Among the 81 questions on the entire CH STRONG survey, survey developers included 39 questions from U.S. national and state-based surveys, including the National Health Interview Survey, National Health and Nutrition Examination Surveys, Behavioral Risk Factor Surveillance System, and American Community Survey. Among the included reproductive health survey questions, one (“Have you ever been pregnant?”) was sourced from National Health Interview Survey. However, CH STRONG was one of the first projects to survey women with CHDs on their sexual and reproductive health within the context of their heart condition, so there were no existing national surveys available from which to draw the remaining CHD-specific questions to assess reproductive health. For these and any other questions unique to CH SRONG, survey developers ensured use of plain language to improve readability.

Three CH STRONG questions inquired about receipt of reproductive health counseling by a clinician: 1) “Has a doctor, nurse, or other health care worker ever talked with you about the safest type of birth control or contraception to use because of your heart problem?”; 2) “Has a doctor, nurse, or other health care worker ever talked with you about special concerns about becoming pregnant because of your heart problem?”; and 3) “Has a doctor, nurse, or other health care worker ever advised you to avoid pregnancy because of your heart problem?”

We also examined three questions on reproductive health concerns and outcomes: 1) “How concerned are you about your ability to have children?”; 2) “Have you ever delayed or avoided getting pregnant because of concerns about your heart problem?”; and 3) “Have you ever been pregnant?”

Sociodemographic characteristics for this analysis included state of birth (Arkansas, Arizona, or Georgia), self-reported race/ethnicity (Hispanic, non-Hispanic Black, non-Hispanic white, or none of the above), insurance status (private, public, other, or none), education level (less than high school, high school graduate, or some college or above), marital status (has ever been married or never married), and, based on census tract–level data for the respondent’s current address, residential rurality (>50% urban vs. ≤50% urban) and percentage of families in census tract with household income below the federal poverty level (<10% of families, 10%–40% of families, or >40% of families). CH STRONG also included a question asking whether the survey was completed by proxy (e.g., by a parent).

We categorized race/ethnicity based on two survey questions: 1) “Do you consider yourself to be Hispanic or Latino?”, with available responses “Yes” or “No”; and 2) “What race or races do you consider yourself to be? One or more categories may be selected” with options “American Indian or Alaskan Native,” “Asian,” “Black or African American,” “Native Hawaiian or Pacific Islander,” and “White.” Eight respondents did not report race or ethnicity. So as not to exclude these cases, we used the maternal race/ethnicity reported on the birth certificate. All those indicating Hispanic ethnicity are categorized as “Hispanic,” regardless of their reported race. To protect the identity of respondents, race/ethnicity categories were aggregated for those containing fewer than 10 respondents. The “None of the Above” race/ethnicity category includes respondents from the Non-Hispanic American Indian or Alaskan Native, Non-Hispanic Asian, Non-Hispanic Native Hawaiian or Pacific Islander, and Non-Hispanic Multiple Races groups.

Health characteristics included presence of cardiac and non-cardiac comorbidities, based on survey report, and documentation of a comorbid chromosomal abnormality (e.g., Down syndrome), derived from data from the birth defects registries. Cardiac comorbidities were identified by survey response to “Have you ever been told by a doctor or other health professional that you had any of the following conditions?: congestive heart failure; cardiac dysrhythmias or irregular heartbeat; a heart attack (also called myocardial infarction); stroke; hypertension, also called high blood pressure.” Non-cardiac comorbidities were identified by survey response to the question: “Have you ever been told by a doctor or other health professional that you had any of the following conditions?: diabetes or sugar diabetes; obstructive sleep apnea; cancer or a malignancy of any kind; a mood disorder or depression; asthma; ulcer (stomach, duodenal, or peptic ulcer); arthritis, gout, lupus, or fibromyalgia.”

