Linguistic Disparities in Child Health and Presence of a Medical Home Among US Latino Children

Abstract

Objectives

The impact of household language on Latino-White and Latino intragroup disparities in child health and having a medical home in the US is poorly understood. This study aimed to examine these disparities (1) between Whites and Latinos (overall and stratified by English-primary-language [EPL] and non-English-primary-language [NEPL] households); (2) within Latinos, stratified by household language; and (3) potential moderation of disparities by social determinants.

Methods

Cross-sectional analysis of nationally representative sample of children 0–17 years old from the 2016–2018 National Survey of Children’s Health. We evaluated associations of child race/ethnicity and household language with child health and presence of a medical home. Multivariable logistic regression was used to compare groups of interest, adjusting for sociodemographic factors and health needs. Moderation was assessed using interaction terms for household income, parental educational attainment, and child insurance coverage.

Results

Among 81,514 children, 13.5% were NEPL Latino, and 19.4% were EPL Latino. Compared with EPL Whites, both EPL and NEPL Latinos had reduced excellent/very good health (aOR:0.70; 95%CI: 0.58–0.84; and aOR:0.42; 95%CI: 0.33–0.53) and presence of a medical home (aOR:0.62; 95%CI: 0.56–0.69; and aOR:0.45; 95%CI: 0.37–0.54), respectively. Among Latinos, NEPL (vs. EPL) was also associated with reduced excellent/very good health (aOR:0.61; 95%CI: 0.46–0.83), and presence of a medical home (aOR:0.66; 95%CI: 0.48–0.78); these associations were magnified by adverse social determinants.

Conclusions

Striking Latino-White and within-Latino medical-home disparities persist in the US, particularly for NEPL Latino children. Interventions should target social determinants and the rich sociocultural and linguistic diversity of the Latino population.

Latinos comprise the largest ethnic and linguistic minority population in the US, surpassing 60 million in 2019.¹ Among Latinos, 73% live in non-English-primary-language (NEPL) households, and, of this population, 44% have limited English proficiency (LEP) defined as the ability to speak English “less than very well”.² Latino children in NEPL households are more likely than those in English-primary-language (EPL) households to live in poverty, lack health insurance and have parents with lower educational attainment.³ Compared to their EPL counterparts, Latino parents in NEPL households are more likely to report suboptimal child health, decreased healthcare access and ease of use, and higher dissatisfaction with the pediatric primary care of their children.⁴

The American Academy of Pediatrics (AAP) recommends that all children receive high-quality primary care meeting criteria for a medical home, which is care that is accessible, family-centered, continuous, comprehensive, coordinated, compassionate, and culturally effective.⁵ Having a medical home is associated with better healthcare utilization patterns (i.e. increased preventive care visits, decreased outpatient and emergency room sick visits) and improved overall child health.^6,7 Medical home disparities, however, persist for Latino children.^8,9

Not enough is known about the contribution of NEPL to Latino-White and within Latino child health and medical home disparities and its interaction with key social determinants of health. Previous studies of Latino-white medical home disparities were conducted almost a decade ago.¹⁰ Since then, the Affordable Care Act was enacted, which fueled multi-sector initiatives to promote the medical-home concept, including payment reform incentives and multi-million dollar grants for safety-net organizations to develop as medical homes.^11–13 Whether these investments have reduced national Latino-white disparities is not known. Moreover, to our knowledge, the contribution of primary household language to medical home disparities within the Latino population has not yet been examined. A comprehensive investigation of the interaction between Latino ethnicity, NEPL, and social determinants of health is needed to inform the development of effective interventions to reduce medical home disparities for Latino children residing in NEPL households, a rapidly expanding US group.

To fill these research gaps, the study aims were to examine: 1) disparities between EPL non-Latino White children and Latino children (overall, and stratified by primary household language); 2) disparities within Latino children, comparing NEPL vs. EPL; and 3) whether social determinants, including household income, parental educational attainment, and child health insurance coverage, modified associations of child race/ethnicity and household language with the primary outcomes.

