Clinical characteristics and healthcare utilization of adults with congenital heart disease on Medicaid insurance in the United States Authors: Lidija B McGrath, Claire Cambron, Robin L Baudier, Abigail M Khan Published: 2026 ABSTRACT AbstractIntroductionPoverty is linked to an increased risk of acquired heart disease, but its impact on outcomes in adults with congenital heart disease (CHD) is not well defined. Medicaid insurance is a strong proxy for poverty in United States. We characterize population characteristics and healthcare utilization among adults with CHD on Medicaid insurance, both overall and by CHD disease complexity.MethodsThe Transformed Medicaid Statistical Information System (T-MSIS) collects comprehensive public health insurance data from all 50 states and territories (Medicaid and Children's Health Insurance Program). We identified 260,535 adults (>18 years of age) with ICD-10 diagnosis codes consistent with CHD across the years 2016-2018. Descriptive statistics were used to compare groups.ResultsOf 260,535 adults with CHD studied, 63.1% were female with a mean age of 40.2 ± 19.1 years. 44.5% had moderate or complex CHD, and 5.8% died during the study period. Medical and cardiac comorbidities were common: rhythm disorders (46.7%), mental health diagnoses (44.9%), hypertension (44.6%), and metabolic syndrome (42.4%). One quarter (24.8%) had a disability, 15.5% had a diagnosis consistent with substance abuse, and 17.7% used nicotine. 23.9% had incidence equivalents of one or more hospitalizations per year, and 32.1% had incidence equivalents of ≥5 emergency department visits in three years.ConclusionWe characterized the population of adults with CHD on Medicaid and found high rates of medical comorbidities, disability, smoking, substance use, and utilization of emergency department services. This vulnerable population may benefit from targeted interventions aimed at reducing comorbid conditions and thus improving long term outcomes.Keywords: Medicaid, Congenital heart disease, Poverty, Comorbidities, Healthcare utilization Background The population of adults with congenital heart disease (CHD) in the United States (US) is currently estimated at almost 2 million. The health system burden of adult congenital heart disease (ACHD) is anticipated to increase over time, related to both increasing population size and the accumulation of acquired comorbidities as the population ages [1]. Understanding the size, clinical characteristics, and healthcare utilization of adults with CHD is essential for modeling the future resource needs of this complex population. Non-medical factors have been shown to impact health outcomes in both CHD and non-CHD populations [2,3]. Specifically, poverty has been linked to worse health outcomes in the general population, including increased rates of cardiovascular events and all-cause mortality [4]. In the US, Medicaid health insurance is a strong proxy for poverty due to income-based enrollment requirements. Prior to 2014, Medicaid was only available to select eligible groups (children, pregnant women, parents and caretakers, seniors, people with disabilities) with low income requirements varying by state, but Medicaid expansion now qualifies anyone with an income ≤138% of the poverty level in states where it has been adopted. A significant proportion of adults with CHD in the US utilize Medicaid, a proportion that was found to increase only slightly from pre-vs post-Medicaid expansion [5], potentially due to CHD with severe symptoms such as cyanosis or secondary pulmonary vascular obstructive disease already qualifying as a disabled group. However, studies of this population have previously been limited by lack of access to comprehensive datasets. We used the Transformed Medicaid Statistical Information System (T-MSIS), an administrative database which collects comprehensive public health insurance data from all 50 US states and territories, to examine the ACHD population on Medicaid. The goal of this analysis was to understand both population characteristics and healthcare utilization patterns and whether it varies by disease severity to inform future attempts at optimizing outcomes for this vulnerable patient population. Methods Study design This retrospective cohort study characterized adults with CHD who were Medicaid recipients in all 50 US states and territories from January 2016 through December 2018. Moderate or complex CHD as defined by the 2025 ACHD guidelines [6] is the exposure of interest in this internal comparison cohort study design. Our objective is to determine whether incidence of death and healthcare utilization, as well as comorbidity prevalence differ from those with mild CHD in the Medicaid population. Data Sources and processing Limited datasets cut