Although consensus guidelines on the management of cardiovascular disease in pregnancy reserve cesarean delivery for obstetric indications, there is a paucity of data to support this approach.
The objective of the study was to compare cardiovascular and obstetric morbidity in women with cardiovascular disease according to the plan for vaginal birth or cesarean delivery.
We assembled a prospective cohort of women delivering at an academic tertiary care center with a protocolized multidisciplinary approach to management of cardiovascular disease between September 2011 and December 2016. Our practice is to encourage vaginal birth in women with cardiovascular disease unless there is an obstetric indication for cesarean delivery. We allow women attempting vaginal birth a trial of Valsalva in the second stage with the ability to provide operative vaginal delivery if pushing leads to changes in hemodynamics or symptoms. Women were classified according to planned mode of delivery: either vaginal birth or cesarean delivery. We then used univariate analysis to compare adverse outcomes according to planned mode of delivery. The primary composite cardiac outcome of interest included sustained arrhythmia, heart failure, cardiac arrest, cerebral vascular accident, need for cardiac surgery or intervention, or death. Secondary obstetric and neonatal outcomes were also considered.
We included 276 consenting women with congenital heart disease (68.5%), arrhythmias (11.2%), connective tissue disease (9.1%), cardiomyopathy (8.0%), valvular disease (1.4%), or vascular heart disease (1.8%) at or beyond 24 weeks’ gestation. Seventy-six percent (n = 210) planned vaginal birth and 24% (n = 66) planned cesarean delivery. Women planning vaginal birth had lower rates of left ventricular outflow tract obstruction, multiparity, and preterm delivery. All women attempting vaginal birth were allowed Valsalva. Among planned vaginal deliveries 86.2% (n = 181) were successful, with a 9.5% operative vaginal delivery rate. Five women underwent operative vaginal delivery for the indication of cardiovascular disease without another obstetric indication at the discretion of the delivering provider. Four of these patients tolerated trials of Valsalva ranging from 15 to 75 minutes prior to delivery. Adverse cardiac outcomes were similar between planned vaginal birth and cesarean delivery groups (4.3% vs 3.0%, P = 1.00). Rates of postpartum hemorrhage (1.9% vs 10.6%, P < .01) and transfusion (1.9% vs 9.1%, P = .01) were lower in the planned vaginal birth group. There were no differences in adverse cardiac, obstetric, or neonatal outcomes in the cohort overall or the subset of women with high-risk cardiovascular disease or a high burden of obstetric comorbidity.
These findings suggest that cesarean delivery does not reduce adverse cardiovascular outcomes and lend support to a planned vaginal birth for the majority of women with cardiovascular disease including those with high-risk disease.
Improvements in cardiovascular care have led to an increased prevalence of cardiovascular disease (CVD) in pregnancy with a rising contribution of CVD to maternal mortality. Women with CVD in pregnancy are at an increased risk of adverse obstetric and cardiac events. While risk factors and scoring systems for adverse events in women with CVD have been described, most studies fail to explore the impact of management strategies on the rates of these outcomes.
Why was this study conducted?
The study was conducted to compare cardiovascular and obstetric morbidity in women with cardiovascular disease according to plan for vaginal birth or cesarean delivery.
Adverse cardiac outcomes were similar between women planning vaginal birth and those planning cesarean delivery (4.3% vs 3.0%, P = 1.00). Rates of postpartum hemorrhage (1.9% vs 10.6%, P < .01) and transfusion (1.9% vs 9.1%, P = .01) were lower in the planned vaginal birth group. All women tolerated a trial of Valsalva without hemodynamic compromise or adverse cardiac events.
What does this add to what is known?
This prospective cohort study from a contemporary patient population managed by a multidisciplinary team with a standardized approach to care demonstrates the safety of attempted vaginal birth including a trial of Valsalva for women with cardiac disease in pregnancy.
Planned mode of delivery is an important decision for pregnant women with CVD and their providers. Consensus guidelines typically reserve cesarean delivery for obstetric indications with specific exceptions for high-risk disease. However, rates of cesarean delivery for women with CVD are higher than those in the general population with a 33% rate of primary cesarean delivery for cardiac indications in recent literature.
