Social Risk Interventions and Health Care Utilization for Pediatric Asthma

Key Points

Question

Are asthma-specific interventions addressing social risks associated with asthma-related health care utilization among children?

Findings

In this systematic review of 38 studies and meta-analysis of 19 studies (comprising 5441 participants), participation in a social risk–based intervention was associated with significantly reduced risk of asthma-related emergency department visits and hospitalizations among children.

Meaning

Findings suggest that it is important to consider universal social risk screening and implementation of social risk–based interventions as elements of pediatric asthma care guidelines.

Abstract

Importance

Social determinants of health (SDOH) correlate with pediatric asthma morbidity, yet whether interventions addressing social risks are associated with asthma outcomes among children is unclear.

Objective

To catalog asthma interventions by the social risks they address and synthesize their associations with asthma-related emergency department (ED) visits and hospitalizations among children.

Data Sources

PubMed, Scopus, PsycINFO, SocINDEX, CINAHL, and references of included full-text articles were searched from January 1, 2008, to June 16, 2021.

Study Selection

Included articles were US-based studies evaluating the associations of interventions addressing 1 or more social risks with asthma-related ED visits and hospitalizations among children. The systematic review included 38 of the original 641 identified articles (6%), and the meta-analysis included 19 articles (3%).

Data Extraction and Synthesis

Data extraction followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses reporting guideline. The SDOH intervention clusters were identified by grouping studies according to the social risks they addressed, using the Healthy People 2020 SDOH framework. Random-effects models created pooled risk ratios (RRs) as the effect estimates.

Main Outcomes and Measures

Patients with ED visits or hospitalizations were the primary outcomes. Subgroup analyses were conducted by an SDOH intervention cluster. Sensitivity analyses were conducted for each, removing outlier studies and studies failing to meet the minimum quality threshold.

Results

In total, 38 studies were included in the systematic review, with 19 of these studies providing data for the meta-analysis (5441 participants). All interventions addressed 1 or more of the health, environment, and community domains; no interventions focused on the economy or education domains. In the primary analysis, social risk interventions were associated with decreased ED visits (RR, 0.68; 95% CI, 0.57-0.81; I² = 70%) and hospitalizations (RR, 0.50; 95% CI, 0.37-0.68; I² = 69%). In subgroup analyses, the health, environment, and community intervention cluster produced the lowest RR for ED visits (RR, 0.53; 95% CI, 0.44-0.64; I² = 50%) and for hospitalizations (RR, 0.33; 95% CI, 0.20-0.55; I² = 71%) compared with other intervention clusters. Sensitivity analyses did not alter primary or subgroup effect estimates.

Conclusions and Relevance

The results of this systematic review and meta-analysis indicate that social risk interventions are associated with decreased asthma-related ED visits and hospitalizations among children. These findings suggest that addressing social risks may be a crucial component of pediatric asthma care to improve health outcomes.

This systematic review and meta-analysis uses the Healthy People 2020 social determinants of health framework to assess whether asthma interventions addressing social risks are associated with decreased asthma-related emergency department visits and hospitalizations among children.

Introduction

Social determinants of health (SDOH) are major drivers of health outcomes and health inequities among both children and adults and may play a much larger role in determining health outcomes than health care.¹ Social determinants of health are “upstream” community conditions that affect people’s quality of life and health outcomes and shape the world in which people live, learn, work, and play.¹ Social risks, or adverse social conditions individuals experience,² are the “midstream” manifestation of SDOH. Policies and laws are needed to address these SDOH.³ Family-level interventions may mediate social risks.² Health care systems and community groups should address social risks because that may improve health outcomes for individuals. Simultaneously, advocating for policy change is needed to address the root causes of health inequities for communities.³ This is underscored in adult populations with chronic diseases. Adults connected with meal delivery services⁴ to address food insecurity and transitional housing units⁵ to address housing instability have decreased health care utilization and costs.