from the main Transformed Medicaid Statistical Information System (T-MSIS) Analytic Files (TAF) are available to Oregon Health & Sciences University (OHSU) investigators through limited dataset agreements with the OHSU Center for Health Systems Effectiveness. TAF Claims Inpatient Files, Other Services (i.e., outpatient), and Long-term Care Files and Demographic and Eligibility Files from January 2016 through December 2018 were requested for recipients if they had a CHD-associated ICD-10-CM diagnosis code (listed in Supplementary Table 1) during that time frame. Data were linked between files using a combination of the submitting state alpha abbreviation, encrypted state assigned beneficiary unique identifier, and reference year. Prior to analysis, claims data were cleaned by removing spaces, periods, and leading zeros from diagnosis codes. Inpatient and other services files were checked for duplicative observations using submitting state alpha abbreviation, encrypted state assigned beneficiary unique identifier, claim ID, service begin and end dates, billing provider ID, and referring provider ID, as well as admission and discharge date and admitting provider for inpatient claims. All available claims files in 2016-2018 were used to determine CHD, comorbidity, and substance use diagnoses. Only claims occurring in months with one or more days of Medicaid coverage as determined by service begin and end date for inpatient and outpatient files, as well as covered days, admission date, discharge date for inpatient files, or with an amount paid by Medicaid for them above zero were included in incidence analyses. Only 0.97% of inpatient visits and 1.56% of outpatient visits did not qualify by this categorization. Missingness in demographic and eligibility files was minimized by checking across repeated observations to replace missing values for date of birth, sex, and race. For participants with more than one demography record in a reference year for reasons such as moving between states mid-year, the file associated with the most enrollment days was used for characteristics in Table 1. Follow-up time was calculated as the total of the sum of days in each month covered across all observations in the same year for each participant, with a maximum allowable value set to the number of days in that month and year. For 2016, which was a leap year, total follow up could be as high as 366 days. Table 1Baseline characteristics of adult Medicaid recipients with congenital heart disease (CHD) by anatomic disease complexity.Table 1CHD Anatomic ComplexityTotal (N = 260,535)Simple (N = 144,598)Moderate or Complex (N = 115,937)CHD DiagnosisEisenmenger, Cyanosis0 (0.0%)2,239 (1.9%)2,239 (0.9%)Single ventricle/Fontan0 (0.0%)6,227 (5.4%)6,227 (2.4%)TGA0 (0.0%)5,165 (4.5%)5,165 (2.0%)Conotruncal abnormality0 (0.0%)15,103 (13.0%)15,103 (5.8%)Coarctation of the aorta0 (0.0%)22,185 (19.1%)22,185 (8.5%)AV septal defect0 (0.0%)8,337 (7.2%)8,337 (3.2%)Pulmonary valve disease8,747 (6.0%)0 (0.0%)8,747 (3.4%)Ebstein anomaly0 (0.0%)1,699 (1.5%)1,699 (0.7%)Anomalous pulmonary venous return0 (0.0%)1,997 (1.7%)1,997 (0.8%)Shunts (ASD/VSD/PDA)135,851 (94.0%)0 (0.0%)135,851 (52.1%)Subaortic stenosis0 (0.0%)2,285 (2.0%)2,285 (0.9%)Anomalous coronary artery0 (0.0%)17,709 (15.3%)17,709 (6.8%)Bicuspid aortic valve0 (0.0%)32,991 (28.5%)32,991 (12.7%)Sex (Biological)Female103,444 (71.5%)61,061 (52.7%)164,505 (63.1%)Male41,153 (28.5%)54,876 (47.3%)96,029 (36.9%)Missing1 (0.0%)0 (0.0%)1 (0.0%)Age (in Yrs)Mean (SD)39.1 (17.9)41.5 (20.4)40.2 (19.1)Race and EthnicityWhite, non-Hispanic58,456 (48.6%)57,289 (60.5%)115,745 (53.9%)Black, non-Hispanic22,566 (18.8%)13,402 (14.2%)35,968 (16.7%)Other, non-Hispanic8,006 (6.7%)5,608 (5.9%)13,614 (6.3%)Hispanic, all races31,191 (25.9%)18,349 (19.4%)49,540 (23.1%)Missing24,379 (16.9%)21,289 (18.4%)45,668 (17.5%)Primary LanguageEnglish94,666 (84.2%)75,750 (87.4%)170,416 (85.6%)Spanish15,079 (13.4%)8,672 (10.0%)23,751 (11.9%)Other2,635 (2.3%)2,290 (2.6%)4,925 (2.5%)Missing32,218 (22.3%)29,225 (25.2%)61,443 (23.6%)Income Relative to the Federal Poverty Level - Latest in Year0-100%50,634 (81.6%)43,259 (82.7%)93,893 (82.1%)101-133%7,299 (11.8%)5,672 (10.8%)12,971 (11.3%)≥134%4,143 (6.7%)3,387 (6.5%)7,530 (6.6%)Missing82,522 (57.1%)63,619 (54.9%)146,141 (56.1%)Census regionMidwest24,820 (21.4%)25,706 (17.8%)50,526 (19.4%)Northeast28,029 (24.2%)31,146 (21.5%)59,175 (22.7%)South32,015 (27.6%)37,410 (25.9%)69,425 (26.6%)West31,073 (26.8%)50,336 (34.8%)81,409 (31.2%)Disability or SSDI Indicated31,466 (21.8%)33,154 (28.6%)64,620 (24.8%)Abbreviations: ACHD, adult congenital heart disease; ASD, atrial septal defect; AV, atrioventricular; PDA, patent ductus arteriosus; TGA, transposition of the great arteries; VSD, ventricular septal defect. Baseline