Annals of Epidemiology

IntroductionSocial determinants of health (SDoH), defined as nonmedical factors that impact health outcomes, have been associated with fatal crash occurrences. Road users who live in communities with negative SDoH may be at increased risk of crash-related mortality, and the risks may be further heightened among geriatric road users and in rural areas. We evaluated the relationship between the county-level measure of SDoH and county-level fatal crash counts among geriatric and non-geriatric road users living in rural, suburban, and urban areas. MethodsFor this ecological study, we pooled data from Fatality Analysis Reporting System (2018 to 2020) and the U.S. Census Bureau (2019 data) and limited our analyses to the 3,108 contiguous US counties. The outcome measures were county-level fatal crash counts involving (1) geriatric (65 years and older) road users (2) non-geriatric road users, and (2) the general population. The predictor variable was the Multidimensional Deprivation Index (MDI), a score that measures the five domains of SDoH - economic quality, healthcare access, education, community, and neighborhood quality. We defined the MDI as a three-level categorical variable: at or below the national average, within two-fold of the national average, and higher than two-fold of the national average. We controlled for county-level demographics and crash characteristics. We performed a Bayesian spatial Poisson regression analysis using Integrated Nested Laplace Approximations and reported the crash fatality rate ratios (plus 95% Credible Intervals (CrI)). ResultsThe median (Q1, Q3) standardized mortality rate ratios among geriatric and non-geriatric road users were 1.3 (0.6, 2.5) and 1.6 (0.9, 2.7), respectively. A total of 283 (9.1%) and 806 (15.9%) counties were classified as very highly deprived and highly deprived, respectively. Clusters of counties with high deprivation rates were identified in the Southern states. Counties classified as very highly deprived and highly deprived had 40% (95% CrI: 1.24 - 1.57) and 25% (95% CrI: 1.17 - 1.34) increased geriatric fatality crash rate ratios and this pattern of association persisted in suburban and urban areas. Also, counties classified as very highly deprived and highly deprived had 42% (95% CI: 1.27 - 1.58) and 32% (95% CI: 1.23 - 1.38) increased fatality crash rate ratios among all road users and this pattern persisted in suburban and urban areas. Counties with more than four-fold increased fatality rate ratios were located commonly in Texas, Oklahoma, Nevada, and Utah. ConclusionDespite older adults being less frequent road users, county-level deprivation measures of the SDoH are equally associated with geriatric and non-geriatric crash-related fatal rate ratios. Policies that improve county-level SDoH may reduce the county-level fatal rate ratios equally among geriatric and non-geriatric road users.

1.ImportanceTracking the direct and indirect impact of the coronavirus disease 2019 (COVID-19) pandemic on all-cause mortality in the United States has been hindered by the lack of testing and by reporting delays. Evaluating excess mortality, or the number of deaths above what is expected in a given time period, provides critical insights into the true burden of the COVID-19 pandemic caused by the novel Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2). Stratifying mortality data by demographics such as age, sex, race, ethnicity, and geography helps quantify how subgroups of the population have been differentially affected. Similarly, stratifying mortality data by cause of death reveals the public health effects of the pandemic in terms of other acute and chronic diseases. ObjectiveTo provide stratified estimates of excess mortality in Colorado from March to September 2020. Design, Setting, and PopulationThis study evaluated the number of excess deaths both directly due to SARS-CoV-2 infection and from all other causes between March and September 2020 at the county level in Colorado. Data were obtained from the Vital Statistics Program at the Colorado Department of Public Health and Environment. These estimates of excess mortality were derived by comparing population-adjusted mortality rates in 2020 with rates in the same months from 2015 to 2019. ResultsWe found evidence of excess mortality in Colorado between March and September 2020. Two peaks in excess deaths from all causes were recorded in the state, one mid-April and the other at the end of June. Since the first documented SARS-CoV-2 infection on March 5th, we estimated that the excess mortality rate in Colorado was two times higher than the officially reported COVID-19 mortality rate. State-level cumulative excess mortality from all causes reached 71 excess deaths per 100k residents ([~]4000 excess deaths in the state); in contrast, 35 deaths per 100k directly due to SARS-CoV-2 were recorded in the same period ([~]1980 deaths. Excess mortality occurred in 52 of 64 counties, accounting for 99% of the states population. Most excess deaths recorded from March to September 2020 were associated with acute events (estimated at 44 excess deaths per 100k residents and at 9 after excluding deaths directly due to SARS-CoV-2) rather than with chronic conditions ([~]21 excess deaths per 100k). Among Coloradans aged 14-44, 1.4 times more deaths occurred in those months than during the same period in the five previous years. Hispanic White males died of COVID-19 at the highest rate during this time ([~]90 deaths from COVID-19 per 100k residents); however, Non-Hispanic Black/African American males were the most affected in terms of overall excess mortality ([~]204 excess deaths per 100k). Beyond inequalities in COVID-19 mortality per se, these findings signal considerable regional and racial-ethnic disparities in excess all-cause mortality that need to be addressed for a just recovery and in future public health crises.

