Cardiovascular disease (CVD) remains a leading cause of maternal mortality in the United States, comprising 26.5% of pregnancy-related deaths. We sought to evaluate trends in CVD during pregnancy and maternal, obstetric, and fetal outcomes in pregnant women with CVD in the Kaiser Permanente Northern California (KPNC) integrated healthcare system. This retrospective cohort study included adult KPNC members with moderate or greater valvular heart disease, cardiomyopathy, congenital heart disease, or ischemic heart disease during pregnancy from 2010 to 2021. Bivariate analyses and multivariable logistic regression were used to evaluate associations between demographic and clinical risk factors and maternal outcomes in pregnant patients with CVD. Of 320,902 pregnancies, 763 (0.24%) were identified with clinically significant CVD. The prevalence of CVD increased from 0.19% to 0.34% over the decade, predominantly due to an increase in prevalence of women with congenital heart disease. Mean gestational age at delivery was 36.7 weeks with 19.5% experiencing preterm delivery. Cesarean section, pre-eclampsia or eclampsia, and postpartum hemorrhage rates were 29.5%, 20.8%, and 9.7%, respectively. Fetal loss beyond the first trimester occurred in 7.1% pregnancies. Adverse maternal cardiac outcomes occurred in 55 (7.2%) of patients, predominantly driven by congestive heart failure hospitalizations. There were 4 deaths (0.5%) during pregnancy or within 1 year postpartum. Pre-existing CVD (aOR 0.20, p = 0.002) and cardiac medication use (aOR 4.13, p <0.001) were significant predictors of adverse maternal outcomes. Higher left ventricular ejection fraction (aOR 0.95, p = 0.018) was associated with lower odds of adverse maternal outcomes. In conclusion, understanding risk factors for adverse pregnancy outcomes in a diverse, contemporary population of patients with CVD can help refine cardio-obstetric risk assessment and preconception counseling.
Women with cardiovascular disease (CVD) during pregnancy are at almost a 20-fold increased risk for long-term cardiovascular sequalae, even years following pregnancy. Current guidelines identify the presence of depressed left ventricular ejection fraction (LVEF), moderate or greater valvular heart disease, pulmonary hypertension, presence of a mechanical valve, prior cardiac event or cardiac medication use prior to pregnancy, and functional status to be strong predictors of adverse outcomes during pregnancy. ,,, Depressed LVEF and pulmonary hypertension during pregnancy are further associated with an increased risk of life-threatening arrythmias and decompensated congestive heart failure (CHF), as well as poor neonatal outcomes (e.g., prematurity, neonatal intensive care unit [NICU] stays, and fetal death). ,,, Risk stratification during preconception counseling is therefore an integral part of management for patients with pre-existing cardiac disease. ,,,
Data from diverse, multiracial/ethnic populations can identify areas for improvement in the care of pregnant women with CVD, and whether current risk estimations are adequate. In this study, we sought to (1) assess the prevalence of clinically significant CVD and the maternal, obstetric, and fetal/neonatal outcomes of these pregnancies and (2) evaluate risk factors associated with adverse maternal cardiac events in women with clinically significant structural heart disease (i.e., valvular heart disease, cardiomyopathy, congenital heart disease, and ischemic heart disease) from a large, integrated health system across a 12-year period. This contemporary data can help refine risk estimates used for preconception and intrapartum counseling and improve cardio-obstetric care.
Methods
This is a retrospective cohort study of maternal, obstetric, fetal, and neonatal outcomes among adult Kaiser Permanente Northern California (KPNC) members with clinically significant heart disease during pregnancy from 2010 to 2021. The KPNC population is demographically representative of the larger community from which membership is drawn. This study was reviewed and approved by the KPNC Regional Institutional Review Board.
Study Population
The study population included adult patients with pregnancy onset between January 1, 2010, and December 31, 2021, with at least 1 year of membership prior to pregnancy and maintained membership until the end of their pregnancy. Patients who had a first trimester miscarriage or termination of pregnancy were excluded. After identifying all pregnancies during the study period, ICD-9 and ICD-10 codes ( Supplementary Table 1 ) were subsequently applied to identify members with valvular heart disease, cardiomyopathy, congenital heart disease, or ischemic heart disease during pregnancy. Patients with primary pulmonary hypertension and primary aortopathy (except for coarctation of the aorta) were excluded from this study due to the high prevalence of concomitant systemic and genetic disease in this group. Given their distinct pathophysiology and requirements for specialized management—which differ from the broader cohort of patients with CVD in pregnancy—these patients were excluded to ensure generalizability of the cohort.
