Includes admissions other than for subsequent delivery. P values comparing equality of survivor functions for complex, noncomplex, and no congenital heart disease (CHD) were calculated using log-rank test (χ2 = 3941; P < .01).
Includes admissions other than for subsequent delivery. CHD indicates congenital heart disease; CHF, congestive heart failure; HR, hazard ratio; and NA, not applicable.
eTable 1. Inclusion and Exclusion Criteria
eTable 2. CHD Diagnoses With Corresponding ICD-9 Codes
eTable 3. Outcomes
eTable 4. Medical Comorbidities
eTable 5. Type of CHD
Customize your JAMA Network experience by selecting one or more topics from the list below.
Hayward RM, Foster E, Tseng ZH. Maternal and Fetal Outcomes of Admission for Delivery in Women With Congenital Heart Disease. JAMA Cardiol. 2017;2(6):664–671. doi:10.1001/jamacardio.2017.0283
How do the risks and outcomes for labor and delivery compare in women with and without congenital heart disease?
In this study capturing nearly every delivery admission in California during a 7-year period, congenital heart disease was associated with incident heart failure, atrial arrhythmias, fetal growth restriction, and readmission; complex congenital heart disease was also associated with ventricular arrhythmias and maternal in-hospital mortality, although these outcomes were rare, even in women with congenital heart disease.
This study highlights some of the peripartum and postpartum risks for women with CHD, findings that may guide monitoring decisions and risk assessment for pregnant women with CHD at the time of delivery.
Women with congenital heart disease (CHD) may be at increased risk for adverse events during pregnancy and delivery.
To compare delivery outcomes between women with and without CHD.
Design, Setting, and Participants
This retrospective study of inpatient delivery admissions in the Healthcare Cost and Utilization Project’s California State Inpatient Database compared maternal and fetal outcomes between women with and without CHD by using multivariate logistic regression. Female patients with codes for delivery from the International Classification of Diseases, Ninth Revision, from January 1, 2005, through December 31, 2011, were included. The association of CHD with readmission was assessed to 7 years after delivery. Cardiovascular morbidity and mortality were hypothesized to be higher among women with CHD. Data were analyzed from April 4, 2014, through January 23, 2017.
Noncomplex and complex CHD.
Main Outcomes and Measures
Maternal outcomes included in-hospital arrhythmias, eclampsia or preeclampsia, congestive heart failure (CHF), length of stay, preterm labor, anemia complicating pregnancy, placental abnormalities, infection during labor, maternal readmission at 1 year, and in-hospital mortality. Fetal outcomes included growth restriction, distress, and death.
Among 3 642 041 identified delivery admissions, 3189 women had noncomplex CHD (mean [SD] age, 28.6 [7.6] years) and 262 had complex CHD (mean [SD] age, 26.5 [6.8] years). Women with CHD were more likely to undergo cesarean delivery (1357 [39.3%] vs 1 164 509 women without CHD [32.0%]; P < .001). Incident CHF, atrial arrhythmias, ventricular arrhythmias, and maternal mortality were uncommon during hospitalization, with each occurring in fewer than 10 women with noncomplex or complex CHD (<0.5% each). After multivariate adjustment, noncomplex CHD (odds ratio [OR], 9.7; 95% CI, 4.7-20.0) and complex CHD (OR, 56.6; 95% CI, 17.6-182.5) were associated with greater odds of incident CHF. Similar odds were found for atrial arrhythmias in noncomplex (OR, 8.2; 95% CI, 3.0-22.7) and complex (OR, 31.8; 95% CI, 4.3-236.3) CHD, for fetal growth restriction in noncomplex (OR, 1.6; 95% CI, 1.3-2.0) and complex (OR, 3.5; 95% CI, 2.1-6.1) CHD, and for hospital readmission in both CHD groups combined (OR, 3.6; 95% CI, 3.3-4.0). Complex CHD was associated with greater adjusted odds of serious ventricular arrhythmias (OR, 31.8; 95% CI, 4.3-236.3) and maternal in-hospital mortality (OR, 79.1; 95% CI, 23.9-261.8).
Conclusions and Relevance
In this study of hospital admissions for delivery in California, CHD was associated with incident CHF, atrial arrhythmias, and fetal growth restriction and complex CHD was associated with ventricular arrhythmias and maternal in-hospital mortality, although these outcomes were rare, even in women with complex CHD. These findings may guide monitoring decisions and risk assessment for pregnant women with CHD at the time of delivery.
