COVID-19 Disruptions to Social Care Delivery: A Qualitative Study in Two Large, Safety-Net Primary Care Clinics.

BACKGROUND: Social care integration refers to the incorporation of activities into health systems that assist patients with health-related social needs (HRSNs) that negatively impact the health outcomes of their patients, such as food insecurity or homelessness. Social care integration initiatives are becoming more common. The COVID-19 pandemic strained health systems while simultaneously increasing levels of unmet social needs. OBJECTIVE: To describe the effects of the COVID-19 pandemic on established social care delivery in a primary care setting. DESIGN: We used qualitative semi-structured interviews of stakeholders to assess barriers and facilitators to social care delivery in the primary care setting during the COVID-19 health emergency. Data was analyzed using a hybrid inductive/deductive thematic analysis approach with both the Consolidated Framework for Implementation Research (CFIR) and the Screen-Navigate-Connect-Address-Evaluate model of social care integration. SETTING: Two safety-net, hospital-based primary care clinics with established screening for food insecurity, homelessness, and legal needs. PARTICIPANTS: Six physicians, six nurses, six members of the social work team (clinical social workers and medical case workers), six community health workers, and six patients (total N = 30) completed interviews. RESULTS: Four major themes were identified. (1) A strained workforce experienced challenges confronting increased levels of HRSNs. (2) Vulnerable populations experienced a disproportionate negative impact in coping with effects of the COVID-19 pandemic on HRSNs. (3) COVID-19 protections compounded social isolation but did not extinguish the sense of community. (4) Fluctuations in the social service landscape led to variable experiences. CONCLUSIONS: The COVID-19 pandemic disrupted established social care delivery in a primary care setting. Many of the lessons learned about challenges to social care delivery when health systems are strained are important considerations that can inform efforts to expand social care delivery.

Associations between inspections for unsafe housing conditions and evictions in New York City public housing buildings.

Poor housing conditions and evictions are both associated with poor physical and mental health outcomes, such as increased risks for cardiovascular disease, depression, and injuries. However, the relationship between these two negative housing outcomes has received little quantitative study, including in public housing where exposure to these factors and to negative health outcomes are elevated. We therefore sought to examine the relationship between evictions and formal housing safety inspections triggered by tenant complains about poor conditions. We estimated a hierarchical logistic regression model assessing associations between housing quality inspections and evictions using data from January 2017 and March 2020 on 3,746 residential buildings within 299 New York City Housing Authority (NYCHA) developments, adjusting for development size, funding type, and area-level social vulnerability indicators. The average Social Vulnerability Index percentile for the buildings included in this study was 0.90 (SD = 0.12), indicating that these buildings were in areas with greater social vulnerability than 90% of other census tracts in the state. Adjusted predicted probabilities of an eviction increased from 34 to 43% in the presence of a rodent inspection and from 34 to 46% in the presence of an indoor environmental inspection (p < 0.001 for both), indicating that inspections for unsafe housing conditions were associated with evictions at the building level. Substandard housing quality and evictions are important public health concerns. Policies to enhance protections for tenants against both of these social ills simultaneously may be needed to improve community health outcomes.

Household food insufficiency, health status and emergency healthcare utilisation among children with and without special healthcare needs.

OBJECTIVE: To compare exposure to household food insufficiency and the relationship between household food insufficiency and both health status and emergency healthcare utilisation among children with and without special healthcare needs (SHCN). DESIGN: Analysing pooled data from the 2016-2017 iterations of the National Survey of Children's Health, we conducted multivariate logistic regressions on household food insufficiency, health status and emergency healthcare utilisation. We assessed interactions between household food insufficiency and children's SHCN status in our models of health status and utilisation. SETTING: United States. PARTICIPANTS: Parents of a nationally representative sample of non-institutionalised children (aged 0-17 years). RESULTS: Children with SHCN were more likely to experience household food insufficiency (70 v. 56 %), non-excellent health status (67 v. 28 %) and emergency healthcare utilisation (32 v. 18 %) than other children. Household food insufficiency was associated with 37 % (children with SHCN) and 19 % (children without SHCN) reductions in the likelihood of having excellent health. Household food insufficiency was associated with a roughly equal (16-19 %) increase in the likelihood of emergency department utilisation across groups. CONCLUSIONS: Compared with other children, children with SHCN have an elevated risk of exposure to household food insufficiency and experiencing greater reductions in health status when exposed.

