Housing Instability Among Families With Young Children With Special Health Care Needs

Data were collected as part of Children’s HealthWatch, an ongoing cross-sectional study monitoring the health and well-being of young children and families in medical centers in 5 US cities: Baltimore, Maryland; Boston, Massachusetts; Little Rock, Arkansas; Minneapolis, Minnesota; and Philadelphia, Pennsylvania. Each site obtained institutional review board approval before study initiation and yearly thereafter. Recruitment for these analyses (2013–2017) began when the SHCNs screener was added to the interview.

Caregivers of children <48 months old were approached in private settings when seeking pediatric care in the primary care sites (Baltimore, MD, and Minneapolis, MN) and emergency departments (Baltimore, MD; Philadelphia, PA; Boston, MA; and Little Rock, AR). Eligibility criteria included the following: index child age <48 months old, caregiver is a state resident, fluency in English or Spanish, and knowledge of child’s health and household. Caregivers of critically ill and/or injured children or those interviewed within the previous 6 months were not approached.

Of 17 467 caregivers approached by a trained research assistant, 15 303 (87.6%) agreed to be screened, and of those screened, 14 188 (92.7%) eligible caregivers gave consent and completed the face-to-face interview.

The interview included the following sample characteristics: caregiver age, race and/or ethnicity, US birth, educational attainment, marital or partnered status, and employment status; child’s health insurance status; household receipt of housing assistance and/or energy assistance (Low Income Home Energy Assistance Program [LIHEAP]) within the past year; and number of children in the household. Child age and sex were obtained via medical records.

The independent variable, child’s SHCN status, was identified in the interview by using the Children With Special Health Care Needs Screener, a validated, 5-item, caregiver-reported screening tool used by the Maternal and Child Health Bureau and National Center for Health Statistics of the Centers for Disease Control and Prevention.^2,21 Children screen positive for SHCN, regardless of medical diagnosis, if at least 1 of the following health circumstances has lasted or is expected to last >12 months: (1) use or need of prescription medication other than vitamins; (2) greater than average use or need of medical, mental health, or educational services; (3) functional age-related limitations; (4) use or need of specialized therapies; and/or (5) treatment or counseling for emotional or developmental problems. Caregivers were asked about current receipt of child SSI, allowing us to categorize children into 3 groups: (1) no SHCNs, (2) SHCNs and no SSI, and (3) SHCNs and receiving SSI.

The dependent variable was housing instability. Families were categorized as experiencing housing instability if they reported (1) being behind on rent or mortgage in the previous year, (2) moving ≥2 times in the past year (multiple moves), and/or (3) homelessness within the child’s lifetime (living in a shelter, motel, temporary or transitional living situation, or scattered-site housing or no steady place to sleep at night).¹¹ 

To assess sample characteristics, we stratified the groups into 3 categories (no SHCNs, SHCNs and no SSI, and SHCNs and receiving SSI). Bivariate associations were assessed via χ² tests or Student’s t tests as appropriate. Continuous variables all met normality assumptions. The descriptive characteristics using the 3-level categorization clarified some important within-SHCN group demographic differences (eg, caregiver employment, age, and race and/or ethnicity). Also, given SSI requirements, the group receiving SSI represented a combination of higher developmental and economic risk and need for intervention,¹⁵ and for these reasons, primary multivariable analyses used the 3-level categorization.

We conducted multivariable logistic regressions, adjusted for demographics (site; caregiver age, US birth, race and/or ethnicity, educational attainment, marital or partnered status, and employment status; child’s age, sex, and public or private health insurance; and number of children in the household), comparing associations between the SHCN predictor and housing instability circumstances. Children without SHCNs were the referent group. We then compared associations between exposure groups (no SHCNs, SHCNs and no SSI, and SHCNs and receiving SSI) and housing instability circumstances.

Publicly financed housing or rental assistance (eg, Section 8 or Housing Choice program vouchers, public housing, or other rental assistance) and energy assistance (LIHEAP) may help families with low income meet housing-related needs.^12,22 To determine if receipt of housing assistance and/or LIHEAP attenuated the association between SHCNs and housing instability, a combined housing assistance–LIHEAP variable was added to the adjusted models.

In all models, we reported adjusted odds ratios (aORs) and 95% confidence intervals (CIs). Analyses used 2-sided tests and SAS software (version 9.4; SAS Institute, Inc, Cary, NC). Significance was set at P < .05.

