INTRODUCTION: Disparities in exposure to and harm associated with pesticides are established. However, high resolution spatial data on exposure to pesticides are lacking. PURPOSE: To describe the development of a pesticide metric for Washington State and to evaluate the associations between pesticide exposure and sociodemographic characteristics of census tracts. METHODS: We used the US cropland data set to quantify the location and size of agricultural land. These data were overlaid with state- and county-level estimated annual agricultural pesticide use to estimate pesticide exposure at the census tract level. Sociodemographic characteristics of census tracts were from the US Centers for Disease Control and Prevention's Social Vulnerability Index (SVI). Generalized additive models evaluated the associations between pesticide exposure and each of the SVI variables. RESULTS: The median exposure to pesticides was 1.5 lbs/sq mi. Significant associations were observed between exposure to pesticides and a higher percentage of population below poverty, populations unemployed, populations 65 and older, non-Hispanic white populations, those with limited English language proficiency, mobile homes, and group quarters. CONCLUSIONS: The results inform public health and policy efforts to identify areas and populations that are most vulnerable to pesticide exposure and improve the health and well-being of farmworkers and populations residing near agricultural areas.

OBJECTIVE:By limiting household expenditures on rent and utilities and connecting individuals to health services, federal housing assistance programs could facilitate health care access among low-income cancer survivors. We examined the association between receipt of rental assistance and health care access among low-income adult cancer survivors. METHODS:We used 2019-2023 National Health Interview Survey data on adults aged ≥20 years (1) with a history of cancer diagnosis, (2) with a family income-to-poverty ratio <2, and (3) who were renters. We used propensity score weighting to address differences in observed demographic characteristics between rental assistance recipients and nonrecipients. We used logistic regression analyses to examine the associations of rental assistance receipt with outcome variables. RESULTS:Lack of receipt of rental assistance was significantly associated with higher odds of delaying medical care due to cost (odds ratio [OR] = 1.90; 95% CI, 1.07-3.40) and experiencing medical financial hardship (OR = 1.85; 95% CI, 1.21-2.80), as well as lower odds of being covered by health insurance (OR = 0.04; 95% CI, 0.01-0.18). CONCLUSION:Receipt of rental assistance may help improve health care access among low-income adult cancer survivors. Our findings are important in relation to a shortage of affordable housing in the United States and highlight the need for efforts to expand housing assistance.

PURPOSE:To examine the association between telehealth utilization and mammogram receipt by rurality, and to characterize geographic patterns of concurrent increases in telehealth and mammography use in urban and rural communities. METHODS:The sample included women, aged 50 years or older, receiving services in primary care or gynecology/women's health practices at MultiCare Health System from 2018 to 2023. For the difference-in-difference (DiD) analyses, women who had only in-person visits during the prepandemic period but had at least one telehealth visit in each year of the postpandemic period were defined as the treatment group. Women who had only in-person visits during both pre- and postpandemic periods were defined as the control group. For the spatial analyses, we analyzed ZIP Codes in which both telehealth and mammography utilization increased from 2018-2019 to 2020-2022. FINDINGS:DiD analyses suggested that telehealth use was significantly associated with a greater probability of mammogram receipt among rural women; however, this association was not found in urban women. Spatial analyses indicated that concurrent increases in telehealth and mammogram use occurred in both rural and urban settings but exhibited greater spatial concentration in urban areas and comparatively more dispersed patterns across rural ZIP Code Tabulation Areas. CONCLUSIONS:The findings suggest that telehealth may be particularly beneficial for improving mammography screening among rural individuals, who typically experience greater structural barriers to accessing health care. While urban areas may exhibit clustered patterns driven by interconnected care systems, rural areas may experience more diffuse but meaningful gains as telehealth expands access across large geographic distances.

BACKGROUND:The objective of the study was to understand barriers to telehealth use among women and to examine mammography uptake in relation to these barriers. For this purpose, we implemented a survey to explore barriers to telehealth and use of cancer-preventative services. METHODS:Recruitment was based on electronic health record data from MultiCare Health System. Recruitment and enrollment of patients was initiated in 2024 and ended in March 2025, when the target sample size of 1,000 was reached. Women, aged between 50 and 75 years, were surveyed. Women with a personal history of breast cancer were excluded from the analyses. Among those included, 614 women had received virtual care in the previous 5 years, and 265 women had not received virtual care. Women who had 5 or more mammograms in the last 10 years were classified as 'optimal mammogram utilizers'. Those reporting 0-4 mammograms were identified as 'mammogram under-utilizers'. Logistic regression analyses examined the association of telehealth use with mammogram utilization. RESULTS:Compared to women who received virtual health care visits, women whose providers did not offer virtual care had significantly lower odds of optimal mammogram utilization. Lack of receipt of virtual care due to other reasons (i.e., technology issues and lack of preference/need) was not significantly associated with mammogram utilization. CONCLUSIONS:Study findings suggest that greater provider acceptance and use of telehealth for delivery of health care services could be associated with better mammogram utilization. Understanding barriers to provider acceptance of telehealth could inform efforts to expand telehealth.