IntroductionThe Affordable Care Act (ACA), whose major provisions were Medicaid expansion, state insurance exchanges, and allowing dependents to remain on their parents insurance until age 26, was implemented fully in 2014. Increased access to care might have improved access to lung/bronchus cancer diagnosis, treatment, and outcomes. We hypothesized differential changes in lung/bronchus mortality among Mississippi (which did not expand Medicaid) and New York (that did expand Medicaid). MethodsWe utilized the CDC Wonder database to compare lung/bronchus cancer mortality rates in Mississippi, New York, and the United States as a whole, comparing such rates 5 years prior (2009-2013) to 5 years after (2015-2019) ACA enactment. Statistical significance was deemed at p <0.05. ResultsIn the 5 years prior to ACA implementation (2009-2013), there was no statistically-significant difference between Mississippi, New York, or the total U.S. in the percent of population dying from lung/bronchus cancer (p >0.7). In the 5 years following ACA implementation (2015-2019), there was a statistically-significant decrease between Mississippi, New York, and the total U.S. in the percent of population dying from lung/bronchus cancer (p <0.002). Comparing the years 2009-2013 (5 years prior to ACA) and the years 2015-2019 (5 years after ACA), there was a statistically-significant difference in the decrease in percent of population dying from lung and bronchus cancers in New York, Mississippi, and the total United States (p <0.002), with New York having the greatest percent decline. In New York, the percentage of mortalities from lung and bronchus cancers decreased from 0.004691% to 0.004088% (p <0.0001). In Mississippi, the percentage of mortalities from lung and bronchus cancers decreased from 0.006537% to 0.006282% (p = 0.0061). In the total US, the percentage of mortalities from lung and bronchus cancers decreased from 0.05097% to 0.0473% (p <0.0001). ConclusionState-level participation in the ACAs Medicaid expansion was associated with disproportionate improvement in lung/bronchus mortality compared with non-participation and with the U.S. in total.

BackgroundThe SARS-CoV-2 pandemic has disproportionately affected racial and ethnic minority communities across the United States. We sought to disentangle individual and census tract-level sociodemographic and economic factors associated with these disparities. Methods and FindingsAll adults tested for SARS-CoV-2 between February 1 and June 21, 2020 were geocoded to a census tract based on their address; hospital employees and individuals with invalid addresses were excluded. Individual (age, sex, race/ethnicity, preferred language, insurance) and census tract-level (demographics, insurance, income, education, employment, occupation, household crowding and occupancy, built home environment, and transportation) variables were analyzed using linear mixed models predicting infection, hospitalization, and death from SARS-CoV-2. Among 57,865 individuals, per capita testing rates, individual (older age, male sex, non-White race, non-English preferred language, and non-private insurance), and census tract-level (increased population density, higher household occupancy, and lower education) measures were associated with likelihood of infection. Among those infected, individual age, sex, race, language, and insurance, and census tract-level measures of lower education, more multi-family homes, and extreme household crowding were associated with increased likelihood of hospitalization, while higher per capita testing rates were associated with decreased likelihood. Only individual-level variables (older age, male sex, Medicare insurance) were associated with increased mortality among those hospitalized. ConclusionsThis study of the first wave of the SARS-CoV-2 pandemic in a major U.S. city presents the cascade of outcomes following SARS-CoV-2 infection within a large, multi-ethnic cohort. SARS-CoV-2 infection and hospitalization rates, but not death rates among those hospitalized, are related to census tract-level socioeconomic characteristics including lower educational attainment and higher household crowding and occupancy, but not neighborhood measures of race, independent of individual factors.

