Universal Difference-in-Differences for Causal Inference in Epidemiology.

Difference-in-differences is undoubtedly one of the most widely used methods for evaluating the causal effect of an intervention in observational (i.e., nonrandomized) settings. The approach is typically used when pre- and postexposure outcome measurements are available, and one can reasonably assume that the association of the unobserved confounder with the outcome has the same absolute magnitude in the two exposure arms and is constant over time; a so-called parallel trends assumption. The parallel trends assumption may not be credible in many practical settings, for example, if the outcome is binary, a count, or polytomous, as well as when an uncontrolled confounder exhibits nonadditive effects on the distribution of the outcome, even if such effects are constant over time. We introduce an alternative approach that replaces the parallel trends assumption with an odds ratio equi-confounding assumption under which an association between treatment and the potential outcome under no treatment is identified with a well-specified generalized linear model relating the pre-exposure outcome and the exposure. Because the proposed method identifies any causal effect that is conceivably identified in the absence of confounding bias, including nonlinear effects such as quantile treatment effects, the approach is aptly called universal difference-in-differences. We describe and illustrate both fully parametric and more robust semiparametric universal difference-in-differences estimators in a real-world application concerning the causal effects of a Zika virus outbreak on birth rate in Brazil. A supplementary digital video is available at: http://links.lww.com/EDE/C90.

Single proxy control.

Negative control variables are sometimes used in nonexperimental studies to detect the presence of confounding by hidden factors. A negative control outcome (NCO) is an outcome that is influenced by unobserved confounders of the exposure effects on the outcome in view, but is not causally impacted by the exposure. Tchetgen Tchetgen (2013) introduced the Control Outcome Calibration Approach (COCA) as a formal NCO counterfactual method to detect and correct for residual confounding bias. For identification, COCA treats the NCO as an error-prone proxy of the treatment-free counterfactual outcome of interest, and involves regressing the NCO on the treatment-free counterfactual, together with a rank-preserving structural model, which assumes a constant individual-level causal effect. In this work, we establish nonparametric COCA identification for the average causal effect for the treated, without requiring rank-preservation, therefore accommodating unrestricted effect heterogeneity across units. This nonparametric identification result has important practical implications, as it provides single-proxy confounding control, in contrast to recently proposed proximal causal inference, which relies for identification on a pair of confounding proxies. For COCA estimation we propose 3 separate strategies: (i) an extended propensity score approach, (ii) an outcome bridge function approach, and (iii) a doubly-robust approach. Finally, we illustrate the proposed methods in an application evaluating the causal impact of a Zika virus outbreak on birth rate in Brazil.

Higher-order evidence.

Higher-order evidence is evidence about evidence. Epidemiologic examples of higher-order evidence include the settings where the study data constitute first-order evidence and estimates of misclassification comprise the second-order evidence (e.g., sensitivity, specificity) of a binary exposure or outcome collected in the main study. While sampling variability in higher-order evidence is typically acknowledged, higher-order evidence is often assumed to be free of measurement error (e.g., gold standard measures). Here we provide two examples, each with multiple scenarios where second-order evidence is imperfectly measured, and this measurement error can either amplify or attenuate standard corrections to first-order evidence. We propose a way to account for such imperfections that requires third-order evidence. Further illustrations and exploration of how higher-order evidence impacts results of epidemiologic studies is warranted.

Heat-related fatalities in North Carolina 1999-2017.

OBJECTIVES: Research shows the highest rates of occupational heat-related fatalities among farm laborers and among Black and Hispanic workers in North Carolina (NC). The Hispanic population and workforce in NC have grown substantially in the past 20 years. We describe the epidemiology of heat-related fatal injuries in the general population and among workers in NC. METHODS: We reviewed North Carolina death records and records of the North Carolina Office of the Chief Medical Examiner to identify heat-related deaths (primary International Classification of Diseases, Tenth Revision diagnosis code: X30 or T67.0-T67.9) that occurred between January 1, 1999, and December 31, 2017. Decedent age, sex, race, and ethnicity were extracted from both the death certificate and the medical examiner's report as well as determinations of whether the death occurred at work. RESULTS: In NC between 1999 and 2017, there were 225 deaths from heat-related injuries, and 25 occurred at work. The rates of occupational heat-related deaths were highest among males, workers of Hispanic ethnicity, workers of Black, multiple, or unknown race, and in workers aged 55-64. The highest rate of occupational heat-related deaths occurred in the agricultural industry. CONCLUSIONS: Since the last report (2001), the number of heat-related fatalities has increased, but fewer were identified as workplace fatalities. Rates of occupational heat-related deaths are highest among Hispanic workers. NC residents identifying as Black are disproportionately burdened by heat-related fatalities in general, with a wider apparent disparity in occupational deaths.