The COVID-19 pandemic led to stay-at-home orders, resulting in sudden changes in movement patterns and social restrictions that impacted mental health. This study examined changes in individual behaviors during the pandemic using detailed smartphone-based activity data and determined whether behaviors and green space exposure buffered against mental health concerns. A total of 224 twins from the Washington State Twin Registry provided smartphone location data and completed a baseline survey and up to seven waves of follow-up surveys on mental health outcomes (ie, anxiety, depression, and stress). Objective measures of activities, locations, and time spent in green space were collected using Google Location History. Green space exposures were assessed using all location data. Associations between activities, locations, green space, and each mental health outcome were assessed using linear mixed models. There were changes in activities and locations from baseline to wave 1; by wave 7, the direction and magnitude of changes varied across measures, with some remaining above and others below their baseline levels. Mental health outcomes were associated with a subset of locations, activities, and green space measures although associations were not observed consistently across all exposure-outcome combinations examined.

Background: Formerly incarcerated people with serious mental illnesses (SMI) experience the criminal legal system unequally and have elevated rates of recidivism, homelessness, general medical problems, and substance use disorders. Permanent supportive housing (PSH) can be used during reentry, but it has limited resources for addressing community integration, a key component of reentry. PSH are often located in high-poverty environments with increased criminogenic risk. The geography of PSH also includes public spaces, which are associated with positive outcomes. The risk environment framework provides a structure for understanding the geography of PSH through its focus on the physical, social, economic, and policy influences on the micro and macro environments of reentry. Methods: This is a novel QUAL + QUAN (spatial) concurrent mixed-methods study that will examine how individual, interpersonal, and environmental factors interact with public and private spaces to inform reentry wellbeing. Eighty multimethod interviews (i.e., qualitative, quantitative, and participatory mapping methods) will be conducted with formerly incarcerated clients with SMI. Go-along interviews will be conducted with 20 of these participants. Participatory mapping will be geocoded and sites identified as places of importance, frequent participation, and belonging will be evaluated in relation to objective features of spaces to develop a community resilience index. Findings will ultimately be integrated into an intervention development codesign process with a community advisory board. Discussion: During reentry, individual, interpersonal, and environmental factors can interact with these environments to produce or reduce risk. If addressed, these factors can contribute to reentry wellbeing, through improved community participation and treatment engagement and reduced psychiatric distress and substance use.

We investigated associations between neighborhood walkability and physical activity using twins (5477 monozygotic and same-sex dizygotic pairs) as "quasi-experimental" controls of genetic and shared environment (familial) factors that would otherwise confound exposure-outcome associations. Walkability comprised intersection density, population density, and destination accessibility. Outcomes included self-reported weekly minutes of neighborhood walking and moderate-to-vigorous physical activity (MVPA) and days per week using transit services (eg, bus, commuter rail). There was a positive association between walkability and walking, which remained significant after controlling for familial and demographic factors: a 1% increase in walkability was associated with a 0.42% increase in neighborhood walking. There was a positive association between walkability and MVPA, which was not significant after considering familial and demographic factors. In twins with at least 1 day of transit use, a 1-unit increase in log (walkability) was associated with a 6.7% increase in transit use days; this was not significant after considering familial and demographic factors. However, higher walkability reduced the probability of no transit use by 32%, considering familial and demographic factors. Using a twin design to improve causal inference, walkability was associated with walking, whereas walkability and both MVPA and absolute transit use were confounded by familial and demographic factors.This article is part of a Special Collection on Environmental Epidemiology.