Asthma is a common pediatric chronic medical condition.⁶ A former Health and Human Services Secretary rhetorically stated,³ “How can a mother with an asthmatic son really improve his health if it’s their living environment that’s driving his condition?” Asthma-related health care utilization among children correlates with SDOH. Emergency department (ED) visits and hospitalizations for asthma are associated with adverse neighborhood measures of crime, educational attainment, and household income.^7,8 Length and cost of asthma-related hospitalizations increase with worsening neighborhood deprivation.⁹ This body of research highlights the association between children’s environment and their health.¹⁰ Asthma is the logical pediatric disease to evaluate the ability of interventions to address social risks.²

Two recent reviews cataloged pediatric asthma interventions using SDOH frameworks.^11,12 One narrative review examined risk factors by level of associated effect (individuals, neighborhoods, or health care systems) and then cataloged interventions by location (school, community or home).¹¹ The authors highlighted the need to address factors unique to high-risk neighborhoods, including navigating limited access to care.¹¹ A scoping review used the World Health Organization’s SDOH framework to examine associations of structural and intermediary determinants with asthma outcomes.¹² Their results reinforced the association between SDOH and asthma morbidity.¹²

To the best of our knowledge, asthma interventions have not been systematically analyzed using the Healthy People (HP) 2020 SDOH framework. This framework is composed of 5 SDOH domains: economic stability (economy), education (education), social and community context (community), health and health care (health), and neighborhood and the built environment (environment).¹⁰ In addition, asthma interventions have not been evaluated for their ability to effect children’s health utilization by addressing social risks. We designed this meta-analysis to begin to address this gap in the literature. We sought to respond to the call by the American Academy of Pediatrics to catalog and develop interventions addressing the pillars of poverty to improve pediatric health outcomes.¹³

The objectives of this meta-analysis are to (1) systematically review pediatric asthma interventions by the social risks addressed using the HP 2020 framework and (2) synthesize the association of social risk–based interventions for pediatric asthma with asthma-related utilization outcomes.

Methods

Study Eligibility

We created a review protocol identifying studies that examined social risk interventions for pediatric asthma using the HP SDOH 2020 framework to classify the focus of each intervention. Studies in the United States were included if (1) outcome variables included asthma-related ED visits or hospitalizations, (2) 1 or more social risks (categorized by HP 2020 framework) were addressed by the intervention (exposure), (3) an intervention was evaluated, (4) the study population reported separate outcomes for children 18 years of age or younger, and (5) the study was published during or after 2008. This start date was the publication year for the US-based HP 2020 framework. This systematic review and meta-analysis was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline.

Studies were excluded from the systematic review and meta-analysis if they did not separate results between children and adult participants, were not conducted in the US, were published before 2008, did not evaluate an intervention, or reported a composite measure of acute health care utilization that included scheduled or unscheduled clinic encounters.

Information Sources and Search Strategy

All authors, including a professional medical librarian (S.K.), created the search strategy (eAppendix in the Supplement) and conducted the initial literature search. Five major databases were searched from January 1, 2008, to the present: PubMed, Scopus, PsychINFO, SocINDEX, and CINAHL. References of included full-text articles were searched for additional studies to include. Databases were last checked on June 16, 2021.

Study Selection

Two of us (J.T. and K.P.) independently completed a title and abstract review using the Covidence online program.¹⁴ Disagreements were resolved through discussion and consensus agreement. One of us (J.T.) examined the reference lists of the included full-text articles for additional studies that met inclusion criteria.

Systematic Review Data Extraction

One of us (J.T.) extracted data using a Microsoft Excel–based form. Extracted variables included study characteristics (authors, publication year, and study design); study population (number of participants and age); intervention details (intervention components, SDOH domains of the addressed social risk factors, setting, target audience, and duration); and outcome variables (outcome measurement, time from intervention completion to outcome measurement, and reporting method). Multiple outcome measurements were included in the systematic review: (1) patients with ED visits or hospitalizations, (2) total number of ED visits and hospitalizations by study group, (3) length of stay, and (4) hospitalization charges. If outcomes data for multiple time periods were provided, data for all periods were extracted.