Forty-year trends in fatal occupational injuries in North Carolina.

BACKGROUND: We describe progress in the control of deaths on-the-job due to fatal occupational injury in North Carolina over the period 1978-2017. METHODS: Forty years of information on fatal occupational injuries in North Carolina has been assembled from medical examiners' reports and death certificates, supplemented by newspaper and police reports. Cases were defined as unintentional fatal occupational injuries among adults. Annual estimates of the population at risk were derived from US Census data, and rates were quantified using Poisson regression methods. RESULTS: There were 4434 eligible deaths. The unintentional fatal occupational injury rate at the beginning of the study period was more than threefold the rate at the end of the study. The fatal occupational injury rate among men declined from 9.6 per 100,000 worker-years in the period 1978-1982 to 3.1 per 100,000 worker-years in the period 2013-2017. The fatal occupational injury rate among women declined from 0.3 per 100,000 worker-years in the period 1978-1981 to 0.1 per 100,000 worker-years in the period 2013-2017. Declines in rates were observed for young adults as well as older workers and were observed across all major industry categories. Average annual declines in rates were greatest in those industries and occupations that had the highest fatal injury rates at the start of the study period. CONCLUSIONS: The substantial decline in fatal injury rates underscores the importance of injury prevention and demonstrates the ability to make meaningful reductions in unintentional fatal injury.

Forty years of struggle in North Carolina: Workplace segregation and fatal occupational injury rates.

OBJECTIVE: To assess workplace segregation in fatal occupational injury from 1992 to 2017 in North Carolina. METHODS: We calculated occupational fatal injury rates within categories of occupation, industry, race, age, and sex; and estimated expected numbers of fatalities among Black and Hispanic male workers had they experienced the rates of White male workers. We also estimated the contribution of workforce segregation to disparities by estimating the expected number of fatalities among Black and Hispanic male workers had they experienced the industry and occupation patterns of White male workers. We assessed person-years of life-lost, using North Carolina life expectancy estimates. RESULTS: Hispanic workers contributed 32% of their worker-years and experienced 58% of their fatalities in construction. Black workers were most overrepresented in the food manufacturing industry. Hispanic males experienced 2.11 (95% CI: 1.86-2.40) times the mortality rate of White males. The Black-White and Hispanic-White disparities were widest among workers aged 45 and older, and segregation into more dangerous industries and occupations played a substantial role in driving disparities. Hispanic workers who suffered occupational fatalities lost a median 47 life-years, compared to 37 among Black workers and 36 among White workers. CONCLUSIONS: If Hispanic and Black workers experienced the workplace safety of their White counterparts, fatal injury rates would be substantially reduced. Workforce segregation reflects structural racism, which also contributes to mortality disparities. Root causes must be addressed to eliminate disparities.

25-Year fatal workplace suicide trends in North Carolina: 1992-2017.

BACKGROUND: Suicide is a serious public health problem in the United States, but limited evidence is available investigating fatal suicides at work. There is a substantial need to characterize workplace suicides to inform suicide prevention interventions and target high-risk settings. This study aims to examine workplace suicide rates in North Carolina (NC) by worker characteristics, means of suicide used, and industry between 1992 and 2017. METHODS: Fatal workplace suicides were identified from records of the NC Office of the Chief Medical Examiner system and the NC death certificate. Sex, age, race, ethnicity, class of worker, manner of death, and industry were abstracted. Crude and age-standardized homicide rates were calculated as the number of suicides that occurred at work divided by an estimate of worker-years (w-y). Rate ratios and 95% confidence intervals (CIs) were calculated, and trends over calendar time for fatal workplace suicides were examined overall and by industry. RESULTS: 81 suicides over 109,464,430 w-y were observed. Increased rates were observed in workers who were male, self-employed, and 65+ years old. Firearms were the most common means of death (63%) followed by hanging (16%). Gas service station workers experienced the highest fatal occupational suicide rate, 11.5 times (95% CI: 3.62-36.33) the overall fatal workplace suicide rate, followed by Justice, Public Order, and Safety workers at 3.23 times the overall rate (95% CI: 1.31-7.97). CONCLUSION: Our findings identify industries and worker demographics that were vulnerable to workplace suicides. Targeted and tailored mitigation strategies for vulnerable industries and workers are recommended.

