Publications

Rural Latinx immigrants experienced disproportionately negative health and economic impacts during the COVID-19 pandemic. They contended with the pandemic at the intersection of legal status exclusions from the safety net and long-standing barriers to health care in rural regions. Yet, little is known about how rural Latinx immigrants navigated such exclusions. In this qualitative study, we examined how legal status stratification in rural contexts influenced Latinx immigrant families' access to the safety net. We conducted interviews with first- and second-generation Latinx immigrants (n = 39) and service providers (n = 20) in four rural California communities between July 2020 and April 2021. We examined personal and organizational strategies used to obtain economic, health, and other forms of support. We found that Latinx families navigated a limited safety net with significant exclusions. In response, they enacted short-term strategies and practices - workarounds - that met immediate, short-term needs. Workarounds, however, were enacted through individual efforts, allowing little recourse beyond immediate personal agency. Some took the form of strategic practices within the safety net, such as leveraging resources that did not require legal status verification; in other cases, they took the form of families opting to avoid the safety net altogether.

The suicide rate among adolescents and young adults in the United States increased 57% between 2007 and 2018, from 6.8 to 10.7 deaths per 100 000 individuals. Recent research characterized as alarming the increases in overall suicide rates among young Black and other racial/ethnic minority populations. To assess the temporal trends in overall suicide and firearm suicide mortality rates among non-Hispanic Black young adults, we conducted a sex-specific Joinpoint regression analysis to identify changing trends in these rates between 1999 and 2019. Data were obtained from the Centers for Disease Control and Prevention's Web-based Injury Statistics Query and Reporting System. Results showed an 84.5% increase in the firearm suicide rate among young Black men and a 76.9% increase among young Black women between 2013 and 2019. Additional research is needed to investigate potential population-level exposures during or before 2013 that may have influenced suicide and firearm suicide risk.

CONTEXT: Segregation has been linked to unequal life chances.  Individuals from marginalized communities experience more crime, higher levels of poverty, poor health, and less civic engagement; and segregated metropolitan regions witness inequality in access to basic services. This paper builds on this previous work by linking segregation to infection and deaths from COVID-19.

METHODS: Using Census data matched to COVID infection and death statistics at the county level, we offer a theoretical basis for our choice of segregation measures and predictions for different racial groups. We analyze the relationship between two dimensions of segregation, racial isolation and racial unevenness, and COVID outcomes for different racial and ethnic groups.

FINDINGS: We find that in counties where Black and Latino residents live in more isolated neighborhoods, they were much more likely to contract COVID-19. This pattern was exacerbated in counties with a high proportion of front-line workers. We also find that racial segregation increased COVID-19 death rates for Black, Latino, and white residents.

CONCLUSIONS: Our findings suggest that devastating outcomes of the coronavirus pandemic were linked to a long history of racial marginalization and entrenched discrimination produced by structural inequalities embedded in our geographies and should inform public health planning going forward.

OBJECTIVE: Patient presentations to the emergency department (ED) for alcohol-involved injury represent a growing public health burden, but their characteristics and sequelae remain understudied. This study examined mortality rates among ED patients presenting with alcohol-involved injuries and assessed how mortality varied by injury intent and other characteristics.

METHOD: This retrospective cohort study used statewide, longitudinally linked ED patient record and mortality data from California. Participants comprised all residents presenting to a licensed ED in 2009-2012 with a nonfatal injury that involved comorbid diagnosis of alcohol use disorder (AUD; n = 261,222; 59.3% male). Injury intent was defined using International Classification of Diseases, 9th Revision, Clinical Modification external cause-of-injury codes. Cox regression was used to investigate factors associated with 12-month all-cause mortality rates. Age-, sex-, and race/ethnicity-adjusted standardized mortality ratios (SMRs) were calculated using statewide mortality data.

RESULTS: Most ED injury visits involving an AUD diagnosis were coded as unintentional (75.9%). Following the index ED visit, all-cause mortality among AUD-involved injury patients was 5,205 per 100,000 person-years, five times higher than the demographically matched population (SMR = 5.3; 95% confidence interval [5.2, 5.4]). Adjusted Cox regression models indicated that patients whose index injury was unintentional, and whose AUD was for acute intoxication, had significantly higher mortality. Most deaths among unintentionally injured patients were from natural causes, whereas external-cause deaths were relatively more common in the other patient groups.

CONCLUSIONS: AUD-involved injury presentations to the ED in California are common and associated with high patient mortality burden, which varies by injury intent. Interventions are needed to reduce excess mortality in these patients.