AO_SCPLOWBSTRACTC_SCPLOWO_ST_ABSBackgroundC_ST_ABSPrior research has identified higher rates of COVID-19 mortality among people of color (relative to non-Hispanic whites) and populations in high-poverty neighborhoods (relative to wealthier neighborhoods). It is unclear, however, whether non-Hispanic whites in high-poverty neighborhoods experience elevated mortality, or whether people of color living in wealthy areas are relatively protected. Exploring socioeconomic position in combination with race/ethnicity can lead to a more detailed understanding of the specific processes that result in COVID-19 inequities. Methods and FindingsWe used census and individual-level mortality data for the non-Hispanic white, non-Hispanic Black, and Hispanic/Latinx populations of Cook County, Illinois, USA. We excluded deaths related to nursing homes and other institutions. We calculated age and gender-adjusted mortality rates by race/ethnicity, census tract poverty quartile, and age group (0-64 and [&ge;]65 years). Within all racial/ethnic groups, COVID-19 mortality rates were greatest in the highest-poverty quartile and lowest in the lowest-poverty quartile. The mortality rate for younger non-Hispanic whites in the highest-poverty quartile was 13.5 times that of younger non-Hispanic whites in the lowest-poverty quartile (95% CI: 8.5, 21.4). For young people in the highest-poverty quartile, the non-Hispanic white and Black mortality rates were similar. Among younger people in the lowest-poverty quartile, non-Hispanic Black and Hispanic/Latinx people had mortality rates nearly three times that of non-Hispanic whites. For the older population, the mortality rate among non-Hispanic whites in the highest-poverty quartile was less than that of lowest-poverty non-Hispanic Black and Hispanic/Latinx populations. ConclusionsOur findings suggest racial/ethnic inequalities in COVID-19 mortality are partly, but not entirely, attributable to the higher average socioeconomic position of non-Hispanic whites relative to the non-Hispanic Black and Hispanic/Latinx populations. Future research on health equity in COVID-19 outcomes should collect and analyze individual-level data on the potential mechanisms driving population distributions of exposure, severe illness, and death.

Infant mortality (IM), or death prior to the first birthday, is a key public health metric that increases with neighborhood structural inequities. However, neighborhood exposures shift as communities undergo gentrification, a pattern of neighborhood change defined by increasing affluence (in wealth, education, and housing costs). Gentrification has inconsistent associations with infant health outcomes like IM, which may be due to differing relationships between its composite measures and such outcomes. We designed a retrospective cohort analysis of all births and deaths from 2010-2019 across 4 metropolitan areas in Michigan to determine how gentrification and its neighborhood-change components are associated with risk of IM, using multilevel multivariable logistic regression models. Among 672,432 infants, 0.52% died before 1 year. IM was not associated with gentrification. Census tracts with greater increases in income and education had lower rates of IM, but tracts with greater increases in rent costs had higher rates of IM. In unadjusted models, odds of IM were 40% and 15% lower for infants living in tracts in the top quartile increase in household income and college completion, respectively, compared to infants from tracts with the least amount of change. Odds of IM were also increased 29% in infants from tracts with the most increases in rent, though these differences were attenuated when adjusting for individual social factors. Indicators of increasing community affluence have opposing relationships with IM. Policies and interventions that address rising housing costs may reduce IM.

We used 2017 Behavioral Risk Factor Surveillance System (BRFSS) data (N=444,649) to estimate the proportion of US adults who report comorbidities that suggest heightened risk of complications from COVID-19. Co-morbidities included cardiovascular disease, chronic obstructive pulmonary disease (COPD), diabetes, asthma, hypertension, and/or cancer other than skin, based on data from China. Overall 45.4% (95% CI 45.1-45.7) of adults reported any of the 6 comorbidities, increasing from 19.8% (19.1-20.4) for ages 18-29 years to 80.7% (79.5-81.8) for ages 80+ years. State rates ranged from 37.3% (36.2-38.5) in Utah to 58.7% (57.0-60.4) in West Virginia. Rates also varied by race/ethnicity, health insurance status, and employment. Excluded were residents of nursing homes or assisted living facilities. Although almost certainly an underestimate of all adults at risk due to these exclusions, these results should help in estimating healthcare needs for adults with COVID-19 complications living in the community. Article Summary LineOverall, 45.4% of US adults were estimated to be at heightened risk of COVID-19 complications due to co-morbidities, increasing from 19.8% for ages 18-29 years to 80.7% for ages 80+ years, with state-to-state variation.

INTRODUCTIONThe Affordable Care Acts Medicaid expansion aimed to enhance healthcare access for low-income individuals and minority groups, promoting early screening and treatment to improve health equity. OBJECTIVEThis study examines the impact of Medicaid expansion on lung cancer-specific survival (CSM) and overall mortality (OS) by comparing outcomes in Texas (non-expansion of ACA) and California (expansion of ACA). METHODOLOGYWe conducted a retrospective study using data from SEER cancer registry (2007-2021) to evaluate the impact of Medicaid expansion on lung cancer survival in California (expansion) vs. Texas (non-expansion). The study included adults aged 18-64, with periods split into pre-ACA (2007-2013), one-year washout (2014), and post-ACA (2015-2021). We utilized a DID design and adjusted for important covariates. RESULTSAmong 119,937 individuals with Lung cancer, 52.1% were in California (62,521), while 47.8% were in Texas (57,416). The pre-ACA period included 60,010 individuals (53.1% in California and 46.9% in Texas), and 59,927 patients were in the post-ACA period (51.2% in California and 48.8% in Texas). Overall, Medicaid expansion was associated with a 1.12-point (- 1.12, 95% CI -1.46 to -0.77) reduction in the hazard of cancer-specific mortality. The policy was also associated with a 0.81point reduction in the hazard of overall mortality (-0.81, 95% CI -1.06 to -0.57). CONCLUSIONMedicaid expansion was associated with a significant improvement in lung cancer outcomes among individuals with lung cancer in California, which implemented the policy in 2014, compared to Texas, which has not yet implemented the policy.