Understanding the Role of Past Health Care Discrimination in Help-Seeking and Shared Decision-Making for Depression Treatment Preferences.

As a part of a larger, mixed-methods research study, we conducted semi-structured interviews with 21 adults with depressive symptoms to understand the role that past health care discrimination plays in shaping help-seeking for depression treatment and receiving preferred treatment modalities. We recruited to achieve heterogeneity of racial/ethnic backgrounds and history of health care discrimination in our participant sample. Participants were Hispanic/Latino (n = 4), non-Hispanic/Latino Black (n = 8), or non-Hispanic/Latino White (n = 9). Twelve reported health care discrimination due to race/ethnicity, language, perceived social class, and/or mental health diagnosis. Health care discrimination exacerbated barriers to initiating and continuing depression treatment among patients from diverse backgrounds or with stigmatized mental health conditions. Treatment preferences emerged as fluid and shaped by shared decisions made within a trustworthy patient-provider relationship. However, patients who had experienced health care discrimination faced greater challenges to forming trusting relationships with providers and thus engaging in shared decision-making processes.

Food insecurity patterns before and after initial receipt of Supplemental Security Income.

OBJECTIVE: To assess patterns of food insecurity before and after initial receipt of Supplemental Security Income (SSI) benefits. DESIGN: We analysed data from a nationally representative sample. We estimated two difference-in-difference models comparing food insecurity patterns among eventual SSI recipients with patterns among eligible non-recipients during two time frames. The first model assessed changes in food insecurity immediately before SSI benefits were first received and the second model assessed changes in food insecurity after programme entry. SETTING: 2008 panel of the Survey of Income and Program Participation.ParticipantsNon-institutionalized population of the USA. RESULTS: The percentage of eventual SSI recipients experiencing food insecurity rose from 18 to 30 % in the year before programme entry, compared with a change from 17 to 18 % for eligible non-recipients. Adjusting for sociodemographic covariates, the difference-in-difference estimator for this comparison was statistically significant (P=0·01). Additionally, the percentage of recipients experiencing food insecurity fell from 28 % in the year before programme entry to 16 % in the year after entry, compared with a change from 16 to 17 % for eligible non-recipients. Adjusting for sociodemographic covariates, the difference-in-difference estimator for this comparison was marginally significant (P=0·07). CONCLUSIONS: Food insecurity rises prior to SSI entry but may be alleviated by programme benefits. Greater nutritional supports for SSI applicants awaiting decisions may reduce the burden of food insecurity in this population and improve health outcomes.

Inpatient Medicaid Usage and Expenditure Patterns After Changes in Supplemental Nutrition Assistance Program Benefit Levels.

INTRODUCTION: Food insecurity worsens health outcomes and is associated with increased health care usage and expenditures. The Supplemental Nutrition Assistance Program (SNAP) reduces but does not eliminate recipients' food insecurity. We sought to determine whether inpatient Medicaid usage and expenditure patterns responded to an April 2009 increase in SNAP benefit levels and a subsequent November 2013 decrease. METHODS: Interrupted time series models estimated responses to the 2009 and 2013 SNAP changes in the Medicaid population, compared responses between Medicaid and Medicare recipients, and compared responses between Medicaid recipients with different likelihoods of having a disability. Analyses used 2006 through 2014 Healthcare Cost and Utilization Project National (previously Nationwide) Inpatient Sample data. RESULTS: After the 2009 SNAP increase, Medicaid admission growth fell nationally from 0.80 to 0.35 percentage points per month (a difference of -0.45; 95% CI, -0.72 to -0.19), adjusting for enrollment. After the 2013 SNAP decrease, admission growth rose to 2.42 percentage points per month (a difference of 2.07; 95% CI, 0.68 to 3.46). Inflation-adjusted monthly Medicaid expenditures followed similar patterns and were associated with $26.5 billion (in 2006 dollars) in reduced expenditures over the 55 months of the SNAP increase, and $6.4 billion (in 2006 dollars) in additional expenditures over the first 14 months after the SNAP decrease. Effects were elevated for Medicaid compared with Medicare recipients and among people with a high likelihood of having a disability. CONCLUSION: Although alternative causal explanations warrant consideration, changes in SNAP benefit levels were associated with changes in inpatient Medicaid usage and cost patterns.

Massachusetts Inpatient Medicaid Cost Response to Increased Supplemental Nutrition Assistance Program Benefits.

