Pulmonary Hypertension and Pregnancy Outcomes: Insights From the National Inpatient Sample
Background Pregnant women with pulmonary hypertension (PH) are at risk for adverse cardiac outcomes, particularly at the time of labor and delivery. The purpose of this study is to define the impact of PH on pregnancy outcomes and the risk of major adverse cardiac events (MACE).
Methods and Results The National Inpatient Sample was screened for hospital admissions of women delivering during the years 2003 to 2012. The primary outcome was MACE, a composite of death, cardiac arrest, cardiogenic shock, myocardial infarction, respiratory failure, arrhythmia, stroke, and embolic event. Data on 1519 patients with PH and 6 757 582 without heart disease or PH were available. There were 59.6% with isolated PH; 10.7% with PH and congenital heart disease; 18.1% with PH and valvular heart disease; 3% with PH and valvular heart disease and congenital heart disease; 6.6% PH and cardiomyopathy; and 1.9% with PH and cardiomyopathy and valvular heart disease. Compared with women without heart disease or PH, women with PH experienced significantly higher MACE (24.8 versus 0.4%, P<0.0001). Among the subsets of women with PH, the highest MACE was noted in women with the combination of PH and cardiomyopathy and valvular heart disease, and PH and cardiomyopathy, primarily because of heart failure and arrhythmia. Women with PH were significantly more likely to experience eclampsia syndromes, preterm delivery, and intrauterine fetal demise (P<0.0001 for all). PH subtype was significantly associated with MACE in multivariable analysis (P<0.001).
Conclusions In a contemporary data set of pregnant women in the United States, PH was associated with an increase in MACE during the hospitalization for delivery, with an exceptionally elevated risk among women with associated cardiomyopathy.
What Is New?
Pregnant women with pulmonary hypertension (PH) are at risk for adverse cardiac outcomes, particularly at the time of labor and delivery.
PH was associated with an increase in major adverse cardiac events during the hospitalization for delivery compared with those without, largely related to heart failure and arrhythmia.
Maternal mortality was low overall, and lower than previously reported, but multifold higher than women without PH. PH associated with left heart disease experience substantially higher adverse cardiac events than isolated PH or PH associated with congenital heart disease.
What Are the Clinical Implications?
Women with PH are at substantial risk during pregnancy and delivery.
The etiology of PH as well as its association with other forms of heart disease, particularly left heart disease, should be included in risk stratification before and during pregnancy.
Women with PH who become pregnant warrant a multidisciplinary approach for management of pregnancy, labor, and delivery.
Prepregnancy counseling can help inform patients and providers regarding their reproductive choices and substantial risks involved.
Pulmonary hypertension (PH) is a disease characterized by debilitating symptoms and an overall diminished life expectancy because of narrowing of the pulmonary vasculature, often leading to right heart failure. Pregnancy in women with PH has long been regarded as high risk for maternal and neonatal complications, including maternal death,1, 2, 3 particularly during labor and delivery and the period immediately postpartum. Often, PH is considered a contraindication to pregnancy.4, 5, 6 This is largely because of an inability of the right ventricle to adapt to the volume shifts and hemodynamic changes associated with pregnancy, labor, and delivery, such as the marked increase in plasma volume and cardiac output, decrease in systemic vascular resistance, and increased susceptibility to thromboembolic events because of hypercoagulability.7 Recent data from the United States suggest that incidence of pregnancy in women with PH is increasing,8 a rather alarming finding given the risks in pregnancy. Women with PH do become pregnant,1, 8 whether planned or unplanned, and may choose to continue with a pregnancy often despite adequate counseling. A substantial number of women are opting to enter into such pregnancies, likely because of recent advancements in diagnosis and medical therapies that have allowed more women with PH to live through child‐bearing age, as well as a growing number of adults with congenital heart disease who may be complicated by PH.
PH can be primary in pathogenesis or the result of congenital heart disease (CHD), valvular heart disease (VHD), cardiomyopathy, or related to chronic thromboembolic disease or other systemic disease.9 We aimed to characterize the adverse maternal cardiac outcomes in pregnant women with PH resulting from multiple different causes or associated conditions during the hospitalization for delivery in the United States and to determine the risk factors associated with adverse maternal cardiac events.