Labor, delivery, and the postpartum period are a time of increased arrhythmia and heart failure incidences, even in patients without cardiovascular disease. Possible explanations include altered hemodynamics with myocardial irritability owing to increases in intravascular volume, hypokalemia, autonomic changes, and hormonal influences on adrenergic receptor number and behavior.1-3 Maternal arrhythmias increase during labor and delivery, possibly owing to volume shifts and hypertension.4 In women with arrhythmias before pregnancy, recurrence during pregnancy is common and is associated with an increased risk for fetal complications.5 Despite these risks, a small series including women with implantable cardioverter defibrillators followed up during subsequent pregnancies has revealed no increase in implantable cardioverter defibrillator shocks.6,7 Likewise, congestive heart failure (CHF) is more common during and after pregnancy in women without a history of cardiovascular disease.8
Owing to improvements in the treatment of congenital heart disease (CHD), most patients with CHD now survive to adulthood, and more than 1 million adults are living with CHD in the United States.9,10 As patients with CHD live longer, arrhythmias are an increasing cause of morbidity and mortality.11,12 In addition, more women with CHD are becoming pregnant, and these patients are at increased risk for arrhythmias and CHF.13,14
We sought to evaluate contemporary maternal and fetal outcomes in women with and without CHD during admissions for delivery in California. We used the Health Care Cost and Utilization Project (HCUP), a group of databases that have already been used to study CHD hospitalizations and procedures,15-19 to evaluate cardiac and other complications of labor and delivery for women with and without CHD from 2005 through 2011.
Admissions for female patients with a diagnosis code from the International Classification of Diseases, Ninth Revision (ICD-9), a procedure code from the International Classification of Diseases, Ninth Revision, Clinical Modification, or a diagnosis related group code for delivery20,21 from January 1, 2005, through December 31, 2011, were identified from HCUP’s California State Inpatient Database. If an ICD-9 code for cesarean delivery was not present and the delivery type was not specified, delivery was categorized as vaginal. The details of the ICD-9 codes used are listed in eTable 1 in the Supplement. Identifying patient information was anonymized before analysis, and certification to use these deidentified HCUP data was obtained from the Committee on Human Research at the University of California, San Francisco.
Diagnoses of CHD were determined by ICD-9 or Current Procedural Terminology codes used by others previously (eTable 2 in the Supplement).17,22-24 If a CHD diagnosis code was present for any encounter during the study period, the patient was assigned that diagnosis for all visits. Complex CHD included endocardial cushion defects, hypoplastic left heart syndrome, tetralogy of Fallot, transposition of the great arteries, truncus arteriousus, and univentricular heart defects. All other CHD defects were classified as noncomplex.22-24 Patients with complex and noncomplex CHD diagnoses were assigned to the complex CHD group.
Maternal cardiovascular outcomes included in-hospital atrial arrhythmias (including atrial fibrillation, atrial flutter, or paroxysmal supraventricular tachycardia), serious ventricular arrhythmias (including paroxysmal ventricular tachycardia, ventricular fibrillation, ventricular flutter, cardiac arrest, and sudden death due to an unknown cause during the puerperium), eclampsia or preeclampsia, and CHF (eTable 3 in the Supplement). Other maternal outcomes included length of stay and in-hospital mortality. Maternal and fetal outcomes, defined by others previously,25 included preterm labor, anemia complicating pregnancy, placental abnormalities (previa, abruptio, or accreta), infection during labor, fetal growth restriction, fetal distress, and fetal death.
Outcomes after discharge included readmission to the hospital (after excluding admissions for subsequent delivery) for CHF, atrial arrhythmias, and serious ventricular arrhythmias at 1 year. Women with a primary residence outside California were excluded from this analysis. Because these outcomes were uncommon and because we found no significant differences between the complex and noncomplex CHD groups, both were analyzed as a single group for the 1-year follow-up.
Medical comorbidities were classified based on ICD-9 diagnosis codes for coronary artery disease, CHF, cerebrovascular disease, chronic kidney disease, hypertension, diabetes, and depression (eTable 4 in the Supplement).19,25 Patients were identified as having the comorbidity of interest if the diagnosis code was recorded for any health care encounter from January 1, 2005, to the admission date for delivery. Thus, these comorbidities included comorbidities diagnosed before pregnancy and those diagnosed during pregnancy but before the admission for delivery.
