Background

Innovation and technical advancement have revolutionized the field of pediatric and adult congenital heart disease (CHD) over the past century. As clinical outcomes have improved dramatically over time, however, healthcare disparities have persisted for the most vulnerable populations. Structural cardiac defects are the most common birth defect, affecting approximately 0.8–1% of the population.1,2,3 These birth defects range in complexity and occur across all the socioeconomic groups. With CHD mortality in infancy and childhood decreasing substantially with the evolution of advanced surgical and catheter-based interventions, >90% of children with CHD now survive into adulthood, and this large population of adults with CHD continues to grow with time.4 As a result, there are now more adults than children living with CHD in the US. Although survival has improved, CHD patients continue to face major socioeconomic and demographic disparities in outcomes at all ages.3

Social determinants of health (SDH) are conditions in which people live and grow up within the wider context of systems and influences shaping daily life.5 SDH include poverty, lack of insurance, housing instability, parental educational attainment, immigration status, food insecurity, and transportation barriers. These factors contribute to poor clinical outcomes, healthcare inequities, and escalating healthcare costs. The central importance of the association of SDH with health outcomes specifically in the context of cardiovascular diseases was underscored by the American Heart Association and American College of Cardiology in their 2019 guidelines for clinical risk assessment.6 There are no published systematic reviews, however, of the associations of SDH with major CHD outcomes across the lifespan, including fetal diagnosis; incidence and prevalence; infant mortality; post-surgical outcomes; access to care, loss to follow-up, and hospital readmissions; neurodevelopmental outcomes and quality of life (QOL); and adult CHD. The study aim, therefore, was to conduct a systematic review of the association of SDH with CHD outcomes.

Methods

SDH: definitions

The following SDH were included in this analysis: poverty, uninsurance, housing instability, parental educational attainment, immigration status, food insecurity, and transportation barriers. These were chosen because they are the domains addressed in a recently published SDH screening instrument used for interventions effective in reducing social risks and improving child and caregiver health.7 For articles in which there was no assessment of socioeconomic status (SES), Medicaid coverage was used as a proxy for low income.

Outcomes

The following CHD outcome categories across the lifespan were evaluated: fetal diagnosis; incidence and prevalence; infant mortality; post-surgical outcomes; access to care, loss to follow-up, and hospital readmissions; neurodevelopmental outcomes and QOL; and adult CHD.

Inclusion and exclusion criteria

Inclusion criteria consisted of published, original research on the associations of SDH with CHD. Exclusion criteria included: (1) letters to the editor, commentaries, editorials, viewpoints, perspectives, opinion pieces, case reports, book chapters, author or keyword indexes, and review articles; (2) publications that did not address clinical outcomes in patients with CHD; (3) articles focusing on acquired pediatric cardiac diseases, variants of normal, patent foramen ovale, primary arrhythmias, cardiovascular complications of connective tissue disorders, and pulmonary hypertension in the absence of structural CHD; (4) studies that classified SDH as race/ethnicity, maternal stress, environmental exposures, or vitamin or drug/alcohol exposures; (5) animal-only studies; (6) analyses of the association of CHD (as an independent variable) with SDH (as the dependent variable); and (7) articles on populations outside of the US or Canada (because the focus was on SDH in developed countries in North America with comparable healthcare systems).

Literature search

Using sentinel articles to harvest and test search terms, the following search strategy was developed for PubMed/MEDLINE to retrieve all records using natural language and controlled vocabulary (when available) relating to the association of SDH with CHD (Table 1). This strategy then was translated and adapted for the other databases. The following databases were searched from date of inception through May 18, 2020: PubMed MEDLINE (including Pre-MEDLINE and non-MEDLINE; 1945 to May 2020), Scopus (Elsevier; 1966 to May 2020), Cochrane Central Register of Controlled Trials (Wiley; through May 2020), CINAHL (Ebsco; 1981 to May 2020), PsycInfo (Ebsco; 1872 to May 2020), Social Interventions Research & Evaluation Network (SIREN) Evidence & Resource Library (University of California, San Francisco; through May 2020), and SocIndex (Ebsco; 1895 to May 2020). No filters were used for language or publication date. ProQuest RefWorks (Legacy version) was used to de-duplicate and manage all citations.

