Keywords
neuraxial analgesia - very preterm infants - neonatal outcomes - necrotizing enterocolitis
- severe grade intraventricular hemorrhage
Neuraxial analgesia is one of the safest and most common methods of pain management
in labor, used by more than 70% of women who deliver in the United States.[1] These methods include spinal anesthesia, epidural analgesia, and combined spinal–epidural
analgesia, and have been associated with improved neonatal outcomes in comparison
to systemic opioids, including higher Apgar scores and lower incidence of naloxone
administration for neonatal respiratory depression.[2]
The maternal and neonatal safety of neuraxial analgesia has been well established
in both obstetric and anesthesia literature, with previous studies demonstrating that
maternal use of neuraxial analgesia is not associated with an increase in neonatal
morbidity in term neonates.[3]
[4]
[5]
[6]
[7] In a recent Cochrane review, neuraxial analgesia appeared to have no immediate adverse
effect on neonatal status by Apgar scores or neonatal intensive care unit (NICU) admissions.[8] Many studies have shown similar results for neonatal outcomes regardless of the
type of neuraxial analgesia used.[9]
Other studies have demonstrated a potential benefit of maternal neuraxial analgesia
for neonates delivering at term.[2]
[10]
[11]
[12]
[13] This effect is hypothesized to be related to the association between labor pain
and maternal hyperventilation, stress, and anxiety.[14] Maternal hyperventilation resulting from pain can lead to maternal hypoxemia and
respiratory alkalosis, thus shifting the oxyhemoglobin dissociation curve leftward
and increasing the affinity of oxygen for maternal hemoglobin, which subsequently
impairs the placental transfer of oxygen to the fetus.[14] Maternal respiratory alkalosis can result in uteroplacental vasoconstriction, subsequently
decreasing uterine blood flow and fetal oxygenation.[15] Neuraxial analgesia has been shown to alleviate the adverse hypoxemic effects associated
with labor pain, resulting in improved fetal oxygen delivery.[16]
[17]
[18] Furthermore, increased maternal stress and anxiety due to labor pain may result
in elevated levels of circulating maternal catecholamines, thereby lowering fetal
oxygenation and increasing fetal acidosis.[19] Neuraxial analgesia has been shown to decrease maternal concentrations of circulating
catecholamines.[20]
Although use of neuraxial analgesia has been shown to improve uteroplacental blood
flow and maternal and fetal hemodynamics related to labor pain, possibly improving
immediate outcomes in term neonates, there is little existing literature on the association
between neuraxial analgesia use and outcomes in preterm neonates. This raises a clinically
important question as this high-risk population may benefit even more from the physiologic
effects of neuraxial analgesia on improved uteroplacental blood flow, as compromised
neonates are already at increased risk of hypoxemia in the setting of prematurity.[21]
[22] The objective of this study is to evaluate the association between maternal use
of neuraxial analgesia for labor and/or delivery and neonatal outcomes in very preterm
infants.
Materials and Methods
This was a retrospective cohort study of women delivering singleton neonates between
23 and 32 weeks' gestation at a large volume academic center between 2012 and 2016.
Inclusion criteria included women who delivered singleton live births. Exclusion criteria
included women with multifetal gestations, major fetal anomalies, no intent to resuscitate,
intrauterine fetal demise, receipt of general anesthesia, and contraindications to
neuraxial analgesia including disseminated intravascular coagulopathy.
Women who received neuraxial analgesia encompassed those who received epidural analgesia,
spinal anesthesia, or combined spinal epidural analgesia for delivery. Neuraxial analgesia
at our institution is typically initiated with combined spinal epidural technique
using intrathecal bupivacaine and fentanyl. Maintenance labor analgesia is a standardized
infusion of bupivacaine and fentanyl administered through the epidural catheter via
either a continuous infusion or programmed intermittent epidural bolus. All data were
abstracted from the electronic medical record and reviewed by the coauthors.