For all patients found to have pregnancies complicated by structural heart diseases, the investigator team (CS, EP, MM, SP) manually reviewed electronic health records (EHR) and where available, cardiac imaging data to determine if clinically significant CVD existed. Categories were as follows: Congenital heart disease included Fontan/single ventricle, cor pulmonale, Eisenmenger, ventricular septal defect, atrial septal defect, congenital pulmonic stenosis or regurgitation (moderate or greater in severity), patent ductus arteriosus, Shone complex, coarctation of the aorta, coronary anomaly, Tetralogy of Fallot, other or multiple congenital lesions. Primary valvular heart disease included rheumatic and nonrheumatic acquired valvular heart disease, moderate or greater in severity. Cardiomyopathy included ischemic and nonischemic cardiomyopathies. Ischemic heart disease included obstructive coronary artery disease (CAD), myocardial infarction with no obstructive CAD (MINOCA), and spontaneous coronary artery dissection (SCAD).
Demographic and clinical characteristics were extracted from the EHR and Kaiser Permanente Division of Research databases, including: age, self-identified race/ethnicity, census tract median household income, education, parity, multiple gestation, body mass index (BMI), substance use, hypertension, diabetes mellitus, CHF, percutaneous coronary intervention, coronary artery bypass grafting, and the presence of pre-existing CVD (as defined by presence of disease prior to pregnancy). Where available, echocardiogram and cardiac magnetic resonance imaging (MRI) data within 1 year prior to or during pregnancy were reviewed; extracted data included severity of valvular disease and LVEF. Cardiac medication use was collected both prior to and during the pregnancy.
Outcomes
Maternal outcomes were collected from pregnancy initiation to 12 months postpartum, including hospitalization for acute decompensated CHF, aortic dissection, stroke/transient ischemic attack, myocardial infarction (MI), sustained ventricular tachycardia/fibrillation, sustained supraventricular tachycardia, and maternal death. Obstetric outcomes included preterm delivery (defined as prior to 37 weeks gestation), pre-eclampsia/eclampsia, postpartum hemorrhage, Cesarean section delivery, and need for emergent Cesarean section delivery. Fetal outcomes included intrauterine fetal loss (14 to 19 weeks gestational age), intrauterine fetal demise (20 weeks or greater in gestational age), low APGAR score, small or large for gestational age, and NICU stay. Neonatal deaths were also extracted.
Statistical Analysis
We report the prevalence of CVD (overall and by category) in pregnancies each year from 2010 to 2021. Wald 95% confidence intervals were calculated for each annual proportion and the linear trend in proportions over time was assessed using a Cochran–Armitage test. We report the frequencies and percentages of adverse maternal, obstetric, and fetal outcomes in all pregnancies with CVD.
Baseline characteristics of women with CVD during pregnancy are presented using means and standard deviations for continuous variables and frequencies and percentages for categorical variables. Among women who had more than 1 pregnancy during the study timeline, CVD status and baseline characteristics were those at the time of their first pregnancy. Chi-square tests were used to assess the association between women’s baseline characteristics and adverse maternal outcomes.
Multivariable logistic regression was used to describe risk factors associated with maternal outcomes in pregnant patients with CVD. Among patients who had more than 1 pregnancy during the study timeline, their first pregnancy was included in the regression analysis. Age, race/ethnicity, and clinical and demographic factors that had a significant bivariate association (p <0.05) with the outcome were included in a multivariable model. Factors other than age and race/ethnicity that did not have a multivariable adjusted association with the outcome (p ≥0.20) were removed from each model through backwards elimination.
P-values were considered significant at <0.05. All statistical analyses were conducted in SAS 9.4 for Windows.
