating room 1 (0.7) 2 (1.5) 2.02 (0.19–22.12) .55 1 (0.7) .53
Emergent operative delivery 0 1 (0.7) — .31 0 .22
Adverse event (pulmonary edema) 0 0 — — 0 —

FHR, fetal heart rate; IV, intravenous; RR, relative risk.


Data are n (%) unless otherwise specified.
* Pairwise comparison of study intervention groups using x2 or Fisher exact test where appropriate.

Three-group comparison using Kruskal-Wallis test. If three-group analysis was statistically significant, pairwise comparison between all
groups was performed with indicating reference group significantly different than both study intervention groups (‡).

Indicating reference group significantly different than both study intervention groups.
§
Hypotension defined a priori as greater than 20% decrease in systolic or diastolic blood pressure from baseline admission value. Outcome
includes all patients with any blood pressure measurements in the 60 minutes after epidural placement meeting criteria for hypotension.

500-mL groups, respectively, P,.01 for three-group hypotension. Incidental placement of combined
and pairwise comparisons). In three-group compari- spinal–epidural analgesia occurred in six patients
sons, there were no significant differences in rates of (four in the 500-mL group, one in the 1,500-mL
interventions for an abnormal FHR or maternal hypo- group, and one in the reference group), a rate that
tension. However, the composite rate of interventions was not significantly different among study groups
for an abnormal FHR or maternal hypotension was (P5.37 for study intervention groups). Only two pri-
significantly lower for the reference group compared mary outcome events occurred among those women
with the 500-mL study intervention group (15.2% who received combined spinal–epidural analgesia,
compared with 44.2%, P,.001). This included signif- one in each of the two study intervention groups. In
icant differences in individual outcomes of intentional a per-protocol analysis excluding patients receiving
position changes (8.0% compared with 29.5%), sup- combined spinal–epidural analgesia, the results were
plemental oxygen (5.8% compared with 18.7%), and similar to the intention-to-treat analysis (data not
additional IV fluid boluses (10.9% compared with shown). There were no cases of pulmonary edema or
27.3%) (P,.001 for each). other adverse events related to the study intervention.
Two patients developed a category 3 FHR pattern Trends in systolic and diastolic BP, pulse pres-
in the hour after epidural placement (one in the sure, and mean arterial pressure between baseline
500-mL group and one in the reference group). In (before IV fluid bolus) and 60 minutes after the
both cases, the FHR pattern was transient, normalized epidural test dose are displayed in Figure 2. As antic-
with resuscitative interventions, and did not require ipated, systolic BP, diastolic BP, and mean arterial
operative delivery. A single emergent cesarean deliv- pressure all decreased after epidural placement. In
ery after epidural placement occurred in a patient in response to the IV fluid bolus, women allocated to
the 1,500-mL study intervention group after a pro- 1,500 mL had a significantly greater increase in pulse
longed FHR deceleration in the absence of maternal pressure than those allocated to 500 mL, and this

1372 Lappen et al Pulse Pressure and Postepidural Complications OBSTETRICS & GYNECOLOGY

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
increase was sustained across the 60 minutes after the DISCUSSION
epidural test dose (Fig. 2C). This increase in pulse In this randomized trial, we found that the adminis-
pressure resulted primarily from a statistically signifi- tration of a 1,500-mL IV fluid bolus at the time of
cant and sustained increase in systolic BP (Fig. 2A). epidural placement for women with a narrow pulse
Additionally, during the 60 minutes after epidural pressure on admission for delivery decreased the
placement, women in the reference group demon- subsequent occurrence of FHR abnormalities when
strated significantly higher systolic BP and pulse pres- compared with a 500-mL IV fluid bolus. In addition,
sure than the 500-mL, but not the 1,500-mL, study a 1,500-mL IV fluid bolus also resulted in less
intervention group (Figs. 2A, C). No significant inter- frequent postepidural maternal hypotension and
group differences in diastolic BP or mean arterial pres- fewer resuscitative obstetric and anesthetic interven-
sure were evident (Figs. 2B, D). Mean maternal heart tions. Admission pulse pressure, as a peripheral
rate increased after epidural placement with peak surrogate of central volume status, may be used to
heart rate occurring concurrently to the BP nadir; individualize intrapartum fluid management at the
however, no significant intergroup differences were time of initiation of neuraxial labor analgesia.
evident (data not shown). This trial demonstrates a physiology-based
Data regarding obstetric and neonatal outcomes approach to fluid management at the initiation of
are displayed in Table 3. We found no differences in neuraxial labor analgesia and highlights the use of
the frequency of cesarean delivery, operative vaginal pulse pressure as a peripheral indicator of central
delivery, chorioamnionitis, or neonatal morbidities volume status. Observational studies have demon-
between study groups. strated that changes in pulse pressure and stroke