Maternal demographic and clinical characteristics were analyzed by maternal use of
neuraxial analgesia before delivery. These characteristics included age at delivery,
nulliparity, obesity as defined by body mass index more than or equal to 30 kg/m2, maternal race and ethnicity, insurance status, gestational age at delivery, antepartum
admission, preterm labor, preterm premature rupture of membranes, preeclampsia and/or
hemolysis, elevated liver enzymes, low platelet count (HELLP) syndrome, gestational
diabetes or pregestational diabetes, delivery via cesarean including planned cesarean
delivery, and need for emergent delivery. Additional characteristics included receipt
of antibiotics prior to delivery for group B streptococcus prophylaxis, intra-amniotic
infection, or preterm premature rupture of membranes; receipt of magnesium prophylaxis
for fetal neuroprotection prior to delivery; receipt of a full course of antenatal
corticosteroids for fetal lung maturity prior to delivery; neonatal sex; neonatal
birthweight; clinical diagnosis of placental abruption; intra-amniotic infection;
presence of meconium; and induction of labor.
Neonatal outcomes of very preterm neonates born to women who used neuraxial analgesia
were compared to those born to women who did not receive neuraxial analgesia. These
outcomes included cord umbilical artery pH less than 7.0; 1 minute Apgar score less
than 7; 5 minute Apgar score less than 7; hypothermia (defined as core body temperature < 36.5 degrees
Celsius); need for vasopressors; NICU prolonged length of stay defined as greater
than 60 days; blood transfusion needed in the first week of life; prolonged mechanical
ventilation (defined as mechanical ventilation needed for more than 7 days); bronchopulmonary
dysplasia; necrotizing enterocolitis (NEC); late onset sepsis (defined as culture
positive sepsis after 3 days); severe grade III/IV intraventricular hemorrhage (IVH);
and neonatal death after delivery.
Maternal and neonatal characteristics and outcomes were compared using the Student's
t test or Mann–Whitney U test where applicable for continuous variables, and χ2 or Fisher's exact test where applicable for categorical variables. Multivariable
linear and logistic regression were used to assess the independent association between
maternal receipt of neuraxial analgesia and neonatal outcomes after adjusting for
demographic and clinical characteristics that were significant at p-value less than 0.05 on bivariable analysis. Two sensitivity analyses were performed.
In the first, propensity score analysis was performed between 10 and 90% of the overall
cohort in order to evaluate for outcomes data after controlling for outliers. In the
second, the same analyses were performed in the subgroup of patients who received
obstetric interventions, defined as the receipt of magnesium prophylaxis for fetal
neuroprotection, a full course of antenatal corticosteroids for fetal lung maturity,
and antibiotics prior to delivery. All tests were two-tailed and p-value less than 0.05 denoted significance. All statistical analyses were performed
with SPSS Statistical software (Version 22; IBM Corp, Armonk, NY). This study was
approved by the Institutional Review Board of Northwestern University.
Results
A total of 478 eligible women delivered very preterm singleton neonates during this
4-year study period. Of these, 352 (73.6%) received neuraxial analgesia prior to delivery.
Women who had neuraxial analgesia were more likely to deliver at a later preterm gestational
age, to have developed preeclampsia and/or HELLP, to have a higher neonatal birthweight,
to have undergone induction of labor, and to have delivered via cesarean delivery,
and were less likely to have a clinical diagnosis of abruption. These women were also
more likely to have received magnesium prophylaxis for fetal neuroprotection, a full
course of antenatal corticosteroids, and antibiotics prior to delivery ([Table 1]). There were otherwise no differences in the remaining maternal demographic and
clinical characteristics between the two groups.