OBJECTIVES: To investigate the impact of an increase in Supplemental Nutrition Assistance Program (SNAP) benefits on Medicaid costs and use in Massachusetts. METHODS: Using single and multigroup interrupted time series models, I examined the effect of an April 2009 increase in SNAP benefits on inpatient Medicaid cost and use patterns. I analyzed monthly Medicaid discharge data from 2006 to 2012 collected by the Massachusetts Center for Health Information and Analysis. RESULTS: Inpatient costs for the overall Massachusetts Medicaid population grew by 0.55 percentage points per month (P < .001) before the SNAP increase. After the increase, cost growth fell by 73% to 0.15 percentage points per month (-0.40; P = .003). Compared with the overall Medicaid population, cost growth for people with the selected chronic illnesses was significantly greater before the SNAP increase, as was the decline in growth afterward. Reduced hospital admissions after the SNAP increase drove the cost declines. CONCLUSIONS: Medicaid cost growth fell in Massachusetts after SNAP benefits increased, especially for people with chronic illnesses with high sensitivity to food insecurity.

Eliciting patient past experiences of healthcare discrimination as a potential pathway to reduce health disparities: A qualitative study of primary care staff.

OBJECTIVE: To understand whether and how primary care providers and staff elicit patients' past experiences of healthcare discrimination when providing care. DATA SOURCES/STUDY SETTING: Twenty qualitative semi-structured interviews were conducted with healthcare staff in primary care roles to inform future interventions to integrate data about past experiences of healthcare discrimination into clinical care. STUDY DESIGN: Qualitative study. DATA COLLECTION/EXTRACTION METHODS: Data were collected via semi-structured qualitative interviews between December 2018 and January 2019, with health care staff in primary care roles at a hospital-based clinic within an urban safety-net health system that serves a patient population with significant racial, ethnic, and linguistic diversity. PRINCIPAL FINDINGS: Providers did not routinely, or in a structured way, elicit information about past experiences of healthcare discrimination. Some providers believed that information about healthcare discrimination experiences could allow them to be more aware of and responsive to their patients' needs and to establish more trusting relationships. Others did not deem it appropriate or useful to elicit such information and were concerned about challenges in collecting and effectively using such data. CONCLUSIONS: While providers see value in eliciting past experiences of discrimination, directly and systematically discussing such experiences with patients during a primary care encounter is challenging for them. Collecting this information in primary care settings will likely require implementation of multilevel systematic data collection strategies. Findings presented here can help identify clinic-level opportunities to do so.

Associations between Teacher Wellbeing and Social Support for Physical Activity in Two Urban School Districts during the early stages of the SARS-CoV-2 Pandemic.

Given physical activity's protective effects on mental health and the potential for school districts to support teachers in this area, we explore teacher wellbeing protective factors including social support for exercise. Specifically, we measured the association between social support for exercise and teacher wellbeing in racially and ethnically diverse urban school districts. Based on a prior partnership with 19 schools across 5 districts, we obtained approval from two districts to outreach to teachers (n=206) and invite them to complete the Teacher Subjective Wellbeing Questionnaire and the Social Support and Exercise Survey during the 2020-2021 academic year. We applied linear regression models for continuous variables with teacher wellbeing as the dependent variable and social support as the independent variable adjusting for teacher- and school-level factors. Teachers (n=121) across eight K-8th grade schools completed the survey. The majority of teachers identified as female (77%) and non-White (84%). In the adjusted analysis (n = 104), there was a positive association between family social support for exercise and teacher wellbeing (ß = 0.31; P Value < 0.05). Thus, for every unit increase in family social support for exercise, a small 0.31 unit increase in teacher wellbeing was predicted. Additional research is needed to better understand this relationship in marginalized school districts as it may yield insights to be applied through multiple channels. District representatives have an opportunity to positively influence teacher wellbeing, an important component to supporting student success, building educational equity, and closing the achievement gap.

SNAP Participation and Emergency Department Use.