characteristics of adult Medicaid recipients with congenital heart disease (CHD) by anatomic disease complexity. Abbreviations: ACHD, adult congenital heart disease; ASD, atrial septal defect; AV, atrioventricular; PDA, patent ductus arteriosus; TGA, transposition of the great arteries; VSD, ventricular septal defect. Study population Medicaid recipients included in the data pull described in the above Data Sources section were eligible for inclusion in this retrospective cohort if they met the following criteria: they were 18 years or older during one or more of the study years; they had one or more Medicaid enrollment days during the study years; they had a CHD diagnosis as defined in Supplementary Table 1. Participants were excluded if they did not meet these inclusion criteria or if criteria were unable to be assessed either through not having demographic and eligibility files available during study years or having missing values for both age and date of birth. For CHD diagnoses determined by multiple codes, those codes were not required to occur at the same time point or in the same file. Complexity of CHD diagnosis was determined as described in Supplementary Table 1. As reflected in the flow diagram in Fig. 1, 260,535 met these inclusion criteria. Medicaid Adult Congenital Heart Disease (ACHD) study flow diagram. Participants were eligible for study inclusion if they had a CHD associated ICD-10 diagnosis code included in the pull list in the years 2016-18, were ≥18 years of age in the year(s) with coverage, had greater than zero Medicaid enrollment days, and had a diagnosis code or combination of codes indicating ACHD diagnosis. Outcomes Death was determined from the TAF Demography and Enrollment file. Healthcare utilization outcomes were determined using TAF Inpatient Files. Inpatient claims with overlapping dates for the same Medicaid recipient were grouped as single hospital admissions. An admission was considered a primary cardiovascular admission if the primary diagnosis code (or secondary code if primary is a CHD code listed in Supplementary Table 1) in an associated claim is listed in Supplementary Table 2. An admission was considered a cardiovascular contributor if the diagnosis code is present in a hospitalization associated claim but is not listed as the primary code. Emergency department (ED) visit was determined from the TAF Inpatient Files and TAF Other Services files if participants had an emergency inpatient admission type code or if they had revenue center code values of 0450 – 0459 (Emergency Room) or 0981 (Professional fees – Emergency Room). The number of outpatient ED visits was determined by the number of unique outpatient ED dates for each participant. CHD diagnosis CHD diagnosis was determined hierarchically in the order listed in Supplementary Table 1; i.e., if a participant had more than one diagnosis, they were assigned the diagnosis listed higher on the table. This categorization of CHD diagnoses as well as disease complexity has previously been validated and used by our group [7]. Comorbidities Substance use was determined using the ICD-10-CM codes listed in Supplementary Table 3. ICD-10-CM codes used to categorize comorbidities listed in Supplementary Table 4 were identified using a combination of categorizations the investigators have used in previous studies [7] and codes associated with Clinical Classification Software (CCS) categories identified using the R package CCS (v. 0.0.0.9) diagnosis dataset, which uses CCS developed and maintained by the US Agency for Healthcare Research and Quality (AHRQ) contemporaneous to the study time frame. Demographics Demographics included are participants' sex, age, race and ethnicity, primary language, income relative to federal poverty level, immigration status, and disability. Geographic regions were categorized by state using the U.S. Census Bureau's regional categorizations [8]. Statistical analysis Characteristics of ACHD Medicaid recipients in each year and at baseline were summarized as mean (SD) for continuous and n (%) for categorical variables. Missingness for each variable is shown in tables except when none were missing. Crude incidence (i.e., event rate) of death, hospital admissions (primary cardiovascular, cardiovascular contributor, and any admissions), and ED visits (outpatient and ED admissions) were compared across CHD anatomic complexity groups via Poisson distribution with an offset of the log of follow-up time using the PROC GENMOD function in SAS. The event rate ratio of death adjusting for age was compared similarly but used individual person (rather than aggregate) follow-up time. Comorbidities were compared across CHD complexity groups using the Chi-square test, with effect size estimates obtained using Cramer's