PHW251 Fall 2023 Project- Group 19

Problem Statement

Asthma remains a significant health concern in California, affecting approximately 5.2 million residents. In 2014, 13.8% of adults and 13.7% of children had a lifetime asthma diagnosis, with 8.1% of adults and 9.4% of children currently experiencing asthma [1] [2]. The Children’s Health Study, funded by California Air Resources Board (CARB), revealed a correlation between elevated PM2.5, ozone, carbon monoxide, NO2, and NOx exposure and increased emergency department visits and hospital admissions in children [3]. Disparities exist, with higher asthma prevalence among Blacks and AI/AN populations [4]. Hispanic and African American children, particularly those without private insurance, often reside in areas with heightened traffic-related air pollution [5].

In our study, we aim to delve into the intricate relationship between environmental exposures, socio-demographic factors, and asthma-related emergency department visits in California. By understanding these associations, we aspire to identify vulnerable populations and develop targeted interventions, thereby reducing the overall burden of asthma and fostering more equitable healthcare outcomes across diverse communities.

Methods

Three publicly available datasets served as the primary source data for this project. Datasets are accessible through the California Office of Environmental Health Hazard Assessment (OEHHA) and the California Health and Human Services Agency (CalHHS). 

The dataset, CalEnviroScreen 4.0, contains 50 columns including detailed environmental measures reported by California census tracts. The second dataset reports an OEHHA developed score, Cal Enviro Screen (CES). The objective for this scoring system is to recognize and systematically assess the impact of the complex factors that contribute to population health. The dataset contains 15 covariates reported by census tracts. CES 4.0 score is a composite of 21 indicators linked with that range from air and water pollutants to social determinants of health such as poverty and unemployment. OEHHA measures and scoring are from 2021. The CalHHS dataset includes county level hospital utilization data from 2015-2020. Our work focuses on the age-adjusted rate for asthma emergency department (ED) visits/rates for all age groups, stratified by the five race and/or ethnicity categories reflected in these datasets.

In the first stage of data cleaning, variable names were standardized, NA values were removed, numeric values rounded, observations were deduplicated, and census tract observations were grouped by “county.” For each county, the median CES 4.0 Score was calculated for composite census tracts. Raw population counts per county for “race/ethnicity” were calculated using census tract level variables (total population and percentage distributions for race/ethnicity). Raw scores for PM 2.5 informed a new variable “pm2.5_level.” The three datasets were then joined by county name to form a single dataset with only relevant covariates selected.

Results

Interactive Bubble Chart showing the relationship between Annual PM 2.5 concentration and Asthma Emergency Department Visit Rates by California County

Interpretation: The data from 29 California counties in 2021 doesn’t show a clear relationship between Annual Particulate Matter 2.5 concentration, population size, and Asthma Emergency Department Visit rates. This lack of a definitive relationship suggests the possibility of confounding factors, including socio-demographic and environmental elements, contributing to high visit rates in certain counties despite small population size and low PM 2.5 concentration.

Interactive Table Showing Environmental and Social Data of California Counties.

Interpretation: This interactive chart shows data from all 58 California counties including average PM 2.5 levels, median CES score, and average housing burden. PM 2.5 levels were defined as Good or Moderate based on EPA standards [6], CES scores were defined as Low (0-30), Moderate (30-50), and High (50+) based on data from CDPH [7], and housing burden levels were defined as low (0-12%), medium (12-16%), and high (16+%), with cutoff points based on information from The Justice Gap [8].

Interactive Table Showing Demographic Data of California Counties.

Interpretation: This interactive chart shows data from all 58 California counties on racial breakdowns of each county. This table can be compared to table 1 to review any associations between environmental and social factors in different counties and racial breakdown.

Faceted Scatterplots showing Associations between the Percentage Population of Six Racial Categories and Average CES 4.0 Scores by County-level Data.