Application domains such as environmental health science, climate science, and geosciences—where the relationship between humans and the environment is studied—are constantly evolving and require innovative approaches in geospatial data analysis. Recent technological advancements have led to the proliferation of high-granularity geospatial data, enabling such domains but posing major challenges in managing vast datasets that have high spatiotemporal similarities. We introduce the Hierarchical Grid Partitioning (HierGP) framework to address this issue. Unlike conventional discrete global grid systems, HierGP dynamically adapts to the data’s inherent characteristics. At the core of our framework is the Map Point Reduction (MPR) algorithm, designed to aggregate and then collapse data points based on user-defined similarity criteria. This effectively reduces data volume while preserving essential information. The reduction process is particularly effective in handling environmental data from extensive geographical regions. We structure the data into a multilevel hierarchy from which a reduced representative dataset can be extracted. We compare the performance of HierGP against several state-of-the-art geospatial indexing algorithms and demonstrate that HierGP outperforms the existing approaches in terms of runtime, memory footprint, and scalability. We illustrate the benefits of the HierGP approach using two representative applications: analysis of over 289 million location samples from a registry of participants and efficient extraction of environmental data from large polygons. While the application demonstration in this work has focused on environmental health, the methodology of the HierGP framework can be extended to explore diverse geospatial analytics domains.

BackgroundAddressing key behavioral risk factors for chronic diseases, such as diet, requires innovative methods to objectively measure dietary patterns and their upstream determinants, notably the food environment. Although GIS techniques have pushed the boundaries by mapping food outlet availability, they often simplify food access dynamics to the vicinity of home addresses, possibly misclassifying neighborhood effects. Leveraging Google Location History Timeline (GLH) data offers a novel approach to assess long-term patterns of food outlet utilization at an individual level, providing insights into the relationship between food environment interactions, diet quality, and health outcomes.MethodsWe leveraged GLH data previously collected from a sub-set of participants in the Washington State Twin Registry (WSTR). GLH included more than 287 million location records from 357 participants. We developed methods to identify visits to food outlets using outlet-specific buffer zones applied to the InfoUSA data on food outlet locations. This methodology involved the application of minimum and maximum stay durations, along with revisit intervals. We calculated metrics from the GLH data to detect frequency of visits to different food outlet classifications (e.g. grocery stores, fast food, convenience stores) important to health. Several sensitivity analyses were conducted to examine the robustness of our food outlet metrics and to examine visits occurring within 1 and 2.5 km of residential locations.ResultsWe identified 156,405 specific food outlet visits for the 357 study participants. 60% were full-service restaurants, 15% limited-service restaurants, and 16% supermarkets. Mean visits per person per month to any food outlet was 12.795. Only 8, 10 and 11% of full-service restaurants, limited-service restaurants, and supermarkets, respectively, occurred within 1 km of residential locations.ConclusionsGLH data presents a novel method to assess individual-level food utilization behaviors.

The evidence linking urban greenspace to individual’s physical activity (PA) levels is mixed. This study examines relationships between street-level and satellite-derived greenspace measures with PA outcomes. Our sample included 7855 adult twins enrolled in the Washington State Twin Registry from 2009 to 2020 living in urban areas; 14,095 total survey observations were analyzed. We applied a deep learning segmentation algorithm to Google Street View images sampled from 100 m around residential addresses to quantify street-level greenspace. Bouts and duration of PA, including moderate to vigorous PA and neighborhood walking were self-reported. We applied mixed-effects linear regression models to determine relationships between greenspace measures and PA outcomes, overall and stratified by residential population density. Adjusted models included age, body mass index, sex, race, education, income, neighborhood deprivation, urban sprawl, and seasonality. A series of sequential models was constructed to test associations between various greenspace exposures and PA outcomes. Overall, we found no consistent associations between greenspace exposures and PA outcomes. We found that the summer normalized difference vegetation index was associated with an increase in moderate to vigorous PA in low population density areas, but this was not significant when controlling for seasonality. Both Google Street View and normalized difference vegetation index were associated with lower total walking for those residing in areas with high population density only. Findings highlight the importance of seasonality and the need to address where PA is actually done.

This study examined the association of rental assistance receipt with cost-related medication nonadherence (CRN) engagement in low-income adults with diabetes. Using National Health Interview Survey (NHIS) data from 2016 through 2019 and 2020 through 2022, we included low-income adults who were 1) diagnosed with diabetes, 2) prescribed medications, and 3) renters. Propensity score weighting approach created a sample in which receipt of rental assistance was independent of observed sociodemographic characteristics. Logistic regression examined the association of rental assistance receipt with CRN, respectively. Lack of receipt of rental assistance was significantly associated with higher odds of CRN engagement in NHIS 2016-2019 (Odds ratio=2.32; 95% confidence interval=(1.59, 3.37); p&lt;.0001) and NHIS 2020-2022 (Odds ratio=1.74; 95% confidence interval=(1.04, 2.91); p=.03). Given the shortage of affordable housing in the United States, findings suggest that expansion of affordable housing could be critical for improving health outcomes in low-income adults with diabetes.