For women attempting vaginal birth, the role of operative vaginal delivery to minimize Valsalva may actually increase maternal morbidity. These consensus guidelines are based primarily on expert opinion with a paucity of data to support the recommendations.
Although decreasing rates of primary cesarean delivery is a national priority, the impact of mode of delivery on current and subsequent pregnancies for women with CVD is largely unexplored. We undertook this study to examine the association between planned mode of delivery and maternal morbidity in a population of women with CVD. We hypothesize that in these women planned vaginal birth confers no increased risk of obstetric or cardiovascular morbidity compared with planned cesarean delivery.
Materials and Methods
We prospectively enrolled pregnant women with CVD receiving care at the Brigham and Women’s Hospital (Boston, MA) from September 2011 through December 2016 in the Standardized Outcomes In Reproductive Cardiovascular Care initiative. The Institutional Review Boards of the Brigham and Women’s Hospital and Boston Children’s Hospital approved this protocol, and each patient provided written informed consent.
Data were collected prospectively at each clinic visit, during all pregnancy admissions, and through delivery and postpartum care. We categorized women according to their underlying CVD into 1 of the following groups: congenital heart disease, connective tissue disease, cardiomyopathy, valvular disease, vascular disease, and arrhythmia.
Women were classified as having high-risk CVD with a history of any of the following criteria or if they developed these high-risk features during pregnancy: New York Heart Association function class >II, oxygen saturation <90%, systemic ejection fraction <40%, left ventricular outflow tract (LVOT) peak gradient >30 mm Hg, subpulmonary ejection fraction <40%, or aortic conditions associated with connective tissue disease.
We discussed each woman at a monthly multidisciplinary conference that included representatives from maternal-fetal medicine, cardiology, obstetric anesthesia, and nursing. Delivery discussions included location (labor and delivery, cardiac intensive care unit [ICU], cardiac catheterization laboratory, hybrid cardiac operating room), monitoring (telemetry, invasive hemodynamic monitoring, continuous pulse oximetry), timing, and mode of delivery (vaginal delivery, cesarean delivery, or vaginal delivery with assisted second stage with vacuum or forceps).
Our practice is to encourage vaginal birth in women with CVD unless there is an obstetric indication for cesarean delivery. The ultimate mode of delivery was at the discretion of the delivering obstetrician.
For this study we classified women according to their planned mode of delivery into 1 of 2 groups: planned vaginal birth or planned cesarean delivery. Patients in the planned cesarean delivery group were further classified by obstetric eligibility for planned vaginal birth. Women with placenta previa, malpresentation, prior uterine surgery, or maternal or fetal instability prohibiting labor were considered ineligible for planned vaginal birth.
Peripartum records were independently reviewed by 2 obstetricians (S.R.E., C.E.R.) with complete ascertainment as to both the planned and actual modes of delivery. Actual mode of delivery was defined as vaginal or cesarean delivery. Vaginal delivery was divided into spontaneous vaginal delivery or operative vaginal delivery (either vacuum- or forceps-assisted vaginal delivery). The indication for the planned cesarean deliveries, operative vaginal deliveries, and unplanned cesarean deliveries in the attempted vaginal birth group were noted.
Maternal cardiac, obstetric, and neonatal covariates were collected using standard obstetric definitions unless otherwise noted. We noted a history of an adverse event in the pregnancy of interest based on our primary outcome definition in the following text. To better characterize the burden of medical comorbidity in the cohort, patient comorbidities were collected and tallied using the obstetric comorbidity index developed by Bateman et al. This validated index accounts for medical and obstetric comorbidities with higher scores predictive of admission to the ICU and adverse maternal outcomes.
For the present analysis, the primary outcome of interest was a composite outcome of peripartum cardiovascular morbidity. The composite cardiac outcome consisted of 1 or more of the following: congestive heart failure (diagnosed by physical examination and requiring diuresis), sustained symptomatic arrhythmia (>30 seconds in duration or requiring therapy), a cerebral vascular event, new or worsening valvar dysfunction, endocarditis, aortic dissection, need for cardiac intervention, cardiac arrest, and cardiac death.