Estimating the Effect of a Treatment When There Is Nonadherence in a Trial.

Randomized trials offer a powerful strategy for estimating the effect of a treatment on an outcome. However, interpretation of trial results can be complicated when study subjects do not take the treatment to which they were assigned; this is referred to as nonadherence. Prior authors have described instrumental variable approaches to analyze trial data with nonadherence; under their approaches, the initial assignment to treatment is used as an instrument. However, their approaches require the assumption that initial assignment to treatment has no direct effect on the outcome except via the actual treatment received (i.e., the exclusion restriction), which may be implausible. We propose an approach to identification of a causal effect of treatment in a trial with 1-sided nonadherence without assuming exclusion restriction. The proposed approach leverages the study subjects initially assigned to control status as an unexposed reference population; we then employ a bespoke instrumental variable analysis, where the key assumption is "partial exchangeability" of the association between a covariate and an outcome in the treatment and control arms. We provide a formal description of the conditions for identification of causal effects, illustrate the method using simulations, and provide an empirical application.

Inverse Probability Weighting to Estimate Exposure Effects on the Burden of Recurrent Outcomes in the Presence of Competing Events.

Recurrent events-outcomes that an individual can experience repeatedly over the course of follow-up-are common in epidemiologic and health services research. Studies involving recurrent events often focus on time to first occurrence or on event rates, which assume constant hazards over time. In this paper, we contextualize recurrent event parameters of interest using counterfactual theory in a causal inference framework and describe an approach for estimating a target parameter referred to as the mean cumulative count. This approach leverages inverse probability weights to control measured confounding with an existing (and underutilized) nonparametric estimator of recurrent event burden first proposed by Dong et al. in 2015. We use simulations to demonstrate the unbiased estimation of the mean cumulative count using the weighted Dong-Yasui estimator in a variety of scenarios. The weighted Dong-Yasui estimator for the mean cumulative count allows researchers to use observational data to flexibly estimate and contrast the expected number of cumulative events experienced per individual by a given time point under different exposure regimens. We provide code to ease application of this method.

Generalized Difference-in-Differences.

Difference-in-differences (DID) analyses are used in a variety of research areas as a strategy for estimating the causal effect of a policy, program, intervention, or environmental hazard (hereafter, treatment). The approach offers a strategy for estimating the causal effect of a treatment using observational (i.e., nonrandomized) data in which outcomes on each study unit have been measured both before and after treatment. To identify a causal effect, a DID analysis relies on an assumption that confounding of the treatment effect in the pretreatment period is equivalent to confounding of the treatment effect in the post treatment period. We propose an alternative approach that can yield identification of causal effects under different identifying conditions than those usually required for DID. The proposed approach, which we refer to as generalized DID, has the potential to be used in routine policy evaluation across many disciplines, as it essentially combines two popular quasiexperimental designs, leveraging their strengths while relaxing their usual assumptions. We provide a formal description of the conditions for identification of causal effects, illustrate the method using simulations, and provide an empirical example based on Card and Krueger's landmark study of the impact of an increase in minimum wage in New Jersey on employment.

Disparities in Fatal Occupational Injury Rates in North Carolina, 1978-2017: Comparing Nonmanagerial Employees to Managers.

BACKGROUND: We examined fatal occupational injuries among private-sector workers in North Carolina during the 40-year period 1978-2017, comparing the occurrence of fatal injuries among nonmanagerial employees to that experienced by managers. METHODS: We estimated a standardized fatal occupational injury ratio by inverse probability of exposure weighting, taking nonmanagerial workers as the target population. When this ratio measure takes a value greater than unity it signals settings in which nonmanagerial employees are not provided as safe a work environment as that provided for managers. RESULTS: Across all industries, nonmanagerial workers in North Carolina experienced fatal occupational injury rates 8.2 (95% CI = 7.0, 10.0) times the rate experienced by managers. Disparities in fatal injury rates between managers and the employees they supervise were greatest in forestry, rubber and metal manufacturing, wholesale trade, fishing and extractive industries, and construction. CONCLUSIONS: The results may help focus discussion about workplace safety between labor and management upon equity, with a goal of providing a work environment for nonmanagerial employees as safe as the one provided for managers.