Table 1 Strategies for database searches (up to May 18, 2020).
Full size table

Once articles were identified and compiled by the search criteria described above and duplicates removed, vetting was performed by title and abstract by two authors using strict inclusion and exclusion criteria (Fig. 1). For studies for which there was lack of clarity regarding whether or not inclusion criteria were met, final decisions were made by reaching consensus among at least three authors. Once title and abstract vetting was completed, a full-text review was performed using the inclusion and exclusion criteria to determine final inclusion in the systematic review. Consensus opinion with regards to inclusion of studies was again reached when questions arose.

Fig. 1
figure 1

Flow diagram of study selection (adapted from PRISMA), including study identification, screening, eligibility, and final inclusion procedures and numbers.

Full size image

PROSPERO registration

This systematic review was registered in PROSPERO (CRD42020169253).

Results

The initial search generated 3992 citations. A total of 88 studies met inclusion criteria (Fig. 1 and Table 2). Study designs were variable and included retrospective chart reviews, retrospective and prospective cohort studies, cross-sectional studies, and prospective case–control studies. No studies were identified that examine housing instability.

Table 2 Summary of the included studies on the association of social determinants of health with congenital heart disease (CHD).
Full size table

The sections that follow report the findings on SDH associations with seven major CHD outcomes across the lifespan: fetal diagnosis; incidence and prevalence; infant mortality; post-surgical outcomes; access to care, loss to follow-up, and hospital readmissions; neurodevelopmental outcomes and QOL; and adult CHD.

Fetal diagnosis of CHD

Four articles were identified that examined SDH associations with the fetal diagnosis of CHD (Table 2).8,9,10,11 Three studies documented that poverty or low SES is associated with a significantly lower likelihood of a prenatal CHD diagnosis; one study also found that low maternal educational attainment and public insurance were SDH risk factors for no prenatal CHD diagnosis. An analysis of 444 patients presenting to Boston Children’s Hospital with critical CHD (defined as surgical or catheter intervention required in infants ≤30 days old) revealed that only 35% of those in the lowest SES composite-score quartile received a prenatal CHD diagnosis, vs. 62% of those in the highest SES quartile. A retrospective study of >2.5 million infants born in California revealed that the lowest income tertile, public insurance, and low maternal educational attainment were associated with a significantly higher likelihood of CHD. In a study of 535 patients presenting to Children’s Hospital of Wisconsin with CHD, residing in a higher poverty zip code was associated with a significantly lower odds of a prenatal CHD diagnosis. One study of 100 Cincinnati infants with CHD found no association of family income, parental educational attainment, or insurance coverage with prenatal CHD diagnosis, although the sample size (100) was limited and multivariable analyses were not performed.

CHD incidence and prevalence

Fifteen articles examined the association of SDH with CHD incidence and prevalence (Table 2).12,13,14,15,16,17,18,19,20,21,22,23,24,25,26 Poverty generally was found to be significantly associated with CHD incidence and prevalence, but such associations were either equivocal or lacking for other SDH examined, including food insecurity, immigration, and parental educational attainment.