SDOH Intervention Classification

Studies were cataloged according to the HP 2020 SDOH domain of the social risks they addressed. The HP 2020 SDOH domains are economy, education, community, health, and environment. Studies that addressed social risks belonging to the same SDOH domains were grouped together to identify clusters of interventions inherent to this review. For example, studies that addressed 1 or more social risks in 1 domain (eg, health) were grouped as a single domain intervention cluster. Studies that addressed social risks from multiple domains (eg, health and environment) were grouped into a multidomain intervention cluster. Interventions incorporating asthma education were cataloged under health literacy (health). Interventions using community health workers were cataloged under social cohesion (community).

Quality Assessment Methods

Study quality was assessed for all studies included in the systematic review and meta-analysis with an established tool used in a prior meta-analysis of asthma interventions.¹⁵ This tool assesses 6 domains: (1) intervention description, (2) participants, (3) study measures, (4) strength of findings, (5) accounting for data variation, and (6) ethics. A threshold of minimum quality in 4 of 6 domains was used.

Statistical Analysis

Asthma-related ED visits and hospitalizations were chosen as 2 patient-centered outcome measures. These events represent life-threatening exacerbations that are costly and potentially preventable.¹⁶ In addition, 4 HP 2030 goals focus on reducing asthma-related ED visits and hospitalizations among children.¹⁷ Extracted variables for these analyses included (1) the number of patients with ED or hospitalization events in the control (preintervention) and intervention (postintervention) groups (numerator) and (2) the number of total participants in each group (denominator). If studies did not provide patients with events (eg, provided only mean differences or total events per group), authors were contacted. Of 18 authors contacted, 6 responded, and 2 provided meta-analysis data. The 16 studies for which data were not provided were excluded from the meta-analysis.

Pooled risk ratios (RRs) with corresponding 95% CIs were the effect estimates for each outcome. We used RevMan software¹⁸ to create those estimates with random-effects models and inverse-variance analysis. Study heterogeneity was assessed using the I² statistic.

Subgroup and Sensitivity Analyses

Subgroup analyses were conducted for each identified SDOH intervention cluster. Two sensitivity analyses were completed: (1) removal of outliers (individual studies with 95% CIs that did not overlap with the overall estimated 95% CI)¹⁹ and (2) removal of studies that failed to meet minimum quality standards in 4 of 6 domains.

Results

Study Selection

The initial search identified 641 articles. Of these, 66 were included in the full-text review. We identified 24 additional articles from the reference lists of the included full-text articles. In total, 38 studies were included in the systematic review, and 19 studies were included in the meta-analysis (5441 participants) (Figure 1). The most frequent reasons for exclusion from the systematic review and meta-analysis were study design (most commonly noninterventional, retrospective studies) and study outcomes (most commonly process outcomes or outpatient utilization outcomes).

Figure 1. Flowchart of Study Identification and Selection.

Study Characteristics

All included studies addressed social risks belonging to the health (n = 35), environment (n = 21), and community (n = 15) domains (Table 1).^{20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57} Some studies addressed social risks in more than 1 domain category; thus, the total number does not equal the number of included studies. No included studies addressed social risks in the economy or education domains. Most interventions addressed social risks belonging to multiple domains (n = 23) rather than to a single SDOH domain (n = 15). Table 2^{20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57} provides intervention descriptions. The intervention details, including study design, duration, setting, and outcomes, are given in eTable 1 in the Supplement.

Table 1. Classification of Articles by Healthy People 2020 Social Determinants of Health Domains^a.