The carcinogenicity of opium consumption: a systematic review and meta-analysis.

The carcinogenicity of opium consumption was recently evaluated by a Working Group convened by the International Agency for Research on Cancer (IARC). We supplement the recent IARC evaluation by conducting an extended systematic review as well as a quantitative meta-analytic assessment of the role of opium consumption and risk for selected cancers, evaluating in detail various aspects of study quality on meta-analytic findings. We searched the published literature to identify all relevant studies on opium consumption and risk of selected cancers in humans through 31 October, 2022. Meta-relative risks (mRRs) and associated 95% confidence intervals (CIs) were estimated using random-effects models for studies of cancer of the urinary bladder, larynx, lung, oesophagus, pancreas, and stomach. Heterogeneity among studies was assessed using the I2 statistic. We assessed study quality and conducted sensitivity analyses to evaluate the impact of potential reverse causation, protopathic bias, selection bias, information bias, and confounding. In total, 2 prospective cohort studies and 33 case-control studies were included. The overall pooled mRR estimated for 'ever or regular' versus 'never' use of opium ranged from 1.50 (95% CI 1.13-1.99, I2 = 0%, 6 studies) for oesophageal cancer to 7.97 (95% CI 4.79-13.3, I2 = 62%, 7 studies) for laryngeal cancer. Analyses of cumulative opium exposure suggested greater risk of cancer associated with higher opium consumption. Findings were robust in sensitivity analyses excluding studies prone to potential methodological sources of biases and confounding. Findings support an adverse association between opium consumption and cancers of the urinary bladder, larynx, lung, oesophagus, pancreas and stomach.

Trends in fatal occupational injury rates among older workers before and after the Great Recession of 2008.

BACKGROUND: Older workers experience higher rates of fatal occupational injury than younger workers worldwide. In North Carolina, the population of older workers more than doubled between 2000 and 2017. In 2008, the Great Recession changed occupational patterns among all age groups. We examined annual rates and distribution of fatal occupational injuries experienced by older workers, comparing the pre-recession period (2000-2007) to the post-recession period (2009-2017). METHODS: Detailed information on all fatal occupational injuries during the period between 1 January 2000 and 31 December 2017 were abstracted from the records of the North Carolina Office of the Chief Medical Examiner and the office of vital records. The decennial Census and American Community Survey were used to estimate the population at risk and derive annual rates of fatal occupational injury. RESULTS: During the study period, 537 occupational fatalities occurred among workers 55+ years of age. The rate of fatal occupational injury among older workers declined 2.8% per year, with a 7.7% yearly decline in the pre-recession period compared with a 1.4% increase per year in the post-recession period. Workers 65+ years of age experienced rate increases in both periods. The highest rates of unintentional fatal occupational injury (injuries that were not purposefully inflicted) were observed in forestry, fishing hunting and trapping, and wood building manufacturing. Intentional fatal occupational injury rates (homicide, suicide) were highest in transportation, gas/service stations and grocery/food stores. CONCLUSIONS: Older workers have persistently high rates of fatal occupational injury in North Carolina before and after the Great Recession.

Fatal occupational injuries in North Carolina, 1992-2017.

OBJECTIVES: After declining for several decades, fatal occupational injury rates have stagnated in the USA since 2009. To revive advancements in workplace safety, interventions targeting at-risk worker groups must be implemented. Our study aims to identify these at-risk populations by evaluating disparities in unintentional occupational fatalities occurring in North Carolina (NC) from 1992 to 2017. METHODS: Our retrospective cohort study drew on both the NC Office of the Chief Medical Examiner system and the NC death certificate data system to identify unintentional fatal occupational injuries occurring from 1992 to 2017. Unintentional fatal occupational injury rates were reported across industries, occupations and demographic groups, and rate ratios were calculated to assess disparities. RESULTS: Among those aged 18 and older, 2645 unintentional fatal occupational injuries were identified. Fatal occupational injury rates declined by 0.82 injuries/100 000 person-years over this period, falling consistently from 2004 to 2009 and increasing from 2009 to 2017. Fatal injury rates were highest among Hispanic workers, who experienced 2.75 times the fatal injury rate of non-Hispanic White workers (95% CI 2.42 to 3.11) and self-employed workers, who experienced 1.44 times the fatal injury rate of private workers (95% CI 1.29 to 1.60). We also observed that fatal injury rates increased with age group and were higher among male relative to female workers even after adjustment for differential distributions across occupations. CONCLUSIONS: The decline in unintentional fatal occupational injury rates over this period is encouraging, but the increase in injury rate after 2009 and the large disparities between occupations, industries and demographic groups highlight the need for additional targeted safety interventions.