V. The Bonferroni adjustment was used to control for multiple comparisons, however as very large sample sizes can result in the significance of very small differences between groups, we emphasize effect size differences instead in our reporting of results. Euler diagrams of comorbidities stratified by gender and age were created using the R package euler (v7.0.2). Analyses were performed using SAS (v9.4) and R (v4.3.2). Ethical considerations This study was reviewed by the institutional review board at Oregon Health & Science University (OHSU) and determined to be IRB exempt. The T-MSIS datasets used in this study were available to the investigators through a re-use limited dataset data use agreement with OHSU's Center for Health Systems Effectiveness, the holder of the primary data use agreement with Medicaid's Research and Data Center. Results Study inclusion characteristics are shown in Fig. 1. 1,037,823 unique individuals were identified with an ICD-10 code consistent with CHD. After eliminating those not linked to a demography file (N = 6718), individuals with missing age information (N = 23,557), those <18 years of age (N = 715,435), those with no Medicaid enrollment days (N = 1817) and those without CHD based on our categorizations shown in Supplementary Table 1 (N = 29,707), there were 260,535 individuals eligible for inclusion in the study. Demographic and geographic characteristics at time of enrollment Baseline demographic characteristics are shown by CHD anatomic complexity in Table 1. 63.1% were female with a mean age of 40.2 ± 19.1 years. 55.5% (N = 144,598) had simple CHD and 44.5% (N = 115,937) had moderate or complex CHD. Specific CHD diagnoses are shown in Table 1. Female recipients comprised 71.5% of the simple CHD population, compared to 52.7% of the moderate or complex population, a weak but significant association (Table 2, Cramer's V 0.19, p = 0.19). 53.9% (N = 115,745) of individuals with available race and ethnicity data identified as White and non-Hispanic. There were missing race and ethnicity data in 17.5% (N = 45,668). 85.6% (N = 170,416) of those with primary language identified (missing data in 23.6%) were primarily English-speaking. The largest proportion of study participants lived in the West (31.2%, N = 81,409), followed by the South (26.6%, N = 69,425). Table 2Comorbidities of adult Medicaid recipients with congenital heart disease (CHD) in 2016-18 by CHD complexity.Table 2CHD Anatomic ComplexityTotal (N = 260,535)P ValueaCramer's VSimple (N = 144,598)Moderate or Complex (N = 115,937)Sex (Biological)F103,444 (71.5%)61,061 (52.7%)164,505 (63.1%)<0.0010.19M41,153 (28.5%)054,876 (47.3%)96,029 (36.9%)Age category18-2437,150 (25.7%)32,168 (27.7%)69,318 (26.6%)<0.0010.1025-3436,069 (24.9%)22,401 (19.3%)58,470 (22.4%)35-4420,496 (14.2%)13,820 (11.9%)34,316 (13.2%)45-5417,715 (12.3%)14,457 (12.5%)32,172 (12.3%)55-6418,819 (13.0%)16,202 (14.0%)35,021 (13.4%)65+14,349 (9.9%)16,889 (14.6%)31,238 (12.0%)Disabilityb31,466 (21.8%)33,154 (28.6%)64,620 (24.8%)<0.0010.08Cardiovascular comorbidities Rhythm disorder62,417 (43.2%)59,281 (51.1%)121,698 (46.7%)<0.0010.08 Hypertension59,788 (41.3%)56,291 (48.6%)116,079 (44.6%)<0.0010.07 Coronary artery disease28,215 (19.5%)35,306 (30.5%)63,521 (24.4%)<0.0010.13 Heart failure24,054 (16.6%)27,704 (23.9%)51,758 (19.9%)<0.0010.09 Stroke25,691 (17.8%)8,569 (7.4%)34,260 (13.1%)<0.0010.15Other comorbiditiesMore common in simple CHD Mental health disorder65,367 (45.2%)051,613 (44.5%)116,980 (44.9%)0.0080.01 Metabolic syndrome63,900 (44.2%)046,599 (40.2%)110,499 (42.4%)<0.0010.04 Nicotine use27,724 (19.2%)018,513 (16.0%)46,237 (17.7%)<0.0010.04 Substance abuse24,719 (17.1%)0015,698 (13.5%)40,417 (15.5%)<0.0010.05 Cancer14,727 (10.2%)0010,125 (8.7%)24,852 (9.5%)<0.0010.02More common in mod-complex CHD Renal disease23,975 (16.6%)20,753 (17.9%)44,728 (17.2%)<0.0010.02 Sleep apnea16,047 (11.1%)15,007 (12.9%)31,054 (11.9%)<0.0010.03 Liver disease13,743 (9.5%)11,917 (10.3%)25,660 (09.8%)<0.0010.01 Hypercholesterolemia41,997 (29.0%)37,867 (32.7%)79,864 (30.7%)<0.0010.04 Diabetes39,832 (27.5%)32,654 (28.2%)72,486 (27.8%)0.0080.01aBonferroni adjusted for multiple comparisons.bDisability or Social Security Disability Insurance indicated. Comorbidities of adult Medicaid recipients with congenital heart disease (CHD) in 2016-18 by CHD complexity. Bonferroni adjusted for multiple comparisons. Disability or Social Security Disability Insurance indicated. 43.9% (N = 114,394) of participants had income data available. 