Table 1
Cohort demographic and clinical characteristics associated with receipt of neuraxial
analgesia
|
Neuraxial analgesia (n = 352)
|
No neuraxial analgesia
(n = 126)
|
p-Value
|
|
Maternal age (years)
|
30.37 ± 6.53
|
31.36 ± 5.93
|
0.13
|
|
Nulliparous
|
198 (56.3%)
|
62 (49.2%)
|
0.18
|
|
Obese (BMI ≥ 30 kg/m2)
|
139 (39.5%)
|
44 (34.9%)
|
0.36
|
|
Race/ethnicity
|
0.16
|
|
Non-Hispanic White
|
122 (34.7%)
|
44 (34.9%)
|
|
|
Non-Hispanic Black
|
115 (32.7%)
|
29 (23.0%)
|
|
|
Hispanic
|
47 (13.4%)
|
21 (16.7%)
|
|
|
Other/unknown
|
68 (19.3%)
|
32 (25.4%)
|
|
|
Public insurance
|
213 (60.5%)
|
77 (61.1%)
|
0.91
|
|
Gestational age (weeks)
|
29.38 ± 2.30
|
28.9 ± 2.52
|
0.05
|
|
Preeclampsia/HELLP
|
111 (31.5%)
|
24 (19.0%)
|
0.008
|
|
Gestational or pregestational diabetes
|
23 (6.5%)
|
8 (6.3%)
|
0.94
|
|
Antepartum admission
|
285 (81.0%)
|
94 (74.6%)
|
0.13
|
|
Antibiotics received before birth
|
331 (94.0%)
|
110 (87.3%)
|
0.015
|
|
Magnesium prophylaxis
|
320 (90.9%)
|
100 (79.4%)
|
< 0.001
|
|
Full betamethasone course
|
310 (88.1%)
|
98 (77.8%)
|
0.005
|
|
Preterm labor
|
102 (29.0%)
|
38 (30.2%)
|
0.80
|
|
PPROM
|
120 (34.1%)
|
47 (37.3%)
|
0.52
|
|
Induction of labor
|
66 (18.8%)
|
8 (6.3%)
|
< 0.001
|
|
Cesarean delivery
|
154 (43.8%)
|
0 (0.0%)
|
< 0.001
|
|
Emergent delivery
|
22 (6.3%)
|
6 (4.7%)
|
0.76
|
|
Clinical diagnosis of abruption
|
40 (11.4%)
|
29 (23.0%)
|
< 0.001
|
|
Intra-amniotic infection
|
37 (10.5%)
|
10 (7.9%)
|
0.41
|
|
Meconium
|
16 (4.5%)
|
4 (3.2%)
|
0.51
|
|
Male infant
|
176 (50.0%)
|
56 (44.4%)
|
0.23
|
|
Birthweight (grams)
|
1,348.47 ± 481.44
|
1,256.56 ± 388.58
|
0.05
|
Abbreviations: BMI, body mass index; HELLP, hemolysis, elevated liver enzymes, low
platelet count; PPROM, preterm premature rupture of membranes.
Data reported as n (%) or mean ± standard deviation.
On bivariable analysis of neonatal outcomes, infants born to women who used neuraxial
analgesia had a decreased incidence of NEC, severe grade IVH, cord umbilical artery
pH less than 7.0, and NICU length of stay for more than 60 days ([Table 2]). After adjusting for maternal demographic and clinical characteristics, use of
maternal neuraxial analgesia remained independently associated with a decreased odds
of NEC (adjusted odds ratio [aOR]: 0.28, 95% confidence interval [CI]: 0.12–0.62)
and grade III/IV IVH (aOR: 0.33, 95% CI: 0.13–0.87). These associations remained significant
even after adjusting for cesarean delivery, which was considered a clinically and
statistically significant confounder, as all patients who underwent cesarean delivery
received neuraxial anesthesia ([Table 2]).