METHODS

Data Source

The data source was the 2016–2018 National Survey of Children’s Health (NSCH), a publicly available dataset supported by the Maternal and Child Health Bureau and administered by the National Center for Health Statistics, that provides nationally representative estimates of measures of child health and primary care for children 0–17 years old.¹⁴ Since 2016, the survey has included an annual self-administered web/paper-based survey utilizing an address-based sampling frame. A screener instrument was used to identify households with children, with oversampling of children with special healthcare needs and young children 0–5 years old. A follow-up instrument was used to collect age-appropriate, detailed information about one child randomly selected from the household. The respondent was an adult parent or caregiver (“parent” hereafter) familiar with the child’s health and healthcare utilization. NSCH instruments were available in English and Spanish. Overall response rates for the years included was 40.7%, 37.4% and 43.1% for 2016, 2017 and 2018, respectively. Survey weighting accounted for sampling weights as well as screener and follow-up instrument non-response with further adjustment to population controls at the household level to yield reliable estimates. Full details of the methodology are reported elsewhere.¹⁵ The Boston University School of Medicine/Boston Medical Center Institutional Review Board determined that this study was exempt from human-studies review.

Study Population

A total of 102,341 NSCH surveys were completed in 2016–2018. In the NSCH, parent-reported categories for race and ethnicity of the child include Latino, non-Latino White, non-Latino Black, non-Latino Asian and multiracial/other. In this study, we included non-Latino White (White hereafter) children in EPL households and Latino children of any race. Latino children were further stratified by the primary language spoken at home. We excluded non-Latino children of other races and children missing data for primary household language (Fig. 1). The 2016–2018 NSCH dataset used for analysis contained imputed missing values for household income and child race/ethnicity (missing values for race/ethnicity less than 1%).

Figure 1.

Participant selection.

Measures

Independent variables were three mutually exclusive racial/ethnic and linguistic groups: White children in EPL households (EPL White), and Latino children in EPL (EPL Latino) and NEPL households (NEPL Latino). Primary household language was measured by the question, “what is the primary language spoken in this child’s home?” with the option English and non-English. Dependent variables included: 1) overall child health status, and 2) presence of a medical home, and 3) four medical-home subcomponents, including having a usual source of care, getting needed referrals, family-centered care, and effective care coordination. Overall child health was assessed using the parent’s answer to the question, “in general, how would you describe this child’s health?” As in previous studies using NSCH data,¹⁶ we dichotomized the answers to the health status question as “excellent or very good” vs. “good, fair, or poor”. Presence of a medical home was defined using the standardized composite measure of the Child and Adolescent Health Measurement Initiative (CAHMI),¹⁷ which is based on the AAP medical-home concept.⁵ NSCH survey questions and criteria used to construct the CAHMI medical home composite measure and its subcomponents are detailed in Supplementary Table 1. In brief, to qualify as having a medical home, children must meet criteria for having a usual source of care and family-centered care. Additionally, children who needed referrals or care coordination in the previous 12 months must also meet criteria for those subcomponents. Per NSCH definition, “children with a valid, positive parental response to at least one subcomponent of the medical home, but with the remainder of the subcomponents missing or legitimately skipped (i.e. no need for referrals or care coordination in the previous 12 months), are categorized as having a medical home.”¹⁷

Potential modifiers in this study included adverse social determinants such as low income, low parental educational attainment, and lack of insurance or sporadic insurance coverage. Low-income was defined as a household income <200% of the federal poverty threshold (FPT) (vs. ≥ 200% FPT). Low parental educational attainment was defined as not obtaining a high-school degree (vs. at least having a high-school degree). Uninsured or sporadically insured was defined as lacking insurance at the time of the survey or having any gaps in insurance coverage during the previous 12 months (vs. currently insured and consistently insured over that period).

Variables hypothesized to confound the relationship between independent and dependent variables were included as covariates. These variables, consistent with prior studies using NSCH data,^10,18 included child sociodemographic characteristics (age and sex), health status (except when modeling health status), special healthcare needs, and the social determinants detailed above, in addition to foreign birth of the child, and insurance type (public only, private only, a combination of public and private, or uninsured). To define special healthcare needs, we used the CAHMI composite measure which identifies children with either chronic functional limitations and/or greater healthcare needs than their peers.¹⁷

Analysis

To obtain multi-year estimates, individual year survey weights were adjusted for, following the procedures published by the US Census Bureau.¹⁹ To account for the complex survey design, all analyses were performed using the survey procedures in STATA 14.0 (Stata Corp, College Station, TX).