Fig. 2. Trend in mean maternal vital sign parameters from before intravenous (IV) fluid bolus to 60 minutes after the epidural
test dose. Systolic blood pressure (A), diastolic blood pressure (B), pulse pressure (C), and mean arterial pressure (D).
Postbolus vital sign data were obtained only if the IV fluid bolus was completed before the epidural test dose. Therefore,
mean values at the postbolus time are based on n558 (500-mL study intervention group), n565 (1,500-mL study inter-
vention group), and n537 (reference group). Statistical analysis performed using two-way analysis of variance with Tukey
adjustment for multiple comparisons. *P,.01 between the 500-mL group and both the 1,500-mL and reference groups.
†P,.01 between the reference group and both the 500-mL and 1,500-mL groups. ‡P,.01 between the reference group and

500-mL group. §P,.01 between the 500-mL and 1,500-mL groups.


Lappen. Pulse Pressure and Postepidural Complications. Obstet Gynecol 2017.

VOL. 130, NO. 6, DECEMBER 2017 Lappen et al Pulse Pressure and Postepidural Complications 1373

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
Table 3. Maternal and Neonatal Labor Outcomes

Study Intervention Groups Reference Group


(Pulse Pressure Less Than 45 mm Hg) (Pulse Pressure 50 mm Hg or Greater)
500-mL Bolus 1,500-mL Bolus 500-mL Bolus
Outcome (n5139) (n5137) P* (n5138) P†

Mode of delivery .91 .87


Spontaneous vaginal 118 (84.9) 118 (86.1) 115 (83.3)
Operative vaginal 11 (7.9) 9 (6.6) 16 (11.6)
Cesarean 10 (7.2) 10 (7.3) 7 (5.1)
Chorioamnionitis 6 (4.3) 8 (5.4) .60 5 (3.6) .68
Birth weight (g) 3,2376401 3,2486482 .89 3,2126437 .79
Apgar score less than 7 at 5 min 2 (1.44) 0 .50 0 .99
Cord gas
Arterial pH less than 7.10 2 (1.4) 4 (2.9) .45 1 (0.7) .18
Base excess less than 212.0 1 (0.7) 1 (0.7) 1.0 1 (0.7) .99
NICU admission 10 (7.2) 11 (8.0) .79 6 (4.4) .24
Duration (d) 6 (4–8) 8 (5–10) .52 9 (5–10) .78
Mechanical ventilation 1 0 1.0 0 .99
FHR, fetal heart rate; NICU, neonatal intensive care unit.
Data are n (%), mean6SD, or median (interquartile range) unless otherwise specified.
* Pairwise comparison of study intervention groups using x2, Fisher exact, or Wilcoxon rank-sum tests where appropriate.

Three-group comparison using Kruskal-Wallis test or analysis of variance.

volume are correlated with stroke volume falling as effect of providing a larger volume of IV fluid to
pulse pressure narrows.15 This relationship highlights women with a narrow pulse pressure. However, we
the underlying pathophysiology reflected by a narrow also acknowledge that various hemodynamic charac-
pulse pressure; namely, that lower central volume sta- teristics facilitate physiologic compensation in the
tus results in reduced cardiac preload and therefore setting of acute changes in vascular tone, sympathetic
decreased stroke volume by Starling mechanics. As outflow, and intravascular volume. As such, the
such, intrapartum women with a narrow pulse pres- relationship among maternal volume status, hemody-
sure may have relative hypovolemia with cardiac namic profile, and postepidural outcomes is complex
function on a steeper part of the Frank-Starling curve. and may not be comprehensively characterized by
Relative hypovolemia may become problematic in any single clinical predictor or variable.
the setting of pregnancy and neuraxial labor analge- The prevalence of postepidural FHR abnormalities
sia, in which uterine autoregulation of blood flow is in our trial was approximately twofold higher than that
limited6,7 and sympathetic blockade blunts the regu- seen by Miller et al.18 Miller defined postepidural FHR
lation of vascular tone. This confluence of physiologic abnormalities as “recurrent late decelerations and/or
factors likely predisposes women with a narrow pulse prolonged decelerations” in the 60 minutes after epidu-
pressure to postepidural FHR abnormalities or hypo- ral placement, whereas our primary outcome included
tension. Therefore, as supported by this trial, women any category 2 or 3 FHR pattern within the 60 minutes
with a narrow pulse pressure may improve their abil- after epidural dosing. We acknowledge that our pri-
ity to maintain systemic BP and uterine perfusion with mary outcome definition may result in an increased
greater volume loading at the time of initiation of prevalence of postepidural FHR abnormalities and
neuraxial analgesia. may be less specific for the identification of a clinically
The vital sign trajectories in this trial provide significant FHR pattern that requires resuscitation or
further evidence to support this underlying physiol- intervention; however, our definition improves gener-
ogy. Women with a narrow admission pulse pressure alizability and reproducibility. Variation in the rates of
who received 1,500 mL of fluid demonstrated signif- the primary outcome between studies could also reflect
icantly higher postepidural pulse pressure and systolic a difference in study design, anesthetic technique,
BP values as compared with those who received 500 patient characteristics, or regional variation in labor
mL. Notably, the pulse pressure and systolic BP management practices.
trends did not differ between the 1,500-mL group Our study has several strengths, including the
and the reference group, which implies a restorative prospective data collection and randomized study