BACKGROUND:Given limited data regarding the spatial epidemiology of overdose among women amid the current overdose crisis, we evaluated (1) changes in spatiotemporal clustering of overdose over time, (2) the association between residential proximity to overdose clusters and recent nonfatal overdose, and (3) the association between 'risk environment' features and residential proximity to overdose clusters. METHODS:Questionnaire data were from a merged community-based cohort of marginalized women who use drugs in Vancouver, Canada (09/2014-08/2022). Emerging hotspot analysis was used to classify residential proximity to spatiotemporal clusters of nonfatal overdose and kernel density estimation was used to visualize the spatiotemporal distribution of nonfatal overdose clustering over the 8-year study. Statistical analyses drew on bivariate and multivariable logistic regression using generalized estimating equations (GEE). FINDINGS:Over eight years, among 650 participants (3461 observations), 37·2 % experienced a nonfatal overdose at least once. Annual period prevalence of nonfatal overdose increased from 9·1 % in 2014-15 to 25·6 % in 2021-2022. The highest-density clusters were in Vancouver's Downtown Eastside/Strathcona neighborhoods, where clusters became larger and more dispersed from 2016-onwards. Residential proximity to overdose clusters was associated with higher odds of recent nonfatal overdose. 'Risk environment' features of unstable housing, unsafe sleeping environments, and physical violence were associated with elevated odds of residential proximity to overdose clusters. INTERPRETATION:Marginalized women face a high and rising burden of nonfatal overdose, which is influenced by the 'risk environments' in which they reside. Scale-up of geographically tailored overdose prevention services, harm reduction, and programs addressing violence and housing are needed.

Objective:Early detection of breast cancer through routine screening improves survival. Increased engagement with healthcare services may promote mammography uptake. This study examined whether greater telehealth use is associated with higher mammogram screening rates. Methods:We analyzed data from January 1st 2018 to December 31st 2023 from the MultiCare Health System in Washington State. For each year, we determined whether a patient received a screening mammogram during that year or the following year. We assessed the association between type of encounter (≥1 telehealth visit vs. in-person only vs. no encounters) and the likelihood of having a mammogram, adjusting for race/ethnicity, insurance type, age, and pre-/post-COVID-19 period. Results:Among 140,390 female patients (609,061 patient-year observations), those with no encounters were least likely to undergo mammography. Women who used telehealth were less likely to be screened than those with in-person visits but more likely than those with no visits (rural: Odds Ratio (OR) = 0.50, 95 % Confidence Interval (CI) = 0.43,0.60; urban: OR = 0.68, 95 % CI = 0.65,0.70). Telehealth rose from 0.04 % to 4 % post-COVID-19, while mammography rates increased from 0 % to an average of 13 %. Conclusions:Expanding telehealth access may increase routine mammography, particularly among previously unscreened populations.

INTRODUCTION:Research has focused on the built environment (e.g., neighborhood walkability) that supports or hinders physical activity because it is potentially modifiable. This study investigated the associations between changes in neighborhood walkability and changes in physical activity in an adult twin cohort. METHODS:Longitudinal data (2009-2020) from 7,439 identical and fraternal twins comprising 2,800 complete pairs from a community-based registry were analyzed. Participants were free of mobility limitations and resided at their current residential location for at least 1 year. A series of phenotypic (nongenetically informed) models were used to test the effect of walkability change on change in physical activity. These were re-estimated in a series of quasi-causal models by leveraging the genetically informed nature of the twin design to test the effect of walkability change on change in physical activity while controlling for genetic and shared environmental confounds. RESULTS:Change in neighborhood walkability was associated with change in neighborhood walking but not in moderate-to-vigorous physical activity, which held after controlling for genetic and shared environmental confounding, plus standard demographic covariates, length of follow-up, and moving status. A 1-unit increased change in neighborhood walkability was associated with a 2.7-minute increased change in neighborhood walking per week, independent of familial confounds and covariates. Moving to a neighborhood that is 5.5 units greater in walkability could increase neighborhood walking by about 15 minutes per week. CONCLUSIONS:This study supports a quasi-causal relationship between changes in neighborhood walkability and changes in neighborhood walking, extending previous cross-sectional findings in the same twin cohort by establishing temporality.