Of the sample of 14 188 participants (Table 1), 80.4% of children did not have SHCNs, 15.5% had SHCNs and no SSI, and 4.1% had SHCNs and were receiving SSI. The 3 most prevalent medical conditions among SHCNs were asthma (46.4%), speech and/or language problems (20.8%), and developmental delay (18.2%). Children’s mean age was 19.6 months (SD = 14); 46% were girls. Most caregivers (91.1%) were birth mothers; 5.7% were fathers and 3.2% were other (ie, adoptive or foster parent or grandparent). Caregivers’ mean age was 27.8 years (SD = 6.1); 74.7% were born in the United States, 46.5% were African American and non-Hispanic, 33.6% were Hispanic, 16.2% were white, 29.8% were married or partnered, and 48.0% employed. Housing assistance was reported by 17.8% of participants, and LIHEAP was reported by 19.2%. Being behind on rent or mortgage was reported by 20.6%, multiple moves were reported by 4.7%, and homelessness was reported by 8.0%. Families of children with (regardless of SSI receipt) versus without SHCNs were significantly (P < .001) more likely to report each of the adverse housing-instability circumstances: behind on rent or mortgage (24.1% vs 19.7%), ≥2 moves in the past year (6.2% vs 4.3%), and lifetime homelessness (9.9% vs 7.5%).

Bivariate characteristic differences by 3-level SHCN stratification (Table 1) included interview site, caregiver characteristics (US born, age, race and/or ethnicity, education, married or partner status, and employment status), child characteristics (age, sex, and health insurance), and household characteristics (number of other children and receipt of housing assistance and LIHEAP). As expected, rates of children with SHCNs and receiving SSI were highest in the Little Rock site, the only tertiary children’s hospital in Arkansas serving children with chronic and subspecialty care needs.

In bivariate analyses, each of the adverse housing circumstances differed by SHCN status (P = .001; Table 2), with the highest prevalence seen among children with SHCNs and not receiving SSI.

Table 3 compares housing instability circumstances in households of children with and without SHCNs (2 group comparison, regardless of SSI receipt). Controlling for sample characteristics, households of children with SHCNs had significantly greater odds of being behind on rent or mortgage (aOR 1.27 [95% CI 1.14–1.41]; P < .001) and homelessness during the child’s lifetime (aOR 1.39 [95% CI 1.18–1.65]; P < .001). The odds of experiencing multiple moves were nonsignificant (aOR 1.22 [95% CI 1.00–1.48]; P = .05). With the addition of housing assistance and/or LIHEAP as a covariate to the analyses, odds of housing instability circumstances did not substantially change.

Table 4 compares housing instability circumstances in households of children with SHCNs and no SSI and those with SHCNs and receiving SSI to households with children without SHCNs. Controlling for sample characteristics, children with SHCNs and no SSI had significantly greater odds of each of the 3 housing instability circumstances: behind on rent or mortgage (aOR 1.28 [95% CI 1.14–1.44]; P < .001), multiple moves (aOR 1.29 [95% CI 1.05–1.59]; P = .01), and ever homeless (aOR 1.44 [95% 1.20–1.72]; P < .001). There were no substantial changes with the addition of housing subsidies and LIHEAP. Households of children with SHCNs and receiving SSI were not significantly different from households of children without SHCNs in the circumstances of being behind on rent or mortgage, experiencing multiple moves, or ever being homeless. The addition of housing assistance and LIHEAP as a covariate to the analyses did not substantially change the odds of any of the 3 housing instability circumstances.

This study identified that families of children <48 months old with SHCNs are more likely to report 2 circumstances of housing instability than families of children without SHCNs: behind on rent or mortgage and homelessness during the child’s lifetime. The third indicator of housing instability, multiple moves, was nonsignificant and the least prevalent across groups at 4.7%. By including SSI status in the SHCN stratification, we found that compared with families of children without SHCNs, families of children with SHCNs and no SSI had significantly increased odds of experiencing all 3 adverse circumstances of housing instability. The likely explanation is that regardless of the reason for SSI nonreceipt, families of children with SHCNs and no SSI still may have had medical and/or educational expenses that financially hinder their housing stability.⁹ Compared with families without SHCNs, families of children with SHCNs and receiving SSI did not experience greater housing instability. This finding is particularly noteworthy given that children with SHCNs receiving SSI are more likely than those with SHCNs and no SSI to have severe disabilities, increased needs for early intervention and multiple medical and developmental interventions, and higher financial pressures as documented by meeting federal disability criteria and statutory financial and asset limitations.^13,14 

Our results extend previous findings on housing instability among children with disabilities by focusing on young children (<48 months old), using a broad definition of children with SHCNs, rather than disabilities, and 3 adverse circumstances of housing instability.^9,20,23 Using the Children With Special Health Care Needs Screener, we increased our understanding of the needs of young children at risk for long-term medical conditions who may not have received a diagnosis at these early ages, although families may be incurring increased medical and/or educational expenses. Young children, often excluded from or underrepresented within national studies, may differ from older cohorts as their symptoms either abate or intensify.^9,23,24 For example, a child with beginning signs of autism spectrum disorder or developmental delay may not qualify for SSI at 4 years old but may qualify at school age.