Source	Health and health careb	Social and community contextb	Neighborhood and the built environmentb
Atherly et al,20 2009c	X
Horner and Brown,21 2014c	X
Riera et al,22 2015d	X
Robinson et al,23 2008c	X
Tapp et al,24 2017c	X
Zorc et al,25 2009c	X
Bollinger et al,26 2010d	X
Celano et al,27 2012c	X
Eakin et al,28 2012d	X
McQuaid et al,29 2012d	X
Halterman et al,30 2018d	X
Magzamen et al,31 2008d	X
Butz et al,32 2019d			X
Colton et al,33 2015c			X
Matsui et al,34 2017d			X
Coughey et al,35 2010c	X	X
Findley et al,36 2011c	X	X
Fisher et al,37 2009c	X	X
Flores et al,38 2009c	X	X
Otsuki et al,39 2009c	X	X
Gruber et al,40 2016d	X		X
Kennedy et al,41 2017c	X		X
Largo et al,42 2011c	X		X
Mankikar et al,43 2016d	X		X
McClure et al,44 2017c	X		X
Sweet et al,45 2014d	X		X
Turcotte et al,46 2014d	X		X
Shuler et al,47 2015d	X		X
Bryant-Stephens et al,48 2008d	X	X	X
Butz et al,49 2011d	X	X	X
Campbell et al,50 2015d	X	X	X
Janevic et al,51 2016d	X	X	X
Lara et al,52 2013c	X	X	X
Leibel et al,53 2020d	X	X	X
Rapp et al,54 2018c	X	X	X
Shani et al,55 2015d	X	X	X
Turyk et al,56 2013c	X	X	X
Woods et al,57 2016c	X	X	X

Abbreviation: X, domain addressed by the study.

^a Two social determinants of health domains, economic stability and education, were not addressed by the included studies.

^b Social risks addressed in each social determinants of health domain include access to primary care, access to health care, and health literacy (health and health care); social cohesion (social and community context); environmental conditions and housing quality (neighborhood and the built environment).¹⁰

^c Articles included in the systematic review and meta-analysis.

^d Articles included in systematic review only.

Table 2. Intervention Descriptions by Social Determinants of Health Intervention Cluster.

