![[ Visit Client Website ]](images/banner.png)
BACKGROUND: There are relatively few U.S. data addressing asthma in rural versus urban areas. Poor geospatial resolution (such as county-level data for urban areas) may leave important drivers of asthma prevalence and health care utilization obscured. To examine factors related to geography that extend beyond traditional descriptors of disparity, this analysis examines the association between asthma emergency department (ED) visit rates in Wisconsin and rural-urban geography based on zip code tabulation areas (ZCTA), accounting for prevalence and potential explanatory factors.
METHODS: We examined rural-urban status using ED visits for asthma and BRFSS data for 2009-2011. Each ZCTA was classified into six rural-urban (R/U) categories based on census data for size and population density. Average annual population-based ED visit rates, risk-based rates for persons with asthma and risk ratios were calculated separately for total, child and adult populations. Decomposed crude rates showed the contribution of each R/U category to the overall state rate. We also modeled statewide rates of asthma ED visits by ZCTA using multivariable analysis of covariance to analyze mean annual (2010) asthma ED rates by R/U category, while adjusting for distance from hospital and aggregate census data for racial makeup, poverty, gender, age and employment in agriculture.
RESULTS: Population-based rates for both children and adults in the most urban area, Milwaukee (123.6/10,000 and 65.9/10,000), were more than three times higher than in the most rural area, R1 (33.1/10,000 and 19.1/10,000). Urban non-Milwaukee areas had rates 30-40% higher than R1. Adjusting for higher prevalence in urban areas explained some of the difference. The R/U effect differed between children and adults in risk-based but not population-based rates. In univariate analysis, the R/U difference in mean ED rates was significant (P<0.0001), with Milwaukee having the highest mean and no significant difference between R1 and urban non-Milwaukee areas (P=0.752). In multivariate models, the R/U difference in adjusted means remained significant (P=0.002), indicating that if areas had similar distributions of covariates, then R1 would actually have higher rates than urban non-Milwaukee areas (P=0.01). Other variables associated with higher ED rates included minority populations, poverty and shorter distance to treatment.
CONCLUSIONS: In Wisconsin, asthma health care utilization varies significantly by R/U classification. Disparities persist in Milwaukee even after adjusting for race and poverty. As such, new education and healthcare interventions focused on other factors present in urban areas (e.g., outdoor air quality, exposure to indoor asthma triggers, differences in access to care or quality of care received, comorbidities, etc.) may be warranted.