OBJECTIVES: To examine whether Supplemental Nutrition Assistance Program (SNAP) participation is associated with emergency department use among low-income children and whether any such association is mediated by household food hardship and child health status and/or moderated by special health care needs (SHCN) status. We hypothesized SNAP to be associated with reduced likelihoods of emergency department use, with greater effect sizes for children with SHCN and mediation by food hardship and health status. METHODS: In this secondary analysis, we estimated a bivariate probit model (with state-level SNAP administrative policies as instruments) within a structural equation modeling framework using pooled cross-sectional samples of children in low-income households from the 2016 to 2019 iterations of the National Survey of Children's Health (n = 24 990). RESULTS: Among children with and without SHCN, respectively, SNAP was associated with: 22.0 percentage points (pp) (95% confidence interval [CI] 12.2-31.8pp) and 17.1pp (95% CI 7.2-27.0pp) reductions in the likelihood of household food hardship exposure (4.8pp difference-in-differences, 95% CI 2.3-7.4pp), 9.7pp (95% CI 3.9-15.5pp) and 7.9pp (95% CI 2.2-13.6) increases in the likelihood of excellent health status (1.9pp difference-in-differences, 95% CI 0.7-3.0pp), and 7.7pp (95% CI 2.9-12.5pp) and 4.3pp (95% CI 1.0-7.6pp) reductions in the likelihood of emergency department use (3.4pp difference-in-differences, 95% CI 1.8-5.1pp). CONCLUSIONS: We found SNAP participation was associated with lower likelihoods of emergency department use, that better food hardship and health statuses mediated this association, and that effect sizes were larger among children with SHCN. Food hardship relief may improve outcomes for vulnerable children and the health systems serving them.

Cross-sectional study examining household factors associated with SARS-CoV-2 seropositivity in low-income children in Los Angeles.

OBJECTIVES: This study aims to quantify the degree to which an underserved, Hispanic population in Los Angeles is impacted by SARS-CoV-2, and determine factors associated with paediatric seropositivity. DESIGN: Cross-sectional. SETTING: AltaMed, a Federally Qualified Health Center in Los Angeles. PARTICIPANTS: A random sample of households who had received healthcare at AltaMed Medical Group was invited to participate. Households with at least one adult and one paediatric participant between 5 and 17 years of age were eligible to participate. Consented participants completed a survey on social determinants of health and were tested for antibodies using Abbott Architect SARS-CoV-2-IgG and SARS-CoV-2-IgM tests. PRIMARY OUTCOME MEASURE: Seropositive status. RESULTS: We analysed 390 adults (mean age in years, 38.98 (SD 12.11)) and 332 paediatric participants (11.26 (SD 3.51)) from 196 households. Estimated seropositivity was 52.11% (95% CI 49.61% to 55.19%) in paediatric participants and 63.58% (95% CI 60.39% to 65.24%) in adults. Seropositivity was 11.47% (95% CI 6.82% to 14.09%) lower in paediatric participants, but high relative to other populations. A household member with type 2 diabetes (OR 2.94 (95% CI 1.68 to 5.14)), receipt of food stamps (OR 1.66 (95% CI 1.08 to 2.56)) and lower head-of-household education (OR 1.73 (95% CI 1.06 to 2.84)) were associated with paediatric seropositivity. CONCLUSIONS: SARS-CoV-2 seropositivity is high in Hispanic children and adolescents in Los Angeles. Food insecure households with low head-of-household education, and at least one household member with type 2 diabetes, had the highest risk. These factors may inform paediatrician COVID-19 mitigation recommendations. TRIAL REGISTRATION NUMBER: NCT04901624.

Typologies of mental healthcare discrimination experiences and associations with current provider care ratings: A latent class analysis.