Initially, we examined the prevalence of sociodemographic characteristics by CHD severity using χ² tests. In a previous analysis, CH STRONG respondents largely reflected the eligible population. However, respondents were slightly more commonly non-Hispanic white and had severe CHDs compared with non-respondents (Glidewell et al., 2018). To minimize response bias and improve generalizability, we standardized prevalence estimates of respondents to the entire population of eligible women (n = 4,515) by maternal race/ethnicity and CHD severity, as well as birth year and site. Maternal race/ethnicity reported on the birth certificate was used for standardization, as this was available for all eligible participants, versus reported race/ethnicity, which was only available for respondents. Standardized prevalence estimates and 95% confidence intervals (CIs) were calculated by CHD severity for all reproductive health counseling and outcomes. Mean proportions of those with severe CHDs and those with non-severe CHDs were compared using z-scores and corresponding p-values. Additionally, the percentage of women in CH STRONG reporting ever being pregnant was compared with the general population of women ages 19 to 38 years who reported ever being pregnant using data from the 2017 National Health Interview Survey (NHIS) (National Center for Health Statistics, 2017). To account for complex sampling, NHIS-specified weighting was used to generate a population-based estimate and 95% CI.

Adjusted prevalence ratios (aPRs) and 95% CIs for reproductive health counseling and outcomes were calculated using multivariable Poisson regression. Models for reproductive health counseling and outcomes included all considered variables, except no women with chromosomal abnormalities reported ever being pregnant, so this variable was removed from the model. Two sensitivity analyses were conducted: 1) excluding all individuals with proxy-reported survey responses, and 2) excluding those with known chromosomal abnormalities. All analyses were performed in R version 4.2.1 (R Core Team, 2021).

Of the 894 female respondents, 14.4% (n = 129) were excluded from analysis due to missing data. The 765 women included in the analysis differed from those excluded with respect to pregnancy history, state of birth, insurance status, education, and proxy report (Supplemental Table 1). Of the analytic sample, 28.5% (n = 218) had severe CHDs, but proportions differed across sites (Table 1). A larger percentage of women with severe CHDs had chromosomal abnormalities and proxy-reported surveys. No other characteristics differed by CHD severity.

After standardization, 21.3% of the sample reported receiving counseling on safe contraceptive methods. Compared with women with non-severe CHDs, more women with severe CHDs reported receiving safe contraceptive counseling (44.0% vs. 13.7%), clinician counseling on pregnancy (63.3% vs. 16.5%), and clinician advice to avoid pregnancy (32.0% vs. 6.4%; all p < .01; Figure 1; Supplemental Table 2). Overall, 33.6% of women with CHDs were concerned about their ability to have children, and concerns were more prevalent among women with severe CHDs than those with non-severe CHDs (40.9% vs. 31.2%, p = .01). Accordingly, more women with severe CHDs reported avoiding or delaying pregnancy than did those with non-severe CHDs (26.6% and 10.7%, p ≤ .001). Although a history of pregnancy was lower among the women with severe CHDs (30.0% vs. 37.2%, p = .07), this did not meet statistical significance. Additionally, a lower percentage of women in CH STRONG (35.4%, 95% CI [32.2, 38.7%]) reported ever being pregnant compared with similarly aged women in the general population (54.1%; 95% CI [51.8, 56.4%]; p < .001).