82.1% (N = 93,893) had an income equal to or below the federal poverty line, which in 2016 was $28,300 for a family of 4. An additional 11.3% (N = 12,971) had an income between 101 and 133% of the federal poverty line. 24.8% (N = 64,620) were disabled or receiving Social Security Disability insurance, with a higher percentage of those with moderate or complex CHD having a disability (28.6%; N = 33,154) than those with simple CHD (21.8%; N = 31,466) (Table 2). Prevalence of comorbidities Rates of medical comorbidities varied based on anatomic disease complexity (Table 2), but the differences were negligible in almost all cases (Cramer's V 0-0.1). Coronary artery disease was weakly associated with disease complexity, although the percentage difference in prevalence was more than 10% (moderate or complex CHD: 30.5% vs. simple 19.5%, p < 0.001; Cramer's V 0.13). In the overall population, the most common cardiovascular comorbidity was rhythm disorders (46.7%, N = 121,698), followed by hypertension (44.6%, N = 116,079). The most common non-cardiac comorbidities were mental health disorders (44.9%, N = 116,980), followed by metabolic syndrome (42.4%, N = 110,499). Nicotine use was present in 17.7% (N = 46,237; simple CHD 19.2% [N = 27,724], moderate or complex 16.0% [N = 18,513]) and substance abuse in 15.5% (N = 40,417; simple CHD 17.1% [N = 24,719], moderate or complex 13.5% [N = 15,698]). Healthcare utilization and mortality Healthcare utilization and mortality are stratified by CHD complexity in Table 3. Over a median of 3.0 years (IQR for simple: 2.1, 3.0; IQR for moderate or complex: 2.2, 3.0), death occurred in 5.2% of those with simple CHD (N = 7459) and 6.6% of those with moderate or complex disease (N = 7640). The age-adjusted event rate ratio for those with moderate or complex CHD as compared with simple CHD was 1.01 (0.98, 1.04). Given that bicuspid aortic valve (BAV) encompasses a heterogeneous spectrum of disease severity, we performed an analysis including BAV in the simple complexity group. In this analysis, moderate or complex CHD was associated with death (age-adjusted event rate ratio 1.12 [1.08, 1.16]) (Supplementary Table 5). Table 3Incidence of death and of healthcare utilization of adult Medicaid recipients with congenital heart disease (CHD) in 2016-18 by anatomic disease complexity.Table 3SimpleModerate or ComplexTotalP valuea(N = 144,598)(N = 115,937)N = 260,535Median [IQR] follow up, years3.0 [2.1, 3.0]3.0 [2.2, 3.0]3.0 [2.2, 3.0]Death# of events (%)7,459 (5.16%)7,640 (6.59%)15,099 (5.80%)Median [IQR] age at death61 [51, 73]68 [53, 84]63 [52, 79]Event rate per 100 person-years (95% CI)2.08 (2.03, 2.12)2.62 (2.56, 2.68)2.32 (2.28, 2.36)<0.001Event rate ratio adjusted for age(reference)1.01 (0.98, 1.04)->0.99Primary cardiac hospital admissions# of events27,38527,67755,062Event rate per 100 person-years (95% CI)7.62 (7.53, 7.71)9.49 (9.40, 9.60)8.46 (8.39, 8.53)<0.001Cardiac contributor hospital admissions# of events63,59753,932117,529Event rate per 100 person-years (95% CI)17.70 (17.56, 17.83)18.49 (18.34, 18.65)18.05 (17.97, 18.15)<0.001Any hospital admissions# of events248,199149,514397,713Event rate per 100 person-years (95% CI)69.06 (68.79, 69.34)51.27 (51.01, 51.53)61.09 (60.98, 61.21)<0.001ED visits (ED to admission or outpatient)# of events652,614427,8231,080,437<0.001Event rate per 100 person-years (95% CI)181.60 (181.16, 182.04)146.71 (146.27, 147.15)165.97 (165.66, 166.28)High utilizers, n (%)≥1 hospital admissions/yearb39,125 (27.06%)23,105 (19.93%)62,230 (23.89%)≥5 ED visits/3 yearsb51,241 (35.44%)32,486 (28.02%)83,727 (32.14%)Abbreviations: CI: confidence interval; IQR: interquartile range.aBonferroni adjusted for multiple comparisons.bTo account for differences in follow-up time, high utilizers of inpatient admissions and ED visits were determined using incidence rounded to two decimal places as # of admissions/follow-up time≥1 and # of ED hospital admissions and outpatient ED visits/follow-up time≥1.67. Incidence of death and of healthcare utilization of adult Medicaid recipients with congenital heart disease (CHD) in 2016-18 by anatomic disease complexity. Abbreviations: CI: confidence interval; IQR: interquartile range. Bonferroni adjusted for multiple comparisons. To account for differences in follow-up time, high utilizers of inpatient admissions and ED visits were determined using incidence rounded to two decimal places as # of admissions/follow-up time≥1 and # of ED hospital admissions and outpatient ED visits/follow-up time≥1.67. Hospital admissions with a primary cardiovascular diagnosis were more common in those with moderate or complex CHD than in those with simple CHD (event rate per 100 person-years in those with simple CHD: 7.62 [7.53, 7.71], moderate or complex CHD event rate per 100 person-years: 9.49 [9.40, 9.60], p < 0.001). In contrast to cardiac-specific admissions, overall hospital admissions were more common in those with simple CHD (event rate per 100 person-years in those with simple CHD: 69.06 [68.79, 69.34], moderate or