Source	Intervention description
Health
Atherly et al,20 2009	3 Education sessions focused on asthma medications and pathophysiology; how to improve asthma control; constructive coping strategies and communication skills
Horner and Brown,21 2014	16 Education sessions: in part A, children reviewed self-efficacy skills for inhaler technique, coping, problem-solving skills with symptoms; in part B, parents received Home Asthma booklet and had 1 home visit to review individualized education and AAP
Riera et al,22 2014	1 Education fair with stations reviewing (1) medication, (2) peak flow, and (3) action plan
Robinson et al,23 2008	Weekly 3-h literacy and asthma education session with refresher courses every 4-6 wk over 6 mo; opportunity to attend 5-d asthma camp
Tapp et al,24 2017	Minimum 1 shared decision visit during which asthma education and parent perception of asthma control was reviewed and goals of care and treatment plan were negotiated
Zorc et al,25 2009	1-Time viewing of asthma education video in ED; letter about child’s asthma sent from ED to PCP
Bollinger et al,26 2010	Mobile asthma clinic providing medical care and education every 4-12 wk based on child’s asthma severity
Celano et al,27 2012	4-6 Home visits during 4 mo that included an asthma counselor, respiratory therapist, and psychology fellow to review asthma triggers, control, resources, MDI technique, and family identified goals
Eakin et al,28 2011	3 Intervention groups: mobile asthma clinic (medical care and asthma education) or facilitated asthma communication intervention (home visit with asthma educator and review of how to communicate with PCP)
McQuaid et al,29 2012	3 Home visits reviewing smoking cessation counseling and asthma education (medication and inhaler use; triggers and avoidance; and symptoms and control)
Halterman et al,30 2018	3 School-based telemedicine visits for asthma care throughout 1 school year in addition to directly observed therapy for preventative asthma medications
Magzamen et al,31 2008	4 Weekly 50-min school-based asthma education sessions (asthma basics, triggers, medications, and problem solving) provided by an asthma nurse for 1 mo
Environment
Butz et al,32 2019	3 Visits (2 home) providing environmental remediation based on the child’s allergen sensitization results; if applicable, brief motivational interview to implement a total home smoking ban
Colton et al,33 2015	Natural intervention arose when Boston Housing Authority reconfigured half of public housing as green housing
Matsui et al,34 2017	Integrated pest management provided in 2 visits 1-2 wk apart with reassessment every 3 mo
Health and community
Coughey et al,35 2010	Telephone-based care coordination and service integration program staffed by 2 bilingual asthma care coordinators who lived in the same community on which this intervention focused and had experience in health education, social services, and community outreach. Services included: asthma education and connection with various medical, educational, and social resources
Findley et al,36 2010	2 Education sessions by the Northern Manhattan Asthma Basics for Children Initiative that reviewed asthma medications, triggers and avoidance, asthma signs, and communication skills with clinicians. Supplemental education activities were available for parents and children. Provided at the child’s childcare center. Parent asthma mentors mobilized parents to participate and provided Q&A sessions for enrolled parents. Parent mentors received 24 h of asthma training and needed to be bilingual and show control of their child’s asthma.
Fisher et al,37 2009	2 In-person visits followed by telephone calls every 2 wk for 3 mo then monthly for 2 y. Health coaches were from similar neighborhoods to participants and received 3 mo training on asthma education and behavior change methods
Flores et al,38 2009	During group sessions, 15 asthma-related topics were reviewed during the course of the intervention. Parent mentors received 2 d of training on asthma, how to be a peer mentor, and how to address social risks for families; monthly group visits included peer support and networking with other families. Mentors provided 24 h, 7 d/wk of support by telephone as needed
Otsuki et al,39 2009	5 Education visits reviewed asthma medications, AAP, health care access barriers, and concerns about asthma. Asthma educators partnered with families to set asthma control goals and provided positive reinforcement
Health and environment
Gruber et al,40 2016	Education provided about asthma triggers. Housing assessment with remediations as indicated; if none were indicated, participants provided with HEPA cleaner and allergen covers
Kennedy et al,41 2017	4 Counseling sessions provided about asthma topics individualized to the family. One environment assessment to identify triggers and allergens
Largo et al,42 2011	3 Education sessions provided for asthma triggers and injury prevention. Basic products and custom remediation plans created and implemented
Mankikar et al,43 2016	Education provided about asthma prevention and management, injury prevention, and lead poisoning prevention; 2 Home visits 2 mo apart providing an environmental assessment and intervention (with supplies or referrals as needed)
McClure et al,44 2017	Home visits with nursing students to review medications, asthma symptoms, and AAP. Home assessment with 1 or more products provided (eg, allergen cover, vacuum with HEPA filter)
Sweet et al,45 2014	Education provided about asthma management and trigger reduction; caregiver understanding of child's asthma care was reviewed. Home assessment completed with provision of bedding and cleaning supplies; structural issues addressed if identified
Turcotte et al,46 2014	Education provided about asthma triggers and safety hazards. Custom remediation plans included integrated pest management, cleaning, cleaning supplies, or structural intervention
Shuler et al,47 2015	Asthma education materials provided; medications and AAP reviewed during visits; children received individualized asthma education session. Home assessed and remediation provided (families linked to services, provided resources for irritant reduction, and housing code violations reported)
Health, environment, and community
Bryant-Stephens et al,48 2008	5 Home visits reviewed asthma knowledge, medications, inhaler techniques, triggers, and avoidance. Home assessment completed; supplies provided for irritant or allergen reduction and cleaning at each visit. Lay health educators lived in the same geographic area as participants and were required to have high school education and 3 y of experience. They received didactic training on asthma, and study design and showed knowledge through mock scenarios and practical assessments
Butz et al,49 2011	Children taught how to avoid SHS 2 Air cleaners provided for home Health coach provided motivational interviewing and goal setting for secondhand smoke reduction
Campbell et al,50 2015	4 Visits, combination of home and telephone, providing asthma education and support Cleaning, pest management, and irritant reduction products were provided The CHWs shared ethnic and cultural similarities with participants, had family experience with asthma, and lived in the same geographic location. They received 40 h of initial training and 10-20 h of additional training throughout the year.
Janevic et al,51 2016	During 1-4 home visits with or without telephone follow-up,a asthma education was provided; environmental assessments were performed; supplies for mitigation were provided (eg, mattress cover). Asthma care coordinators provided home visits; they were trained in culturally relevant care and had various professional backgrounds (asthma program nurses, community health educators, or CHWs).
Lara et al,52 2013	Education on recognition and reduction of allergens and irritants provided Safe sleeping zone module with irritant reduction products provided; cockroach mitigation techniques reviewed The CHWs provided 2-3 home visits
Leibel et al,53 2020	Environmental assessment was performed during visit; cleaning recipes and trigger mitigation techniques reviewed. The CHWs provided 1 home visit to reinforce asthma knowledge provided in the severe asthma clinic. Social determinants of health screening performed and referrals made as needed to community resources with telephone follow-up during the ensuing 12 mo
Rapp et al,54 2018	Asthma educators, certified by the National Asthma Educator Certification Board, provided clinic-based education tailored to each participant. For participants with uncontrolled asthma, CHW provided a home visit, assessment, and remediation recommendations
Shani et al,55 2015	Education included medication review, proper inhaler use, trigger reduction. Asthma-friendly environmental products provided and integrated pest management skills taught. Peer educators delivered home visits (their training not described)
Turyk et al,56 2013	Home visits included asthma education (eg, medications, AAP, barriers to care, and triggers) and care coordination. Irritant reduction supplies provided; minor remediation tools given; and referrals provided for more extensive remediation. Community health educators with community ties received 1 wk of training
Woods et al,57 2016	Home visits included asthma education and care coordination. Home assessment, remediation, cleaning supplies or irritant reduction products, and integrated pest management. The CHW provided home visits and case management

