Racial and Ethnic Disparities in Testing of Hepatitis C Virus–Exposed Children Across the United States

Abstract

Background

Despite rising hepatitis C virus (HCV) prevalence among pregnant individuals in the United States, HCV testing among exposed infants remains low. Although recent guidelines recommend early ribonucleic acid (RNA) testing for HCV-exposed children to help improve testing rates, national studies describing factors associated with HCV testing and the type of testing completed are lacking.

Methods

In this retrospective national study, we characterized HCV testing and care among HCV-exposed infants born between 2010 and 2020 captured in the electronic health record-based TriNetX Research Network. We analyzed factors associated with appropriate HCV testing completion (negative or positive HCV RNA testing or negative HCV antibody testing at any age through study end in 2022) and with RNA compared with antibody testing using univariable and multivariable logistic regression with clustered standard errors by healthcare organization.

Results

Of 8516 HCV-exposed children, 45.8% completed any HCV testing and 42.1% completed appropriate testing (25% of whom had RNA testing only). A total of 182 (5.1% of appropriately tested children) had evidence of HCV infection. Of 104 treatment-eligible children, 14.4% were treated. Black (odds ratio [OR]: 0.38, 95% confidence interval [CI]: 0.26–0.55), Asian/Pacific Islander (OR: 0.06, 95% CI: 0.03–0.11), and Hispanic/Latinx (OR: 0.56, 95% CI: 0.36–0.88) children had lower odds of appropriate testing compared with White and non-Hispanic/Latinx children.

Conclusions

Fewer than half of HCV-exposed children in this national sample were tested for HCV, with lower testing odds among Black, Asian/Pacific Islander, and Hispanic/Latinx children. Substantial work to increase testing and treatment and decrease disparities in testing among HCV-exposed children is needed to help reach US HCV elimination goals.

BACKGROUND

Hepatitis C virus (HCV) prevalence rose 16-fold among US pregnancies from 1998 to 2018 [1]. HCV is vertically transmitted in approximately 8% of deliveries to women with HCV infection, resulting in a substantial rise in HCV-exposed infants [2]. Despite this increase, HCV testing of exposed infants has been poor across the United States with only 5%–61% of infants screened [3–5].

Historically, guidelines from the Centers for Disease Control and Prevention (CDC) and the American Academy of Pediatrics recommended HCV antibody testing in exposed infants after 18 months of age [6, 7]. Some guidelines also allowed for ribonucleic acid (RNA) testing as early as 2 months, as an alternative. RNA testing was less preferred given its expense and reportedly high false positive rates and because early identification does not allow earlier treatment as HCV treatment is not available for children under age 3 years [6, 8]. However, accumulating evidence demonstrates poor screening rates, and high specificity of current RNA testing [3, 9]. Additionally, performing HCV RNA testing as early as 2 months of age is cost-saving compared with antibody testing at 18 months given reduced loss to follow-up of exposed infants [10]. In response, in November 2023, the CDC recommended RNA testing for all HCV-exposed infants between 2 and 6 months of age and antibody testing only for at-risk infants identified after 18 months of age [3].

Prior to the 2023 updated CDC guidance, some pediatric HCV experts advocated for early RNA testing as a preferable approach [4, 11, 12], but it remained an alternative recommendation [6, 8], and national trends in HCV testing of perinatally exposed infants remain uncharacterized. Furthermore, no previous multistate studies report data on negative test results or examine disparities associated with testing. We sought to characterize HCV testing and the care cascade prior to the 2023 CDC recommendations to better understand characteristics associated with disparate testing rates in perinatally exposed infants to help formulate plans to best apply the new testing guidance.

METHODS

Study Design, Data Source, and Ethical Statement

This retrospective cohort study used US data from the TriNetX Research Network, a global health-collaborative clinical research platform gathering real-time electronic medical record (EMR) data (Supplementary Exhibit 2). Upon download, the network included approximately 113 million patients across 55 healthcare organizations (HCOs) serving both insured and uninsured patients with data through November 2022.

The current study was determined not to be human subject research by the Boston University Medical Center Institutional Review Board. We followed STROBE reporting guidelines (Supplementary Exhibit 1).

Patient Population

We included infants born between 2010 and 2020 with viral hepatitis exposure (International Classification of Diseases, Tenth Revision [ICD-10] code Z20.5) and ≥1 TriNetX-captured encounter. We excluded those with hepatitis B virus (HBV) testing but no HCV testing to minimize the risk of misclassifying HBV cases as HCV and those born after 2020 (as they may not have been 18 months old by the study end).