Radon and lung cancer in the pooled uranium miners analysis (PUMA): highly exposed early miners and all miners.

OBJECTIVES: Radon is a ubiquitous occupational and environmental lung carcinogen. We aim to quantify the association between radon progeny and lung cancer mortality in the largest and most up-to-date pooled study of uranium miners. METHODS: The pooled uranium miners analysis combines 7 cohorts of male uranium miners with 7754 lung cancer deaths and 4.3 million person-years of follow-up. Vital status and lung cancer deaths were ascertained between 1946 and 2014. The association between cumulative radon exposure in working level months (WLM) and lung cancer was modelled as the excess relative rate (ERR) per 100 WLM using Poisson regression; variation in the association by temporal and exposure factors was examined. We also examined analyses restricted to miners first hired before 1960 and with <100 WLM cumulative exposure. RESULTS: In a model that allows for variation by attained age, time since exposure and annual exposure rate, the ERR/100 WLM was 4.68 (95% CI 2.88 to 6.96) among miners who were less than 55 years of age and were exposed in the prior 5 to <15 years at annual exposure rates of <0.5 WL. This association decreased with older attained age, longer time since exposure and higher annual exposure rate. In analyses restricted to men first hired before 1960, we observed similar patterns of association but a slightly lower estimate of the ERR/100 WLM. CONCLUSIONS: This new large, pooled study confirms and supports a linear exposure-response relationship between cumulative radon exposure and lung cancer mortality which is jointly modified by temporal and exposure factors.

Bespoke Instruments: A new tool for addressing unmeasured confounders.

Suppose that an investigator is interested in quantifying an exposure-disease causal association in a setting where the exposure, disease, and some potential confounders of the association of interest have been measured. However, there remains concern about residual confounding of the association of interest by unmeasured confounders. We propose an approach to account for residual bias due to unmeasured confounders. The proposed approach uses a measured confounder to derive a "bespoke" instrumental variable that is tailored to the study population and is used to control for bias due to residual confounding. The approach may provide a useful tool for assessing and accounting for bias due to residual confounding. We provide a formal description of the conditions for identification of causal effects, illustrate the method using simulations, and provide an empirical example concerning mortality among Japanese atomic bomb survivors.

Amplification of Bias Due to Exposure Measurement Error.

Observational epidemiologic studies typically face challenges of exposure measurement error and confounding. Consider an observational study of the association between a continuous exposure and an outcome, where the exposure variable of primary interest suffers from classical measurement error (i.e., the measured exposures are distributed around the true exposure with independent error). In the absence of exposure measurement error, it is widely recognized that one should control for confounders of the association of interest to obtain an unbiased estimate of the effect of that exposure on the outcome of interest. However, here we show that, in the presence of classical exposure measurement error, the net bias in an estimate of the association of interest may increase upon adjustment for confounders. We offer an analytical expression for calculating the change in net bias in an estimate of the association of interest upon adjustment for a confounder in the presence of classical exposure measurement error, and we illustrate this problem using simulations.

A Bespoke Instrumental Variable Approach to Correction for Exposure Measurement Error.

A covariate-adjusted estimate of an exposure-outcome association may be biased if the exposure variable suffers measurement error. We propose an approach to correct for exposure measurement error in a covariate-adjusted estimate of the association between a continuous exposure variable and outcome of interest. Our proposed approach requires data for a reference population in which the exposure was a priori set to some known level (e.g., 0, and is therefore unexposed); however, our approach does not require an exposure validation study or replicate measures of exposure, which are typically needed when addressing bias due to exposure measurement error. A key condition for this method, which we refer to as "partial population exchangeability," requires that the association between a measured covariate and outcome in the reference population equals the association between that covariate and outcome in the target population in the absence of exposure. We illustrate the approach using simulations and an example.

Association Between Long-term Ambient PM2.5 Exposure and Cardiovascular Outcomes Among US Hemodialysis Patients.