Eleven articles analyzed an association between low SES and an increased CHD incidence or prevalence, and most found that SES was significantly associated with CHD incidence or prevalence. SES definitions, however, varied among the studies, with low SES defined as individual poverty, low family income, neighborhood poverty, maternal educational attainment, parental employment, operator/laborer occupation, crowding, rental occupancy, or some combination thereof.12,13,14,15,17,18,19,20,21,25,26 A study of 1.9 million children born in Ontario, Canada, revealed that birth in low SES areas was associated with significantly higher CHD rates (rate ratio = 1.20; 95% confidence interval [CI] = 1.15–1.24).12 A population-based study of 2.4 million live-born infants in California documented that those residing in neighborhoods with the lowest SES composite score had a significantly higher CHD incidence vs. those from the highest SES neighborhoods (adjusted odds ratio (OR) = 1.31; 95% CI, 1.21–1.41).25 Low SES also was found to be associated with a significantly higher CHD incidence risk in studies using cardiology clinic registries13 and national databases.17

Two articles used the Nationwide Inpatient Sample (NIS) to examine secular trends in CHD prevalence, but reached different conclusions. One study demonstrated that those in the upper income quartile experienced a significantly greater temporal decrease in the prevalence of severe CHD vs. those in the lowest income quartile,19 whereas another found that mild CHD prevalence significantly increased only in the high SES group.20 Another population-based study using the NIS reported that the overall CHD incidence was actually significantly lower in the lowest SES group, although the authors speculated that this may have been due to lower access to hospitals with better diagnostic tools.21

A study on food insecurity as a risk factor for conotruncal heart defects reported that food insecurity was associated with higher adjusted odds of d-transposition of the great arteries, but only among normal-weight and underweight mothers (and not those who were overweight or obese); no association of food insecurity, however, was found with tetralogy of Fallot.16 An analysis of the National Birth Defects Prevention Study revealed that having immigrant parents was associated with significant lower odds of certain CHDs, with the greatest number of significantly protective adjusted ORs noted for foreign-born parents residing ≤5 years vs. >5 years in the US.22 Another study, however, found no association of maternal birthplace with left ventricular outflow-tract malformations.24 Two studies found no association of maternal educational attainment with CHD prevalence.23,24

Infant mortality

Nine articles analyzed associations between SDH and infant mortality in CHD patients (Table 2). Poverty, low parental educational attainment, uninsurance, transportation barriers, and immigration status were significantly associated with infant mortality.1,2,3,27,28,29,30,31,32

A study of 229 children with hypoplastic left heart syndrome (HLHS) identified via the Metropolitan Atlanta Congenital Defects Program documented survival rates that were almost three times worse for those residing in high-poverty (9%) vs. low-poverty (25%; P < 0.001) neighborhoods.32 An analysis of data from birth-defect surveillance programs from four states (Arizona, New York, New Jersey, and Texas) on almost 10,000 infants with CHD revealed that poverty was associated with about double the adjusted odds of infant mortality.3

Low maternal educational attainment was associated with a significantly higher risk of CHD infant mortality in three studies.1,3,30 For example, one study of coarctation of the aorta revealed a mortality rates of 27% for infants of mothers who had not completed high school vs. 5% for those who at least completed high school (P = 0.004).30 Research on 4390 infants with CHD also documented that lower paternal educational was associated with a 62% increased risk of infant mortality.31

An analysis of the Texas Birth Defects Registry revealed that uninsured infants with critical and noncritical CHDs had approximately triple and double the risk of neonatal mortality, respectively, compared with infants with private insurance.29 Another Texas study found that residing in a county bordering Mexico was associated with higher adjusted odds of CHD infant mortality.28

Post-surgical outcomes

A total of 25 articles evaluated the association of SDH with post-surgical outcomes in CHD patients (Table 2). Poverty and low SES were consistently associated with adverse post-operative outcomes, including worse HLHS survival,33 increased in-hospital mortality and resource utilization after orthotopic heart transplant for single-ventricle vs. cardiomyopathy patient cohorts,34 higher inter-stage mortality in the Single Ventricle Reconstruction Trial,35 higher mortality following congenital heart surgery,36 worse 1-year transplant-free survival after the Norwood procedure (stage I palliation for single-ventricle CHD),37 unplanned readmission in the first 90 days after congenital heart surgery,38 longer length of stay and higher resource utilization in patients who underwent congenital heart surgery,39,40 and lower school functioning and QOL.41