Data were analyzed from April 4, 2014, through January 23, 2017. Continuous variables are presented as mean (SD); categorical variables, as percentages. Continuous variables that were not normally distributed are presented as median (interquartile range [IQR]). We compared outcomes using multivariate logistic regression with the following covariates: age, race, cesarean delivery, multiple birth, maternal history of CHF, hypertension, coronary artery disease, chronic kidney disease, diabetes, and cerebrovascular disease. We assessed the association of CHD with readmission to the hospital for reasons other than subsequent delivery by using survival analysis, with failure defined as readmission to the hospital to 7 years after delivery. Women without a readmission were administratively censored on December 31, 2011. The association of CHD with the hazard of readmission was evaluated with Cox proportional hazards modeling. Beginning with an a priori set of predictors, backward selection was used to remove covariates with P < .1 in adjusted analyses. Two-tailed P < .05 was considered to be statistically significant. Owing to HCUP publication requirements,26 cell sizes of less than 10 are not reported, although the exact number is used in the analyses. Statistical analyses were performed using STATA/SE software (version 13.1; StataCorp).
Of a total of 27 907 535 inpatient hospitalizations in California from January 1, 2005, through December 31, 2011, we identified 3 642 041 admissions for women with a delivery diagnosis. After including codes for multiple births, we identified 3 702 838 live births accounting for 98.4% of all live births in California during the 7-year period.27
Among these admissions for delivery, we identified 3189 women with noncomplex CHD (mean [SD] age, 28.6 [7.6] years) and 262 women with complex CHD (mean [SD] age, 26.5 [6.8] years) (Table 1) with a variety of congenital defects (eTable 5 in the Supplement). The remaining 3 638 590 women without CHD (mean [SD] age, 28.3 [6.6] years) comprised the comparison group. A history of CHF was more common in women with CHD (19 [7.3%] with complex CHD, 111 [3.5%] with noncomplex CHD, and 4517 [0.1%] without CHD; P < .001). A history of arrhythmias was rare in all groups but somewhat more common in women with CHD.
Women with CHD were more likely to undergo cesarean delivery (121 [46.2%] with complex CHD, 1236 [38.8%] with noncomplex CHD, and 1 164 509 [32.0%] without CHD; P < .001). For tetralogy of Fallot, 56 of 102 women with CHD (54.9%) underwent cesarean delivery (eTable 5 in the Supplement).
Median length of stay was longer for women with CHD compared with women without CHD (3 [IQR, 2-4] vs 2 [IQR, 2-3] days; P < .001). In addition, 29 women with complex CHD (11.1%) and 191 women with noncomplex CHD (6.0%) were admitted for longer than 7 days, compared with only 49 089 women without CHD (1.3%). After multivariate adjustment, CHD was associated with greater odds of admissions longer than 1 week (Table 2). Preeclampsia or eclampsia was observed in 19 women with complex CHD (7.3%), 182 women with noncomplex CHD (5.7%), and 122 539 women without CHD (3.4%). After multivariate adjustment, noncomplex CHD was significantly associated with increased odds of preeclampsia or eclampsia (OR, 1.3; 95% CI, 1.1-1.5; P = .003); the association for complex CHD was not significant (OR, 1.5; 95% CI, 0.9-2.4; P = .12).
Incident CHF, atrial arrhythmias, serious ventricular arrhythmias, and maternal mortality were uncommon during hospitalization, each occurring in fewer than 0.5% of women with CHD (Table 2). After multivariate adjustment, noncomplex CHD (OR, 9.7; 95% CI, 4.7-20.0; P < .001) and complex CHD (OR, 56.6; 95% CI, 17.6-182.5; P < .001) were associated with greater odds of CHF. Similar associations for noncomplex CHD (OR, 8.2; 95% CI, 3.0-22.7; P < .001) and complex CHD (OR, 31.8; 95% CI, 4.3-236.3; P = .001) were found for atrial arrhythmias. After multivariate adjustment, only complex CHD was associated with greater odds of serious ventricular arrhythmias (OR, 35.3; 95% CI, 13.4-93.5; P < .001) or maternal in-hospital mortality (OR, 79.1; 95% CI, 23.9-261.8; P < .001) (Table 2).
Fetal growth restriction was reported in 14 women with complex CHD (5.3%), 76 women with noncomplex CHD (2.4%), and 48 765 women without CHD (1.3%). After multivariate adjustment, noncomplex CHD (OR, 1.6; 95% CI, 1.3-2.0; P < .001) and complex CHD (OR, 3.5; 95% CI, 2.1-6.1; P < .001) were associated with greater odds of fetal growth restriction. Fetal death occurred in no deliveries for women with complex CHD, 22 deliveries for women with noncomplex CHD (0.7%), and 13 017 deliveries for women without CHD (0.4%). After multivariate adjustment, noncomplex CHD was associated with greater odds of fetal death compared with women without CHD (OR, 1.7; 95% CI, 1.1-2.6; P = .02).