We utilized data from the 2003 to 2012 National Inpatient Sample (NIS), collected by the Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project, which is the largest all‐payer inpatient publicly available database in the United States. The NIS provides annual information on approximately 8 million inpatient stays estimated from a stratified sample of hospitals (2003–2011) or hospital discharges (2012) that are nationally representative of all hospital discharges from US acute hospitals. Discharge records related to labor and delivery from years 2003 to 2012 were included in the study. International Classification of Diseases, Ninth Revision (ICD‐9) codes were utilized to ascertain hospitalizations for delivery, defined as any discharge record with a normal delivery or other indications for care in pregnancy‐, labor‐, and delivery‐related diagnoses or delivery‐related procedure as previously described (Data S1).10 Our analysis was limited to delivery‐associated hospitalizations. Clinical Classifications Software codes for ICD‐9 were utilized to identify the VHD and CHD cohorts11; for PH and cardiomyopathy, ICD‐9 codes were used. The remaining codes utilized in the study can be found in Data S1. Because the NIS consists of publicly available de‐identified data files, this current study was exempted from Institutional Review Board approval, per the US Department of Health and Human Services guidelines.
From 2003 to 2012, a total of 8 438 681 records were extracted for delivery‐related analysis (Figure 1). Deliveries with missing data on race or insurance were excluded (n=1 679 574). Deliveries with subtype PH+CHD+cardiomyopathy or PH+CHD+cardiomyopathy+VHD were also excluded because NIS does not allow for reporting of events with n<10 so as to protect individuals’ privacy, as <10 patients were available for study. Similarly, for records with multiple codes for eclampsia status, eclampsia status was assigned according to the priority rule: eclampsia complication>severe preeclampsia>mild preeclampsia>preeclampsia/eclampsia with preexisting hypertension, in order to divide the levels of preeclampsia present in cases where multiple diagnoses were coded.
Patient Characteristics and Outcome Measures
All patient and hospital characteristics were obtained from the NIS. Demographics, medical history, and other covariate data associated with hospital admission extracted included maternal age, race, and insurance status. Delivery at teaching hospital, location and region of hospital, length of stay, and total hospital charges, which represent absolute total hospital charge, were extracted. The primary outcome of interest was major adverse cardiac events (MACE), defined as a composite of outcomes that relate to cardiac complications of delivery and include in‐hospital death, acute myocardial infarction, heart failure, arrhythmia, cerebrovascular events, pulmonary embolism, arterial embolism, atheroembolism, obstetrical pulmonary embolism, and cardiac complications of anesthesia or other sedation in labor and delivery as previously described.10 Potential confounding factors for MACE included diabetes mellitus, multiple gestation, transient hypertension of pregnancy, eclampsia status (including mild preeclampsia, severe preeclampsia, eclampsia complicating pregnancy/childbirth, preeclampsia/eclampsia with preexisting hypertension, postpartum hemorrhage, acute renal failure, cardiorespiratory failure or shock, respiratory failure, and cesarean delivery.
Data were summarized by descriptive statistics. For all descriptive analysis, χ2 tests with exact P values based on Monte Carlo simulation were used to compare unadjusted marginal differences for categorical variables among groups. Welch's ANOVA tests were used to compare unadjusted marginal differences for continuous variables such as age, length of stay, and total charges. Predictors of adverse cardiac outcomes examined included variables related to patients’ demographic information, specific comorbidities, and pregnancy‐related complications. Variables with a P<0.1 in univariate descriptive analyses, together with age, race, and PH type were included in the multivariable logistic regression. Firth bias correction was used because of sparse count issue.12 A P value <0.05 was considered statistically significant and all analysis was performed using SAS 9.3.13
Data on 1519 patients with PH and 6 757 582 without heart disease or PH were available for study (Figure 2). Among the 1519 pregnant women diagnosed with PH, 906 had isolated PH (59.7%). The remainder of the sample had combined disease patterns as follows: 163 (10.7%) had PH with CHD (PH+CHD); 275 (18.1%) had PH with VHD (PH+VHD); 101 (6.6%) had PH with cardiomyopathy; 45 (3.0%) had PH along with VHD and CHD (PH+VHD+CHD); and 29 (1.9%) had PH along with cardiomyopathy and VHD.