One study found that lower maternal educational attainment was associated with a lower Mental Developmental Index score in children who underwent the Norwood procedure,42 but another study found that maternal educational attainment was not associated with neurodevelopmental outcomes in multivariable analyses.43 Two studies found that distance to the hospital was actually associated with a lower risk of readmission post-operatively.38,44 One study also found no association of uninsurance with post-surgical outcomes.44

Access to care, loss to follow-up, and hospital readmissions

Nine articles examined the association of SDH with access to care, loss to follow-up, and hospital admissions (Table 2). Poverty/low SES, transportation barriers, parental educational attainment, and immigrant status were significantly associated with these outcomes. Eight studies documented significant associations of poverty/low SES with these outcomes, including increased risk of missed appointments, loss to follow-up, and hospital readmissions, as well as increases over time of the proportion of admissions to and bed days in pediatric cardiology specialty-care centers. For example, a study of 1034 patients in a large urban pediatric hospital in the Midwest revealed that Medicaid coverage was associated with a significant higher adjusted odds of missing at least one scheduled annual cardiology clinic follow-up visit.45 A study of nearly 800 patients showed that Medicaid coverage and lower median household income were associated with double the unadjusted odds of missed appointments for cardiac magnetic resonance imaging.46 A matched case–control study on risk factors for loss to cardiology clinic follow-up among children and young adults with CHD documented a 1.2 times greater odds of loss to follow-up for every $10,000 reduction in family income.47

Two studies examined transportation barriers and found having to travel ≥200 miles was associated with missed appointments (significantly in bivariate analysis, but with a non-significant trend in multivariable analysis),48 and residence in rural poor communities was associated with the longest mean drive time (69 min) to cardiology clinics.49 One study of a pediatric cardiology outreach clinic for immigrant and refugees found a no-show rate that was higher than the national benchmark.50

Neurodevelopmental outcomes and QOL

Eight articles examined the association of SDH with neurodevelopmental outcomes and QOL in children with CHD and their parents (Table 2). Poverty, parental educational attainment, and transportation barriers were significantly associated with worse neurodevelopmental outcomes and QOL in most studies. Five articles found that poverty/low income was significantly associated with adverse neurodevelopmental or QOL outcomes, including decreased intelligence quotient (IQ), socialization, adaptive behavior, cognition, parental perceived cognitive problems, genetic knowledge, grade-level literacy and math proficiency, memory, and family QOL. For example, an analysis of Arkansas data on children who had CHD surgery at <1 year old found that poverty was associated with double the adjusted odds of not achieving grade-level proficiency in literacy and triple the adjusted odds of not achieving grade-level proficiency in math.51

Lower maternal educational attainment was significantly associated with lower child performance IQ, socialization, adaptive behavior, and cognition in one study52 and with lower grade-level proficiency in literacy in another study.51 Two studies, however, found no association of maternal educational attainment with grade-level proficiency in math or with screening positive on a measure of autism spectrum disorder.51,53

A recent study of 140 parents of young children found that, even when accounting for the severity of the child’s CHD defect (ranging from an innocent murmur to CHD treatment necessitating cardiopulmonary bypass), low income was associated with a significantly lower family QOL.54 Another study found that SES was not associated with parental stress.55

One study showed that a greater travel distance to the hospital was associated with double the adjusted odds of not achieving grade-level proficiency in literacy, but no such association was found for math proficiency.51

Adult congenital heart disease (ACHD)