Table 2
Neonatal outcomes associated with maternal receipt of neuraxial analgesia
|
Neuraxial analgesia (n = 352)
|
No neuraxial analgesia
(n = 126)
|
p-Value
|
aOR[a]
|
95% CI
|
|
Cord umbilical artery pH < 7.0
|
87 (24.7%)
|
44 (34.9%)
|
0.03
|
0.70
|
0.44–1.13
|
|
1 minute Apgar score < 7
|
191 (54.3%)
|
80 (63.5%)
|
0.07
|
|
|
|
5 minutes Apgar score < 7
|
75 (21.3%)
|
37 (29.4%)
|
0.07
|
|
|
|
Hypothermia (core temperature < 36.5 °C)
|
18 (5.1%)
|
7 (5.6%)
|
0.85
|
|
|
|
Need for vasopressors
|
18 (5.1%)
|
10 (7.9%)
|
0.25
|
|
|
|
NICU length of stay > 60 days
|
125 (35.5%)
|
61 (48.4%)
|
0.01
|
0.59
|
0.33–1.05
|
|
Blood transfusion needed during 1st week of life
|
76 (21.6%)
|
36 (28.6%)
|
0.11
|
|
|
|
Mechanical ventilation > 7 days
|
57 (16.2%)
|
27 (21.4%)
|
0.19
|
|
|
|
Bronchopulmonary dysplasia
|
86 (24.4%)
|
36 (28.6%)
|
0.36
|
|
|
|
Necrotizing enterocolitis
|
16 (4.5%)
|
16 (12.7%)
|
0.002
|
0.28
|
0.12–0.62
|
|
Late onset sepsis
|
18 (5.1%)
|
9 (7.1%)
|
0.40
|
|
|
|
Grade III/IV IVH
|
11 (3.1%)
|
11 (8.7%)
|
0.01
|
0.33
|
0.13–0.87
|
|
Neonatal death
|
10 (2.8%)
|
5 (4.0%)
|
0.53
|
|
|
Abbreviations: aOR, adjusted odds ratio; CI, confidence interval; HELLP, hemolysis,
elevated liver enzymes, low platelet count; IVH, intraventricular hemorrhage; NICU,
neonatal intensive care unit.
Note: Data reported as n (%) or mean ± standard deviation.
a Adjusted for factors with p < 0.05 on bivariable analyses, including gestational age, preeclampsia/HELLP, antibiotics
received before birth, magnesium prophylaxis received before birth, full betamethasone
course received before birth, birthweight, clinical diagnosis of abruption, induction
of labor, and cesarean delivery.
Finally, on sensitivity analysis that was performed on 10 to 90% of the overall cohort
to further control for outlier data, neuraxial analgesia remained significant in predicting
decreased incidence of NEC (aOR: 0.17, 95% CI: 0.03–0.84) and grade III/IV IVH (aOR:
0.29, 95% CI: 0.02–0.89). An additional sensitivity analysis was performed for the
83.8% of patients who received obstetric interventions. Neuraxial analgesia remained
significantly associated with decreased odds of NEC (aOR: 0.41, 95% CI: 0.14–0.77)
and grade III/IV IVH (aOR: 0.43, 95% CI: 0.12–0.84) in this restricted population
of women for whom precipitous delivery did not occur. Neuraxial analgesia did not
remain significantly associated with decreased incidence of cord umbilical artery
pH less than 7.0 or NICU length of stay for more than 60 days on either of the sensitivity
analyses performed.
Comment
Neuraxial analgesia is one of the most commonly employed techniques for pain control
in labor, and understanding its role in the very preterm parturient is essential.
In this observational study of women who delivered before 32 weeks, maternal use of
neuraxial analgesia may be associated with improved neonatal outcomes, such as decreased
odds of NEC and severe grade IVH, even after controlling for existing interventions
for prematurity and mode of delivery. Prior work has suggested such effects may be
due to improved neonatal acid–base status from changes in maternal physiology as a
result of adequate labor analgesia.[16]
[17]
[18]
Although there were some differences in clinical characteristics between women who
used neuraxial analgesia and those who did not, the majority of these differences
in potential confounders can be clinically explained. As expected, all women who underwent
cesarean delivery received neuraxial epidural or spinal anesthesia as general anesthesia
was excluded in this study. Neuraxial analgesia use is also expected to be higher
in women undergoing induction of labor compared to those who present in active spontaneous
labor. Women admitted to the antepartum service with a planned iatrogenic preterm
delivery may be encouraged to ask for neuraxial analgesia during induction of labor
in anticipation of prolonged labor or need for emergent cesarean delivery. Furthermore,
antepartum patients who have the time to receive neuraxial analgesia prior to delivery
may have longer labor courses allowing time for more neonatal protective measures
such as the administration of antibiotics, magnesium, and corticosteroids. Although
women who received neuraxial analgesia were less likely to have a clinical diagnosis
of abruption, this is likely because abruption may preclude epidural use due to concern
for disseminated intravascular coagulation, even if not overtly diagnosed.