1374 Lappen et al Pulse Pressure and Postepidural Complications OBSTETRICS & GYNECOLOGY

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
design with specific attention to control for potentially REFERENCES
confounding factors, particularly those related to 1. Osterman MJ, Martin JA. Epidural and spinal anesthesia use
maternal volume status such as BMI and the volume during labor: 27-state 257 reporting area, 2008. Natl Vital Stat
Rep 2011;59:1–13, 16.
of fluid received before epidural dosing; the inclusion
of a reference group as a negative control24; the exclu- 2. Traynor AJ, Aragon M, Ghosh D, Choi RS, Dingmann C,
Tran ZV, et al. Obstetric anesthesia workforce survey:
sion of abnormal FHR patterns before epidural place- a 30-year update. Anesth Analg 2016;122:1939–46.
ment; the determination of the primary outcome by
3. Anim-Somuah M, Smyth RM, Jones L. Epidural versus non-
an evaluator masked to study intervention arm; and epidural or no analgesia in labour. The Cochrane Database of
the use of objective and reproducible secondary out- Systematic Reviews 2011, Issue 12. Art. No.: CD000331.
come measures. We also acknowledge several limita- 4. Wolfler A, Salvo I, Sortino G, Bonati F, Izzo F. Epidural analgesia
tions of our study. As a pragmatic study, masking the with ropivacaine and sufentanil is associated with transient fetal
patients and the care teams to the study intervention heart rate changes. Minerva Anesthesiol 2010;76:340–5.
was not feasible. Notably, the rates of primary and 5. Nielsen PE, Erickson JR, Abouleish EL, Perriatt S, Sheppart C.
secondary outcome measures were concordant with Fetal heart rate changes after intrathecal sufentanil or epidural
bupivacaine for labor analgesia: incidence and clinical signifi-
observed vital sign trends, which supports the primary cance. Anesth Analg 1996;83:742–6.
study findings. Although the majority of women in the 6. Venuto RC, Cox JW, Stein JH, Ferris TF. The effect of changes in
study received an IV bolus as a coload (with epidural perfusion pressure on uteroplacental blood flow in the pregnant
initiation), some received the bolus as a preload (com- rabbit. J Clin Invest 1976;57:938–44.
pleted just before the epidural test dose). Data regard- 7. Greiss FC Jr. Pressure-flow relationship in the gravid uterine
ing optimal timing of the fluid bolus are inconsistent, vascular bed. Am J Obstet Gynecol 1966;96:41–7.
and a meta-analysis did not demonstrate any differ- 8. Warwick D, Ngan Kee BHB. Uteroplacental blood flow. In:
ence in the frequency of hypotension for women Chestnut DH, Wong CA, Tsen LC, Warwick D, Ngan Kee BHB,
Bellin Y, et al, editors. Chestnut’s obstetric anesthesia: princi-
receiving a coload compared with a preload before
ples and practice. 5th ed. Philadelphia (PA): Elsevier Saun-
cesarean delivery.25 Therefore, although we cannot ders; 2014;39–54.
exclude confounding, differences in bolus timing 9. Collins KM, Bevan DR, Beard RW. Fluid loading to reduce
unlikely affected our study results and this clinical abnormalities of fetal heart rate and maternal hypotension dur-
variation improves the generalizability of our study. ing epidural analgesia in labour. BMJ 1978;1460–1.
Additionally, we were unable to perform echocardi- 10. Check TG, Samuels P, Miller F, Tobin M, Gutsche BB. Normal
ography or other noninvasive assessments of volume saline i.v. fluid load decreases uterine activity in active labour.
status that would allow us to derive contemporaneous Br J Anaesth 1996;77:632–5.
estimates of stroke volume and cardiac output that 11. Kinsella SM, Pirlet M, Mills MS, Tuckey JP, Thomas TA. Ran-
domized study of intravenous fluid preload before epidural
could validate the relationship between pulse pressure analgesia during labour. Br J Anaesth 2000;85:311–3.
and these cardiac parameters and confirm pulse pres-
12. Kubli M, Shennan AH, Seed PT, O’Sullivan G. A randomised
sure as a surrogate for intravascular volume. How- controlled trial of fluid pre-loading before low dose epidural
ever, in a prospective cohort study of healthy, analgesia for labour. Int J Obstet Anesth 2003;12:256–60.
intrapartum women, we demonstrated a greater 13. Hofmeyr G, Cyna A, Middleton P. Prophylactic intravenous
hemodynamic response of the carotid artery to auto- preloading for regional analgesia in labour. The Cochrane Data-
transfusion after a passive leg raise among those with base of Systematic Reviews 2004, Issue 4. Art. No.: CD000175.
a narrow admission pulse pressure compared with 14. American Society of Anesthesiologists Task Force on Obstetric
a normal pulse pressure.26 Lastly, our study had lim- Anesthesia. Practice guidelines for obstetric anesthesia: an up-
dated report by the American Society of Anesthesiologists Task
ited power to detect any differences in the rates of Force on Obstetric Anesthesia. Anesthesiology 2007;106:843–63.
emergent operative delivery or adverse fetal out-
15. Convertino VA, Cooke WH, Holcomb JB. Arterial pulse pres-
comes, which are rare events in obstetric practice. sure and its association with reduced stroke volume during pro-
In conclusion, we found that the administration of gressive central hypovolemia. J Trauma 2006;61:629–34.
a 1,500-mL IV fluid bolus at the time of epidural 16. Convertino VA, Ryan KL, Rickards CA, Salinas J, McManus JG,
placement to women with a narrow admission pulse Cooke WH, et al. Physiological and medical monitoring for en
pressure significantly reduces the likelihood of post- route care of combat casualties. J Trauma 2008;64(4 suppl):S342–53.
epidural FHR abnormalities, maternal hypotension, 17. Bentzer P, Griesdale DE, Boyd J, MacLean K, Sirounis D, Ayas
NT. Will this hemodynamically unstable patient respond to
and resuscitative interventions. Admission pulse pres-
a bolus of intravenous fluids? JAMA 2016;316:1298–309.
sure, as a peripheral surrogate of central volume
18. Miller NR, Cypher RL, Nielsen PE, Foglia LM. Maternal pulse
status, can identify women who may benefit from pressure at admission is a risk factor for fetal heart rate changes
a larger volume infusion than is commonly adminis- after initial dosing of a labor epidural: a retrospective cohort
tered at the time of epidural placement. study. Am J Obstet Gynecol 2013;209:382.e1–8.