The objective of the present study was to determine the impact of attempted mode of delivery on this primary composite cardiovascular outcome. Therefore, the primary outcome was considered present only if it occurred in the peripartum or 6 week postpartum period and was not present at the time of admission for delivery.
Secondary outcomes of interest included the mode of delivery, a composite outcome of obstetric morbidity, maternal ICU admission, the occurrence of severe maternal morbidity, admission to the neonatal intensive care unit, and a composite outcome of neonatal morbidity.
The composite obstetric outcome was based on the primary outcome set forth in the Maternal-Fetal Medicine University Networks Assessment of Perinatal Excellence (APEX) study with some modifications. The composite obstetric outcome included venous thromboembolism (including deep venous thrombosis), postpartum hemorrhage, and peripartum infection during the first 6 weeks postpartum.
Severe maternal morbidity was defined using the recent consensus guidelines set forth by the American College of Obstetricians and Gynecologists and the Society for Maternal-Fetal Medicine. As with the primary outcome, both severe maternal morbidity and maternal ICU admission were considered present only if they occurred in the peripartum period and were absent at the time of admission for delivery. We did not include third- or fourth-degree perineal lacerations involving the anal sphincter in the composite obstetric outcome despite their association with short- and long-term morbidity in line with using the recent consensus guidelines set forth by the American recommendations and the APEX composite outcome definition. These rates are reported separately. The composite neonatal outcome was similarly based on the work of the APEX study and, along with neonatal intensive care unit admission, limited to term, singleton, nonanomalous fetuses.
The associations between intended mode of delivery and maternal, fetal, or obstetric characteristics were evaluated with the χ 2 test or Fisher exact test for categorical variables or the Wilcoxon rank-sum test for continuous variables. Statistical significance was defined with a 2-tailed value of P < .05.
The heterogeneity of CVD and maternal comorbidity in the cohort coupled with the low rate of adverse outcomes limited our statistical power to perform adjusted analyses controlling for multiple covariates. We therefore conducted multiple preplanned sensitivity analyses examining outcomes for the subset of women with CVD and for those with an obstetric comorbidity index above 7. All analyses were performed with Statistical Analysis Software (SAS), version 9.4 (Copyright 2013; SAS Institute, Inc, Cary NC).
Figure 1 shows the identification and categorization of patients for the current study. Two-hundred seventy-six women met inclusion criteria and delivered in the time period of interest. Of these women, 76.1% (n = 210) planned a vaginal birth and 23.9% (n = 66) planned cesarean delivery. As shown in Figure 1 , the majority of patients in the cohort had congenital heart disease (68.5%), followed by arrhythmias (11.2%), connective tissue disease (9.1%), cardiomyopathy (8.0%), valvular disease (1.4%), or vascular heart disease (1.8%) The specific cardiac diagnoses are detailed in the Supplemental Table . There was no difference in the type of heart disease between women attempting vaginal birth and those planning cesarean delivery ( P = .69). After reviewing each patient in our multidisciplinary conference, no patients had a plan for cesarean delivery specifically for the indication of cardiac disease.
The demographic and clinical characteristics of eligible women according to planned mode of delivery are shown in Table 1 . Women attempting vaginal birth had lower rates of elevated LVOT peak gradients compared with those with planning cesarean delivery (4.8% vs 13.6%, P = .02). Five of the 9 women in the group of patients with elevated LVOT peak gradients planning cesarean delivery were elective repeat cesarean deliveries after being denied a trial of labor at other institutions during their first pregnancies. Rates of adverse cardiac events before and during pregnancy were similar between groups.