Neuraxial analgesia may also be recommended by providers to women with preeclampsia
or HELLP without coagulopathy, as it is generally preferred over general anesthesia
due to the potential for difficult airways from soft tissue or laryngeal edema in
patients with hypertensive disorders of pregnancy.[23] Patients with preeclampsia or HELLP may be encouraged to ask for epidural analgesia
earlier in their labor course due to the risk of developing more severe forms of coagulopathy
as the disease progresses.
Importantly, maternal use of neuraxial analgesia remained independently associated
with decreased odds of NEC and grade III/IV IVH in very preterm neonates after controlling
for differences in clinical characteristics and mode of delivery on multivariable
analysis, and remained significant on sensitivity analyses, which were performed to
attempt to limit the effects of outliers or the potential unmeasured confounding introduced
by precipitous delivery. As discussed previously in the literature on term neonates,
maternal neuraxial analgesia may improve neonatal outcomes such as Apgar scores by
alleviating maternal pain-related hypoxemia and improving fetal oxygenation. These
hemodynamic benefits may extrapolate to even more clinically relevant outcomes in
preterm neonates as this population may be more predisposed or at higher risk of hypoxemia
due to the etiology and/or consequences of their prematurity.[21]
[22]
This study is unique in that it is one of the first to address the association between
maternal neuraxial analgesia and outcomes in very preterm neonates. A strength of
this analysis is that it was performed at a large academic center with a diverse,
high-risk patient population. Additionally, antepartum care for women at risk of delivering
very preterm neonates is managed via standardized protocols with regard to use of
existing interventions to improve neonatal outcomes of prematurity, such as administration
of antenatal corticosteroids, magnesium prophylaxis for fetal neuroprotection, and
intrapartum antibiotics. Moreover, unlike large-scale population databases, our data
include highly granular details about parturients and their neonates, with the ability
to corroborate all details directly from the electronic medical record.
Our study must be examined within the context of its limitations. This study was conducted
at a large academic center with around the clock in-house availability of obstetric
anesthesia services, and thus may not be generalizable to different practice settings.
Because this is an observational study, it is impossible to fully account for all
potential confounders. While we attempted to control for potential confounders, such
as receipt of existing interventions for prematurity, gestational age at delivery,
and mode of delivery, it is possible that there are other factors that we did not
account for which may improve neonatal outcomes independently of maternal neuraxial
analgesia use. For example, although there was no difference in emergent delivery
for maternal or fetal compromise between the two groups, it is possible that mothers
who did not use neuraxial analgesia delivered more compromised neonates due to precipitous
preterm labor or inadequate prenatal and antenatal care. Consequently, women who did
not use neuraxial analgesia may have been less likely to receive magnesium prophylaxis,
complete a full course of antenatal corticosteroids, and receive antibiotics due to
precipitous preterm delivery as a potential unmeasured confounder, which we attempted
to account for in our sensitivity analysis. Furthermore, the effect of neuraxial analgesia
may also be influenced by unmeasured clinical characteristics such as the overall
duration of preterm labor, the etiology of preterm birth, the degree of maternal pain,
and the use of additional analgesic therapies such as intravenous opioids. Additionally,
there was limited power to detect differences in some rare outcomes. Finally, as a
retrospective cohort study, an inherent limitation of this work lies in its inability
to determine a causal relationship between these factors. Nonetheless, since randomization
of women with elevated risk of very preterm birth to receive versus not receive neuraxial
analgesia would likely be unacceptable, we propose that this observational investigation
represents an important perspective on an important unanswered question, despite these
limitations.
Further research is required to prospectively investigate this association in very
preterm neonates, although the known relationship between neuraxial analgesia and
improved uteroplacental blood flow suggests there may be a physiological benefit for
neonatal oxygenation. Our study provides initial support for the potential added benefit
of neuraxial analgesia on improved neonatal outcomes in this high-risk population.
