Analysis and Methods
Goal
Estimate the number of children at risk for hospitalization due to asthma.
Method
The researchers in this study used regression analysis to calculate the association of hospitalized children due to an asthma-related event and surrounding environmental determinants. The final model was created by combining CalEnvironScreen's pollution burden score, population characteristics score, and cumulative disadvantage of social and environmental risks at the zip code level, with California’s Office of Statewide Health Planning and Development's Hospital discharge data for children under 14 years of age.
3 Models
Model 0: Examined if asthma hospitalizations significantly varied at the zip code level after adjusting for individual-level risk factors (i.e. asthma, cardiovascular disease and low birth weights).
Model 1: Introduced the Composite CalEnviroScreen Score (CES).They found that CES had a positive effect on asthma hospitalizations, with a coefficient (b) of 0.016 and statistical significance (p < 0.001). For every unit increase in CES, the Risk Ratio (RR) for mean hospitalizations was expected to increase by 1.016 (with a confidence interval of 1.014 to 1.018, statistically significant and positive) times the mean rate- an expected 9 additional hospitalizations per 10,000 individuals in the population at risk.
Model 2: Omitted the Composite CalEnviroScreen Score and instead used the Pollution Burden Score (PBS) and the Population Characteristics Score (PCS). The effect of PCS (b = 0.103, p = 0.009) was more than twice as large as the effect of PBS (b = 0.016, p = 0.008). This model suggests that population characteristics (like health and socioeconomic factors) may play a bigger role in the outcome than pollution exposure alone.
Model 3: Represented all of the environmental pollutant and population characteristic indicators that remained significant and omitted individual-level risk factors to generalize a population average. This model found among the pollutant indicators of the PBS, Diesel Particulate Matter (DPM) (b = 0.002, p < 0.001) was the only environmental pollutant that remained significant after accounting for other variables. This finding suggests with an increasing level of DPM, an increased risk of asthma hospitalizations at the zip code level can be expected. In regards to the PCS in this model, increasing levels of linguistic isolation were significantly associated with lower asthma hospitalizations at the zip code level (b = -0.013, p = 0.002), while increasing levels of poverty (b = 0.003, p < 0.001) and racial/ethnic minorities (b = 0.011, p < 0.001) were associated with higher asthma hospitalizations at the zip code level.
Variables
Major findings
Boys (53%) used hospital services more than girls did (47%)
Youngest children (43%) used the majority of hospital services, children aged 5 to 9 (30%) and children aged 10 to 14 (27%) followed.
White (33%) and Latino (31%) populations had the majority of hospital discharges, while African Americans(11%) had the smallest proportion.
African-American children had more than three times the risk for hospitalization than white children (14 vs. 4 per 10,000, respectively)
Children on Medi-Cal were about three times at risk for hospitalization than privately insured children (10 vs. 3 per 10,000, respectively).
Children with private insurance had a higher discharge rate (55%) compared to those covered by Medi-Cal (45%).
There are significant differences in mean rates of asthma hospitalizations between zip codes, even after accounting for individual-level characteristics, particularly for those with high concentrations of diesel particulate matter, poverty, and racial/ethnic minorities.
Higher levels of linguistic isolation at the zip code level were associated with reduced asthma hospitalizations yet clustering of immigrants, with fewer educational attainment opportunities and households with non-English-speaking adults may explain this effect.
Communities with higher poverty levels and racial/ethnic segregation had increased rates of hospitalizations.
Social factors, both at the individual and community levels, played a significant role in asthma morbidity, even after adjusting for environmental pollutants.
The geographical distribution of diesel particulate matter was associated with increased mean rates of asthma hospitalizations, with data suggesting the relationship between diesel particulate matter and asthma hospitalization is strongest in zip codes near highway 99 in the San Joaquin Valley.
The study found that the CalEnviroScreen effectively identifies communities with high asthma hospitalization rates, particularly for those exposed to high levels of diesel particulate matter, poverty, and who are racial/ethnic minorities.
Gaps
The study faced limitations in linking CalEnviroScreen data to hospital records, with only 47% of the data successfully linked. This may have biased the results, as the included zip codes have higher CES values, potentially overlooking lower CES scores.
The data were limited in accounting for in-home asthma triggers such as smoking and mold.
There is a possibility that linguistically isolated communities may experience underreporting or misdiagnosis of asthma, either because symptoms are not recognized or medical care is not sought out due to language or cultural barriers.
The total score of the CalEnviroScreen does not capture the specific historical, contextual, systemic, and other environmental risks present at the local level. This is because different regions within California have distinct profiles shaped by varying and changing environmental pollutants and populations.