Abbreviations: AAP, asthma action plan; CHWs, community health workers; ED, emergency department; HEPA, high-efficiency particulate air; MDI, metered-dose inhaler; PCP, primary care physician; SHS, secondhand smoke.

^a This multisite intervention implemented in 4 institutions had some variation across sites with respect to number of home visits, use of telephone follow-up, and inclusion of AAP review and goal setting in visits.

SDOH Intervention Clusters

Five intervention clusters were identified (Table 1). Twelve studies addressed health only. Seven studies addressed health literacy through asthma education provided to children or their caregivers^{20,21,22,23,24,25,31}; 4 studies incorporated health literacy with access to care by delivering asthma education or asthma care in the child’s home,^26,27,28,29 and 1 study implemented school-based telemedicine visits for asthma care.³⁰ Three interventions addressed only environment. Two studies incorporated home remediation,^32,34 and 1 study examined living in green vs traditional public housing.³³ Eight studies addressed health and environment by combining asthma education and home remediation.^{40,41,42,43,44,45,46,47} Five studies addressed health and community by incorporating asthma education and peer support in the child’s home or community or by telephone-based care coordination.^{35,36,37,38,39} Ten interventions addressed health, environment, and community through a combination of asthma education, home environmental assessment or remediation, and community health workers.^{48,49,50,51,52,53,54,55,56,57}

Assessment of the Quality of Study Methods

Studies included in the systematic review and meta-analysis were assessed for study quality (eTable 2 in the Supplement). All studies provided clear rationales for their intervention and outcomes. A total of 36 studies provided a replicable intervention description. Most studies used before and after intervention (n = 18) and randomized group (n = 14) designs. The remainder of the studies used a cohort with a nonrandomized or historic control design (n = 5). All studies met the minimum quality standards in 4 of 6 domains.

Social Risk Interventions and Risk of Utilization

A total of 19 studies provided data for the meta-analysis. The overall RR for ED visits following a social risk–based intervention was 0.68 (95% CI, 0.57-0.81; I² = 70%). The overall RR for hospitalizations was 0.50 (95% CI, 0.37-0.68; I² = 69%). Figure 2 provides the data for the ED visit RRs, and Figure 3 gives the hospitalization RRs.

Figure 2. Risk of an Emergency Department Visit Following Participation in a Social Risk–Based Intervention.

Pooled risk ratios (RRs) with corresponding 95% CIs were the effect estimates (with random-effects models and inverse-variance analysis) for each outcome.

Figure 3. Risk of Hospitalization Following Participation in a Social Risk–Based Intervention.

Pooled risk ratios (RRs) with corresponding 95% CIs were the effect estimates (with random-effects models and inverse-variance analysis) for each outcome.