Interpretation: These scatter plots map the Mean CES Score on the x-axis and percentage of county-level population data by race/ethnicity category, as indicated on the y-axis and title for each plot. These six plots allow comparisons to be made between the x and y covariates, and CalEnviroScreen is the source for data categories listed for race/ethnicity. “Unknown Population,” used here, is analogous to the CalEnviroScreen variable name “other_multiple.” A detailed explanation of CES Score components can be found here: https://oehha.ca.gov/calenviroscreen/scoring-model

Discussion

The results from the analysis do not allow members to infer associations between the Asthma ED rates and the variables under study. In the interactive bubble chart, which assessed asthma ED visit rates, mean PM 2.5 and county, we do not see a clear linear relationship between asthma ED rates and mean PM 2.5 levels. This lack of a definitive relationship suggests the possibility of confounding factors, including sociodemographic and environmental elements, contributing to high visit rates in certain counties despite small population size and low PM 2.5 concentration.

However, the interactive tables and scatter plots show more definitive relationships between race, CES score, PM 2.5 level, and housing burden. In particular, we see percentage of Hispanic population increasing strongly as median CES score increases per the faceted scatter plot. If we review the information in tables, the three counties with the highest Hispanic population all have high mean housing burden and high median CES score. One has moderate mean PM 2.5 level and the others have good mean PM 2.5 level. However, there was not much variation of the mean PM 2.5 level within the data set overall.

We recommend that further research explore the association between race and asthma ED visit rates per county. The data provided had information on asthma ED visits by race and age, but did not provide as much information about racial breakdown per county and ED access within different counties. A limitation of this report is that there may be populations of individuals who do not go to the emergency department when facing a medical issue due to lack of insurance, residency status, and/ or proximity to a hospital, compared with the population of individuals who can access the emergency department easily.

Data Dictionary

Data Dictionary: Environment and Asthma in California
Variable Name Description
Variable Name California county that the census tract falls within
Variable Name Annual mean PM 2.5 concentrations
Variable Name Percent housing burdened low income households
PM 2.5 Level Rating of average PM 2.5 level per CA county based on mean county value compared to EPA pm 2.5 level ratings
CES 4.0 Level Category of median CES 4.0 level per CA county based on CDPH groupings
Housing Burden Level Category of mean percent housing burdened low income households based on cutoff points set by The Justice Gap
Ai/An visits Asthma Emergency visit rates per 10,000 Native American residents
Asian visits Asthma Emergency visit rates per 10,000 Asian residents
Black visits Asthma Emergency visit rates per 10,000 African American residents
Hispanic visits Asthma Emergency visit rates per 10,000 Hispanic residents
White visits Asthma Emergency visit rates per 10,000 White residents
Age adjusted ed visit rate Age Adjusted Asthma Emergency Department Visit Rate for each County
CES 4.0 Score CalEnviroScreen Score, Pollution Score multiplied by Population Characteristics Score
California County California county that the census tract falls within
Total Population 2019 ACS population estimates in census tracts
Hispanic (%) 2019 ACS population estimates of the percent per county of those who identify as Hispanic or Latino
White (%) 2019 ACS population estimates of the percent per county of those who identify as non-Hispanic white
African American (%) 2019 ACS population estimates of the percent per county of those who identify as non-Hispanic African American or black
Native American (%) 2019 ACS population estimates of the percent per county of those who identify as non-Hispanic Native American
Asian American (%) 2019 ACS population estimates of the percent per county of those who identify as non-Hispanic Asian or Pacific Islander
Other/Multiple (%) 2019 ACS population estimates of the percent per county of those who identify as non-Hispanic "other" or as multiple races
median_ces_4_0_score median of the census tract level CalEnviroScreen Score, also called CES 4.0 Score, for each county

References

  1. The Burden of Asthma in California: A Surveillance Report - Public Health Institute
  2. Asthma Prevalence in California: A Surveillance Report
  3. https://ww2.arb.ca.gov/sites/default/files/classic/research/apr/past/94-331c.pdf
  4. https://www.cdph.ca.gov/Programs/CCDPHP/DEODC/EHIB/CPE/CDPH%20Document%20Library/Asthma_Surveillance_in_CA_Report_2017.pdf
  5. https://www.cdc.gov/asthma/stateprofiles/asthma_in_ca.pdf 
  6. https://www.epa.gov/sites/default/files/2016-04/documents/2012_aqi_factsheet.pdf
  7. https://oehha.ca.gov/calenviroscreen/report/calenviroscreen-40
  8. https://justicegap.lsc.gov/resource/section-2-todays-low-income-america/