PurposeLow-dose computed tomography lung cancer screening is effective for reducing lung cancer mortality. It is critical to understand the lung cancer screening practices for screen-eligible individuals living in Alabama and Georgia where lung cancer is the leading cause of cancer death. High lung cancer incidence and mortality rates are attributed to high smoking rates among underserved, low income, and rural populations. Therefore, the purpose of this study: (1) to define sociodemographic and clinical characteristics of patients who were screened for lung cancer at an Academic Medical Center (AMC) in Alabama and a Safety Net Hospital (SNH) in Georgia. MethodsA retrospective cohort study of patient electronic health records who received lung cancer screening between 2015 to 2020 was performed to identify the study population and outcome variable measures. Chi-square tests and Student t-tests were used to compare screening uptake across patient demographic and clinical variables. Bivariate and multivariate logistic regressions determined significant predictors of lung cancer screening uptake. ResultsAt the AMC, 67,355 were identified as eligible for LCS and 1,129 were screened. In bivariate analyses, there were several differences between those who were screened and those who were not screened. Screening status in the site at Alabama varied significantly by age (P<0.01), race (P<0.001), marital status (P<0.01), smoking status (P<0.01) health insurance (P<0.01), median income (P<0.01), urban status (P<0.01) and distance from UAB (P<0.01). Those who were screened were more likely to have lesser comorbidities (2.31 vs. 2.53; P<0.001). At the SNH, 11,011 individuals were identified as screen-eligible and 500 were screened. In the site at Georgia, screening status varied significantly by race (P<0.01), health insurance (P<0.01), and distance from site (P<0.01). At the AMC, the odds of being screened increased significantly if the individual was a current smoker compared to former smoker (OR=3.21; P<0.01). At the SNH, the odds of being screened for lung cancer increased significantly with every unit increase in co-morbidity count (OR = 1.12; P=0.01) ConclusionThe study provides evidence that LCS has not reached all subgroups and that additional targeted efforts are needed to increase lung cancer screening uptake. Furthermore disparity was noticed between adults living closer to screening institutions and those who lived farther.

BackgroundHistorical redlining policies enacted in the 1930s and 1940s that restricted investment in Black neighborhoods shaped neighborhood conditions that may contribute to inequities in health and mortality among older adults today. Areas "redlined" by the Home Owners Loan Corporation (HOLC) in the 1930s against Black neighborhoods are associated with worse present-day area-level health outcomes. We examined whether early, personal exposure to redlining close to when the maps were drawn is associated with individual-level mortality hazard (survival time ratio) and self-rated health in older adults. MethodsWe used mapped 1940 census enumeration districts to assign 1930s HOLC redlining categories (green A ("best"), blue B ("still desirable"), yellow C ("definitely declining"), and red D ("hazardous")) to Health and Retirement Study participants based on 1940 census residence. We applied survey weights and ran a survival analysis with a parametric normal distribution maximum likelihood estimation to account for survivorship bias, and logistic regression on self-rated health, and included analyses stratified by race. Results1940 HOLC-categorized yellow C (0.62 times the survival time, 95% CI: 0.41, 0.92) and red D (red: 0.59, 95% CI: 0.40, 0.87) exposures were significantly associated with reduced survival time compared to green A in both unadjusted and adjusted models. In stratified analyses, both Black and white residents of redlined areas had worse survival time ratios compared to green A, though the magnitude of effect was larger for Black residents than for white residents. Yellow C (Odds Ratio: 1.94, 95% CI: 1.16, 3.23) and red D (2.34, 95% CI: 1.37, 3.98) areas were also associated with increased odds of worse self-rated health compared to green A areas. DiscussionLiving in redlined areas in the 1940s is associated with worse mortality survival for both Black and white older adults and with decreased self-rated health in older adults between 1992 and 2018. These findings extend beyond broader prior research demonstrating present-day area-level associations of redlining with worse health and are consistent with prior research on individual-level exposure to redlining. Associations with worse mortality in both Black and white residents (with stronger effects in Black residents) are consistent with theory and research demonstrating that structural racism degrades health for all communities.