Early childhood is a developmentally sensitive period in which contextual stresses and environmental instabilities could particularly jeopardize children’s current and future health and well-being.²⁵ Disruptions of routine may result from multiple moves and homelessness, threatening the stability of day care, medical home, and other medical and educational services and resources.²⁶ For children with SHCNs, disruptions may mean service gaps or changes in clinical providers, including early intervention, resulting in being lost to follow-up and detrimental access to care effects.^19,27 Each of the adverse circumstances of housing instability examined also negatively affects parental supports and increases parental stresses and depression,¹¹ all of which are known to negatively affect parent-child relationships and child well-being.^28,29 

Access to both housing assistance and LIHEAP are limited nationally. Only 1 in 4 renters who are eligible for assistance receives federal housing assistance.^30,31 Already inadequate, LIHEAP benefit availability is further limited by “first come, first served” distribution to eligible households.³² In each SHCN group, there were important gaps between families reporting ≥1 adverse housing circumstance and families receiving housing assistance or LIHEAP. For example, in the SHCNs–receiving-SSI group, 30.4% reported ≥1 adverse housing circumstance, whereas only 16.1% received housing assistance. Statistical control for receipt of housing assistance and/or LIHEAP did not substantially change the odds of housing instability. However, housing and/or LIHEAP assistance receipt possibly benefited families in ways not measured in this study, such as housing quality and consistent household utilities, both of which are related to health.^28,29,33 Additionally, because families with low income may need to trade-off allocating financial resources among basic needs, future studies of households with children with SHCNs should evaluate if housing assistance and/or LIHEAP receipt decreases trade-offs among other basic needs.³⁴ 

This study has methodologic limitations. Cross-sectional analyses do not capture dynamic household changes over time or evaluate causality. Homelessness was reported within the child’s lifetime, whereas other variables were current (eg, SSI and housing assistance and/or LIHEAP) or occurred in the past year (eg, behind on rent and multiple moves). Most study variables were caregiver reported and therefore could be influenced by reporting bias and shared method variance. There may be unmeasured household confounders, including possibilities of other children with SHCNs or other household members receiving SSI. Although we assumed that children with SHCNs and receiving SSI had severe conditions and associated expenses, children’s medical and educational expenses vary and were not measured. In addition, SSI receipt associated with diagnoses, family finances, and SSI review changes across time. Our data are sentinel, not nationally representative, but reflect low-income households from 5 urban medical centers in different states primarily serving families using public health insurance. Our data do show an incidence of SHCNs similar to reports in 2009 and 2010, wherein 23.1% of US children <5 years old with SHCNs lived below 200% of the federal poverty level.⁴ 

This study has important public health implications. Families of children with SHCNs are more likely to experience housing instability than similar families with children without SHCNs. Housing instability, particularly being behind on rent or mortgage or moving twice or more in 1 year, may not be as apparent to health care and social service providers as homelessness. Therefore, our results suggest that agencies serving children with SHCNs ask all 3 housing instability screening questions and provide intervention referrals as needed. In addition, our results suggest that clinicians provide SSI information to parents of children with SHCNs who may be eligible for SSI.

Our study adds to the body of evidence that the SSI program provides critical support in meeting basic needs to families of children with SHCNs and should thus be maintained and strengthened.²⁰ Potential improvements to SSI that would further increase family resources needed to maintain housing stability may include (1) raising the income limit, indexing it to inflation so families are able to maintain at least partial work and earn wages necessary to better afford stable housing; (2) increasing asset limits so that families can save for unexpected expenses that might otherwise destabilize the family’s housing; and (3) increasing availability to families of flexible funds for emergency stabilization and eviction prevention as strategies for preventing homelessness, especially for children with SHCNs.

Most US communities lack enough safe and affordable homes, increasing strain on families, particularly families with children with SHCNs.^26,30,33 Policies that incentivize additional production of housing for families with low income and increased assistance for market-rate housing costs might improve families’ housing stability and the health and well-being of all children.³⁵ 

Families of children with SHCNs are more likely to experience housing instability than families of children without SHCNs. SSI may help protect families of children with SHCNs from all 3 adverse circumstances of housing instability: behind on rent, multiple moves, and homelessness. Protecting families with children from housing instability is an important investment in the health and well-being of all families, particularly families of children with SHCNs, and should be a public policy priority.

aOR

adjusted odds ratio

CI

confidence interval

LIHEAP

Low Income Home Energy Assistance Program

SHCN

special health care need

SSI

Supplemental Security Income