Dust storms and cardiorespiratory emergency department visits in three Southwestern United States: application of a monitoring-based exposure metric.

Climate change is projected to increase the risk of dust storms, particularly in subtropical dryland, including the southwestern US. Research on dust storm's health impacts in the US is limited and hindered by challenges in dust storm identification. This study assesses the potential link between dust storms and cardiorespiratory emergency department (ED) visits in the southwestern US. We acquired data for 2005-2016 from eight IMPROVE (Interagency Monitoring of PROtected Visual Environments) sites in Arizona, California, and Utah. We applied a validated algorithm to identify dust storm days at each site. We acquired patient-level ED visit data from state agencies and ascertained visits for respiratory, cardiovascular, and cause-specific subgroups among patients residing in ZIP codes within 50 km of an IMPROVE site. Using a case-crossover design, we estimated short-term associations of ED visits and dust storms, controlling for temporally varying covariates. During 2005-2016, 40 dust storm days occurred at the eight IMPROVE sites. Mean PM 10 and PM 2.5 levels were three to six times greater on dust storm days compared to non-dust storm days. Over the study period, there were 2 524 259 respiratory and 2 805 925 cardiovascular ED visits. At lags of 1, 2, and 3 days after a dust storm, we observed 3.7% (95% CI: 1.0%, 7.6%), 4.9% (95% CI: 1.1%, 8.9%), and 5.0% (95% CI: 1.3%, 8.9%) elevated odds of respiratory ED visits compared to non-dust storm days. Estimated associations of dust storm days and cardiovascular disease ED visits were largely consistent with the null. Using a monitoring-based exposure metric, we observed associations among dust storms and respiratory ED visits. The results add to growing evidence of the health threat posed by dust storms. The dust storm metric was limited by lack of daily data; future research should consider information from satellite and numerical models to enhance dust storm characterization.