Subgroup Analyses by SDOH Intervention Cluster

The results for the subanalyses of ED visits and hospitalizations are given in eFigures 1 and 2 in the Supplement, respectively. Six studies provided outcome data for ED visits and hospitalizations for the health intervention cluster.^{20,21,23,24,25,27} For the environment intervention cluster subgroup analysis, only 1 study³³ provided outcome data. Two studies^42,44 in the health and environment intervention cluster provided outcome data for ED visits, and 3 studies^41,42,44 included data for hospitalizations. These interventions incorporated asthma education and home environment remediation. This subgroup analysis tied with the health and environment and community intervention cluster in producing the lowest RR for hospitalizations (0.33; 95% CI, 0.21-0.50; I² = 0%). In the health and community intervention cluster, 4 studies^35,36,37,38 provided outcomes for ED visits, and 5 studies^{35,36,37,38,39} provided outcomes for hospitalizations. Health and environment and community were the only 3-domain intervention cluster. Interventions incorporated asthma education, home environment assessment or remediation, and community health workers. Four studies provided outcome data for ED visits and hospitalizations for this intervention cluster.^52,54,56,57 This subgroup analysis produced the lowest RR for ED visits (0.53; 95% CI, 0.44-0.64; I² = 50%) and tied with the health and environment intervention cluster for lowest RR for hospitalizations (0.33; 95% CI, 0.20-0.55; I² = 71%).

Sensitivity Analyses

When outlier studies for ED visits^37,57 and hospitalizations^38,57 were removed, the significance of the overall and subgroup effect estimates did not change, but heterogeneity was substantially reduced. All studies met minimum quality threshold. No additional sensitivity analyses were completed.

Discussion

This meta-analysis, to the best of our knowledge, is the first to leverage the HP 2020 SDOH framework to synthesize existing literature and quantify the association of social risk–based interventions with asthma-related health care utilization among children. Our results support the core vision of an overarching HP goal, which is rooted in improving a child’s living conditions to facilitate improved health.¹⁷ The findings of this meta-analysis support 4 HP 2030 SDOH objectives focused on reducing ED visits and hospitalizations among children with asthma.¹⁷ The present study takes an important step by highlighting the value of addressing individual-level social risks while, in parallel, the pediatric community advocates for policies to address population-level SDOH.

Our review showed that asthma interventions addressing social risks were associated with reduced ED visits and hospitalizations. We found that existing asthma interventions addressed social risks in only 3 of 5 HP 2020 SDOH domains (health, community, and environment). No interventions addressed risks in the economy or education domains. In our subgroup analyses, the health, environment, and community intervention cluster was associated with the greatest reduction in ED visits. These interventions combined place-based asthma education delivery, home environment remediation, and peer support. For hospitalizations, the health and environment intervention cluster (eg, only asthma education and environmental remediation) provided an RR equal to that of the health, environment, and community intervention cluster. This finding suggests that the added benefit of community, through the implementation of peer support with community health workers or health coaches, may have a greater association with reduced ED visits compared with hospitalizations. The health and community intervention cluster that did not incorporate environment was associated with the least reductions in both ED visits and hospitalizations. This finding supports our interpretation that health (asthma education) and environment (environmental remediation) are the 2 most integral components associated with effective social risk–based asthma interventions.