ImportancePersistent racial and ethnic disparities in breast and prostate cancer mortality are well documented. Most prior studies emphasize between-group differences and rely on population averages or single composite measures of social disadvantage, which can obscure high-need communities within groups. How socio-behavioral determinants of health vary within groups across local gradients of cancer mortality remains incompletely characterized. A framework that combines race- and cancer-specific mortality with local, domain-level socio-behavioral profiles may help identify where burden is greatest and which specific barriers warrant prioritization. ObjectiveTo determine how socio-behavioral risk relates to breast and prostate cancer mortality within racial and ethnic groups and to characterize domain-specific behavioral profiles across low-, moderate- and high-mortality counties to inform targeted, equity-oriented cancer control strategies. DesignCross-sectional study of U.S. counties. Setting United States, county-level analysis. Participants3,141 U.S. counties, stratified within Non-Hispanic White, Non-Hispanic Black, and Hispanic populations. ExposuresCounty-level socio-behavioral determinants of health measured using a composite index comprising seven domains: community solidarity; education, health literacy, and digital connectivity; quality of care; housing and environmental risk; economic livelihoods; lifestyle behaviors; and touchpoints with care. Main outcomes and measuresRace/ethnicity-specific, age-adjusted breast and prostate cancer mortality rates (2018-2022) and county-level socio-behavioral risk scores. Counties were grouped into mortality tertiles within each race/ethnicity-by-cancer-stratum. ResultsAcross groups, higher socio-behavioral risk was associated with higher breast and prostate cancer mortality. For breast cancer, socio-behavioral risk increased monotonically across mortality tertiles for all groups, with the largest within-group increases among Hispanic and Non-Hispanic Black women. For prostate cancer, risk generally increased across mortality tertiles for all groups. Although Hispanic populations had lower population-average mortality, high-mortality Hispanic counties exhibited pronounced risk in lifestyle behaviors, economic livelihoods, and touchpoints with care. Domain patterns associated with high mortality varied by race, ethnicity, and cancer type, with touchpoints with care and economic livelihoods consistently prominent. Conclusions and relevanceWithin-group heterogeneity in socio-behavioral risk is substantial across U.S. counties. Linking population-specific, domain-level socio-behavioral profiles to cancer mortality may support more precise and equity-oriented cancer control strategies than reliance on group averages or composite indices. Key pointsO_ST_ABSQuestionC_ST_ABSWithin racial and ethnic groups, how do socio-behavioral determinants of health vary across US counties with low, moderate, and high breast and prostate cancer mortality? FindingsIn this cross-sectional study, higher county-level socio-behavioral risk was associated with higher breast and prostate cancer mortality across racial and ethnic groups. Race/ethnicity-specific, domain-level profiles revealed within-group heterogeneity, including persistently elevated risk among Non-Hispanic Black populations and pronounced domain-specific gaps in high-mortality Hispanic counties. MeaningLinking population-specific socio-behavioral profiles to local cancer mortality can guide more precise and equity-oriented prioritization of intervention domains and geographies than reliance on group averages or composite indices.

ImportanceCOVID-19 racial disparities have gained significant attention yet little is known about how age distributions obscure racial-ethnic disparities in COVID-19 case fatality ratios (CFR). ObjectiveWe filled this gap by assessing relevant data availability and quality across states, and in states with available data, investigating how racial-ethnic disparities in CFR changed after age adjustment. Design/Setting/Participants/ExposureWe conducted a landscape analysis as of July 1st, 2020 and developed a grading system to assess COVID-19 case and death data by age and race in 50 states and DC. In states where age- and race-specific data were available, we applied direct age standardization to compare CFR across race-ethnicities. We developed an online dashboard to automatically and continuously update our results. Main Outcome and MeasureOur main outcome was CFR (deaths per 100 confirmed cases). We examined CFR by age and race-ethnicities. ResultsWe found substantial variations in disaggregating and reporting case and death data across states. Only three states, California, Illinois and Ohio, had sufficient age- and race-ethnicity-disaggregation to allow the investigation of racial-ethnic disparities in CFR while controlling for age. In total, we analyzed 391,991confirmed cases and 17,612 confirmed deaths. The crude CFRs varied from, e.g. 7.35% among Non-Hispanic (NH) White population to 1.39% among Hispanic population in Ohio. After age standardization, racial-ethnic differences in CFR narrowed, e.g. from 5.28% among NH White population to 3.79% among NH Asian population in Ohio, or an over one-fold difference. In addition, the ranking of race-ethnic-specific CFRs changed after age standardization. NH White population had the leading crude CFRs whereas NH Black and NH Asian population had the leading and second leading age-adjusted CFRs respectively in two of the three states. Hispanic populations age-adjusted CFR were substantially higher than the crude. Sensitivity analysis did not change these results qualitatively. Conclusions and RelevanceThe availability and quality of age- and race-ethnic-specific COVID-19 case and death data varied greatly across states. Age distributions in confirmed cases obscured racial-ethnic disparities in COVID-19 CFR. Age standardization narrows racial-ethnic disparities and changes ranking. Public COVID-19 data availability, quality, and harmonization need improvement to address racial disparities in this pandemic. Key PointsO_ST_ABSQuestionC_ST_ABSWhat are the racial-ethnic disparities in COVID-19 case fatality ratios (CFR) across states after adjusting for age? FindingsWe conducted direct standardization among 391,991 COVID-19 cases and 17,612 deaths from California, Illinois and Ohio to compare age-adjusted CFR across race-ethnicities. The racial-ethnic disparities in CFR narrowed and the ranking changed after age standardization. MeaningAge distributions in confirmed cases obscured racial-ethnic disparities in COVID-19 CFR.