Results
We identified 320,902 pregnancies that occurred during the study period, and 763 pregnancies (0.24% prevalence) in 557 women were complicated by clinically significant CVD. Of these 763 pregnancies, 546 (71.5%) had congenital heart disease, 99 (12.9%) cardiomyopathy, 89 (11.7%) primary valvular heart disease, and 29 (3.8%) ischemic heart disease. From 2010 to 2021, the prevalence of CVD during pregnancy increased significantly from 0.19% in 2010 to 0.34% in 2021 (p <0.001, Figure 1 ). As expected, this increased prevalence was primarily driven by an increase in pregnancy in women with congenital heart disease (p <0.001, Figure 2 ). When stratified by age, increased prevalence of pregnancies complicated by significant CVD was seen between ages 25 to 34 (p <0.001) ( Supplementary Figure 1 ). When stratified by self-identified race/ethnicity, a statistically significant increase in pregnancies was seen in White (p = 0.002), other Asian (p = 0.016), and unclassified (race/ethnicity unknown or multiethnic classification, p = 0.008) groups. nonstatistically significant increases were seen in Hispanic (p = 0.184), South Asian (p = 0.459), Filipino (p = 0.381), and Black (p = 0.329) groups ( Supplementary Figure 2 ).
Prevalence of cardiovascular disease during pregnancy (2010 to 2021) at Kaiser Permanente, Northern California Annual prevalence of cardiovascular disease during pregnancy is shown between 2010 and 2021 at Kaiser Permanente, Northern California. Cochran–Armitage trend test p <0.001.
Prevalence of cardiovascular disease during pregnancy (2010 to 2021) by disease category, Kaiser Permanente Northern California Annual prevalence of cardiovascular disease during pregnancy is shown between 2010 and 2021, Kaiser Permanente Northern California, by disease categories. Cochran–Armitage trend test p = 0.176 for cardiomyopathy, p <0.001 for congenital heart disease, p = 0.093 for ischemic heart disease, p = 0.765 for valvular heart disease.
Cardiovascular Characteristics
Baseline characteristics for the women with clinically significant CVD during pregnancy (n = 557) are shown in Table 1 . Mean age was 30.8 years and 36.0% were White, 24.0% Hispanic, 10.0% Black, and 23.0% Asian (including 5.0% South Asian and 5.0% Filipino). Based on geographic census tract data, the proportion of presumed college graduates was 43.0% and 96.0% spoke English as a primary language. In this cohort, 481 patients had a cardiac diagnosis prior to pregnancy and the remainder were diagnosed during pregnancy. Beta blockers were the most used medication before and during pregnancy (n = 206, 37.0%), followed by antiplatelets medication (n = 181, 33.0%), and diuretics (n = 133, 24.0%). Mean LVEF in this cohort was 59.7% with valvular pathology including 3 mechanical valves, 24 bioprosthetic valves, and 11 with significant mitral or aortic stenosis. Over half of the patients included in the cohort were nulliparous at their index pregnancy (n = 312, 56.0%) and 13 (2.0%) had multiple gestation pregnancies. Most of the cohort included 1 pregnancy per woman (432, n = 78.0%), with 17.0% having 2 pregnancies, and 6.0% having 3 or more pregnancies included in the analysis.
Table 1
Baseline characteristics of women with cardiovascular disease during pregnancy (n = 557)
| Demographics | n (%) or mean (SD) |
|---|---|
| Age, mean (SD) | 30.8 (6.2) |
| Race/ethnicity | |
| White | 203 (36) |
| Black | 53 (10) |
| Hispanic | 135 (24) |
| South Asian | 29 (5) |
| Filipino | 27 (5) |
| Other Asian | 70 (13) |
| Other/Unknown | 40 (7) |
| Social determinants of health | |
| English as primary language | 537 (96) |
| Proportion of college graduates, based on census tract, mean (SD) | 0.43 (0.19) |