Nineteen articles examined associations of SDH with ACHD outcomes. Poverty/low income (13 studies), uninsurance (5 studies), educational attainment (3 studies), and transportation barriers (3 studies) were significantly associated with adverse ACHD outcomes (Table 2). Poverty/low income was significantly associated with a variety of adverse ACHD outcomes, including hospital admissions,56 hospital readmission57 and death after ACHD surgery,58 higher inpatient resource utilization,59 physical QOL,60 worse neurocognitive test performance,61 endocarditis-related hospitalizations,62 surgical complications,63 and missed clinic appointments.64 For example, analyses of national databases documented double the odds of inpatient death for low-income (Medicaid) patients after ACHD surgery58 and that patients in the lowest income quartile had significantly higher adjusted odds of hospitalization for infective endocarditis vs. the next income quartile.62

All five studies on uninsurance found significant associations with adverse ACHD outcomes, including significantly greater odds of hospitalization,56,65 outpatient loss to follow-up,66,67 unsuccessful transfer of care from pediatric to adult congenital cardiology care,66 and hospitalization for infective endocarditis.62 For example, one study found that uninsured ACHD patients were significantly less likely to have their pediatric care transferred to ACHD cardiologists, at only 8%, and most likely to have no follow-up, at 74%.66

Three studies found that ACHD patient educational attainment was significantly associated with adverse ACHD outcomes, including lower purpose-of-life scores,68 decreased exercise frequency,69 and residing farther from an ACHD center.70 Three studies also examined the association of transportation barriers with adverse ACHD outcomes. One found that transportation barriers were significantly associated with performance of ACHD surgery outside of an ACHD specialty center.71 Another study revealed that uninsurance, poverty, and lower educational attainment were significantly associated with ACHD patients with >6-h drive to the nearest ACHD center.70 The third study, however, found no association of driving distance with attendance at ACHD outpatient clinic appointments.72

Discussion

This systematic review documented that a wide variety of SDH are significantly associated with adverse outcomes across the lifespan of CHD patients, from prenatal diagnosis to ACHD. Indeed, the study findings dramatically underscore that SDH are significantly associated with many of the most important and serious CHD outcomes, including a lower likelihood of prenatal diagnosis, increased CHD incidence, higher infant mortality, worse post-surgical outcomes, greater inpatient resource utilization, more missed clinic appointments, increased loss to follow-up, lower performance IQ, worse cognition, decreased grade-level proficiency in literacy and math, reduced family QOL, a higher risk for ACHD endocarditis, more ACHD hospitalizations and hospital readmissions, unsuccessful transfer of care from pediatric to adult congenital cardiology care, and increased odds of complications and death after ACHD surgery.

These study findings indicate that an urgent priority and one of the most important interventions for CHD patients would be routinely screening for SDH, with referrals to appropriate services for those who screen positive. The study results suggest that this SDH screening and referral should occur in all CHD care settings, including prenatal visits, neonatal intensive care units and pediatric intensive care units, primary-care and specialty practices, and ACHD clinics. Major national organizations, including the American Academy of Pediatrics, American College of Cardiology, American Academy of Family Physicians, and National Academy of Sciences, Engineering, and Medicine, have all endorsed SDH screening and referral to appropriate services.6,73,74,75,76 Research shows that patients and caregivers are comfortable with SDH screening.73,77,78,79,80 A recent study showed that SDH screening and referral can reduce the number of SDH and improve child health.7

Parent mentors are an evidence-based intervention that has the potential to prove effective in both reducing SDH and improving outcomes for children with CHD and their families. Parent mentors are a special category of community health workers who already have a child with a particular condition (such as CHD) who then receive training to help other parents with children with that condition, including obtaining appropriate healthcare and addressing SDH. A randomized, controlled trial (RCT) of the effects of parent mentors on children with asthma and their families revealed that parent mentors were associated with significant reductions in wheezing, asthma exacerbations, emergency-department visits, and missed parental work days, while improving parental self-efficacy, and saving money.81 Another RCT of a parent-mentor intervention to enroll uninsured children documented that parent mentors are significantly more effective than traditional Medicaid/CHIP outreach and enrollment methods in insuring uninsured minority children; obtaining insurance faster; renewing coverage; improving access to primary, dental, and specialty care; reducing unmet needs and out-of-pocket costs; achieving parental satisfaction and care quality; and sustaining long-term coverage; they also saved $6045 per insured child per year, an 850% return on investment.82 This RCT resulted in federal legislation in the 2018 CHIP Reauthorization bill83 and $120 million in Centers for Medicare and Medicaid Services funding for parent mentors.84,85,86 Thus, parent mentors could analogously prove to be highly effective in addressing SDH in children with CHD and their families.

