Dust Storms and Respiratory Emergency Department Visits in the Southwestern United States

Abstract

BACKGROUND AND AIM: Climate change is projected to increase the risk of dust storms, particularly in the southwestern US. Research on dust storm impacts is hindered by challenges in dust storm identification. This study assesses the potential link between dust storms and respiratory-related emergency department (ED) visits in the southwestern US. METHODS: We acquired data for 2005-2016 from IMPROVE (Interagency Monitoring of PROtected Environments) monitoring sites in California (n=2), Arizona (n=5), Nevada (n=2), and Utah (n=1). Applying a validated algorithm that identifies dust events of natural origin, dust storm days at each IMPROVE site were identified. We acquired patient-level ED visit data from state agencies and selected visits for respiratory diseases, asthma, and chronic obstructive pulmonary disease (COPD) among patients residing within 50 km of a monitoring site. Using a case-crossover design, we estimated associations of ED visits and dust storms, controlling for meteorology, ozone, and holidays. RESULTS: During 2005-2016, 37 dust storm days occurred at the 10 IMPROVE sites. Mean PM10 and PM2.5 levels were 3-6 times greater on dust compared to non-dust storm days. We observed associations of respiratory ED visits and dust storms, with strongest associations at lag 3 [odds ratio of 1.049 (95% confidence interval: 1.002-1.098)]. Results were driven by Arizona, with an odds ratio for respiratory ED visits of 1.068 (95% confidence interval: 1.008-1.133). Associations for asthma were similar, although with larger uncertainty given fewer ED visits. Associations for COPD were consistent with the null. CONCLUSIONS: Using a monitoring-based exposure metric, we observed associations among respiratory ED visits and dust storms. The results add to growing evidence of the health threat posed by dust storms. The dust storm metric was limited by lack of daily monitoring data; future research will incorporate satellite and models for enhanced dust storm characterization. KEYWORDS: Climate, Dust, PM, Morbidity