|Characteristics||Total (n = 276)||Planned vaginal birth (n = 210)||Planned cesarean delivery (n = 66)||P value a|
|Maternal age, y b||32.6 (29.2–35.4)||32.4 (29.2–35.0)||33.5 (29.4–37.5)||.06|
|White, not Hispanic||193 (69.9)||142 (67.6)||51 (77.3)|
|Black||21 (7.6)||17 (8.1)||4 (6.1)|
|Hispanic, white||30 (10.9)||23 (10.9)||7 (10.6)|
|Hispanic, Nonwhite||3 (1.1)||3 (1.4)||0 (0)|
|Asian||12 (4.4)||12 (5.7)||0 (0)|
|Other/declined||17 (6.2)||13 (6.2)||4 (6.1)|
|Smoking||20 (7.3)||14 (6.7)||6 (9.1)||.59|
|Prepregnancy BMI ≥30 kg/m 2||62 (22.5)||44 (21.0)||18 (27.3)||.28|
|Chronic hypertension||12 (4.4)||7 (3.3)||5 (7.6)||.17|
|Pregestational diabetes||3 (1.1)||2 (0.95)||1 (1.5)||.56|
|Prior adverse event||39 (14.1)||30 (14.3)||9 (13.6)||.89|
|High-risk cardiac disease c|
|Connective tissue disease||25 (9.1)||18 (8.6)||7 (10.6)||.62|
|NYHA class >II||6 (2.2)||3 (1.4)||3 (4.5)||.15|
|Oxygen saturation <90%||1 (0.37)||0 (0)||1 (1.5)||.24|
|Systemic EF <40%||8 (2.9)||5 (2.4)||3 (4.6)||.40|
|LVOT peak gradient >30 mm Hg||19 (7.0)||10 (4.8)||9 (13.6)||.02|
|Adverse event in pregnancy d||15 (5.4)||8 (3.8)||7 (10.6)||.06|
a P value was calculated by a χ 2 test or Fisher exact test for categorical variables unless otherwise noted
b Continuous variables are presented as median (interquartile range) with the P value calculated by Wilcoxon rank-sum test
c High-risk CVD was defined as 1 or more of the following risk factors: prior adverse event, NYHA class >II, oxygen saturation <90%, systemic EF <40%, and LVOT peak gradient >30 mm Hg, subpulmonary EF <40%, or connective tissue disease
d Adverse event in pregnancy includes heart failure, sustained symptomatic arrhythmia, new or worsening valvar dysfunction, endocarditis, aortic dissection, need for cardiac intervention, cardiac arrest, and cardiac death not associated with delivery.
Table 2 demonstrates the obstetric and intrapartum characteristics for women in the cohort. Rates of nulliparity (52.9% vs 28.8%, P < .01) were higher in the vaginal delivery group compared with the cesarean delivery group. Women attempting vaginal birth were less likely to be preterm (11.9% vs 28.8%, P < .01) compared with their cesarean delivery counterparts. Other obstetric features including rates of preeclampsia or gestational hypertension, gestational diabetes, fetal growth restriction, and fetal anomalies were similar between the 2 groups. The planned cesarean delivery group had a higher median obstetric comorbidity index score with similar rates of obstetric comorbidity indices greater than 7.
|Characteristics||Total (n = 276)||Planned vaginal birth (n = 210)||Planned cesarean delivery (n=66)||P value a|
|Nulliparous||130 (47.1)||111 (52.9)||19 (28.8)||< .01|
|Gestational age <37 wks||44 (15.9)||25 (11.9)||19 (28.8)||< .01|
|Spontaneous conception||247 (89.5)||189 (90.0)||58 (87.9)||.62|
|Preeclampsia or gestational hypertension||25 (9.1)||18 (8.6)||7 (10.6)||.62|
|Gestational diabetes||19 (6.9)||14 (6.7)||5 (7.6)||.78|
|Intrauterine growth restriction||38 (13.8)||28 (13.3)||10 (15.2)||.69|
|Fetal anomalies||21 (7.6)||16 (7.6)||5 (7.6)||1.00|
|Multiple gestation||3 (1.1)||2 (0.95)||1 (1.5)||.56|
|Neuraxial analgesia||261 (94.6)||199 (94.8)||62 (93.9)||.76|
|Obstetric comorbidity index b||4 (3–5)||4 (3–5)||5 (3–6)||< .01|
|Obstetric comorbidity index >7||11 (4.0)||6 (2.9)||5 (7.6)||.14|