When women were followed up after discharge, 30 155 women without CHD (1.1%) and 166 women with CHD (5.6%) were readmitted to the hospital within 30 days (P < .001); the readmission rate exclusive of subsequent deliveries remained higher for women with CHD at 1 year (OR, 3.6; 95% CI, 3.3-4.0; P < .001) (Table 3) and to 7 years after discharge (χ2 = 3941, P < .001, log-rank test for comparing complex CHD, noncomplex CHD, and no CHD) (Figure 1). At 1 year, readmissions for CHF (OR, 6.7; 95% CI, 4.0-11.1; P < .001), atrial arrhythmias (OR, 10.9; 95% CI, 7.0-17.1; P < .001), and serious ventricular arrhythmias (OR, 6.2; 95% CI, 3.5-11.2; P < .001) were more common for women with CHD than for those without (Table 3). For readmissions within 1 year of delivery, the 10 most common diagnoses were noncardiac for women without CHD, whereas for women with CHD, 3 of the top 10 were cardiac (CHF, ostium secundum, atrial septal defect, and chest pain). After multivariate analysis, noncomplex CHD (hazard ratio, 3.88; 95% CI, 3.66-4.12) and complex CHD (hazard ratio, 4.51; 95% CI, 3.70-5.49) had the largest association with increasing the hazard of readmission (Figure 2).
In this study of more than 3.6 million admissions for delivery in California during a 7-year period that accounted for more than 98% of all live births, women with CHD were more likely to have a history of CHF, undergo cesarean delivery, and have longer lengths of stay. Incident maternal CHF, arrhythmias, and mortality were uncommon for all groups of women, but the odds of these outcomes were substantially greater for women with complex CHD. The adjusted odds of CHF, atrial arrhythmias, and maternal death were also increased for women with noncomplex CHD. At 1 year, readmissions for CHF, atrial arrhythmias, and serious ventricular arrhythmias were more common for women with CHD. After multivariate analysis, CHD had a greater associaton with increasing the hazard of readmission than did diabetes, chronic kidney disease, or hypertension. Despite these associations, the absolute risks for adverse cardiovascular outcomes among women with CHD were still low. The risks for incident CHF, supraventricular arrhythmias, serious ventricular arrhythmias, and maternal in-hospital mortality were less than 0.5% among women with CHD. The magnitude of the increases in relative risk among women with CHD appears to be high in part owing to comparisons against the rare levels of these complications in the general population.
Women with CHD more often underwent cesarean delivery, but the underlying reasons for this difference are unclear. Cesarean delivery is more commonly used in mothers who have delivered by cesarean delivery previously and in cases of breech presentation, dystocia, and fetal distress.28-30 Women with CHD were not more likely to have codes for fetal distress; therefore, increased use of cesarean delivery does not seem to be attributable to fetal instability. Cesarean deliveries may have been more likely to be planned owing to physician and/or patient preference because of the belief that the procedure has less effect on maternal physiology, although at present no definitive data on this issue exist. We found a higher rate of cesarean delivery among women with CHD (46.2% for complex CHD and 38.8% for noncomplex CHD) than did other studies. In a multicenter cohort of mothers with CHD from more than a decade ago, Siu and colleagues31 found that 27% of births were by cesarean delivery and 96% of these were for obstetric indications (compared with maternal cardiac status indication in only 4%).
We also found significantly greater odds of preeclampsia or eclampsia in women with noncomplex CHD and increased odds in women with complex CHD. Reduced placental perfusion may cause preeclampsia,32 which could be more common in women with CHD. Studies have shown that compared with the general population and women with repaired defects, women with an unrepaired atrial septal defect or ventricular septal defect have an increased risk for preeclampsia.33,34 In addition, even in women with CHD, preeclampsia or eclampsia is uncommon and only occurred in 7.3% of women in our complex CHD group.
Compared with the studies by Karamlou and colleagues35 and Thompson and colleagues,36 our study captured nearly every birth in California, as opposed to a representative sample of the United States obtained from the HCUP Nationwide Inpatient Sample. All 3 studies found higher rates of cesarean delivery and cardiac complications among women with CHD. Because of differences between the databases, the present study was also able to ascertain serious complications requiring readmission after discharge.