The chi-square test of independence was used to analyze categorical variables and the adjusted Wald test was used for continuous variables to compare racial/ethnic and linguistic groups with respect to child characteristics, primary outcomes, and study covariates. For model building, all covariates with a P ≤ .1 in bivariate analyses were included, as well as covariates hypothesized a priori to confound the relationship between child race/ethnicity and household language with the primary outcomes.

To assess aim 1 (Latino-White disparities), weighted multivariable logistic regression was performed to compare Latino children (overall and stratified by primary household language) with White children (the reference group). To assess aim 2 (within Latino disparities), multivariable logistic regression was performed for Latino children only, comparing NEPL with EPL (the reference group). For aim 3 (moderation), separate multivariable models were constructed, with introduction of interaction terms for the following social determinants: household income, parental educational attainment, and child health insurance coverage. We then performed stratified models by each social determinant for interaction terms with a P < .05.

Sensitivity analyses excluding adolescents (12–17 years old) also were conducted, since English-proficient adolescents residing in NEPL households are likely to be less dependent on their parents or guardians for their primary care, which may influence their medical-home outcomes. For all main analyses, the threshold for statistical significance was a two-sided P < .05.

RESULTS

The final study population consisted of 81,514 children, representing an estimated 54.2 million children in the U.S. (Fig. 1). Of these, 70,149 (67.1%) were White, and 11,365 (32.9%) were Latino. Among Latino children, 41% resided in NEPL households. Regarding child sociodemographic characteristics (Table 1), NEPL Latino children were more likely to be foreign-born than EPL Latinos. The proportion of children with special healthcare needs was lower in NEPL vs. EPL Latino children. Latino children were significantly more likely than White children; and NEPL Latino more likely than EPL Latino children, to reside in low-income households, have parents who are not high-school graduates, and to be uninsured or sporadically insured. Regarding primary outcomes (Table 2), 93% percent of White, 89% of EPL Latino, and 82% of NEPL Latino children were reported to be in excellent/very good health; and 48% of White, 32% of EPL Latino, and 20% of NEPL Latino children met criteria for the presence of a medical home. NEPL Latino children were least likely to meet criteria for presence of a medical home in all individual subcomponents.

Table 1.

Selected Characteristics of US Children 0–17 years old in the study sample.

Characteristic	EPL White (n=70, 149 [67.11%])	Latino (n=11, 365 [32.9%])	Latino
			EPL (n=8, 110 [19.40%])	NEPL (n=3, 255 [13.49%])
	Weighted %	Weighted %	Weighted %	Weighted %
Child characteristics
Age, mean, y	8.58	8.71	8.57	8.93
Male %	51.4	50.6	50.7	50.5
Nativity
Child foreign-born %	1.0	6.2*	2.0*	12.2*◊
Special healthcare needs
No CSHCN	80.6	83.7*	80.8	87.9*◊
CSHCN less complex	6.0	3.8*	5.0	2.1*◊
CSHCN more complex	13.4	12.5*	14.2	10.0*◊
Child’s insurance coverage
Public only	19.6	44.8*	37.1*	55.9*◊
Private only	71.8	37.9*	50.0*	20.5*◊
Both public and private	3.6	4.8*	4.3*	5.4*◊
Uninsured	4.2	10.4*	7.2*	15.0*◊
Child insurance continuity
Continuously insured	94.2	86.8*	90.1*	91.8*◊
Uninsured or sporadically insureda	5.8	13.2*	9.9*	8.2*◊
Socioeconomic characteristics
Household income
<200% FPT	28.7	60.4*	49.5*	76.1*◊
≥200% FPT	71.4	39.6*	50.5*	23.9*◊
Parental educational attainment
Not high-school graduate	2.6	22.4*	9.6*	40.8*◊
High-school degree or above	97.4	77.6*	90.4*	59.2*◊

NEPL: Non-English primary language. EPL: English primary language. CSHCN: Children with special healthcare needs. FPT: Federal poverty threshold Pairwise comparisons using chi-square test for categorical variables and the adjusted Wald test for continuous variables;

^aUninsured at the time of the survey or gaps in insurance over the previous 12 months

^*P <.01 comparing overall Latino, EPL and NEPL Latino vs. EPL White children.