PURPOSE: This study examined the epidemiology of self-harm emergency department (ED) visits among Asian American and Pacific Islander (AAPI) youth, and associated factors.

METHODS: We used California ED visit records in 2010 and 2011 to calculate incidence rates of self-harm ED visits for AAPI versus non-Hispanic White (NHW) patients aged 10-29 years. Demographic and clinical characteristics were compared for AAPI versus NHW patients presenting with self-harm. We used modified Poisson regression models to estimate the relative risk of recurrent ED self-harm visits for AAPI versus NHW patients and examined the association of insurance type and gender with recurrent self-harm among AAPIs.

RESULTS: Rates of self-harm ED visits for young AAPI patients were 38 and 26 per 100,000 among females and males, respectively. Although AAPI patients presenting with self-harm were equally or less likely than NHW patients to have comorbid psychological and substance use diagnoses at their index visit, they were 25% more likely to be admitted to hospital. However, they were 40% less likely to have a recurrent ED self-harm visit. Among AAPI patients, those who used Medicaid were significantly more likely than those with other insurance to be admitted as inpatients.

DISCUSSION: Young AAPI patients presenting to EDs with deliberate self-harm have different sociodemographic and clinical profiles compared to NHW patients. Our study also demonstrates significant heterogeneity in risk of recurrent self-harm by gender and insurance type among AAPI patients. This information may be useful for future intervention programs among self-harming AAPI youth.

U.S. food insecurity rates rapidly increased during the COVID-19 pandemic, with disproportionate impacts on Latino immigrant households. We conducted a qualitative study to investigate how household food environments of rural Latino immigrants were affected during the COVID-19 pandemic. Thirty-one respondents (42% from low food security households) completed interviews (July 2020-April 2021) across four rural counties in California. A rural household food security conceptual framework was used to analyze the data. Early in the pandemic, food availability was impacted by school closures and the increased consumption of meals/snacks at home; food access was impacted by reduced incomes. Barriers to access included limited transportation, excess distance, and lack of convenience. Key resources for mitigating food insecurity were the Supplemental Nutrition Assistance Program (SNAP), the Pandemic Electronic Benefits Transfer (P-EBT), school meals, charitable food programs, and social capital, although the adequacy and acceptability of charitable food distributions were noted issues. Respondents expressed concern about legal status, stigma, and the public charge rule when discussing barriers to government nutrition assistance programs. They reported that food pantries and P-EBT had fewer access barriers. Positive coping strategies included health-promoting food substitutions and the reduced consumption of meals outside the home. Results can inform the development of policy and systems interventions to decrease food insecurity and nutrition-related health disparities among rural Latino immigrants.

Introduction: Suicide rates are extremely high among emergency department patients seen for deliberate self-harm. Inpatient hospitalization is often recommended for these patients, but evi- dence on the suicide prevention impacts of hospitalization is scarce. Confounding by indication and challenges to implementing randomized designs are barriers to advances in this field. Methods: Investigators used 2009−2012 statewide data on 57,312 self-harm emergency department patients from California, linked to mortality records. Naive 12-month and 30-day suicide risks were estimated among patients who were hospitalized versus those who were discharged. Then, generalized random forest methods were applied to estimate the average treatment impacts of hospitalization on suicide, conditioning on observable covariates. Associations were calculated separately for sex- and age-specific subgroups. Analyses were conducted in February 2019−August 2021. Results: In naive analyses, suicide risk was significantly higher in hospitalized than in discharged patients in each subgroup. In 12-month models accounting for the observed covariates through generalized random forest methods, hospitalized male patients had 5.4 more suicides per 1,000 patients (95% CI=3.0, 7.8), hospitalized patients aged 10−29 years had 2.4 more suicides per 1,000 (95% CI=1.1, 3.6), and those aged ≥50 years had 5.8 more suicides per 1,000 (95% CI=0.5, 11.2) than corresponding discharged patients. Hospitalization was not significantly associated with sui- cide among female patients or patients aged 30−49 years in generalized random forest analyses. Patterns were similar in 30-day generalized random forest models. Conclusions: Emergency department personnel intend to hospitalize self-harm patients with high suicide risk; this study suggests that this goal is largely realized. Analyses that control for confound- ing by observable covariates did not find clear evidence that hospitalization reduces suicide risk and could not rule out the possibility of iatrogenic effects.