| Median household income, based on census tract,in $ (SD) | 77,127 (31,878) |
| Obstetric history | |
| Nulliparous at index pregnancy | 312 (56) |
| Parity at index pregnancy | |
| 0 | 312 (56) |
| 1 | 140 (25) |
| 2 | 73 (13) |
| 3+ | 29 (5) |
| Multiple gestation (twins or triplets) | 13 (2) |
| Number of pregnancies per woman included in this cohort | |
| 1 | 432 (78) |
| 2 | 92 (17) |
| 3+ | 33 (6) |
| Comorbidities prior to pregnancy | |
| Tobacco smoker | 80 (15) |
| Illicit drug use | 20 (4) |
| BMI (kg/m 2), mean (SD) | 26.3 (6.4) |
| Hypertension | 32 (6) |
| Diabetes | 10 (2) |
| Timing of cardiac diagnosis | |
| Prepregnancy PCI | 9 (2) |
| Prepregnancy CABG | 6 (1) |
| Prepregnancy congestive heart failure | 46 (8) |
| Cardiac diagnosis prior to pregnancy | 481 (86) |
| Cardiac medications prior to or during pregnancy | |
| Beta blockers | 206 (37) |
| Calcium channel blockers | 81 (15) |
| Diuretics | 133 (24) |
| Nitrates | 12 (2) |
| Hydralazine | 50 (9) |
| Antiplatelets | 181 (33) |
| Anticoagulants | 39 (7) |
| Antiarrhythmics | 14 (3) |
| Digoxin | 18 (3) |
| Methyldopa | 6 (1) |
| Echocardiographic parameters | |
| Left ventricular ejection fraction (%), mean (SD) | 59.7 (7.2) |
| Mechanical valve | 3 (1) |
| Bioprosthetic valve | 24 (4) |
| Left ventricular outflow tract obstruction | 3 (1) |
| High risk valve disease (significant MS/AS) | 11 (2) |
| Systemic valvular regurgitation | 48 (9) |
| Pulmonary regurgitation | 26 (5) |
| Pulmonary atrioventricular valvular regurgitation | 5 (1) |
Maternal Outcomes
Of the 763 pregnancies complicated by clinically significant CVD, 55 (7.2%) experienced adverse maternal cardiac events, with the most common being CHF hospitalization (n = 47, 6.2%, Table 2 ). The timing of maternal cardiac events was highest intrapartum or within 1 week postpartum (n = 19, 34.5%) followed by 28 weeks or later in pregnancy (n = 11, 20.0%). When evaluating individual outcomes, 37.0% of the CHF exacerbations and 75.0% of maternal deaths in this cohort occurred at the time of delivery or within 1 week postpartum ( Supplementary Table 2 ). The mean gestational age at delivery was 36.7 weeks, with 149 (19.5%) pregnancies complicated by a preterm delivery.
Table 2
Adverse maternal, obstetric, and fetal cardiac event rates in n = 763 pregnancies
| n (%) | |
|---|---|
| Any maternal cardiac event | 55 (7.2) |
| Ventricular tachycardia/ventricular fibrillation | 3 (0.4) |
| Sustained SVT | 3 (0.4) |
| Congestive heart failure hospitalization | 47 (6.2) |
| Aortic dissection | 0 (0.0) |
| Stroke/TIA | 5 (0.7) |
| Myocardial infarction | 4 (0.5) |
| Maternal death | 4 (0.5) |
| Timing of maternal cardiac events | |
| 0-12 weeks | 6 (10.9) |
| 13-27 weeks | 6 (10.9) |
| 28 weeks to antepartum | 11 (20.0) |
| At labor/delivery or <1 week postpartum | 19 (34.5) |
| Postpartum 1-4 weeks | 6 (10.9) |
| Postpartum 1-3 months | 3 (5.5) |
| Postpartum 3-6 months | 1 (1.8) |
| Postpartum 6-12 months | 3 (5.5) |
| Obstetric outcomes | |
| Gestational age at delivery (in weeks), mean (SD) | 36.7 (6.0) |
| Preterm delivery (before 37 weeks) | 149 (19.5) |
| Pre-eclampsia/Eclampsia | 159 (20.8) |
| Caesarean section | 225 (29.5) |
| Emergent Caesarean section | 38 (5.0) |
| Postpartum hemorrhage | 74 (9.7) |
| Immediate postpartum hemorrhage (<24 hours) | 69 (9.0) |
| Delayed postpartum hemorrhage (>24 hours) | 5 (0.7) |
| Fetal outcomes | |
| Pregnancy loss | 54 (7.1) |
| 13-19 weeks | 40 (5.2) |
| ≥ 20 weeks | 14 (1.8) |
| APGAR score at 5 minutes, mean (SD) | 8.78 (0.79) |
| Birth weight in grams, mean (SD) | 3211 (618) |
| Small for gestational age | 82 (10.7) |
| Large for gestational age | 71 (9.3) |
| Need for NICU stay | 86 (11.3) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