^◊P < .01 comparing NEPL vs EPL Latino children.

Table 2.

Child Health Status and Presence of a Medical Home by Child Race/Ethnicity and Primary Household Language

Outcome	Non-Latino EPL White (n=70, 149 [67.11%])	Latino (n=11, 365 [32.9%])	Latino
			EPL (n=8, 110 [19.40%])	NEPL (n=3, 255 [13.49%])
	Weighted %	Weighted %	Weighted %	Weighted %
Child health status
Excellent or very good	93.1	86.2*	89.0*	82.2*◊
Good	5.9	11.6*	8.9*	15.6*◊
Fair/poor	1.0	2.1*	2.1*	2.2*◊
Medical home outcomes
Presence of medical homea	48.0	27.4*	32.5*	20.1*◊
Medical-home subcomponents
Usual source of careb	68.8	49.6*	55.0*	41.7*◊
Personal doctor/nurse	78.1	64.1*	68.5*	57.8*◊
Usual source sick care	84.7	70.3*	74.9*	63.6*◊
Family-centered careb	76.6	56.5*	63.5*	46.5*◊
Listening	96.2	93.6*	94.9*	91.3*◊
Doctor spent enough time	94.4	85.4*	87.6*	81.4*◊
Cultural sensitivity	96.5	92.5*	94.4*	89.0*◊
Comprehensive information	96.7	93.6*	94.5*	91.9*
Partnership in care	95.6	92.4*	94.1*	89.3*◊
Getting needed referralsc	83.0	74.8*	75.6*	73.6*
Effective care coordinationc	57.9	43.4*	50.5*	34.3*◊

NEPL: Non-English primary language. EPL: English primary language.

Pairwise comparisons using chi-square for categorical variables and the adjusted Wald test for continuous variables

^aAs per the definition of the Child & Adolescent Health Measurement Initiative (CAHMI). To qualify as having a medical home, children must meet criteria for the components of usual source of care and family-centered care; and any children who needed referrals or care coordination must also meet criteria for those components.

^bMust meet all sub-components.

^cAmong those with referral or care coordination needs in the previous 12 months.

^*P <.01 comparing overall Latino, EPL and NEPL Latino vs. EPL White children.

^◊P < .01 comparing NEPL vs EPL Latino children.

Compared with White children, Latino children had significantly lower adjusted odds of having excellent/very good health and presence of a medical home (Table 3). Both NEPL and EPL Latino children were less likely to have a usual source of care, family-centered care, and effective care coordination than White children. Among Latinos, those residing in NEPL (vs. EPL) households were less likely to have overall excellent/very good health and to report having a medical home. In analyses of individual subcomponents of the medical home, NEPL Latino children were significantly less likely to have family-centered care and effective care coordination than their EPL Latino counterparts. Among Latino children, primary household language was not associated with having a usual source of care or getting needed referrals.

Table 3.

Associations of Child Race/Ethnicity and Primary Household Language with Child Health and Medical Home

Latino vs. White Children (Referent)

Category

Excellent/Very Good Child Health

Presence of Medical Home

Medical Home Subcomponents

Usual Source of Care

Family-Centered Care

Getting Needed Referrals

Effective Care Coordination

AOR

(95% CI)

AOR

(95% CI)

AOR

(95% CI)

AOR

(95% CI)

AOR

(95% CI)

AOR

(95% CI)

Race/Ethnicity

Latino

0.57

(0.50–0.67)

0.56

(0.52–0.62)

0.62

(0.57–0.68)

0.61

(0.55–0.66)

0.72

(0.56–0.94)

0.79

(0.71–0.88)

White

Referent

Referent

Referent

Referent

Referent

Referent

Race/Ethnicity and Household Language

NEPL Latino

0.42

(0.33–0.53)

0.45

(0.37–0.54)

0.53

(0.45–0.62)

0.47

(0.40–0.56)

0.62

(0.44–0.87)

0.59

(0.49–0.71)

EPL Latino

0.70

(0.58–0.84)

0.62

(0.56–0.69)

0.66

(0.59–0.73)

0.64

(0.57–0.71)

0.74

(0.54–1.02)

0.86

(0.76–0.98)

White

Referent

Referent

Referent

Referent

Referent

Referent

NEPL Latino vs. EPL Latino Children (Referent)