Covariates

We extracted demographics, including race (categorized by TriNetX as White, Black or African American [Black], Asian or Pacific Islander [API], American Indian or Alaska Native [AIAN], and unknown), ethnicity, US census region, and year of birth, follow-up time (years between the first and last TriNetX encounters within the study period), and categorized comorbidities that may affect testing practices, including human immunodeficiency virus (HIV) exposure and HBV diagnosis using ICD-9 and ICD-10 codes (Supplementary Exhibit 3) [11].

Outcomes and Definitions

We analyzed the following HCV care cascade stages occurring anytime between birth and study end (up to age 12 in the oldest children): (1) any HCV testing—HCV antibody or RNA test completed at any age; (2) appropriate HCV testing—negative HCV antibody testing or a single positive or negative HCV RNA test at any age; (3) identified HCV infection—positive HCV RNA test result at any age; (4) treatment eligible—HCV infection that did not spontaneously clear and born 2018 or earlier (as direct-acting antivirals [DAAs] are only approved for children ≥3 years old); (5) linked to HCV care—encounter with a principal diagnosis of HCV; and (6) treatment initiation—EMR evidence of a prescription for a DAA. Supplementary Exhibit 3 reports codes utilized to identify testing, diagnoses, and medications.

Statistical Analysis

We analyzed the number and percentage of exposed infants who achieved each stage of the HCV care cascade and the percentage appropriately tested by birth year. We used univariable and multivariable logistic regression models to assess predictors of appropriate testing with clustered standard errors (SE) by HCO to account for practice differences across institutions. We checked for collinearity using chi-squared and Fisher’s exact tests. We adjusted multivariable models for variables significant in univariable analysis, maternal HIV exposure to account for increased healthcare utilization, and birth cohort to assess temporal changes, particularly given the impact of the COVID-19 pandemic on healthcare utilization. Multivariable models were also adjusted for race and ethnicity a priori, as they often serve as proxies for social determinants of health through which we aimed to identify and understand disparities that could inform targeted interventions to improve healthcare equity. To address potential regional differences, we also stratified the multivariable models by region. All statistical analyses were performed using R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Between 2010 and 2020, 8516 children across 45 HCOs had evidence of HCV exposure. Among those exposed, 62.6% identified as White, 11.6% as Black, 4.3% as API, 0.7% as AIAN, and 20.8% were of unknown race (Table 1). In total, 8.4% identified as Hispanic/Latinx, 62.5% as non-Hispanic/Latinx, and 29.1% had unknown ethnicity. Regionally, 25.1% accessed care at a Northeast HCO, 50.6% at a Southern HCO, 16.6% at a Midwest HCO, and 7.8% at a Western HCO. Over 17% of HCV-exposed infants were also exposed to maternal HIV, and 1.6% of children were diagnosed with HBV. The median follow-up time captured in the dataset was 2.9 years (Interquartile range [IQR] 1.5–4.8 years), and 7053 (82.8%) were followed for ≥1 year (Table 1).

Table 1.

Characteristics of HCV-Exposed Children Born Between 2010 and 2020 Accessing Care at a US TriNetX-Participating Healthcare Organization

Variable	N (%)
Sex
Female	4026 (47.3)
Male	4490 (52.7)
Race
American Indian or Alaska Native	60 (0.7)
Asian or Pacific Islander	364 (4.3)
Black or African American	989 (11.6)
Unknown	1772 (20.8)
White	5331 (62.6)
Ethnicity
Hispanic or Latino	719 (8.4)
Not Hispanic or Latino	5322 (62.5)
Unknown	2475 (29.1)
Region
Midwest (N = 7 HCOs)	1415 (16.6)
Northeast (N = 12 HCOs)	2138 (25.1)
South (N = 22 HCOs)	4305 (50.6)
West (N = 4 HCOs)	658 (7.8)
Year of birth
2010	164 (1.9)
2011	198 (2.3)
2012	216 (2.5)
2013	316 (3.7)
2014	482 (5.7)
2015	861 (10.1)
2016	1090 (12.8)
2017	1319 (15.5)
2018	1392 (16.3)
2019	1301 (15.3)
2020	1177 (13.8)
Birth cohort
2010–2013	894 (10.5)
2014–2017	3752 (44.1)
2018–2020	3870 (45.4)
HIV exposurea
Yes	1463 (17.2%)
No	7053 (82.8%)
HBV diagnosisa
Yes	134 (1.6)
No	8382 (98.4)
Follow-up time
Single encounter  <1 year  ≥1 year Missing	300 (3.5) 1139 (13.4) 7053 (82.8) 24 (0.3)

Abbreviations: HBV, hepatitis B virus; HCO, healthcare organization; HCV, hepatitis C virus; HIV, human immunodeficiency virus; ICD-10, International Classification of Diseases, 10th Edition Code.

