The effect of urban racial residential segregation on ovarian cancer diagnosis, treatment, and survival.

OBJECTIVE: To investigate the effect of racial residential segregation on disparities between Black and White patients in stage at diagnosis, receipt of surgery, and survival. METHODS: Subjects included Black and White patients diagnosed with ovarian cancer between 2005 and 2015 obtained from the Surveillance, Epidemiology, and End Results Program. Demographic data were obtained from the 2010 decennial census and 2013 American Community Survey. The exposure of interest was the index of dissimilarity (IOD), a validated measure of segregation. The outcomes of interest included relative risk of advanced stage at diagnosis and surgery for localized disease, 5-year overall and cancer-specific survival. RESULTS: Black women were more likely to present with Stage IV ovarian cancer when compared to White (32% vs 25%, p < 0.001) and less often underwent surgical resection overall (64% vs 75%, p < 0.001). Increasing IOD was associated with a 25% increased risk of presenting at advanced stage for Black patients (RR 1.25, 95% CI 1.08, 1.45), and a 15% decrease for White patients (RR 0.85, 95% CI 0.73, 0.99). Increasing IOD was associated with an 18% decreased likelihood of undergoing surgical resection for black patients (RR 0.82, 95% CI 0.77, 0.87), but had no significant association for White patients (RR 1.01, 95% CI 0.96, 1.08). When compared to White patients in the lowest level of segregation, Black patients in the highest level of segregation had a 17% higher subhazard of death (HR 1.17, 95% CI 1.07, 1.27), while Black patients in the lowest level of segregation had no significant difference (HR 1.13, 95% CI 0.99, 1.29). CONCLUSION: Our findings demonstrate the direct harm of historical government mandated segregation on Black women with ovarian cancer.

The effect of historic redlining on firearm violence.

BACKGROUND: Firearm homicides disproportionately affect young Black men, which in turn have lasting impact of communities of color as a whole. Previous cross-sectional studies have highlighted the role of discriminatory housing policies on the incidence of urban firearm violence. We sought to estimate the effects of racist housing policies on firearm incidence. METHODS: Firearm incident data were obtained from the Boston Police Department and point locations spatially joined with vector files outlining the original 1930 Home Owner Loan Corporation (HOLC) Redlining maps. A regression discontinuity design was used to assess the increased rate of firearm violence crossing from historically "desirable" neighborhoods (Green) to historically "hazardous" neighborhoods (Red and Yellow) based on HOLC definitions. Linear regression models were fit on either side of the geographic boundaries with firearm incidents graphed at varying distances and the regression coefficient calculated at the boundary. RESULTS: Crossing from desirable to Red hazardous designation there was a significant discontinuity with an increase of 4.1 firearm incidents per 1,000 people (95% CI 0.68,7.55). Similarly, when crossing from desirable areas to the Yellow hazardous designation there was a significant discontinuity and increase of 5.9 firearm incidents per 1,000 people (95% CI 1.85,9.86). There was no significant discontinuity between the two hazardous HOLC designations (coefficient -0.93, 95% CI -5.71, 3.85). CONCLUSIONS: There is a significant increase in firearm incidents in historically redlined areas of Boston. This suggests that interventions should focus on downstream socioeconomic, demographic, and neighborhood detriments of historically discriminatory housing policies in order to address firearm homicides.

Historic redlining, social mobility, and firearm violence.

BACKGROUND: Much of the recorded medical literature focuses on individual-level factors that contribute to firearm violence. Recently, studies have highlighted higher incidence of firearm violence in historically redlined and marginalized areas, but few have gone on to study the downstream associations causing these differences. This study aims to understand the effects of historic redlining and current income mobility on firearm violence. METHODS: Using a retrospective cross-sectional design, shooting incidents were spatially joined with redlining vector files and linked to income mobility data (how much a child makes in adulthood). Participants included all assault and homicide incidents involving a firearm in the city of Boston, between 2016 and 2019. The exposure of interest was redlining designation as outlined by the Home Owner's Loan Corporation (HOLC) in the 1930s and income mobility, stratified by race, defined as the income of a child in their 30s compared with where they grew up (census tract level). The outcome measured was shooting rate per census block. RESULTS: We find that increases in Black income mobility (BIM) and White income mobility (WIM) are associated with significant decreases in rates of firearm incidents in all HOLC designations; however, there is a larger decrease with increasing BIM (relative risk, 0.47 per unit increase in BIM [95% confidence interval, 0.35-0.64]; relative risk, 0.81 per unit increase in WIM [95% confidence interval, 0.71-0.93]). Plotting predicted rates of firearm violence in each HOLC designation at different levels of BIM, there were no significant differences in shooting rates between historically harmful and beneficial classifications above $50,000 of BIM. Despite level of WIM, there were continued disparities between harmful and beneficial HOLC classification. CONCLUSION: These findings highlight the importance of structural racism in the form of redlining and discriminatory housing policies, and the preclusion from economic mobility therein, on the incidence of firearm violence today. LEVEL OF EVIDENCE: Prognostic and Epidemiological; Level IV.