Demographic and clinical characteristics of women with PH and those without heart disease or PH were obtained (Table 1). Compared with women with no heart disease (n=6 757 582), patients with PH were on average older by 2 years, more than twice as likely to be black, had twice as long hospital stays, accumulated greater than 3‐fold charges, were more likely to be insured by Medicare or Medicaid, and were more likely to deliver at a teaching hospital (all P<0.0001). They were also more likely to have multiple gestations and be diabetic. Associated comorbidities are listed in Table 1, and include a higher likelihood of having associated systemic lupus erythematosus, sickle cell disease, chronic thromboembolic disease, obstructive sleep apnea, and preexisting hypertension (all P<0.0001). Chronic thromboembolic disease was a common comorbidity, present in 87.9% of cases. A small group of women had coexisting Eisenmenger syndrome (3%). Of the women with isolated PH, 88 (5.8%) had a diagnosis of primary/idiopathic PH. Among the subclasses of PH, those with PH+cardiomyopathy had the longest length of stay and those with PH+cardiomyopathy+VHD had the highest total inpatient hospital charges.
Prevalence of Maternal Adverse Cardiac Events, Respiratory and Renal Failure
MACE among women with PH and those with no heart disease are compared in Table 2 and Figure 3. MACE was more frequent among patients with PH compared with those with no heart disease (24.8% versus 0.4%); heart failure and arrhythmia were the most common presentations of MACE. Overall maternal mortality was low in the PH cohort, although multifold higher than in the women without any heart disease. Respiratory failure, including use of mechanical ventilation and noninvasive positive pressure ventilation, as well as renal failure were more frequent among women with PH compared with those with no heart disease (P<0.0001).
Among the PH subgroups in Table 2, women with PH+cardiomyopathy+VHD, as well as those with PH+cardiomyopathy, experienced the highest MACE rates (62.1% and 59.4%, respectively), while women with PH+CHD exhibited the lowest MACE rate (15.3%). Again, heart failure and arrhythmia were the most common components of MACE among the PH subgroups. Not unexpectedly, heart failure was most common among the PH+cardiomyopathy+VHD cohort (55.2%, P<0.0001); arrhythmias were most common among the PH+cardiomyopathy cohort (19.8%, P=0.0013). Mechanical ventilation was most utilized among the PH+cardiomyopathy cohort (10.9%), while women with isolated PH developed the highest rate of acute renal failure (1.9%) among all the PH subgroups.
Obstetric Outcomes in Pregnant Women With PH
Women with PH more frequently developed preeclampsia/eclampsia syndrome, including severe preeclampsia (6.7% versus 1.2%), mild preeclampsia (5.3% versus 2.2%), and eclampsia with preexisting hypertension (8.0% versus 0.5%) compared with those women without any heart disease (P<0.0001, Table 3). Women with PH had substantially higher rates of cesarean delivery (49.7% versus 31.7%, P<0.0001). Women with PH had higher rates of obstetric bleeding, postpartum hemorrhage, postpartum infarction, and intrauterine fetal demise (all P≤0.0004). The incidence of preterm delivery was multifold higher than in those without HD (21.5 versus 7.1%, P<0.0001).
Among the PH subclasses in Table 3, women with PH+cardiomyopathy most commonly had severe preeclampsia and eclampsia with preexisting hypertension (9.9% and 11.9%, P<0.01). The incidence of preterm delivery was common, and was highest among the PH+cardiomyopathy+VHD and PH+cardiomyopathy subgroups (37.9 and 57.4%, P=0.0148); the incidence of cesarean delivery was common, and highest among the isolated PH, PH+cardiomyopathy, and PH+cardiomyopathy+VHD subgroups.
Associations With Maternal Adverse Cardiac Events
Multivariable regression analysis examining independent risk factors associated with MACE are shown in Figure 4. Several PH subclasses had a higher risk of MACE as compared with the PH+CHD group (selected since it was the lowest risk group among those studied). Among all patients with PH, those with PH+VHD had an increased odds of MACE (odds ratio [OR] 1.83, 95% confidence interval [CI], 1.08–3.09) compared with those with PH+CHD; PH+VHD+CHD (OR 2.83, 95% CI, 1.32–6.08); PH+cardiomyopathy (OR 6.94, 95% CI, 3.76–12.82); and PH+cardiomyopathy+VHD (OR 6.56, 95% CI, 2.68–16.04). Those with isolated PH had similar risk for MACE as those with PH+CHD group (OR 1.22, 95% CI, 0.76–1.98).