^aBy ICD-10 code (see Supplementary Exhibit 2).

HCV Care Cascade

There were 3898 (45.8%) infants with evidence of any HCV testing, and 3584 (42.1%) infants completed appropriate HCV testing (Figure 1). Of those appropriately tested, 1839 (51.3%) were tested by antibody only, 883 (24.6%) by RNA only, and 862 (24.1%) had both. There were 182 children with positive HCV RNA tests (5.1% of those appropriately tested); 45 (24.7%) of these children spontaneously cleared without evidence of treatment (78% of whom cleared by age 3 years old and 94% cleared by age 5 years old). Of 104 children with persistent HCV infection who were treatment-eligible, 16 (15.4%) were linked to care and 15 (14.4%) were treated. Treated children were aged 3–11; most treatments occurred between 2021 and 2022.

Figure 1.

Care cascade for hepatitis C virus (HCV)-exposed children born between 2010 and 2020. Appropriate HCV testing is defined by a single negative or positive HCV RNA test completed by study end (November 2022). HCV infection indicates positive HCV RNA at any age. Treatment eligible are those born in 2018 or later who did not spontaneously clear without treatment. Linkage to care indicates a visit with a primary diagnosis of HCV. Treated indicates evidence of a direct-acting antiviral prescription.

The percentage of children completing appropriate testing increased from 32%–44% for those born between 2010 and 2013 to 44%–47% for those born between 2014 and 2017 and then decreased to 36%–41% for the 2018–2020 birth cohort (Figure 2).

Figure 2.

Percentage of hepatitis C virus (HCV)-exposed children born in the indicated year who completed appropriate HCV testing by study end (November 2022), as defined by a negative HCV antibody test or a single positive or negative HCV RNA test, at any age.

Univariable Analysis

Univariable logistic regressions demonstrated lower odds of appropriate testing in Black (odds ratio [OR]: 0.40, 95% confidence interval [CI]: 0.28–0.57) and API (OR: 0.06, 95% CI: 0.03–0.011) children compared with White children (Table 2).

Table 2.

Appropriate HCV Testing Completion Among HCV-Exposed Children Born Between 2010 and 2020

	Appropriate Testing (N/%)		Univariable		Multivariable
	Yes	No	ORa	95% CI	ORb	95% CI
Sex
Female	1562 (47.5%)	2310 (47.1%)	1.01	0.93, 1.10
Male	1728 (52.5%)	2591 (52.9%)	Ref	Ref
Race
American Indian or Alaskan Native	25 (0.8%)	31 (0.6%)	0.95	0.57, 1.57	0.93	0.57, 1.51
Asian or Pacific Islander	10 (0.3%)	345 (7.0%)	0.06	0.03, 0.11	0.06	0.03, 0.11
Black or African American	235 (7.1%)	721 (14.7%)	0.40	0.28, 0.57	0.38	0.26, 0.55
Unknown	669 (20.3%)	1053 (21.5%)	0.73	0.52, 1.03	0.73	0.50, 1.07
White	2351 (71.5%)	2751 (56.1%)	Ref	Ref	Ref	Ref
Ethnicity
Hispanic or Latino	226 (6.9%)	479 (9.8%)	0.68	0.46, 1.01	0.56	0.36, 0.88
Not Hispanic or Latino	2014 (61.2%)	3047 (62.2%)	Ref	Ref	Ref	Ref
Unknown	1050 (31.9%)	1375 (28.0%)	1.09	0.71, 1.67	1.11	0.68, 1.81
Region
Midwest	723 (22.0%)	639 (13.0%)	1.42	0.57, 3.55
Northeast	921 (28.0%)	1157 (23.6%)	Ref	Ref
South	1357 (41.2%)	2753 (56.2%)	0.65	0.22, 1.91
West	289 (8.8%)	352 (7.2%)	1.01	0.39, 2.63
Birth cohort
2010–2013	283 (8.6%)	550 (11.2%)	Ref	Ref	Ref	Ref
2014–2017	1575 (47.9%)	2021 (41.2%)	1.39	0.87, 2.21	1.39	0.87, 2.23
2018–2020	1432 (43.5%)	2330 (47.6%)	1.07	0.70, 1.64	1.03	0.67, 1.59
HBV diagnosis
Yes	51 (1.6%)	63 (1.3%)	1.38	0.77, 2.47
No	3239 (98.4%)	4838 (98.7%)	Ref	Ref
HIV exposure
Yes	711 (21.6%)	650 (13.3%)	1.77	0.76, 4.11	1.79	0.75, 4.24
No	2579 (78.4%)	4251 (86.7%)	Ref	Ref	Ref	Ref

Bolding indicates P-value <.05.