BackgroundMultiracial adults represent a growing U.S. population but are often grouped together or reassigned to single-race categories in public health data. Aggregation can obscure important variation across subgroups, limiting opportunities for targeted prevention. MethodsWe analyzed 2014-2023 California Behavioral Risk Factor Surveillance System data (n=100,177) to estimate prevalence of 28 health indicators across racial and ethnic groups, including disaggregated Multiracial subgroups. We categorized participants based on all self-identified races and aggregated subgroups with N<50. We standardized prevalence by age and sex using 2020 California census data, calculated relative standard errors, and used survey-weighted methods to compare prevalence and subgroup differences. ResultsAmong 100,177 participants, Multiracial subgroups had the highest prevalence for 24 of 28 outcomes. American Indian or Alaska Native-Black and Hispanic-Black-White adults had the highest prevalence of chronic conditions, poor general health, and disability. In contrast, Asian Multiracial subgroups (e.g., Asian-Black, Asian-Pacific Islander) more often had the lowest prevalence, though Asian-White adults were not consistently the healthiest subgroup. Differences across Multiracial subgroups exceeded 20 percentage points for nearly half of all outcomes. DiscussionWide health variation among Multiracial adults is masked by common aggregation practices. Subgroups with the highest burden may be overlooked if data are not routinely disaggregated. Public health surveillance systems should expand capacity to collect and report disaggregated race and ethnicity data to better inform prevention strategies.

ObjectivePedestrian mortality in the United States has increased seven times faster than the population growth from 2019 to 2023, according to a Governors Highway Safety Association report. This alarming trend highlights the need to study pedestrian mortality patterns, stratified by gender, race/ethnicity, age group, and state-specific characteristics, along with an exploration of contributing factors driving this surge. MethodsThe Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research (CDC-WONDER) database was used to extract pedestrian mortality data from 1999 to 2020. Age-Adjusted Mortality Rates (AAMRs) per 100,000 population and Annual Percentage Changes (APCs) with 95% confidence intervals (CIs) were calculated. Joinpoint regression analysis was employed to assess the trends across various demographic and regional subgroups. ResultsA total of 140,280 pedestrian deaths occurred in the US between 1999 and 2020. The overall AAMR increased from 2.21 in 1999 to 2.32 in 2020. A steep rise in the APC (3.11) was observed from 2009 to 2020. Men consistently had higher AAMRs than women, while non-Hispanic (NH) American Indians or Alaska Natives had the highest AAMR among races. Individuals aged 35-44 years exhibited the highest APC (6.92) between 2011 and 2020. States in the 90th percentile (Arizona, Florida, New Mexico) had triple AAMRs compared to those in the 10th percentile. Rural areas had the highest APC (3.23) from 2011-2020. ConclusionPedestrian mortality rates in the United States have been rising for over a decade. Enhanced public safety interventions and efforts to address disparities based on race, age, gender, and geographic location are essential to curb the growing burden of pedestrian deaths. O_FIG O_LINKSMALLFIG WIDTH=125 HEIGHT=200 SRC="FIGDIR/small/25331672v1_ufig1.gif" ALT="Figure 1"> View larger version (36K): org.highwire.dtl.DTLVardef@1ce109aorg.highwire.dtl.DTLVardef@1a428d5org.highwire.dtl.DTLVardef@74bff8org.highwire.dtl.DTLVardef@e76547_HPS_FORMAT_FIGEXP M_FIG C_FIG