In a large national contemporary data set of pregnant women in the United States, PH was associated with a high prevalence of maternal adverse cardiac events during the hospitalization for delivery, with an exceptionally elevated risk among women with associated cardiomyopathy. This study is the largest‐known characterization of PH and pregnancy to date. While inpatient maternal mortality was low overall, it was multifold higher than in women with no heart disease. Women with PH associated with left heart disease experienced substantially higher MACE than women with isolated PH or PH associated with CHD. Preeclampsia/eclampsia and preterm delivery were common in women with PH, as were fetal demise, obstetric and postpartum bleeding, and cesarean delivery.
Pregnancy in women with PH has been historically associated with prohibitive maternal mortality, upwards of 50%.1, 3 In the past 15 years, advances in medical therapy and treatment of PH have developed, improving overall quality of life and prognosis.14 The incidence of PH in pregnancy is increasing in the United States,8 a finding that we find particularly alarming given the risk of maternal mortality and morbidity. Recently, there have been reports of lower mortality (3%2), although still multifold higher than what would be considered acceptable to most mothers. The lower mortality herein may be attributable to differences in length of follow‐up, since we measure inpatient mortality from the index hospitalization for delivery, and cannot capture deaths that could have occurred later, a period of particularly high risk for postpartum women with PH because of ongoing hemodynamic changes associated with pregnancy.5 Data suggest these changes persist for up to 6 months postpartum.5 The length of stay and total hospital charges are high, as well as a high prevalence of heart failure and arrhythmia, reflecting complicated and high‐level patient care in the PH group. Herein, we may be capturing a larger number of patients whose degree of PH may not be severe and whose pregnancy outcome may be slightly more favorable. However, the rate of obstetric complications is also high and comparable with existing literature. Nonetheless, it seems that progress in the care of pregnant women with PH has been made, for reasons that are not entirely clear, whether related to improvements in recognition of the disorder and its risk in pregnancy, management in pregnancy, or management of labor and delivery with a multidisciplinary approach.5 However, one should not be overly reassured by the reported lower mortality herein, since women tended to have heart failure and arrhythmias quite frequently (overall MACE ranging between 15% and 62%), particularly when PH was seen in combination with left heart disease such as cardiomyopathy. Recent research from our group8, 10 and that of others9, 15 has shown that cardiomyopathy predicts adverse cardiac events in pregnant women.
Women with PH associated with congenital heart disease had the lowest MACE in our study. This may be because the disease was known and monitored and controlled before pregnancy. The population with PH and CHD is expected to be a highly variable disease population. Eisenmenger syndrome, found in only 3% of the entire PH group, occurs when the pulmonary vascular resistance exceeds the systemic vascular resistance and prompts reversal of the systemic‐to‐pulmonary shunt; it is considered the most hazardous form of pulmonary arterial hypertension in CHD patients, particularly in pregnancy. Women with Eisenmenger syndrome are strongly advised against pregnancy because they are not able to withstand the prolonged hemodynamic stress of pregnancy, including increased blood volume, heart rate, and cardiac output.16 However, paradoxically, our study did not find Eisenmenger syndrome to be an independent predictor of MACE upon multivariable analysis. This may be a function of the cohort not being well represented in this study because the incidence of Eisenmenger syndrome was relatively low in the examined population, though this is reflective of its presence in the general population as Eisenmenger syndrome is uncommon. It is also plausible that the lower than expected MACE rate is because of preservation of right ventricular function.17
Obstetric complications were high in the PH group, including preeclampsia and eclampsia syndromes, preterm delivery, as were fetal demise, obstetric and postpartum bleeding, and the use of cesarean delivery. However, these rates are similar to those found in a recent multinational registry.2 Postpartum hemorrhage, obstetric bleeding, and cesarean section have all been associated with adverse outcomes during pregnancy in women with PH. The optimal mode of delivery remains controversial; however, the literature favors the use of cesarean delivery1, 2 and even recommends it.18 Nonetheless, it is essential that these women be managed optimally by a multidisciplinary team of cardiologists, anesthesiologists, and obstetricians regarding the method and timing of their delivery.