complex CHD: 51.27 [51.01, 51.53], p < 0.001). Emergency department visits were also more common in those with simple CHD (event-rate per 100 person-years 181.60 [181.16, 182.04]) than in those with moderate or complex CHD (event-rate per 100 person-years 146.71 [146.27, 147.15]), p < 0.001. Nearly a third of the overall study population were identified as frequent utilizers of the emergency room, defined as incidence equivalent to ≥5 visits over three years (i.e. # of ED visits/follow-up time ≥1.67) with 32.14% overall prevalence; 35.44% in simple CHD (N = 51,241) and 28.0% in those with moderate or complex CHD (N = 32,486). Approximately one quarter had hospitalization incidence (# of admissions/follow-up time) equivalent to one hospital admission per year (23.9% overall; 27.1% in those with simple CHD and 19.9% in moderate or complex CHD). The prevalence of cardiometabolic comorbidities by age and gender is shown in Fig. 2. Obesity was the most prevalent comorbidity in women <50 years of age (42.2%). Men aged <50 had higher rates of comorbidities than women in the same age group, with rhythm disorders being most common (46.3%). Both men and women over the age of 50 had higher rates of cardiometabolic disorders, with many experiencing the overlap of multiple diagnoses. Euler diagram of co-occurrence of cardiometabolic comorbidities in Medicaid recipients with ACHD by age and gender. In males <50 years (A), rhythm disorder was the most frequently observed cardiometabolic comorbidity (46.3%) followed by obesity (35.8%), hypertension (34.2%), hypercholesterolemia (20.7%), diabetes (16.4%), coronary artery disease (CAD, 15.6%), heart failure (HF, 15.1%), sleep apnea (12.4%), and stroke (8.9%). In males ≥50 years (B), there was much greater overlap of cardiometabolic comorbidities, with hypertension the most prevalent (83.0%) followed by rhythm disorder (63.3%), hypercholesterolemia (61.4%), CAD (57.0%), diabetes (49.4%), obesity (43.4%), HF (39.6%), stroke (26.4%), and sleep apnea (17.7%). In females <50 years (C), obesity was observed most frequently (42.2%) followed by rhythm disorder (35.9%), hypertension (22.5%), diabetes (17.5%), hypercholesterolemia (13.8%), CAD (8.4%), HF (8.3%), sleep apnea (7.3%), and stroke (6.8%). Females ≥50 years (D) had similarly high overlap as males ≥50 years, with hypertension (83.0%), hypercholesterolemia (61.5%), rhythm disorder (61.4%), diabetes (51.3%), obesity (50.4%), CAD (50.1%), HF (39.4%), stroke (24.0%), and sleep apnea (18.4%). Discussion To our knowledge, this is the first comprehensive evaluation of adults with CHD on Medicaid in the US. Our primary findings are as follows: First, the ACHD Medicaid population is a complex one, with high rates of disability, mental health disorders, arrhythmia, and other comorbidities. Second, participants in this study were frequent users of emergency and hospital-based care. Third, 5.8% of participants died over the 3-year study period despite the relatively young age of the study population. This underscores the fact that the ACHD Medicaid population is at high risk of adverse health outcomes. In 2016, 12.7% of the US population was living in poverty, totaling 40.6 million people [9]. This study provides a population level overview of the subset of that population with ACHD who utilized Medicaid health insurance from 2016 to 2018. In the US, Medicaid insurance is a strong proxy for poverty, and poverty has been associated with adverse health outcomes, including heart failure, myocardial infarction, and excess cardiovascular mortality in the general population [[10], [11], [12]]. While the impact of poverty on the adult CHD population has not been extensively studied, the available data suggests that it is linked to poorer outcomes in children with CHD [13]. Adults with CHD have multiple risk factors for poverty, including deficits in cognitive and executive function, challenges finding employment, and high rates of disability [14,15]. Not surprisingly, a significant proportion of these individuals access healthcare via Medicaid [16]. Therefore, this study represents a snapshot of a large and important population within ACHD. We chose to perform analyses in those with simple and moderate or complex CHD separately for multiple reasons. First, based on prior data from our group, we believe that the ICD codes for moderate or complex CHD are more likely to accurately identify adults with CHD [17]. The lack of specificity of simple CHD diagnosis codes makes the simple CHD group more likely to include those without CHD. Second, CHD complexity impacts the risk of poor health outcomes, with a patient with simple CHD (e.g. repaired atrial septal defect) having a different risk profile than a patient with moderate or complex CHD (e.g. Fontan palliation