BACKGROUND:This study evaluates relationships among race, access to endoscopy services, and colorectal cancer (CRC) mortality in Washington state (WA). METHODS:We overlayed the locations of ambulatory endoscopy services with place of residence at time of death, using Department of Health data (2011-2018). We compared CRC mortality data within and outside a 10 ​km buffer from services. We used linear regression to assess the impact of distance and race on age at death while adjusting for gender and education level. RESULTS:Age at death: median 72.9y vs. 68.2y for white vs. non-white (p ​< ​0.001). The adjusted model showed that non-whites residing outside the buffer died 6.9y younger on average (p ​< ​0.001). Non-whites residing inside the buffer died 5.2y younger on average (p ​< ​0.001), and whites residing outside the buffer died 1.6y younger (p ​< ​0.001). We used heatmaps to geolocate death density. CONCLUSIONS:Results suggest that geographic access to endoscopy services disproportionately impacts non-whites in Washington. These data help identify communities which may benefit from improved access to alternative colorectal cancer screening methods.

ImportanceThe impact of cumulative exposure to neighborhood factors on psychosis, depression, and anxiety symptom severity prior to specialized services for psychosis is unknown.ObjectiveTo identify latent neighborhood profiles based on unique combinations of social, economic, and environmental factors, and validate profiles by examining differences in symptom severity among individuals with first episode psychosis (FEP).Design, Setting, and ParticipantsThis cohort study used neighborhood demographic data and health outcome data for US individuals with FEP receiving services between January 2017 and August 2022. Eligible participants were between ages 14 and 40 years and enrolled in a state-level coordinated specialty care network. A 2-step approach was used to characterize neighborhood profiles using census-tract data and link profiles to mental health outcomes. Data were analyzed March 2023 through October 2023.ExposuresEconomic and social determinants of health; housing conditions; land use; urbanization; walkability; access to transportation, outdoor space, groceries, and health care; health outcomes; and environmental exposure.Main Outcomes and MeasuresOutcomes were Community Assessment of Psychic Experiences 15-item, Patient Health Questionnaire 9-item, and Generalized Anxiety Disorder 7-item scale.ResultsThe total sample included 225 individuals aged 14 to 36 years (mean [SD] age, 20.7 [4.0] years; 152 men [69.1%]; 9 American Indian or Alaska Native [4.2%], 13 Asian or Pacific Islander [6.0%], 19 Black [8.9%], 118 White [55.1%]; 55 Hispanic ethnicity [26.2%]). Of the 3 distinct profiles identified, nearly half of participants (112 residents [49.8%]) lived in urban high-risk neighborhoods, 56 (24.9%) in urban low-risk neighborhoods, and 57 (25.3%) in rural neighborhoods. After controlling for individual characteristics, compared with individuals residing in rural neighborhoods, individuals residing in urban high-risk (mean estimate [SE], 0.17 [0.07]; P = .01) and urban low-risk neighborhoods (mean estimate [SE], 0.25 [0.12]; P = .04) presented with more severe psychotic symptoms. Individuals in urban high-risk neighborhoods reported more severe depression (mean estimate [SE], 1.97 [0.79]; P = .01) and anxiety (mean estimate [SE], 1.12 [0.53]; P = .04) than those in rural neighborhoods.Conclusions and RelevanceThis study found that in a cohort of individuals with FEP, baseline psychosis, depression, and anxiety symptom severity differed by distinct multidimensional neighborhood profiles that were associated with where individuals reside. Exploring the cumulative effect of neighborhood factors improves our understanding of social, economic, and environmental impacts on symptoms and psychosis risk which could potentially impact treatment outcomes.

ObjectiveMisinformation and substance use both increased substantially during the COVID-19 pandemic. This study examined potential links between misinformation beliefs and substance use among adults, along with the potential for media literacy to mitigate misinformation's influences on problematic use of widely available substances of misuse.MethodStructural equation modeling (SEM) was used to test a theoretical model of media literacy's effects on substance use, fully mediated by disinformation beliefs, with a nationally representative sample of U.S. adults recruited through a Qualtrics panel of adults using census-based quotas for geographic region, population density, ethnic diversity and gender (N = 1264). The sample was 51.5% male (N = 651); 46.7% female (N = 591); 1.1% nonbinary (N = 13); and 0.7% (N = 9) not reporting.ResultsMedia literacy for source of news positively associated with media literacy for content of news (b = 0.814, p < 0.001). Media literacy for content of news then positively associated with science media literacy (b = 0.192, p < 0.001). Science media literacy then negatively associated with disinformation beliefs (b = -0.586, p < 0.001), and COVID-19 disinformation beliefs associated with an increase in substance use (b = 0.466, p < 0.001). Disinformation beliefs also associated with alcohol and sleep medication co-use (odds = 1.956, p < 0.05).ConclusionsResults demonstrate media literacy's value for substance misuse prevention and effective public health messaging.