Socioeconomic Mediation of Racial Segregation in Pancreatic Cancer Treatment and Outcome Disparities.

OBJECTIVE: To understand the mediating effect of socioeconomic factors on the association between residential segregation and racial disparities in pancreatic cancer (PC). BACKGROUND: Black patients with PC present at a later stage and have worse mortality than White patients. These disparities have been explained by the level of residential segregation. METHODS: Data were obtained from Surveillance, Epidemiology, and End-Results (SEER) and included all Black and White patients who were diagnosed with PC between 2005 and 2015. The primary exposure variable was the Index of Dissimilarity, a validated measure of segregation. County-level socioeconomic variables from the US Census were assessed as mediators. The primary outcomes were advanced stage at diagnosis, surgical resection for localized disease, and overall mortality. Generalized structural equation modeling was used to assess the mediation of each of the socioeconomic variables. RESULTS: Black patients in the highest levels of segregation saw a 12% increased risk [relative risk=1.12; 95% confidence interval (CI): 1.08, 1.15] of presenting at an advanced stage, 11% decreased likelihood of undergoing surgery (relative risk=0.89; 95% CI: 0.83, 0.94), and 8% increased hazards of death (hazard ratio=1.08; 95% CI: 1.03, 1.14) compared with White patients in the lowest levels. The Black share of the population, insurance status, and income inequality mediated 58% of the total effect on the advanced stage. Poverty and Black income immobility mediated 51% of the total effect on surgical resection. Poverty and Black income immobility mediated 50% of the total effect on overall survival. CONCLUSIONS: These socioeconomic factors serve as intervention points for legislators to address the social determinants inherent to the structural racism that mediate poor outcomes for Black patients.

The effect of medicare and screening guidelines on colorectal cancer outcomes.

BACKGROUND: Colorectal cancer screening has been shown effective at reducing stage at presentation, but there is differential uptake of screening based on insurance status. We sought to determine the population-level effect of Medicare and screening guidelines on colorectal screening by race and region. METHODS: Data on Black and white patients with colorectal cancer were obtained from the SEER database. Regression discontinuity was used to assess the causal effect of near-universal health insurance (represented by age 65) and United States Preventive Services Task Force guidelines (age 50) on the proportion of people presenting at advanced stage. This was stratified by race and region. RESULTS: In the Southern United States, Black patients saw a significant decrease in advanced stage at presentation at age 65 (coefficient -0.12, p = 0.003), while white patients did not (coefficient -0.03, p = 0.09). At age 50, neither Black (coefficient 0.09, p = 0.10) nor white patients (coefficient -0.04, p = 0.1) saw a significant decrease in advanced stage. In the Western U.S., neither Black (coefficient 0.02, p = 0.72) or white patients (coefficient -0.02, p = 0.09) saw a significant decrease in advanced stage at age 65; however, both Black (coefficient -0.20, p = 0.008) and white patients (coefficient -0.05, p = 0.03) saw a significant decrease at age 50. CONCLUSIONS: Our data highlight the significant impact that near-universal insurance has on reducing colorectal cancer stage at presentation in areas with poor baseline insurance coverage, particularly for Black patients. To reduce disparities in advanced stage at presentation for colorectal cancer, state-level insurance coverage should be addressed.

Responsible product design to mitigate excessive gambling: A scoping review and z-curve analysis of replicability.

OBJECTIVES: Systematic mapping of evaluations of tools and interventions that are intended to mitigate risks for gambling harm. DESIGN: Scoping Review and z-curve analysis (which estimates the average replicability of a body of literature). SEARCH STRATEGY: We searched 7 databases. We also examined reference lists of included studies, as well as papers that cited included studies. Included studies described a quantitative empirical assessment of a game-based (i.e., intrinsic to a specific gambling product) structural feature, user-directed tool, or regulatory initiative to promote responsible gambling. At least two research assistants independently performed screening and extracted study characteristics (e.g., study design and sample size). One author extracted statistics for the z-curve analysis. RESULTS: 86 studies met inclusion criteria. No tools or interventions had unambiguous evidence of efficacy, but some show promise, such as within-session breaks in play. Pre-registration of research hypotheses, methods, and analytic plans was absent until 2019, reflecting a recent embracement of open science practices. Published studies also inconsistently reported effect sizes and power analyses. The results of z-curve provide some evidence of publication bias, and suggest that the replicability of the responsible product design literature is uncertain but could be low. CONCLUSION: Greater transparency and precision are paramount to improving the evidence base for responsible product design to mitigate gambling-related harm.