of single ventricle). As expected, we found in the current analysis that rhythm disorders and heart failure were more common in moderate or complex CHD, as were liver and renal disease, although the percentage differences were small. These common comorbidities were present in a high proportion of individuals in this study, underscoring the vulnerability of the population and its impact on healthcare utilization. Coronary artery disease (CAD) and hypertension were also more common in higher complexity CHD, with a more than 10% difference in prevalence of CAD between those with moderate or complex CHD and simple CHD. The design of this study does not make it possible to differentiate whether this finding is related to a true difference in prevalence, better screening in those with more complex CHD, gender differences between disease complexity groups, or due to the lack of specificity of simple CHD diagnosis codes making the simple CHD group more likely to include those without true CHD. In contrast, nicotine and substance abuse were more common in those with simple CHD. The lower rates of nicotine use and substance abuse observed in those with moderate or complex CHD suggests one of two things: (1) individuals with moderate or complex CHD may be less likely to use or abuse substances due to concerns about the impact on their heart, and/or (2) they may be less likely to admit their use to healthcare providers. While the rates are lower in those with moderate or complex CHD, they are not insignificant at 13.5% (substance abuse) and 16.0% (nicotine use). Given the detrimental impact of substance use on cardiovascular health, providers caring for adults with CHD should screen their patients regularly and provide appropriate counseling when needed. Moderate or complex CHD represents a minority of CHD in our study, consistent with research done in other populations. While these individuals represent 44.5% of the total ACHD population on Medicaid as measured by our study, the number included was still large at more than 115,937 individuals, underscoring both the strength of our findings and the impact of this group on the US healthcare system. Of note, the size of the moderate or complex CHD population in this study depends on whether BAV is considered a simple or moderate complexity lesion. Isolated congenital aortic valve disease was originally defined as a simple complexity lesion by the Bethesda conference, a classification that was also used in our prior work [18]; however, the 2025 ACC/AHA/HRS/ISACHD/SCAI guidelines have now classified it as a moderate anatomic complexity lesion [6]. To be consistent with the most current guidelines, we included BAV in the moderate or complex category in our analysis. Of note, given the high variability of phenotypic presentation of BAV (ranging from symptomatic children requiring early intervention and subsequent reinterventions in adulthood to asymptomatic, incidentally diagnosed adults without the need for intervention), we did perform analyses with BAV in the simple category as well, which did change the association of disease complexity with mortality. Specifically, disease complexity was associated with mortality when BAV was omitted from the moderate or complex group. This finding suggests that BAV categorization can significantly impact research findings related to clinical outcomes, and future researchers should take this into consideration when making decisions about disease complexity groupings. Interestingly, although those with moderate or complex CHD were slightly more likely to be hospitalized with a primary cardiovascular diagnosis, they were less frequent users of hospital and emergency room services overall. There could be several explanations for this finding. First, the higher rates of substance abuse and mental health disorders in those with simple CHD may contribute to their higher utilization of emergency care. Second, the study population was young and may not yet have reached the age at which hospitalizations are more common. Third, simple CHD is not infrequently diagnosed in individuals engaging with the health care system for other reasons. For instance, patent foramen ovale (PFO) and atrial septal defect (ASD) share an ICD code and are often identified during the evaluation of other disease states, especially stroke. This would be consistent with the finding that those with simple CHD in this study were more likely to have a diagnosis of stroke. Lastly, there was a larger proportion of female Medicaid recipients in the simple CHD category, and given that Medicaid was the source of payment for 42.6% of all births in the US in 2016 [19], this gender difference between disease complexity cohorts could contribute to the higher all cause hospitalization rate in