After adjustment for potential confounders, CHD severity was still associated with reporting clinician counseling about safe contraceptive methods (aPR = 3.0; 95% CI [2.2, 4.0]), clinician counseling on pregnancy (aPR = 3.6; 95% CI [2.7, 4.6]), and advice to avoid pregnancy (aPR = 4.3; 95% CI [2.9, 6.6]; Table 2). Prevalences of clinician counseling about safe contraceptive methods (aPR = 1.5; 95% CI [1.0, 2.0]), clinician counseling on pregnancy (aPR = 1.2; 95% CI [0.9, 1.6]), and advice to avoid pregnancy (aPR = 1.4; 95% CI [0.9, 2.1]) were higher among women with at least one cardiac comorbidity, although the differences were not statistically significant as all three CIs included 1.0. Women living in the highest-poverty census tracts were more likely to report receiving pregnancy counseling about their heart condition (aPR = 1.7; 95% CI [1.0, 2.9]). More women who had ever been married reported receiving contraceptive counseling (aPR = 1.7; 95% CI [1.2, 2.5]), pregnancy counseling (aPR = 1.6; 95% CI [1.1, 2.2]), and clinician advice to avoid pregnancy (aPR = 1.4; 95% CI [0.9, 2.4]), although the 95% CI included 1.0 for the last of these. Women with proxy report were less likely than self-respondents to report receiving clinician counseling on pregnancy (aPR = 0.5; 95% CI [0.3, 0.9]). Women who were identified in MACDP were more likely to report receiving clinician counseling about safe contraceptive methods compared with women identified from the Arkansas birth defect registry (aPR = 1.6; 95% CI [1.1, 2.4]). No other characteristics were associated with reproductive health counseling or advice.

More women with severe CHDs were concerned about their ability to have children (aPR = 1.4; 95% CI [1.1, 1.8]) and delayed or avoided pregnancy due to concerns about heart health (aPR = 2.2; 95% CI [1.5, 3.2]; Table 3). Having at least one non-cardiac comorbidity was associated with concern about ability to have children (aPR = 1.5; 95% CI [1.1, 1.9]), but it was not associated with ever being pregnant (aPR = 1.0; 95% CI [0.7, 1.2]). Women with CHDs were more likely to report pregnancy if they had no insurance (aPR = 1.6; 95% CI [1.1, 2.4]). Those who had ever been married were more likely to have ever been pregnant than their counterparts (aPR = 2.9; 95% CI [2.1, 3.8]). Women with proxy report were less likely than self-respondents to have ever been pregnant (aPR = 0.2; 95% CI [0.1, 0.4]).

Associations did not substantially differ in sensitivity analyses excluding 71 women with chromosomal abnormalities or separately excluding 119 women with proxy-reported survey responses (Supplemental Tables 3 and 4).

This population-based survey highlights gaps in reproductive health counseling. Fewer than 30% of women with CHDs reported that they received guidance from health care providers in this area. Fewer than half of those with severe CHDs, who face elevated risks with hormonal contraceptives, reported receiving clinician counseling about safe contraceptive methods. Yet, one-third of these women reported advice by a health care provider to avoid pregnancy.

Furthermore, more than a third of women with CHDs were concerned about their ability to have children, but less than a third reported receiving clinician counseling on pregnancy related to their heart condition. Women with severe CHDs were more likely to report receiving advice by a health care provider to avoid pregnancy, have concerns about having children, and to delay or avoid pregnancy. Significantly fewer women with CHDs reported at least one pregnancy compared with similarly aged women in the general population.

More than half of women with CHDs surveyed in previous studies conducted at congenital cardiology clinics reported receiving counseling about contraception, but those studies were performed in clinic populations that accessed specialized health care (Andrews et al., 2023; Kovacs et al., 2008; Miner et al., 2017). Consequently, the results of those studies are generalizable to less than half of the CHD population. Additionally, a recent study analyzing perceptions of sexual and reproductive health care among women with CHDs utilized in-depth interviews to identify key themes. Women with CHDs expressed concerns for safe contraceptive options tailored to their CHD and a desire for reproductive health counseling specific to their condition (Stokes et al., 2023). Our study also suggests that health care providers providing care for women with CHDs may not routinely discuss these reproductive health topics with women with CHDs, even though guidelines advise that they do so (ACOG Practice Bulletin No. 212: Pregnancy and Heart Disease, 2019; Canobbio et al., 2017).