Category

Excellent/very good Child Health

Presence of Medical Home

Medical Home Subcomponents

Usual

Source of care

Family-Centered Care

Getting Needed Referrals

Effective Care Coordination

AOR

(95% CI)

AOR

(95% CI)

AOR

(95% CI)

AOR

(95% CI)

AOR

(95% CI)

AOR

(95% CI)

NEPL

0.61

(0.46–0.83)

0.66

(0.53–0.83)

0.82

(0.67–1.0)

0.62

(0.46–0.84)

0.77

(0.57–1.13)

0.61

(0.48–0.78)

EPL

Referent

Referent

Referent

Referent

Referent

Referent

Models adjusted for child’s age, sex, health status (except when modeling health status), special healthcare needs, foreign birth of child, parental educational attainment, low income, child health insurance continuity, and insurance type. NEPL: Non-English primary language. EPL: English primary language.

In moderation analyses (Table 4), a significant interaction by parental educational attainment was identified for the association between being a NEPL Latino child and the medical-home subcomponents of usual source of care and family-centered care. Stratification revealed that, compared with White children, NEPL Latino children had nearly half the odds of having a usual source of care and family-centered care only among those whose parents had lower educational attainment. Similarly, NEPL Latino children had lower odds of having effective care coordination than White children only among those who were uninsured or sporadically insured. In models comparing EPL Latino children with White children, there were no significant interactions by social determinants.

Table 4.

Moderation by Social Determinants of the Associations of Child Race/Ethnicity and Primary Household Language with Child Health and Having a Medical Home

NEPL Latino compared with White Children (Referent)

Potential moderators

Excellent/Very Good Health

Presence of Medical Home

Medical Home Subcomponents

Usual Source of Care

Family-Centered Care

Getting Needed Referrals

Effective Care Coordination

AOR (95% CI)

P

AOR (95% CI)

P

AOR (95% CI)

P

AOR (95 % Cl)

P

AOR (95% CI)

P

AOR (95% CI)

P

By household income

≥200% FPT

0.41 (0.36–0.47)

0.97

0.78 (0.54–1.11)

0.11

0.77 (0.56–1.05)

0.09

0.70 (0.51–0.97)

0.20

1.85 (0.75–4.57)

0.06

0.82 (0.57–1.19)

0.26

<200% FPT

0.42 (0.26–0.68)

0.58 (0.54–0.62)

0.58 (0.54–0.63)

0.57 (0.52–0.61)

0.76 (0.62–0.94)

0.66 (0.60–0.72)

By parental education

High-school grad

0.64 (0.41–0.98)

0.41

0.89 (0.60–1.31)

0.01*

0.78 (0.57–1.06)

0.04*

0.74 (0.54–1.02)

0.004*

1.71 (0.84–3.47)

0.20

0.72 (0.50–1.04)

0.35

Not high-school grad

0.51 (0.36–0.72)

0.46 (0.33–0.65)

0.50 (0.37–0.67)a

0.40 (0.30–0.53)

0.92 (0.44–1.92)

0.57 (0.41–0.80)

By child health insurance

Continuously insured

0.94 (0.57–1.52)

0.78

0.50 (0.27–0.92)

0.65

0.51 (0.44–0.60)

0.15

0.60 (0.41–1.0)

0.05

0.91 (0.61–1.35)

0.14

0.75 (0.47–1.20)

0.02*

Uninsured or sporadically insured

0.86 (0.67–1.12)

0.43 (0.37–0.51)

0.37 (0.24–0.57)

0.37 (0.32–0.44)

0.48 (0.23–1.01)

0.41 (0.34–0.49)

EPL Latino compared with White Children (Referent)

Potential moderators

Excellent/Very Good Health

Presence of Medical Home

Medical Home Subcomponents

Usual Source of Care

Family-Centered Care

Getting Needed Referrals

Effective Care Coordination

AOR (95% CI)

P

AOR (95% CI)

P

AOR (95% CI)

P

AOR (95% CI)

P

AOR (95% CI)

P

AOR (95% CI)

P

By household income

≥200% FPT

0.42 (0.37–0.47)

0.60

0.59 (0.55–0.64)

0.43

0.59 (0.55–0.64)

0.45

0.58 (0.53 0.63)

0.74

1.19 (0.68–2.10)