VOL. 130, NO. 6, DECEMBER 2017 Lappen et al Pulse Pressure and Postepidural Complications 1375

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
19. Executive summary: hypertension in pregnancy. American 23. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG.
College of Obstetricians and Gynecologists. Obstet Gynecol Research electronic data capture (REDCap)—a metadata-
2013;122:1122–31. driven methodology and workflow process for providing
20. Intrapartum fetal heart rate monitoring: nomenclature, inter- translational research informatics support. J Biomed Inform
pretation, and general management principles. ACOG Practice 2009;42:377–81.
Bulletin No. 106. American College of Obstetricians and 24. Arnold BF, Ercumen A. Negative control outcomes: a tool to
Gynecologists. Obstet Gynecol 2009;114:192–202. detect bias in randomized trials. JAMA 2016;316:2597–8.
21. Wilson M, Morganti AA, Zervoudakis I, Letcher RL, Romney BM, 25. Banerjee A, Stocche RM, Angle P, Halpern SH. Preload or
Von Oeyon P, et al. Blood pressure, the renin-aldosterone system, coload for spinal anesthesia for elective cesarean delivery:
and sex steroids throughout normal pregnancy. Am J Med 1980; a meta-analysis. Can J Anaesth 2010;57:24–31.
68:97–104. 26. Lappen JR, Shaman Z, Myers SA, Chien EK. Pulse pressure
22. Vricella LK, Louis JM, Mercer BM, Bolden N. Impact of mor- and carotid artery Doppler velocimetry as indicators of mater-
bid obesity on epidural anesthesia complications in labor. Am J nal volume status: a prospective cohort study. Am J Obstet
Obstet Gynecol 2011;205:370.e1–6. Gynecol 2017;216:S435–6.

1376 Lappen et al Pulse Pressure and Postepidural Complications OBSTETRICS & GYNECOLOGY

Copyright ª by The American College of Obstetricians


and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.