RATIONALE & OBJECTIVE: Ambient PM2.5 (particulate matter with a diameter of 2.5 microns) is a ubiquitous air pollutant with established adverse cardiovascular (CV) effects. However, quantitative estimates of the association between PM2.5 exposure and CV outcomes in the setting of kidney disease are limited. This study assessed the association of long-term PM2.5 exposure with CV events and cardiovascular disease (CVD)-specific mortality among patients receiving maintenance in-center hemodialysis (HD). STUDY DESIGN: Retrospective cohort study. SETTINGS & PARTICIPANTS: 314,079 adult kidney failure patients initiating HD between 2011 and 2016 identified from the US Renal Data System. EXPOSURE: Estimated daily ZIP code-level PM2.5 concentrations were used to calculate each participant's annual average PM2.5 exposure based on the dialysis clinics visited during the 365 days before the outcome. OUTCOME: CV event and CVD-specific mortality were ascertained based on ICD-9/ICD-10 diagnostic codes and recorded cause of death from Centers for Medicare & Medicaid Services form 2746. ANALYTICAL APPROACH: Discrete time hazards models were used to estimate hazards ratios per 1 µg/m3 greater annual average PM2.5, adjusting for temperature, humidity, day of the week, season, age at baseline, race, employment status, and geographic region. Effect measure modification was assessed for age, sex, race, and baseline comorbidities. RESULTS: Each 1 µg/m3 greater annual average PM2.5 was associated with a greater rate of CV events (HR, 1.02 [95% CI, 1.01-1.02]) and CVD-specific mortality (HR, 1.02 [95% CI, 1.02-1.03]). The association was more pronounced for people who initiated dialysis at an older age, had chronic obstructive pulmonary disease (COPD) at baseline, or were Asian. Evidence of effect modification was also observed across strata of race, and other baseline comorbidities. LIMITATIONS: Potential exposure misclassification and unmeasured confounding. CONCLUSIONS: Long-term ambient PM2.5 exposure was associated with CVD outcomes among patients receiving maintenance in-center HD. Stronger associations between long-term PM2.5 exposure and adverse effects were observed among patients who were of advanced age, had COPD, or were Asian. PLAIN-LANGUAGE SUMMARY: Long-term exposure to air pollution, also called PM2.5, has been linked to adverse cardiovascular outcomes. However, little is known about the association of PM2.5 and outcomes among patients receiving dialysis, who are individuals with high cardiovascular disease burdens. We conducted an epidemiological study to assess the association between the annual PM2.5 exposure and cardiovascular events and death among patients receiving regular outpatient hemodialysis in the United States between 2011 and 2016. We found a higher risk of heart attacks, strokes, and related events in patients exposed to higher levels of air pollution. Stronger associations between air pollution and adverse health events were observed among patients who were older at the start of dialysis, had chronic obstructive pulmonary disease, or were Asian. These findings bolster the evidence base linking air pollution and adverse health outcomes and may inform policy makers and clinicians.

Estimating Associations Between Annual Concentrations of Particulate Matter and Mortality in the United States, Using Data Linkage and Bayesian Maximum Entropy.

BACKGROUND: Exposure to fine particulate matter (PM2.5) is an established risk factor for human mortality. However, previous US studies have been limited to select cities or regions or to population subsets (e.g., older adults). METHODS: Here, we demonstrate how to use the novel geostatistical method Bayesian maximum entropy to obtain estimates of PM2.5 concentrations in all contiguous US counties, 2000-2016. We then demonstrate how one could use these estimates in a traditional epidemiologic analysis examining the association between PM2.5 and rates of all-cause, cardiovascular, respiratory, and (as a negative control outcome) accidental mortality. RESULTS: We estimated that, for a 1 log(µg/m3) increase in PM2.5 concentration, the conditional all-cause mortality incidence rate ratio (IRR) was 1.029 (95% confidence interval [CI]: 1.006, 1.053). This implies that the rate of all-cause mortality at 10 µg/m3 would be 1.020 times the rate at 5 µg/m3. IRRs were larger for cardiovascular mortality than for all-cause mortality in all gender and race-ethnicity groups. We observed larger IRRs for all-cause, nonaccidental, and respiratory mortality in Black non-Hispanic Americans than White non-Hispanic Americans. However, our negative control analysis indicated the possibility for unmeasured confounding. CONCLUSION: We used a novel method that allowed us to estimate PM2.5 concentrations in all contiguous US counties and obtained estimates of the association between PM2.5 and mortality comparable to previous studies. Our analysis provides one example of how Bayesian maximum entropy could be used in epidemiologic analyses; future work could explore other ways to use this approach to inform important public health questions.