0.11

0.70 (0.55–0.89)

0.72

<200% FPT

0.38 (0.26–0.54)

0.54 (0.44–0.67)

0.54 (0.44–0.67)

0.56 (0.45–0.69)

0.73 (0.59–0.89)

0.67 (0.61–0.73)

By parental education

High-school grad

0.61 (0.34–1.09)

0.63

0.48 (0.34–0.67)

0.90

0.51 (0.38–0.68)

0.96

0.47 (0.30–0.73)

0.64

2.69 (1.04–6.91)

0.43

0.60 (0.43–0.83)

0.84

Not high-school grad

0.51 (0.36–0.73)

0.46 (0.26–0.79)

0.50 (0.30–0.82)

0.41 (0.31–0.55)

1.68 (0.84–3.37)

0.56 (0.32–0.97)

By child health insurance

Continuously insured

0.93 (0.59–1.45)

0.80

0.45 (0.38–0.52)

0.33

0.67 (0.45–0.98)

0.28

0.38 (0.33–0.45)

0.13

1.24 (0.66–2.36)

0.37

0.42 (0.35–0.50)

0.10

Uninsured or sporadically insured

0.87 (0.67–1.12)

0.36 (0.24–0.54)

0.53 (0.45–0.62)

0.29 (0.20–0.40)

0.89 (0.60–1.31)

0.29 (0.19–0.44)

Models adjusted for child’s age, sex, health status (except when modeling health status), special healthcare needs, foreign birth of child, parental educational attainment, low income, child health insurance continuity, and insurance type. NEPL: Non-English primary language. EPL: English primary language. FPT: federal poverty threshold

^aExample: compared with White children, NEPL Latino children had significantly lower odds of having a usual source of care only among those whose parents had lower educational attainment.

^*P < .05 for interaction term.

In moderation analyses among Latino children only (Table 5), the reduced odds of NEPL children having family-centered care was 42% stronger for NEPL Latino children who resided in low-income households compared with those in households with an income ≥ 200% FPT. Similarly, disparities in family-centered care were 46% larger for NEPL Latino children who were uninsured or had sporadic insurance coverage compared with those who were consistently insured. Within the Latino population, NEPL was associated with less effective care coordination only for NEPL Latino children who were uninsured or sporadically insured, but not for those continuously insured. These findings persisted in sensitivity analyses excluding children 12–17 years (data not shown).

Table 5.

Moderation by Social Determinants of the Associations of Primary Household Language with Child Health and Medical Home Among Latinos

NEPL compared with EPL Latino Children (Referent)
Potential moderators	Excellent/Very Good Health		Presence of Medical Health		Medical Home Subcomponents
					Usual Source of Care		Family-Centered Care		Getting Needed Referrals		Effective Care Coordination
	AOR (95% CI)	P	AOR (95% CI)	P	AOR (95% CI)	P	AOR (95% CI)	P	AOR (95% CI)	P	AOR (95% CI)	P
By household income
≥200% FPT	0.42 (0.26–0.53)	0.69	0.77 (0.54–1.10)	0.07	0.77 (0.56–1.05)	0.05	0.73 (0.44–0.98)	0.003*	1.81 (0.74–4.42)	0.50	0.81 (0.56–1.18)	0.41
<200% FPT	0.37 (0.26–0.53)		0.53 (0.43–0.65)		0.53 (0.43–0.65)		0.31 (0.23–0.42)a		1.26 (0.72 (2.19)		0.67 (0.53–0.85)
By parental education
≥ High school	0.66 (0.43–1.02)	0.86	0.88 (0.59–1.33)	0.06	0.76 (0.55–1.05)	0.15	0.73 (0.53–0.99)	0.10	2.18 (0.85–5.61)	0.11	0.72 (0.50–1.05)	0.36
< High school	0.62 (0.35–1.11)		0.46 (0.26–0.80)		0.49 (0.30–0.82)		0.46 (0.29–0.72)		0.89 (0.42–1.89)		0.53 (0.31–0.92)
By child health insurance
Continuously insured	0.96 (0.59–1.54)	0.99	0.50 (0.27–0.93)	0.34>	0.65 (0.44–0.95)	0.05	0.61 (0.42–0.89)	0.003*	1.17 (0.62–2.23)	0.09	0.77 (0.48–1.23)	0.002*
Uninsured or sporadically insured	0.95 (0.61–1.48)		0.35 (0.24–0.53)		0.36 (0.23–0.56)		0.28 (0.21–0.40)		0.50 (0.23–1.08)		0.29 (0.20–0.44)