The American Heart Association/American College of Cardiology, as well as the American College of Obstetricians and Gynecologists, recommend multidisciplinary reproductive health counseling of women with CHDs by both cardiologists and obstetrician-gynecologists (ACOG Practice Bulletin No. 212: Pregnancy and Heart Disease, 2019; Canobbio et al., 2017). Women with CHDs have also expressed appreciation for multidisciplinary care that included collaboration between their cardiologist and women’s health providers before, during, and after pregnancy (Stokes et al., 2023). Preconception counseling should include an individualized pregnancy plan with consideration of social factors like insurance and distance from specialty care (Canobbio et al., 2017; Miner et al., 2017). Providerfacilitated discussions about safe contraception should account for special considerations for different CHD types (Canobbio et al., 2017). For example, increased thrombotic risk must be considered when prescribing estrogen-based oral contraceptives for women with Fontan palliation, a surgical repair for hypoplastic left heart, a severe CHD (Miner et al., 2017; Pundi et al., 2016; Thorne et al., 2006).

Women with CHDs have higher risk of adverse pregnancy-related outcomes compared with those without CHDs, especially those with cardiac comorbidities; however, published obstetric care recommendations tailored to specific CHD types and status, per World Health Organization (WHO) pregnancy risk category, improve the likelihood of a healthy pregnancy (Canobbio et al., 2017; Downing et al., 2020, 2023; Farr, Downing, et al., 2023; Hardee et al., 2021; Kops et al., 2021; Roos-Hesselink et al., 2019). Pregnancy is contraindicated only for a minority of women with WHO pregnancy category IV status. This cardiac status, which includes severe mitral stenosis, severe aortic stenosis, unrepaired severe coarctation, severe aortic dilation, and previous pregnancy-related cardiomyopathy, is rare and comprises a very small proportion of women with CHDs (Canobbio et al., 2017). Still, one-third of women with severe CHDs in CH STRONG were advised by a health care provider to avoid pregnancy, and just over a quarter of women with severe CHDs chose to delay or avoid pregnancy due to their cardiac diagnosis. Provider preconceptions of women with CHDs may be a reason for the disproportionate percentage of women in CH STRONG reporting provider advice to avoid pregnancy, a phenomenon that has been documented in women with disabilities (Farr, Galindo, et al., 2023; Horner-Johnson et al., 2019).

Shared decision-making between women with CHDs and health care providers plays an important role in advancing reproductive health care. By acknowledging potential biases and prioritizing patients’ preferences, providers can cultivate a more collaborative and patient-centered approach. Evidence-based continuing medical education for clinicians (including cardiologists, primary care providers, and obstetricians/gynecologists) can further enhance access to comprehensive reproductive health services for the growing population of women with CHDs. A valuable resource in this effort is the American Academy of Pediatrics and CDC’s Congenital Heart Defects Toolkit (Congenital Heart Defects Toolkit, 2024).

In CH STRONG, efforts were made to include individuals reflective of the population with CHDs, not just those receiving care; however, the low response rate makes response bias a concern. To minimize this, estimates were weighted according to characteristics of the entire population of recruited women. We excluded 14.4% of respondents missing data on variables of interest. Women with more than a high school education, with insurance coverage, and who self-reported survey answers were more likely included in the analytic sample. Also, counseling could have been conducted but not recalled. However, counseling not recalled by a patient is likely ineffective. Additionally, all women included in this study were of childbearing age, and counseling is recommended throughout a woman’s reproductive life. Site of birth may not reflect current residence. Nevertheless, findings, especially regarding clinical practices, may be limited to Arizona, Arkansas, and Georgia and the surrounding geographic areas and influenced by local policies and practices.

A limited number of reproductive health matters were assessed to keep the survey length manageable. Questions about timing of counseling and reproductive health experiences, aside from pregnancy, were not included, diminishing the ability to analyze outcomes relative to events. Further assessment of age at counseling and detailed pregnancy and birth outcomes in a larger cohort would increase understanding.