Abbreviations: CI, confidence interval; HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; OR, odds ratio.

^aWith clustered standard errors by healthcare organization.

^bAdjusted for race, ethnicity, birth cohort, and HIV diagnosis, with clustered standard errors by healthcare organization.

Multivariable Analysis

Multivariable logistic regressions controlling for race, ethnicity, birth cohort, and HIV exposure demonstrated lower odds of appropriate testing in Black and API children compared with White children (OR: 0.38, 95% CI: 0.26–0.55; OR: 0.06, 95% CI: 0.03–0.11, respectively; Table 2). Hispanic/Latinx children had lower odds of testing compared with non-Hispanic/Latinx children (OR: 0.56, 95% CI: 0.36–0.88).

Region-Stratified Analysis

Region-stratified results were similar to the unstratified analysis, although smaller populations in most regions limited the number of variables possible to include in multivariable analyses, and fewer significant results were observed (Supplementary Exhibit 4). One notable difference was that children exposed to maternal HIV in the Northeast and South had higher odds of testing compared with unexposed children (OR: 2.79, 95% CI: 1.02–7.62, OR: 2.09, 95% CI: 1.07–4.05, respectively).

RNA Testing

Black and Hispanic/Latinx children made up a larger proportion of those tested with RNA compared with antibody testing only (Supplementary Table 5). Univariable regression demonstrated significantly higher odds of HCV RNA testing only in the Midwest compared with the Northeast (OR: 7.31, 95% CI: 1.54–34.68). Multivariable regression adjusted for race, ethnicity, region, and birth cohort demonstrated higher odds of HCV RNA testing in the Midwest compared with the Northeast (OR: 7.17, 95% CI: 2.16–23.77), Black children compared with White children (OR: 1.48, 95% CI: 1.08–2.02), and in those born between 2018 and 2020 compared with children born between 2010 and 2013 (OR: 2.07, 95% CI: 1.27–3.37).

DISCUSSION

To the best of our knowledge, this is the first study to characterize national trends in perinatal HCV testing practices, including disparities by race, ethnicity, and US census region, and to report on the treatment of perinatally infected children. Only 42% of HCV-exposed children in this national sample were appropriately tested for HCV by the study end, and only 14.4% of eligible children with HCV infection were treated. Black, API, and Hispanic/Latinx children had lower odds of testing compared with White and non-Hispanic/Latinx children.

It is important to acknowledge that race and ethnicity are social constructs that reflect underlying social determinants of health. These categories often serve as proxies for various socioeconomic and environmental factors, including healthcare access, quality of care, and systemic inequities. The disparities in testing we observed are likely driven by these broader social determinants rather than inherent differences among racial or ethnic groups. Recognizing this context is essential to addressing the root causes of health inequities, such as improving access to culturally sensitive care, enhancing patient education, and addressing systemic biases within healthcare systems.

Similar to previous studies [11, 13], individuals with HIV exposure demonstrated higher odds of appropriate HCV testing, but this was only significant in the Northeast and South. We expected higher testing rates in this group due to their care in specialized clinics led by pediatric infectious diseases clinicians and streamlined coordination between maternal and infant care. It is possible inclusion of a less specific ICD-9 code (V01.79) or variation across HCOs may have accounted for this not being significant in the full population. Future research should investigate the potential of integrated maternal-infant care models to enhance HCV testing rates and ensure equitable healthcare access for vulnerable populations [14].

We also observed increased odds of RNA testing in Black compared with White children. This may represent an attempt to complete testing prior to loss to follow-up in this group with lower odds of completing appropriate testing. This could also represent unmeasured factors related to geographic variability in laboratory testing, such as reflex RNA testing, following a positive HCV antibody test or inability to order stand-alone RNA testing.

Clustering SE by HCO attenuated associations between demographic variables and testing outcomes, suggesting that site-specific testing practices affect appropriate testing completed and specifically RNA testing outcomes. Interestingly, 25% of appropriately tested infants received RNA testing only, and RNA testing was more frequent in the youngest birth cohort and in the Midwest, demonstrating that many clinicians adopted early HCV RNA testing prior to its formal endorsement by CDC guidelines, likely reflecting recommendations by pediatric HCV experts concerned about loss to follow-up prior to 18 months [4, 11–13].