PURPOSEDisparities of lung cancer incidence exist in Black populations and screening criteria underserve Black populations due to disparately elevated risk in the screening eligible population. Prediction models that integrate clinical and imaging-based features to individualize lung cancer risk is a potential means to mitigate these disparities. PATIENTS AND METHODSThis Multicenter (NLST) and catchment population based (UIH, urban and suburban Cook County) cross-sectional study utilized participants at risk of lung cancer with available lung CT imaging and follow up between the years 2015 and 2024. 53,452 in NLST and 11,654 in UIH were included based on age and tobacco use based risk factors for lung cancer. Cohorts were used for training and testing of deep and machine learning models using clinical features alone or combined with CT image features (hybrid computer vision). RESULTSAn optimized 7 clinical feature model achieved ROC-AUC values ranging 0.64-0.67 in NLST and 0.60-0.65 in UIH cohorts across multiple years. Incorporation of imaging features to form a hybrid computer vision model significantly improved ROC-AUC values to 0.78-0.91 in NLST but deteriorated in UIH with ROC-AUC values of 0.68-0.80, attributable to Black participants where ROC-AUC values ranged from 0.63-0.72 across multiple years. Retraining the hybrid computer vision model by incorporating Black and other participants from the UIH cohort improved performance with ROC-AUC values of 0.70-0.87 in a held out UIH test set. CONCLUSIONHybrid computer vision predicted risk with improved accuracy compared to clinical risk models alone. However, potential biases in image training data reduced model generalizability in Black participants. Performance was improved upon retraining with a subset of the UIH cohort, suggesting that inclusive training and validation datasets can minimize racial disparities. Future studies incorporating vision models trained on representative data sets may demonstrate improved health equity upon clinical use.

BackgroundThe U.S. Southeast has a high burden of SARS-CoV-2 infections and COVID-19 disease. We used public data sources and community engagement to prioritize county selections for a precision population health intervention to promote a SARS-CoV-2 testing intervention in rural Alabama during October 2020 and March 2021. MethodsWe modeled factors associated with county-level SARS-CoV-2 percent positivity using covariates thought to associate with SARS-CoV-2 acquisition risk, disease severity, and risk mitigation practices. Descriptive epidemiologic data were presented to scientific and community advisory boards to prioritize counties for a testing intervention. ResultsIn October 2020, SARS-CoV-2 percent positivity was not associated with any modeled factors. In March 2021, premature death rate (aRR 1.16, 95% CI 1.07, 1.25), percent Black residents (aRR 1.00, 95% CI 1.00, 1.01), preventable hospitalizations (aRR 1.03, 95% CI 1.00, 1.06), and proportion of smokers (aRR 0.231, 95% CI 0.10, 0.55) were associated with average SARS-CoV-2 percent positivity. We then ranked counties based on percent positivity, case fatality, case rates, and number of testing sites using individual variables and factor scores. Top ranking counties identified through factor analysis and univariate associations were provided to community partners who considered ongoing efforts and strength of community partnerships to promote testing to inform intervention. ConclusionsThe dynamic nature of SARS-CoV-2 proved challenging for a modelling approach to inform a precision population health intervention at the county level. Epidemiological data allowed for engagement of community stakeholders implementing testing. As data sources and analytic capacities expand, engaging communities in data interpretation is vital to address diseases locally.

ObjectivesIllness trends are typically monitored by reportable disease and syndromic surveillance systems, but unanticipated health issues might not be captured. Using diagnosis codes, the New York City Health Department developed a data mining method to detect unusual increases in emergency department (ED) visits for any reason. MethodsWe applied the tree-temporal scan statistic in TreeScan software to ICD-10-CM diagnosis codes for ED visits. We searched for unusual citywide increases in ED visits or hospital admissions, over any recent time period, and at any part of and level on the ICD-10-CM tree. We conducted proof-of-concept analyses for March 2020 when COVID-19 emerged, then investigated signals detected in daily, automated analyses during April-August 2025. ResultsIf TreeScan analyses had been in place, then increasing hospital admissions for viral pneumonia (J12) would have triggered a signal on March 13, 2020, two days before widespread COVID-19 community transmission was announced. An extreme heat event in June 2025 triggered a signal for admissions for acute kidney failure (N17), prompting outreach to dialysis networks. A sustained signal for hand, foot, and mouth disease (B08.4) prompted outreach to child care programs. Other signals supported situational awareness, including a seasonal increase for swimmers ear (H60.33) and burns (T30.0) related to consumer fireworks. Practice ImplicationsTreeScan quickly detected credible increases in various diagnoses without pre-specification, from minor to severe, rare to common, acute to sustained, and foreseen to unforeseen. TreeScan can strengthen surveillance for health issues related to new pathogens, non-notifiable conditions, environmental exposures, and mass gatherings.