Study findings on the associations of SDH with ACHD have important implications for practice, research, and policy. CHD has morphed from a critical disease among children to a chronic condition in which the number of ACHD patients (~1.3 million) now exceeds the number children with CHD.9,58,66 Given that at least 85% of children with CHD survive to adulthood, there is an urgent need to provide high-quality specialty care to the growing ACHD population.68 Over time, the number of ACHD hospitalizations has doubled, from ~36,000 in 1998 to >72,000 in 2005.58 Furthermore, the increasing complexity of ACHD has warranted creation of an ACHD subspecialty for centers treating ACHD. SDH screening and appropriate referral to services is thus increasingly critical for ACHD patients. The study results also underscore the importance of consistently considering SES as well as SDH in general when examining health and healthcare outcomes for fetuses, children, and adults with CHD. Furthermore, the study findings suggest that additional research is warranted on the association between SDH and CHD in other developed countries and in developing nations, as well as country comparative studies, particularly regarding the impact of variations in welfare state configurations. Until such research is conducted, caution should be exercised regarding generalizing our study results beyond populations in the US and Canada.

This systematic review revealed several unanswered questions. No published studies were identified on the association of housing instability with CHD outcomes, and a paucity of research was noted on several SDH, including food insecurity, transportation barriers, and lack of health insurance, so more research is need on these topics. The fewest number of studies was noted for fetal diagnosis of CHD, so more investigations are needed of which specific SDH are associated with CHD fetal diagnosis and that  provide a deeper exploration for the reasons behind these associations. Although several studies found associations of low maternal educational attainment with infant mortality and other CHD outcomes, only a single study examined paternal educational attainment, so an ongoing unanswered question is whether and how low paternal educational attainment is associated with CHD outcomes.

Based on the findings of this systematic review, a research agenda is proposed. More studies are needed on the unanswered questions noted above. Research is needed on whether multiple SDH are associated with even worse CHD outcomes and how the various SDH might interact. For example, would an uninsured child with household poverty, food insufficiency, and low parental educational attainment be at especially high risk for adverse CHD outcomes? Studies are needed on whether SDH screening and referral to appropriate services results in reduction of SDH and improved outcomes. RCTs are urgently needed of innovative interventions, such as parent mentors, that might eliminate SDH and achieve better outcomes for children and adults with CHDs and their families. More research also is warranted on interventions tailored to reducing SDH for ACHD patients.

Conclusion

SDH are significantly associated with adverse outcomes across the lifespan of CHD patients, from prenatal diagnosis to ACHD. The study findings dramatically underscore that SDH are significantly associated with many of the most important and serious CHD outcomes, including a lower likelihood of prenatal diagnosis, increased CHD incidence, higher infant mortality, worse post-surgical outcomes, greater inpatient resource utilization, more missed clinic appointments, increased loss to follow-up, lower performance IQ, worse cognition, decreased grade-level proficiency in literacy and math, reduced family QOL, a higher risk for ACHD endocarditis, more ACHD hospitalizations and hospital readmissions, unsuccessful transfer of care from pediatric to adult congenital cardiology care, and increased odds of complications and death after ACHD surgery. SDH screening and referral to appropriate services has the potential to improve outcomes for CHD patients across the lifespan. RCTs are urgently needed of innovative interventions, such as parent mentors, that might eliminate SDH and achieve better outcomes for children and adults with CHDs and their families.