the simple CHD group. Almost 20% more Medicaid recipients with simple CHD were female, compared to moderate or complex CHD group, which had a near even balance between genders. While previous studies have found a higher prevalence of simple CHD in women compared to men [20, 21], the population studied here was 63% female overall, which could be more reflective of the demographic makeup of the US Medicaid population (57.7% female in 2016 [22]) rather than the CHD population. We demonstrate that comorbidities associated with long-term cardiovascular risk tend to cluster in individuals. This clustering varied by both age and gender, with men tending to have more comorbidities at a younger age than women. Men and women over the age of 50 had more cardiometabolic risk factors, significantly increasing their risk of acquired cardiovascular disease and underscoring the importance of effective preventive care. While a minority of individuals in this study had a diagnosis of coronary artery disease, the risk profile of the population suggests that this is likely to change over time. This is important to consider in the development of strategies to effectively care for the aging ACHD population over time. Limitations Our study used an administrative dataset in which cardiovascular diagnoses and comorbidities were identified by ICD-10 codes. As such, the study population includes only those individuals who had at least one healthcare claim between 2016 and 2018. Individuals who did not utilize their Medicaid insurance during the study period were not included in the study. Therefore, this study does not represent the entire ACHD Medicaid population. Specifically, the study population excludes individuals who were healthy enough to forgo healthcare during the study period and those who were unable to access care. Consistent with other administrative datasets, our dataset had high percentages of missing data for race, ethnicity, income and primary language. The information contained for these variables should be considered exploratory and not definitive. In the US, it is common for individuals to move on and off insurance plans. We do not have information about the percentage of individuals who changed or lost their Medicaid insurance during the study period. There are limitations of ICD codes with respect to specificity [17]. Not all types of CHD have their own ICD code, which can lead providers or coders to select a less specific code. As mentioned above, the ICD code for ASD and PFO are the same, meaning that the simple CHD group includes individuals without CHD. We addressed this by performing analyses stratified by disease complexity. Even with this adjustment, ICD codes remain an imperfect identifier of the presence of CHD. For instance, misclassification of the type or presence of CHD can occur when codes are entered by individuals without CHD-specific knowledge, limiting their accuracy in real-world practice. Conclusion The goal of this study was to characterize population characteristics and healthcare utilization patterns among adults with CHD on Medicaid insurance and determine whether they vary by disease complexity. To our knowledge, this is the first study which specifically describes the characteristics and healthcare utilization of the low income ACHD population in the US. We demonstrated that adults with CHD on Medicaid have high rates of both cardiovascular and non-cardiovascular comorbidities, as well as high rates of emergency healthcare utilization. Higher rates of primary cardiac admissions in those with moderate or complex CHD likely reflect the higher cardiac risk profile of this group, while their lower overall hospital and emergency room utilization may be due to less accurate diagnosis coding for identifying simple CHD and baseline gender differences between the simple and moderate or complex groups. Strategies to optimize the cardiovascular health of this population are likely to have long-term impact and should be a priority for ACHD programs. Sources of funding None. CRediT authorship contribution statement Lidija B. McGrath: Conceptualization, Writing – original draft, Writing – review & editing. Claire Cambron: Writing – review & editing. Robin L. Baudier: Data curation, Formal analysis, Writing – original draft, Writing – review & editing. Abigail M. Khan: Conceptualization, Methodology, Writing – original draft, Writing – review & editing. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements Katrina Ramsey, MPH assisted with the design of this project. Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijcchd.2026.100687. Supplementary data The following is the Supplementary data to this article: Multimedia component 1 References Associated Data Supplementary Materials Multimedia component 1