Models adjusted for child’s age, sex, health status (except when modeling health status), special healthcare needs, foreign birth of child, parental educational attainment, low income, child health insurance continuity, and insurance type. NEPL: Non-English primary language. EPL: English primary language. FPT: federal poverty threshold

^aExample: The reduced odds of NEPL children having family-centered care was 42% stronger for NEPL Latino children who resided in low-income households compared with those in households with an income ≥ 200% FPT.

^*P < .05 for interaction term.

DISCUSSION

The study findings demonstrate stark Latino-White disparities in child health status and presence of a medical home. Compared with White children, Latino children were less likely to meet criteria for all subcomponents of the medical home, including having a usual source of care, family-centered care, effective care coordination and getting needed referrals. As with other racial/ethnic minority groups, reasons for suboptimal health and healthcare among Latinos, such as inability to afford care or lack of insurance,²⁰ are partly rooted in socioeconomic disadvantage.²¹ Nevertheless, although income disparities in child health and the medical home are well described,⁸ adjusted stratified analyses by primary household language in this study revealed that Latino children residing in NEPL households are a particularly vulnerable pediatric subgroup.

Unchanged from previous decade-old reports using NSCH data,¹⁰ this study identified a significantly lower medical-home prevalence for Latino children, particularly NEPL Latino children (compared with White children). These results indicate stagnant progress in increasing adoption of the medical home and in reducing linguistic disparities for Latino children, despite over a decade of national and state initiatives fueled by enactment of the ACA in 2010.^11–13 Adding to the existing literature, this study uniquely analyzed disparities by primary household language within the Latino population. Multivariable analyses including only Latino children revealed that residing in NEPL (vs. EPL) households was associated with worse child health and reduced likelihood of having a medical home. In contrast to the Latino-White findings, within-Latino medical-home disparities were largely driven by deficiencies in family-centered care and effective care coordination, two medical-home subcomponents traditionally associated with quality of care.²² Compared with Latino parents in EPL households, Latino parents in NEPL households were significantly more likely, by a large margin, to report that the provider didn’t listen carefully, spend enough time with the child, or provide care that was sensitive to the family’s values and customs. These findings suggest that the quality gap between NEPL and EPL Latino children may be related, at least in part, to linguistic and cultural issues.²³

A central finding of this study is that children in NEPL households experienced the largest disparities in child health and presence of a medical home. Reasons for such disparities are likely multi-factorial. Although language barriers may partly account for why children in NEPL households have worse overall health and medical-home outcomes, almost half of NEPL households have family members that are English proficient.² For the subgroup of NEPL Latino parents with LEP, language barriers may create substantial difficulties in having effective family-provider communication, which in turn affects the ability to establish a partnership in care.²⁴ Clinicians may fail to assess and identify LEP among NEPL Latino families who appear to be proficient in “conversational” English, which can further exacerbate disparities in care.²⁵ Conversely, for Latino children in NEPL households whose parents are English proficient, the observed disparities may be related to non-linguistic factors such as clinicians’ implicit biases and lack of cultural competency, resulting in Latino families experiencing lack of cultural understanding, negative stereotypes, stigma, or discrimination.^26,27 Additionally, undocumented immigrant Latino parents of US citizen children often face multiple additive barriers to quality care, including LEP, lack of insurance, economic vulnerability, and restrictive public policies and programs.²⁸ Consequently, parents’ undocumented status can be associated with adverse outcomes regarding children’s physical and emotional health and use of pediatric health services.²⁹

In this national sample, nearly half of Latino children resided in NEPL households with ~98% of them having Spanish as the household primary language.¹⁷ NEPL Latino children were markedly more likely to reside in low-income households, have parents with lower educational attainment, and be uninsured or have gaps in insurance coverage. Three in four NEPL Latino children resided in low-income households, and two in five of their parents did not graduate high school. Key interactions between NEPL and social determinants were identified. For instance, low-income and lack of continuous health insurance coverage moderated NEPL-EPL disparities in family centered-care among Latino children. Similarly, within-Latino disparities in lacking effective care coordination were significant only among NEPL Latino children who were uninsured or sporadically insured. These results underscore NEPL Latino children’s dual medical and social vulnerability. Therefore, clinicians caring for low-income children-who are overrepresented among racial/ethnic and linguistic minorities-should routinely screen empathetically for social determinants of health,³⁰ and broaden current screeners to include primary household language and parental LEP,³¹ and ensure that at-risk families are connected to appropriate resources.