Background: Discrimination is experienced across demographic attributes (e.g., race and gender) and vantage points (e.g., personal and vicarious), yet few studies have classified these different experiences of discrimination within healthcare systems. Moreover, which discriminatory experiences have greater influence on patient-reported quality outcomes remains poorly understood. To address these gaps, we used latent class analysis (LCA) to identify typologies of past experiences with healthcare discrimination among adults with depression-who experience more frequent and stigmatizing healthcare interactions than the general population-and assess the relationship between class membership and current ratings of patient-reported quality outcomes. Methods: We surveyed a nationally representative sample of adults with depression (n = 803) to assess past experiences of discrimination by medical providers in terms of both the characteristics targeted for discrimination and whether healthcare discrimination was experienced personally or by friends and family members. We conducted an LCA to identify discrimination-exposure classes and a modified Poisson regression to identify associations between class membership and patient-reported quality outcomes (e.g., overall medical provider quality, respect, clear communication, and careful listening), while adjusting for covariates. Results: We identified four latent classes of healthcare discrimination: low discrimination (LD; referent class: 72.2% of total sample), vicarious linguistic discrimination (VL; 13.9%), elevated personal and vicarious racial discrimination (EPVR; 10.5%), and high racial/ethnic discrimination (HRE; 3.4%). Compared to those in the LD class, individuals in the EPVR class had higher rates of reporting their current medical provider's respect and careful listening skills as sometimes or never, (Respect aIRR: 1.90, 95% CI: 1.05-3.42; Listening aIRR: 2.18, 95% CI: 1.29-3.66). Those in the HRE class reported higher rates of reporting their medical provider's quality and communication as poor or fair and lower ratings of careful listening (Quality aIRR: 2.06, 95% CI: 1.08-3.93; Communication aIRR: 1.97, 95% CI: 1.00-3.63; Listening aIRR: 2.41, 95% CI: 1.27-4.59), compared to those in the LD class. Those in the VL class had higher rates of reporting that their medical provider never or sometimes respected or carefully listened to them (Respect aIRR: 2.12, 95% CI: 1.20-3.72; Listening aIRR:1.67, 95% CI:1.03-2.71) than those in, the LD class. Conclusions: Healthcare organizations committed to providing equitable patient care should establish more robust quality improvement approaches to prevent discrimination at the medical provider level as well as structures of accountability to reconcile previously embedded social inequities within the healthcare system.

Linking health education, civic engagement, and research at a large Federally Qualified Health Center to address health disparities.

OBJECTIVE: To develop a framework for patient-centered research in a community health center. STUDY SETTING: Primary organizational case-study data were collected at a large Federally Qualified Health Center (FQHC) in Southern California from 2019 to 2021. STUDY DESIGN: Thirty stakeholders, including patients, community leaders, students, medical providers, and academic partners, participated in community-engagement capacity-building exercises and planning. These activities were guided by Community Based Participatory Principles and were part of an initiative to address health disparities by supporting patient and community-engaged research. DATA COLLECTION: The study included an iterative development process. Stakeholders participated in a total of 44 workgroup meetings and 7 full-group quarterly convenings. The minutes of the meetings from both workgroups and quarterly convenings were used to document the evolution of the initiative. PRINCIPLE FINDINGS: Stakeholders concluded that health equity research needs to be part of a larger engagement ecosystem and that, in some ways, engagement on research projects may be a later-stage form of engagement following patient/community and staff/researcher coeducation and cocapacity building efforts. CONCLUSIONS: Community health center stakeholders viewed successful engagement of community members in patient-centered health equity research as involving a web of longitudinal, evolving internal and external relationships rather than discrete, time-limited, and single-project-based dyadic connections.

Interactions of mouse glycophorin A with the dRTA-related mutant G719D of the mouse Cl-/HCO3- exchanger Ae1.

The AE1 mutation G701D, associated with recessive distal renal tubular acidosis (dRTA), produces only minimal erythroid phenotype, reflecting erythroid-specific expression of stimulatory AE1 subunit glycophorin A (GPA). GPA transgene expression could theoretically treat recessive dRTA in patients and in mice expressing cognate Ae1 mutation G719D. However, human (h) GPA and mouse (m) Gpa amino acid sequences are widely divergent, and mGpa function in vitro has not been investigated. We therefore studied in Xenopus oocytes the effects of coexpressed mGpa and hGPA on anion transport by erythroid (e) and kidney (k) isoforms of wild-type mAe1 (meAe1, mkAe1) and of mAe1 mutant G719D. Coexpression of hGPA or mGpa enhanced the function of meAe1 and mkAe1 and rescued the nonfunctional meAe1 and mkAe1 G719D mutants through increased surface expression. Progressive N-terminal truncation studies revealed a role for meAe1 amino acids 22-28 in GPA-responsiveness of meAe1 G719D. MouseN-cyto/humanTMD and humanN-cyto/mouseTMD kAE1 chimeras were active and GPA-responsive. In contrast, whereas chimera mkAe1N-cyto/hkAE1 G701DTMD was GPA-responsive, chimera hkAE1N-cyto/mkAe1 G719DTMD was GPA-insensitive. Moreover, whereas the isolated transmembrane domain (TMD) of hAE1 G701D was GPA-responsive, that of mAe1 G719D was GPA-insensitive. Thus, mGpa increases surface expression and activity of meAe1 and mkAe1. However, the G719D mutation renders certain mAe1 mutant constructs GPA-unresponsive and highlights a role for erythroid-specific meAe1 amino acids 22-28 in GPA-responsiveness.