Cardiac complications were less commonly observed in this study than in the study by Siu and colleagues,31 who reported a primary cardiac event in 13% of completed pregnancies. However, more than half of these events occurred in the prepartum period, which was not evaluated in the present study. One review13 found cardiac events in 1.9%, CHF in 4.8%, and arrhythmias in 4.5% of deliveries in women with CHD. A Dutch national study of CHD37 identified cardiovascular events at 1 year in 6.4% of pregnancies, and 9 of 11 events were new-onset arrhythmias. In our contemporary statewide study of nearly every delivery to mothers with CHD, arrhythmias occurred in approximately 1% of women.
Prior studies of women with CHD have found that their offspring are at increased risk for premature birth, being small for gestational age, and neonatal mortality.13,38,39 We found greater odds of fetal growth restriction for women with CHD and greater odds of fetal death for women with noncomplex CHD. Patients with complex disease are at particular risk for intrauterine growth restriction, and monitoring for such complications requires vigilance by obstetricians. Owing to coding restrictions, we could not examine prematurity as an end point or as a mediator of the association between maternal CHD and neonatal mortality. Prior studies have also shown that miscarriages and terminations for medical reasons are more common in women with CHD,13,38 but this study did not capture any outcomes before the time of delivery.
This study also found a higher risk for maternal readmission among mothers with CHD at 30 days and 1 and 7 years. Although we examined the cause of readmission at 1 year, we did not do so at 7 years, and the higher readmission rates across long-term follow-up may be attributable to the natural history of CHD and not to complications of pregnancy.
Despite the large number of women with CHD captured in the study, this study has several limitations. First, ICD-9 codes have imperfect sensitivity and specificity,40 and CHD may have been undercoded. However, undercoding is unlikely to bias the results, because even a small number of women with CHD misclassified as not having CHD would not have a sizable effect on summary estimates of more than 3.6 million women. Second, the administrative nature of the database does not allow the medical records of mothers and neonates to be linked to determine the long-term health outcomes of these infants. Third, this analysis captured only women who chose to become pregnant and carried pregnancies to term. We did not evaluate women who chose not to become pregnant or who did not carry pregnancies to term, which may account for approximately 20% of pregnancies among women with CHD.13 Fourth, the number of women with CHD who delivered at low-volume centers was small, and multivariate analysis did not demonstrate an association of hospital delivery volume with outcomes for women with CHD. Fifth, the inpatient nature of this database did not allow us to capture intensity or quality of care before and after delivery. In a recent study of cardiovascular deaths related to pregnancy in California, health care professional factors were found to play a role in most cardiovascular deaths.41 Finally, for the follow-up analysis, only women who were readmitted to a California hospital were captured; therefore, women who moved or died outside the hospital were not captured.
In this study of nearly all women admitted to California hospitals for delivery during a 7-year period, women with CHD were more likely to undergo cesarean delivery and have a longer length of stay. Incident maternal CHF, arrhythmias, and mortality were uncommon in all groups of women, but CHD was significantly associated with incident CHF, atrial arrhythmias, fetal growth restriction, and hospital readmission. Complex CHD was a significant predictor of ventricular arrhythmias and maternal in-hospital mortality. These results may guide monitoring decisions and risk assessment for pregnant women with CHD at the time of delivery. Given the higher rate of cardiac and obstetrical complications, pregnant women with CHD should be treated in a center with expertise in adult CHD.
Corresponding Author: Zian H. Tseng, MD, MAS, Section of Cardiac Electrophysiology, University of California, San Francisco, 500 Parnassus Ave, Box 1354, San Francisco, CA 94143 (email@example.com).
Accepted for Publication: January 25, 2017.
Published Online: April 12, 2017. doi:10.1001/jamacardio.2017.0283
Author Contributions: Drs Hayward and Tseng had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: Hayward.
Drafting of the manuscript: Hayward, Tseng.
Critical revision of the manuscript for important intellectual content: Hayward, Foster.
Statistical analysis: Hayward.
Study supervision: Foster, Tseng.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Hayward reports receiving educational travel grants from Medtronic and Boston Scientific. Dr Foster reports receiving research support from Abbott Vascular. Dr Tseng reports receiving minor honoraria from Biotronik. No other disclosures were reported.
Funding/Support: This work was supported in part by 5R01 HL102090 from the National Heart, Lung, and Blood Institute, National Institutes of Health (Dr Tseng).
Role of the Funder/Sponsor: The sponsor had no role in the design and conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.
Create a personal account or sign in to: