Association between asthma hospitalizations and ozone level in Maricopa County from 2007 through 2012 using time-series analysis

Ozone is and environmental pollutants that can have adverse health effects. However, evaluating these effects is not always a straightforward process because used methods should account for potential confounders including -for example- seasonality of data recorded over time and the fact that this effect may persist after the time of the exposure.  The objective of this study was to evaluate the association with ozone concentration from 2007 through 2012. The study data consisted of daily hospital discharge data (HDD) and hourly meteorological measurements obtained from MC’s Department of Public Health and Department of Air Quality.  A GIS program was used to geocode hospitalizations records and to assign them exposure recordings based on admission date and location. We used time plots and distributed-lag non-linear model (DLNM) to achieve these objectives while accounting for some potential confounders including temperature and day of the week.

 Results: A total of 90,381 asthma hospitalizations were retrieved from the dataset (daily median=39, range: 8-122). Asthma hospitalizations were highest in 2008 (16,949), from November through December, and lowest in 2011 (13213) and in June-July. By contrast, the average daily ozone concentration ranged from 27.05 in 2012 to 30.15 in 2008 and from 13.96 in December to 40.58 in May.  The association between asthma hospitalizations [relative risk (RR/per 10 unit increase of ozone)] start at ~ 1.025; 95%CI(1.006 – 1.046) at lag 0 and diminishes by lag 7.

Our findings suggest exposure to increased ozone level is associated with increased RR of asthma hospitalizations in Maricopa County that last for days after the exposure. This study used recently developed methods that are freely available and could be used to evaluate other health events that are measured over time.