The COVID-19 pandemic exhibits stark social inequities in infection and mortality outcomes. We investigated neighborhood-level inequities across cities in the US and Canada for COVID-19 cumulative case rates (46 cities), death rates (12 cities), testing rates and test positivity (12 cities), using measures that characterize social gradients by race/ethnicity, socioeconomic composition, or both jointly. We found consistent evidence of social gradients for case, death and positivity rates, with the most privileged neighborhoods having the lowest rates, but no meaningful variation in the magnitude of inequities between cities. Gradients were not apparent in testing rates, suggesting inadequate testing in the most deprived neighborhoods. Health agencies should monitor and compare inequities as part of their COVID-19 reporting practices and to guide pandemic response efforts. HIGHLIGHTSO_LIWithin urban regions with available data in the US and Canada, there were strong social gradients for case, death and positivity rates C_LIO_LIThe most racially and/or economically privileged neighborhoods had the lowest rates C_LIO_LISocial gradients were similar for neighborhood-level measures of racial/ethnic composition, income, racialized economic segregation, and racialized occupational segregation C_LIO_LITesting rates did not show consistent social gradients, which suggests that the most deprived neighborhoods have inadequate access to testing relative to their higher disease burden C_LI

ObjectiveTo describe the distribution of alcohol and drug involvement in injurious falls by location and subtype of fall. MethodsUsing the 2019 National Emergency Medical Services Information System (NEMSIS) dataset we identified 1,854,909 patients injured from falls requiring an Emergency Medical Services (EMS) response and determined the fall location (e.g. indoors or on street/sidewalk) and the EMS clinicians notation of alcohol or drug involvement. We analyzed substance involvement by fall subtype, location of fall and patient demographics. ResultsOverall, for 7.4% of injurious falls there was a notation of substance use: 6.5% for alcohol alone, 0.6% for drugs and 0.3% for alcohol and drugs. 21.2% of falls that occurred on a street or sidewalk had a notation of substance use; alcohol use alone for 18.5% of falls, drugs alone for 1.7% of falls and alcohol and drugs for 0.9% of falls. Substance use prevalence was highest, at 30.3%, in the age group 21 to 64 years, for falls occurring on streets and sidewalks, without syncope or heat illness as contributing factors; alcohol use alone for 26.3%, drugs alone for 2.6%, and alcohol and drugs for 1.4%. Reported substance use involvement was more frequent for men compared to women for each location type. ConclusionsOverall, 1-in-5 injurious falls on streets and sidewalks and requiring EMS attention involved substance use, and these numbers likely underestimate the true burden. As cities seek to expand nightlife districts, design strategies to protect pedestrians from falls should be enacted.

ABSTRACT: Importance: Rural urban disparities in chronic disease prevalence are well established; however, the extent to which individual level socioeconomic status (SES) contributes to these disparities remains unclear. Objective: To examine the associations of rurality and SES with the prevalence of five most burdensome chronic diseases among adults. Design: We conducted a retrospective cross sectional study of adults across 27 Upper Midwest counties using the Expanded Rochester Epidemiology Project (E REP) medical record data linkage system to evaluate associations between rurality, SES and chronic disease prevalence. Prevalence of clinically diagnosed asthma, diabetes, hypertension, coronary heart disease, and mood disorders was identified from International Classification of Diseases ICD9/10 codes over a five-year period (2014 to 2019). Setting: Population based Participants: Adults over 18 years residing in the 27 E REP counties, excluding those missing rural urban residence status. Exposure: HOUSES index, an individual level measure of SES, served as the primary measure, while rurality based on Rural Urban Commuting Area (RUCA) codes 4-10 was the secondary measure. Main Outcome: Prevalence of the five clinically diagnosed chronic diseases was identified using ICD9/10 codes from 2014 to 2019. Mixed effect logistic regression models were used and adjusted for demographics and general medical examination receipt, to assess rural urban and SES differences for prevalence of each chronic disease. Results: Among 455,802 adults with available HOUSES index, 42.8% lived in rural areas, 53.8% were female and 87.4% were non-Hispanic White. In the unadjusted analysis, rural and urban populations showed comparable asthma and CHD prevalence, while mood disorders, hypertension, and diabetes were more common in urban areas. After adjusting for demographic factors and healthcare utilization, rural urban differences were no longer statistically significant, whereas SES remained strongly associated with all diseases in a dose response manner (e.g., adjusted Odds Ratio for hypertension (ref: HOUSES index Q4): 1.14, 1.27, and 1.42 for HOUSES index Q3, Q2, and Q1, respectively). Conclusions and Relevance: Individual level SES measured by the HOUSES index, was more strongly associated with chronic disease prevalence than rurality, supporting its integration into population health assessment and risk stratification.