Testing practices overall, however, have unfortunately not improved in recent years. Challenges in healthcare access due to the Coronavirus 2019 (COVID-19) pandemic may have hindered progress as infants in the most recent birth cohort would have become eligible for guideline-appropriate testing during the pandemic. The low proportion linked and treated may also have been due to DAAs only having been approved for children aged 3–11 years old between late 2019 and 2021 during the COVID-19 pandemic. Regardless, significant work is needed to improve overall testing and treatment proportions and disparities in testing by race and ethnicity.

Our results should be interpreted within the context of study limitations. There were substantial missingness of race and ethnicity data (21%–30%) and limited representation of AIAN and API racial groups. While we were able to observe significant differences in testing outcomes by race and ethnicity despite the missingness, it may obscure more nuanced disparities and hinder our ability to fully understand healthcare utilization patterns and outcomes among minoritized communities. Additionally, TriNetX’s de-identification process excludes perinatal conditions and month of birth, limiting our ability to control for neonatal comorbidities, exposures, and measure exact age at testing. This prevented the exclusion of HCV RNA testing between 1 and 2 months, potentially introducing early false negative/positive results. However, data within the first 30 days of life were also excluded, which at least excluded testing occurring at <1 month old. Maternal-infant dyads are not linked, preventing the ability to control for maternal variables or determine whether mothers were viremic. Our transmission rate is likely lower than 7%–12% reported in prospective studies of viremic mothers both because of unknown viremia status and as some infected children may have spontaneously cleared prior to testing. The general viral hepatitis exposure code may also have captured children exposed to hepatitis A, B, or non-perinatal exposures, however; these exposures are relatively rare in US children. Chronic HBV infection is much higher in Asian/Pacific Islander populations, potentially leading to disproportionate misclassification of HBV as HCV in this group and therefore falsely low HCV testing in API children. However, we suspect this was a small proportion misclassified as we excluded those with HBV testing alone to mitigate this [15].

We also likely underestimate exposed infants due to incomplete recording of exposure in EMRs (estimated sensitivity of Z20.5 for perinatal HCV exposure was 80% in 1 study) [16]. One-quarter of children had <18 months of follow-up, and we could not capture care occurring outside TriNetX HCOs; thus, we may be undercounting testing, linkage, or treatment. However, our testing and vertical transmission proportions align with literature-reported figures in other, particularly retrospective, studies [2–4]. Rural vs urban location and local HCV prevalence data were unavailable, hindering assessment of further geographic impacts on racial/ethnic disparities in testing. Finally, while there was an overrepresentation of HCOs in the South, the large and diverse patient populations within these organizations overall can guide strategies for pediatric HCV screening and management across healthcare settings.

This study highlights racial and ethnic disparities in HCV testing among exposed children, which are essential to address for equitable progress toward national HCV elimination. New CDC guidance may help bridge these gaps by capturing more children across race and ethnicity groups before they are lost to follow-up. Proper birth exposure documentation, pediatrician education on new guidelines, and access to lab tests are crucial for achieving these goals and preventing health inequities.

Supplementary Data

Supplementary materials are available at the Journal of the Pediatric Infectious Diseases Society online (http://jpids.oxfordjournals.org).

Notes

Acknowledgments. We thank Benjamin Buzzee for his contributions to data cleaning and TriNetX for their assistance in obtaining the data.

Financial support. This work was supported by the National Institutes of Health, National Institute on Drug Abuse (1K01DA052821 to R. L. E., P30DA040500 through a pilot award to M. R. C.), the National Institute on Allergy and Infectious Diseases (T32AI052074 and T32AI007433 to M. R. C., and through the Providence and Boston Center for AIDS Research - P30AI042853 to R. L. E.), and to the Boston University Clinical and Translational Science Institute (grant number 1UL1TR001430 through a voucher grant to R. L. E.); the Charles A. King Trust Postdoctoral Research Fellowship (to R. L. E.); and a Boston University Chobanian & Avedisian School of Medicine Department of Medicine Career Investment Award (to R. L. E.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Potential conflicts of interest. All authors: No reported conflicts.

Data availability. The data that support the findings of this study are available from the authors but restrictions apply to the availability of these data, which were used under a data use agreement through TriNetX, and so are not publicly available. Data may, however, be available from the authors upon reasonable request and with permission from TriNetX.

Supplement sponsorship. This article appears as part of the supplement “Viral Hepatitis Elimination in Infants, Children, and Pregnant Women,” sponsored by Gilead Sciences.