Epidemiology最新文献

Background: Drug overdose persists as a leading cause of death in the United States, but resources to address it remain limited. As a result, health authorities must consider where to allocate scarce resources within their jurisdictions. Machine learning offers a strategy to identify areas with increased future overdose risk to proactively allocate overdose prevention resources. This modeling study is embedded in a randomized trial to measure the effect of proactive resource allocation on statewide overdose rates in Rhode Island (RI).

Methods: We used statewide data from RI from 2016 to 2020 to develop an ensemble machine learning model predicting neighborhood-level fatal overdose risk. Our ensemble model integrated gradient boosting machine and super learner base models in a moving window framework to make predictions in 6-month intervals. Our performance target, developed a priori with the RI Department of Health, was to identify the 20% of RI neighborhoods containing at least 40% of statewide overdose deaths, including at least one neighborhood per municipality. The model was validated after trial launch.

Results: Our model selected priority neighborhoods capturing 40.2% of statewide overdose deaths during the test periods and 44.1% of statewide overdose deaths during validation periods. Our ensemble outperformed the base models during the test periods and performed comparably to the best-performing base model during the validation periods.

Conclusions: We demonstrated the capacity for machine learning models to predict neighborhood-level fatal overdose risk to a degree of accuracy suitable for practitioners. Jurisdictions may consider predictive modeling as a tool to guide allocation of scarce resources.

The interpretation of vaccine efficacy estimands is subtle, even in randomized trials designed to quantify the immunologic effects of vaccination. In this article, we introduce terminology to distinguish between different vaccine efficacy estimands and clarify their interpretations. This allows us to explicitly consider the immunologic and behavioral effects of vaccination, and establish that policy-relevant estimands can differ substantially from those commonly reported in vaccine trials. We further show that a conventional vaccine trial allows the identification and estimation of different vaccine estimands under plausible conditions if one additional post-treatment variable is measured. Specifically, we utilize a "belief variable" that indicates the treatment an individual believed they had received. The belief variable is similar to "blinding assessment" variables that are occasionally collected in placebo-controlled trials in other fields. We illustrate the relations between the different estimands, and their practical relevance, in numerical examples based on an influenza vaccine trial.

Background: Studies have suggested Medicaid expansion enacted in 2014 has resulted in a reduction in overall cardiovascular disease (CVD) mortality in the United States. However, it is unknown whether Medicaid expansion has a similar effect across race-ethnicity and sex. We investigated the effect of Medicaid expansion on CVD mortality across race-ethnicity and sex.

Methods: Data come from the behavioral risk factor surveillance system and the US Centers for Disease Control's Wide-ranging Online Data for Epidemiologic Research, spanning the period 2000-2019. We used the generalized synthetic control method, a quasi-experimental approach, to estimate effects.

Results: Medicaid expansion was associated with -5.36 (mean difference [MD], 95% confidence interval [CI] = -22.63, 11.91) CVD deaths per 100,000 persons per year among Blacks; -4.28 (MD, 95% CI = -30.08, 21.52) among Hispanics; -3.18 (MD, 95% CI = -8.30, 1.94) among Whites; -5.96 (MD, 95% CI = -15.42, 3.50) among men; and -3.34 (MD, 95% CI = -8.05, 1.37) among women. The difference in mean difference (DMD) between the effect of Medicaid expansion in Blacks compared with Whites was -2.18; (DMD, 95% CI = -20.20, 15.83); between that in Hispanics compared with Whites: -1.10; (DMD, 95% CI = -27.40, 25.20) and between that in women compared with men: 2.62; (DMD, 95% CI = -7.95, 13.19).

Conclusions: Medicaid expansion was associated with a reduction in CVD mortality overall and in White, Black, Hispanic, male, and female subpopulations. Also, our study did not find any difference or disparity in the effect of Medicaid on CVD across race-ethnicity and sex-gender subpopulations, likely owing to imprecise estimates.

Background: Road traffic injury contributes substantially to morbidity and mortality. Canada stands out among developed countries in not conducting a national household travel survey, leading to a dearth of national transportation mode data and risk calculations that have appropriate denominators. Since traffic injuries are specific to the mode of travel used, these risk calculations should consider travel mode.

Methods: Census data on mode of commute is one of the few sources of these data for persons aged 15 and over. This study leveraged a national data linkage cohort, the Canadian Census Health and Environment Cohorts, that connects census sociodemographic and commute mode data with records of deaths and hospitalizations, enabling assessment of road traffic injury associations by indicators of mode of travel (commuter mode). We examined longitudinal (1996-2019) bicyclist, pedestrian, and motor vehicle occupant injury and fatality risk in the Canadian Census Health and Environment Cohorts by commuter mode and sociodemographic characteristics using Cox proportional hazards models within the working adult population.

Results: We estimated positive associations between commute mode and same mode injury and fatality, particularly for bicycle commuters (hazard ratios for bicycling injury was 9.1 and for bicycling fatality was 11). Low-income populations and Indigenous people had increased injury risk across all modes.

Conclusions: This study shows inequities in transportation injury risk in Canada and underscores the importance of adjusting for mode of travel when examining differences between population groups.

Background: Individuals with low socioeconomic position (SEP) experience disproportionately greater alcohol-attributable harm than individuals with high SEP despite similar or less alcohol use (i.e., the alcohol harm paradox). We examined the sex/gender- specific independent and joint effects of education and heavy drinking or volume of alcohol use on 100% alcohol-attributable hospitalization or death.

Methods: We conducted a cohort study among 199,125 current and former alcohol users aged 15-64 years from population-representative Canadian Community Health Surveys (2000-2008) linked to hospitalization and mortality records through 2017. We estimated the sex/gender-specific associations between education and heavy drinking or volume of alcohol use and incident 100% alcohol-attributable hospitalization or death using multivariable Fine and Gray subdistribution hazard models with competing risk (non-100% alcohol-attributable deaths), assessing additive interactions using the Synergy Index (S).

Results: Overall, heavy drinking prevalence and volume of alcohol use were similar or lower in individuals with lower education compared with higher education. Lower education levels compared with a bachelor's degree or above were associated with increased 100% alcohol-attributable hospitalization or death [e.g., less than high school, men: hazard ratio (HR) = 2.78; 95% CI = 2.17, 3.56; women: HR = 2.98; 95% CI = 2.00, 4.44]. We found superadditive joint effects between low education and heavy drinking (men: S = 1.22; 95% CI = 1.14, 1.30; women: S = 1.34; 95% CI = 0.88, 2.04) and low education and higher volume of alcohol use (e.g., excess volume, men: S = 1.30; 95% CI = 1.05, 1.62; women: S = 1.41; 95% CI = 0.77, 2.58), with larger inequities in women than men with similar alcohol use.

Conclusions: Our study is consistent with the hypothesis that increased vulnerability to alcohol use among individuals with lower education partially explains the alcohol harm paradox in Canada.