Current asthma guidelines from the National Heart, Lung, and Blood Institute⁵⁸ and the Global Initiative for Asthma⁵⁹ do not yet advocate for widespread implementation of social risk screening nor social risk–based interventions and policies addressing SDOH. Although SDOH account for up to 55% of all health outcomes,¹ asthma care guidelines only highlight the importance of education and environment in asthma management. Reviews have started to examine pediatric asthma using an SDOH lens.^11,12 Another systematic review provided an extensive classification of interventions addressing SDOH among children and adults across a spectrum of disease.⁶⁰ Our meta-analysis is the first, to our knowledge, to synthesize the association of interventions addressing social risks with pediatric asthma-related health care utilization. We found that interventions responding to social risks from multiple SDOH domains were associated with the greatest reduction in ED visits and hospitalizations. These data support the call to use cross-sector collaboration to reduce asthma-related health disparities among children.¹¹ The effectiveness of multidomain interventions found in this review supports existing research that multicomponent interventions with community partnership are associated with improved pediatric health care utilization.⁶¹ Of particular relevance to the global COVID-19 pandemic and its effects on pediatric health care, our review highlighted innovative methods that may be implemented broadly to increase children’s health care access (health domain) with telephone-based approaches.^30,35 We suggest that assessing families’ social risks and connecting them to programs to address health (increasing asthma-specific health literacy and access to care), environment (remediating home environments), and community (providing peer support) be incorporated into current asthma care guidelines. A tiered approach may be used to first focus on children with uncontrolled or moderate to severe asthma, similar to approaches used by studies examined in this review.^{27,32,40,41,50,52,53}

The absence of asthma-specific interventions that address economy and education in this review highlight future opportunities and directions. Food insecurity is 1 component of the economy domain. One noninterventional study used administrative data to show that receipt of increased Supplemental Nutrition Assistance Program benefits was associated with decreased likelihood of asthma-related ED visits among children.⁶² Two additional pediatric studies, not asthma specific, examined food insecurity interventions.^63,64 The first study found that providing meals in an ED-based Summer Food Service Program was not negatively associated with care delivery and that most caregivers supported the program.⁶³ The second study addressed food insecurity among infants by providing formula, education, and coordination with community resources.⁶⁴ Infants in the intervention group were more likely to be connected with resources and to be up to date with preventative services.⁶⁴ Of notable importance for children with chronic illnesses, such as asthma, there is a growing body of evidence that addressing food insecurity is associated with improved health outcomes among adults with chronic conditions.⁴ Housing instability is another component of the economy domain. One retrospective study examined the association of a 1990s housing policy program with hospital use by the general population.⁶⁵ Children whose families received a housing voucher had decreased hospitalizations compared with those who did not receive a voucher.⁶⁵

One component of the education domain is early childhood education. A non–asthma-specific randomized clinical trial of children found that children who received high-quality education between 0 and 5 years in addition to nutrition and health care during the study period had decreased cardiac and metabolic risk factors in adulthood.⁶⁶ As individual-level interventions and community-wide policies are developed to address economy and education, it is imperative to also evaluate their association with health care utilization outcomes among children.

Limitations

This study has limitations. First, study quality was suboptimal owing to frequent use of before intervention vs after intervention designs and cohort study designs. Most interventions were multicomponent and community based. It may be that studies prioritized having as many children as possible receive the intervention. In future research, it will be important to balance this priority with optimal methodologic rigor to ensure that associations of outcomes with interventions are clearly shown. Second, many included articles were identified from reference lists of included full-text articles. The evolving focus on SDOH research may have limited the utility of the terms used in the initial search. Third, 16 studies for which data were not provided on request were excluded from the meta-analysis. Fourth, in our meta-analysis, there was large study heterogeneity. This was expected and reflects the varied nature of social risk–based interventions. Similarly, these types of interventions will need to be patient centered to be acceptable and will vary when implemented across different communities.

Conclusions

The field of pediatric asthma interventions targeting social risks is varied and evolving. The present systematic review and meta-analysis found that interventions focusing on health, environment, and community were associated with the greatest reduction in asthma-related ED visits and hospitalizations among children. Based on the findings of this meta-analysis, referring families to programs addressing social risks may be a crucial aspect of pediatric asthma care to improve health outcomes. This would be an individual-level solution while the pediatric community simultaneously advocates for implementation of policies addressing these SDOH domains at the community level.

Supplement.

eAppendix. Search Strategy

eTable 1. Study Characteristics by Social Determinant of Health Intervention Cluster

eTable 2. Assessment for Study Quality

eFigure 1. Subgroup Analysis of Risk of an Emergency Department Visit Following Participation in a Social Risk–Based Intervention

eFigure 2. Subgroup Analysis of Risk of Hospitalization Following Participation in a Social Risk–Based Intervention