OBJECTIVE: To estimate the prevalence of pregnancy- associated deaths due to drugs, suicide, and homicide nationwide from 2010 to 2019. METHODS: Using U.S. death certificate records from 2010 to 2019 for 33 states plus the District of Columbia, we identified pregnancy-associated deaths using the pregnancy checkbox and International Classification of Diseases, Tenth Revision codes, calculated pregnancy- associated death ratios, and categorized deaths by cause, timing relative to pregnancy, race or ethnicity, and age. RESULTS: Of 11,782 pregnancy-associated deaths iden- tified between 2010 and 2019, 11.4% were due to drugs, 5.4% were due to suicide, and 5.4% were due to homicide, whereas 59.3% were due to obstetric causes and the remaining 18.5% were due to other causes. Drug-related deaths, suicide, and homicide accounted for 22.2% of pregnancy-associated deaths. All three causes of death increased over the study period, with drug-related pregnancy-associated deaths increasing 190%. Homicide during pregnancy and drug-related deaths, suicides, and homicide in the late postpartum period (43–365 days) accounted for a larger proportion of all deaths in these time periods than the contribution of these causes to all deaths among females of reproduc- tive age. Pregnant and postpartum people identified as non-Hispanic American Indian or Alaska Native were at highest risk of drug-related and suicide death, and peo- ple identified as non-Hispanic Black were at highest risk of homicide. CONCLUSION: Deaths due to drug use, suicide, and homicide constitute more than one fifth of all deaths during pregnancy and the first year postpartum. Drug- related deaths and homicides have increased over the past decade. Substantial racial and ethnic inequities in these deaths exist.

Background. Alcohol-related morbidity and mortality have increased substantially in the U.S. Understanding the population health implications of these concerning trends, including by identifying clinical subgroups of alcohol users at increased risk for potentially preventable acute causes of mortality, is of critical importance. Methods. This retrospective cohort study used statewide, all-payer, longitudinally-linked ED patient record and mortality data from California. Participants comprised all residents presenting to a licensed ED at least once in 2009-2011 with a diagnosis of alcohol use disorder (AUD). Participants were followed for one year after index ED visit to assess acute injury (unintentional poisoning, suicide, homicide, motor vehicle crash, and fall- or fire-related injury) and all-cause mortality rates per 100,000 person-years. Age-, sex-, race/ethnicity-adjusted standardized mortality rates (SMRs) for acute injury causes of death were determined using statewide mortality data. Results. Among 437,855 patients with index non-fatal ED visits for AUD, the 12-month acute injury mortality rate was 608.6 per 100,000 (SMR=8.0; 95% CI=7.7, 8.3), and all-cause mortality was 5,700.7 per 100,000 (SMR=6.5; 95% CI=6.4, 6.6). Unintentional poisoning accounted for 46.5%, and suicide for 19.7%, of acute-injury deaths. Acute injury deaths comprised 71.7% of all-cause mortality among patients aged 10-24 years, but much lower proportions among older patients. Female AUD patients had lower rates for all mortality outcomes. Conclusions. Emergency department patients with a recognized AUD comprise a population at persistently elevated risk for mortality. Age-related AUD patient differences in common causes of death, including drug overdose and suicide, can inform the structure of future clinical interventions.

Objective Improve methodology for equitable suicide death prediction when using sensitive predictors, such as race/ethnicity, for machine learning and statistical methods. Methods Train predictive models, logistic regression, naive Bayes, gradient boosting (XGBoost) and random forests, using three resampling techniques (Blind, Separate, Equity) on emergency department (ED) administrative patient records. The Blind method resamples without considering racial/ethnic group. Comparatively, the Separate method trains disjoint models for each group and the Equity method builds a training set that is balanced both by racial/ethnic group and by class. Results Using the Blind method, performance range of the models sensitivity for predicting suicide death between racial/ethnic groups (a measure of prediction inequity) was 0.47 for logistic regression, 0.37 for naive Bayes, 0.56 for XGBoost and 0.58 for random forest. By building separate models for different racial/ethnic groups or using the equity method on the training set, we decreased the range in performance to 0.16, 0.13, 0.19, 0.20 with Separate method, and 0.14, 0.12, 0.24, 0.13 for Equity method, respectively. XGBoost had the highest overall area under the curve (AUC), ranging from 0.69 to 0.79. Discussion We increased performance equity between different racial/ethnic groups and show that imbalanced training sets lead to models with poor predictive equity. These methods have comparable AUC scores to other work in the field, using only single ED administrative record data. Conclusion. We propose two methods to improve equity of suicide death prediction among different racial/ethnic groups. These methods may be applied to other sensitive characteristics to improve equity in machine learning with healthcare applications.