Over a contemporary period of time, over 1500 women with PH became pregnant and carried a pregnancy to delivery. While there were likely others who miscarried early, or were not captured by this data set for other reasons, it is alarming that this a rather large number of high‐risk group of women who achieved a pregnancy and delivery. This information suggests that there may be gaps in adequate healthcare access and/or preconception counseling in the United States that should be further explored. Because PH pregnancies are extremely high risk, the quality of preconception counseling, contraception methods, and risk counseling made available by the medical community is of paramount importance. One study of adult CHD patients documented that patients often cannot remember ever receiving counseling regarding pregnancy risks or contraception,19 despite this being a Class I recommendation by Society Guidelines.5, 20
Our study has several important limitations. The NIS was especially valuable in studying PH, which has a low incidence in pregnancy, because it allowed for a large, representative US sample. This data set has been utilized in the past to study hospitalization trends and its predictors in patients with heart disease, as well to assess outcomes in PH without pregnancy,14 pregnant women with preexisting cardiomyopathy, CHD, and myocardial disorders.10, 21, 22, 23, 24 However, the data are restricted to inpatient, delivery‐related hospitalizations, and maternal data on medication use, disease severity, and neonatal complications are not available. Late maternal mortality is a real and devastating issue.2, 25 Another important limitation was the exclusion of approximately 20% of data because of missing information on race or insurance status, in order to achieve a more uniform data set; however, exclusion of a large number of records could have had implications on the study results. Despite these limitations, it is validating that our data substantiated existing paradigms about PH, including cardiac and obstetric complications.
Utilizing a national, all‐payer inpatient data set, we found that women with PH endured a significantly higher rate of MACE than women without heart disease, largely related to heart failure and arrhythmia. Maternal mortality was low overall, and lower than previously reported, but multifold higher than women without PH. PH associated with left heart disease experience substantially higher MACE than isolated PH or PH associated with CHD. Preeclampsia/eclampsia and preterm delivery were common in women with PH, as were fetal demise and obstetric and postpartum bleeding. Women with PH who become pregnant warrant a multidisciplinary approach for management of pregnancy, labor, and delivery. Prepregnancy counseling can help inform patients and providers regarding their reproductive choices.
Sources of Funding
Funding was provided by the American Medical Association Foundation Seed Grant and the American Heart Association.
Data S1. ICD9 and CCS Codes Utilized in This Study.
We acknowledge the statistical support provided by the Biostatistical Consulting Core of Stony Brook University School of Medicine.
This article was presented in abstract form at the American Heart Association Scientific Sessions, November 12–16, 2016 in New Orleans, LA.
- ↵Sliwa K, van Hagen IM, Budts W, Swan L, Sinagra G, Caruana M, Blanco MV, Wagenaar LJ, Johnson MR, Webb G, Hall R, Roos‐Hesselink JW; ROPAC investigators . Pulmonary hypertension and pregnancy outcomes: data from the Registry Of Pregnancy and Cardiac Disease (ROPAC) of the European Society of Cardiology. Eur J Heart Fail. 2016;18:1119–1128.
- ↵Weiss BM, Zemp L, Seifert B, Hess OM. Outcome of pulmonary vascular disease in pregnancy: a systematic overview from 1978 through 1996. J Am Coll Cardiol. 1998;31:1650–1657.
- ↵Canobbio MM, Warnes CA, Aboulhosn J, Connolly HM, Khanna A, Koos BJ, Mital S, Rose C, Silversides C, Stout K. Management of pregnancy in patients with complex congenital heart disease: a scientific statement for healthcare professionals from the American Heart Association. Circulation. 2017;135:e50–e87
- ↵Regitz‐Zagrosek V, Blomstrom Lundqvist C, Borghi C, Cifkova R, Ferreira R, Foidart JM, Gibbs JS, Gohlke‐Baerwolf C, Gorenek B, Iung B, Kirby M, Maas AH, Morais J, Nihoyannopoulos P, Pieper PG, Presbitero P, Roos‐Hesselink JW, Schaufelberger M, Seeland U, Torracca L. ESC Guidelines on the management of cardiovascular diseases during pregnancy: the Task Force on the Management of Cardiovascular Diseases during Pregnancy of the European Society of Cardiology (ESC). Eur Heart J. 2011;32:3147–3197.