The study findings have important implications for research and policy. Recognition of primary household language as an important social determinant of racial/ethnic and linguistic minority child health, and its strong interaction with co-occurring socioeconomic risk, are only starting points. Future research should examine whether integrative approaches that address social determinants, as well as cultural and language barriers to care, can effectively reduce Latino-White and within-Latino disparities. Such approaches should include standardized processes to collect data on primary household language and screen for LEP among Latino children and their families, as well as provision of language-concordant care or high-quality interpreter services for those with LEP.^32,33 Local contexts where specific dialects are prevalent require experienced interpreters familiar with those dialects. Comprehensive initiatives to enhance providers’ linguistic and cultural competency starting with medical school training would be an important step forward.²⁶ Similarly, there is a need to understand the barriers and facilitators to implementation and scale-up of community-based initiatives (i.e. Salud al día, Nurse-Family Partnership, Keeping families healthy, Kid’s HELP) that have targeted the specific needs of Latino families to successfully improve Latino child health and healthcare utilization. From a policy perspective, there is a need for more inclusive public policies and evaluation of the performance of Accountable Care Organization models in reducing disparities in the quality of care for linguistic minorities. This requires universal robust data collection of primary household language and LEP, and tracking quality metrics stratified by race/ethnicity and language to identify and address disparities. In addition, some states offer third-party reimbursement for language services through Medicaid and the Children’s Health Insurance Program (which includes federal matching).³⁴ This is currently available in only 13 states and the District of Columbia, however, so the study findings suggest that such reimbursement in all states has the potential to improve the health and outcomes of Latino children and their families.

Strengths of our study include the use of the largest and most diverse pediatric national database to date that contains data on language. This allowed stratification of the heterogeneous Latino population by a key family attribute-primary household language-, which uncovered specific disparities that disproportionally affect Latino children in NEPL households. Certain study limitations also should be noted. The data are cross-sectional, and therefore do not demonstrate causality. In addition, covariates in our multivariable models did not include immigration status, acculturation, or use of interpreter services, which are additional confounders of the relationship between language and healthcare outcomes, but are not available in NSCH. Also not available in NSCH is information about national original, which precluded a more nuanced examination of outcomes among Latino subgroups. Notably, the data are limited by the language variable available in the NSCH survey: primary household language. Previous research has established that LEP is a superior measure of the impact of language barriers on child health, access to care, and use of health services.³⁵ As a result, the disparities identified in this study may have been of greater magnitude had we examined only Latino children in families with LEP. Incorporation of data on the primary language spoken at home and parental LEP in large datasets of child health would contribute to advancing health services research aimed at reducing disparities for ethnic and linguistic minorities.

CONCLUSIONS

Analysis of a nationally representative sample of U.S. children demonstrated marked Latino-White and within-Latino disparities in child health and presence of a medical home that were particularly pronounced for Latino children in NEPL households. Moderation analyses revealed that these disparities were magnified by adverse social determinants, including low income, low parental educational attainment, and lack of or sporadic child health insurance coverage. Consistent with Healthy People 2030 objectives, our results support the need to identify and target the primary language spoken at home as an important social determinant of Latino child health in pediatric care. Interventions to reduce disparities for Latino children should specifically target social determinants and the rich socio-cultural and linguistic diversity of the Latino population.

What’s new:

Despite a decade of efforts to promote the medical home, marked US Latino-white and within Latino disparities persist, particularly for children in NEPL households. Adverse social determinants exacerbate disparities in quality-related medical-home subcomponents, further undermining outcomes of NEPL Latino children.

Abbreviations:

AAP

American Academy of Pediatrics

CAHMI

Child and Adolescent Health Measurement Initiative

EPL

English primary language

LEP

Limited English proficiency

NEPL

Non-English primary language