Depression treatment preferences by race/ethnicity and gender and associations between past healthcare discrimination experiences and present preferences in a nationally representative sample.

BACKGROUND: Depression treatment disparities are well documented. Differing treatment preferences across social groups have been suggested as a cause of these disparities. However, existing studies of treatment preferences have been limited to individuals currently receiving clinical care, and existing measures of depression treatment preferences have not accounted for factors that may be disproportionately relevant to the preferences of disparities populations. This study therefore aimed to assess depression treatment preferences by race/ethnicity and gender in a representative community sample, while accounting for access to healthcare, provider characteristics, and past experiences of discrimination in healthcare settings. METHODS: We conducted a nationally representative study of individuals with depression in and out of clinical care. Treatment preferences (medication versus talk therapy) were elicited through a discrete choice experiment that accounted for tradeoffs with factors related to access and provider characteristics deemed relevant by community stakeholders. Past discrimination was assessed through questions about unfair treatment from medical providers and front desk staff due to personal characteristics (e.g., race, gender). We used conditional logit models to assess treatment preferences by race/ethnicity and gender and examined whether preferences were associated with past experiences of healthcare discrimination. RESULTS: Non-Hispanic white respondents (OR-here, the odds of a talk therapy preference over the odds of a medication preference: 0.80, 95% CI: 0.64, 0.99) and men (OR 0.76, 95% CI: 0.60, 0.96) preferred medication over talk therapy, while non-Hispanic black respondents, Hispanic respondents, and women did not prefer one over the other. Past discrimination in healthcare settings was associated with lower preferences for talk therapy and greater preferences for medication, particularly among non-Hispanic black respondents and women respondents. CONCLUSIONS: Addressing previous methodological limitations yielded estimates for depression treatment preferences by race/ethnicity and gender that differed from past studies. Also, past discrimination in healthcare settings was associated with current treatment preferences.

The Impact Of Medicare's Mental Health Cost-Sharing Parity On Use Of Mental Health Care Services.

Before the implementation of cost-sharing parity in Medicare, beneficiaries faced higher cost sharing for mental health services than for other medical services. The Medicare Improvements for Patients and Providers Act of 2008 phased in cost-sharing reductions in Medicare for outpatient mental health services in the period 2010-14. Using data for 2006-15 from the Medical Expenditure Panel Survey and difference-in-differences analyses, we assessed whether this reduction in mental health cost sharing was associated with changes in specialty and primary care outpatient mental health visits and psychotropic medication fills. We compared people with Medicare and with private insurance before and after parity implementation. Medicare beneficiaries' use of psychotropic medication increased after the implementation of cost-sharing parity, but we did not detect a change in visits. Changes in the use of psychotropic medications were greater among people with probable serious mental illness and among Medicare beneficiaries who did not report having supplemental coverage. The increased medication use could signal improvements in mental health care access among Medicare beneficiaries, especially among the subgroups most likely to benefit from the policy change.

Assessing provider and racial/ethnic variation in response to the FDA antidepressant box warning.

INTRODUCTION: After the 2004 FDA box warning raised concerns about increased suicidal ideation among youth taking antidepressants, antidepressant use decreased for White youth but slightly increased for Black and Latino youth. Better understanding of patient and provider factors contributing to these differences is needed to improve future risk warning dissemination. METHODS: We analyzed antidepressant prescriptions for youth aged 5-17 in 2002-2006 Medicaid claims data from four states (CA, FL, NC, and NY). In multilevel models, we assessed provider- and patient-level contributions to changes in antidepressant use by race/ethnicity and compared responses to the box warning between providers with large (>2/3) and small (<1/3) proportions of minority patients. RESULTS: A significant amount of variance in overall prescribing patterns (calculated by the ICC) was explained at the provider level. Significant provider-level variation was also identified in the differential effect of the box warning by racial/ethnic group. In a test of the influence of provider panel mix, we found that providers with large proportions of minority patients reduced antidepressant prescribing more slowly after the box warning than other providers. DISCUSSION: This study is the first to assess provider- and patient-level variation in the impact of a health care policy change on treatment disparities. Black and Latino youth Medicaid beneficiaries were seen by largely different providers than their White counterparts, and these distinct providers were influential in driving antidepressant prescription patterns following the box warning. Concerted outreach to providers of minority beneficiaries is needed to ensure that risk warnings and clinical innovations diffuse swiftly across racial/ethnic minority groups.