BACKGROUND:Outdoor physical activity (PA) is an important component of overall health; however, it is difficult to measure. Passively collected smartphone location data like Google Location History (GLH) present an opportunity to address this issue. OBJECTIVES:To evaluate the use of GLH data for measuring outdoor PA. METHODS:We collected GLH data for 357 individuals from the Washington State Twin Registry. We first summarized GLH measurements relevant to outdoor PA. Next, we compared accelerometer measurements to GLH classified PA for a subset of 25 participants who completed 2 weeks of global positioning system and accelerometer monitoring. Finally, we examined the association between GLH measured walking and obesity. RESULTS:Participants provided a mean (SD) average 52 (18.8) months of GLH time-activity data, which included a mean (SD) average of 2421 (1632) trips per participant. GLH measurements were classified as the following: 79,994 unique walking trips (11.6% of all trips), 564,558 (81.8%) trips in a passenger vehicle, 11,974 cycling trips (1.7%), and 890 running trips (0.1%). Sixty-two percent of these trips had location accuracy >80%. In the accelerometry evaluation, GLH walking trips had a corresponding mean vector magnitude of 3150 counts per minute, compared with 489 counts per minute for vehicle trips. In adjusted cross-sectional analyses, we observed an inverse association between both walking minutes and trips per month and the odds of being obese (odds ratio = 0.78; 95% CI, 0.60-0.96, and odds ratio = 0.91; 95% CI, 0.82-0.98, respectively). CONCLUSIONS:GLH data provide a novel method for measuring long-term, retrospective outdoor PA that can provide new opportunities for PA research.

Proximity to mental health services is a predictor of timely access to services. The present study sought to investigate whether travel time was associated with engagement in coordinated specialty care (CSC) for early psychosis, with specific attention to whether the interaction of travel time by race and ethnicity had differential impact. Data collected between 2019 and 2022 as part of the New Journeys evaluation, the CSC model in Washington State. This cross-sectional study included a sample of 225 service users with first episode psychosis (FEP) who had received services from New Journeys. Service users’ addresses, and the physical location of CSC were geocoded. Spatial proximity was calculated as travel time in minutes. Scheduled appointments, attendance and program status were captured monthly by clinicians as part of the New Journeys measurement battery. Proximity was significantly associated with the number of appointments scheduled and attended, and program status (graduation/completion and disengagement). Among Hispanic service users with spatial proximity further away from CSC (longer commutes) was associated with a lower likelihood of graduating/completing CSC compared to non-Hispanic service users (p = .04). Non-white services users had a higher risk of disengagement from CSC compared to white service users (p = .03); additionally, the effects of spatial proximity on disengagement were amplified for non-White service users (p = .03). Findings suggest that proximity is associated with program engagement and partially explains potential differences in program status among ethnoracial group.

AimA vital benefit of primary care access is the attenuation of racial and ethnic disparities in mortality rates. The purpose of this study was to examine the associations between race/ethnicity, access to primary care physicians (PCP), and premature mortality among decedents in Washington State between 2015 and 2017.MethodsData on registered deaths and supply of PCPs were obtained from the Washington State Department of Health and American Medical Association, respectively. Multilevel binary logistic regression models were applied to examine the association of access to PCP, race, and premature mortality.ResultsAverage distance to the closest five PCPs was 3 miles. American Indian/Alaska Natives would have had to travel on average 6 miles one-way to reach the nearest PCPs, while people of other racial and ethnic groups would have had to travel between 1 and 3 miles. The odds of premature mortality was 4% higher per unit increase in distance to the closest PCPs. The positive relationship between distance to PCPs and premature mortality was weaker for American Indian/Alaska Natives, stronger for White and Asian people, and not statistically significant for Black and Hispanic people.ConclusionIncreasing access to PCPs, as well as further exploration in understanding, identifying, and alleviating reasons that may lead racial and ethnic minorities to underutilize PCP services are warranted.