- ↵Hemnes AR, Kiely DG, Cockrill BA, Safdar Z, Wilson VJ, Al Hazmi M, Preston IR, MacLean MR, Lahm T. Statement on pregnancy in pulmonary hypertension from the Pulmonary Vascular Research Institute. Pulm Circ. 2015;5:435–465.
- ↵Koutrolou‐Sotiropoulou P, Lima FV, Yang J, Xu J, Stergiopoulos K. Abstract 11980: National Trends and In‐hospital Outcomes during Delivery in Pregnant Women With Heart Disease in the US. Circulation. 2016;134:A11980.
- ↵Siu SC, Sermer M, Colman JM, Alvarez AN, Mercier LA, Morton BC, Kells CM, Bergin ML, Kiess MC, Marcotte F, Taylor DA, Gordon EP, Spears JC, Tam JW, Amankwah KS, Smallhorn JF, Farine D, Sorensen S; Cardiac Disease in Pregnancy I . Prospective multicenter study of pregnancy outcomes in women with heart disease. Circulation. 2001;104:515–521.
- ↵Lima FV, Parikh PB, Zhu J, Yang J, Stergiopoulos K. Association of cardiomyopathy with adverse cardiac events in pregnant women at the time of delivery. JACC Heart Fail. 2015;3:257–266.
- ↵Clinical Classifications Software (CCS) for ICD‐9‐CM. HCUP‐US Tools & Software Page.2015.
- ↵SAS Institute Inc., Version 9.3, Cary, NC: 2011.
- ↵Anand V, Roy SS, Archer SL, Weir EK, Garg SK, Duval S, Thenappan T. Trends and Outcomes of Pulmonary Arterial Hypertension‐Related Hospitalizations in the United States: Analysis of the Nationwide Inpatient Sample Database From 2001 Through 2012. JAMA Cardiol. 2016;1:1021–1029.
- ↵Roos‐Hesselink JW, Ruys TP, Stein JI, Thilen U, Webb GD, Niwa K, Kaemmerer H, Baumgartner H, Budts W, Maggioni AP, Tavazzi L, Taha N, Johnson MR, Hall R; ROPAC Investigators . Outcome of pregnancy in patients with structural or ischaemic heart disease: results of a registry of the European Society of Cardiology. Eur Heart J. 2013;34:657–665.
- ↵Balint OH, Siu SC, Mason J, Grewal J, Wald R, Oechslin EN, Kovacs B, Sermer M, Colman JM, Silversides CK. Cardiac outcomes after pregnancy in women with congenital heart disease. Heart. 2010;96:1656–1661.
- ↵Konstantinides SV. Trends in pregnancy outcomes in patients with pulmonary hypertension: still a long way to go. Eur J Heart Fail. 2016;18:1129–1131.
- ↵Kovacs AH, Harrison JL, Colman JM, Sermer M, Siu SC, Silversides CK. Pregnancy and contraception in congenital heart disease: what women are not told. J Am Coll Cardiol. 2008;52:577–578.
- ↵Warnes CA, Williams RG, Bashore TM, Child JS, Connolly HM, Dearani JA, del Nido P, Fasules JW, Graham TP Jr., Hijazi ZM, Hunt SA, King ME, Landzberg MJ, Miner PD, Radford MJ, Walsh EP, Webb GD, Smith SC Jr., Jacobs AK, Adams CD, Anderson JL, Antman EM, Buller CE, Creager MA, Ettinger SM, Halperin JL, Hunt SA, Krumholz HM, Kushner FG, Lytle BW, Nishimura RA, Page RL, Riegel B, Tarkington LG, Yancy CW; American College of Cardiology, American Heart Association Task Force on Practice Guidelines (Writing Committee to Develop Guidelines on the Management of Adults With Congenital Heart Disease);, American Society of Echocardiography; Heart Rhythm Society; International Society for Adult Congenital Heart Disease; Society for Cardiovascular Angiography and Interventions; Society of Thoracic Surgeons . ACC/AHA 2008 guidelines for the management of adults with congenital heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Develop Guidelines on the Management of Adults With Congenital Heart Disease). Developed in Collaboration With the American Society of Echocardiography, Heart Rhythm Society, International Society for Adult Congenital Heart Disease, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol. 2008;52:e143–e263.
- ↵Collaborators GBDMM . Global, regional, and national levels of maternal mortality, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016;388:1775–1812.