Comparing apples to oranges: Interpreting ozone concentrations from observational studies in the context of the United States ozone regulatory standard

In 2015, the United States Environmental Protection Agency (US EPA) set the ozone National Ambient Air Quality Standards (NAAQS) at 0.070 parts per million (ppm), for an annual 4th highest daily 8-hour (h) maximum average concentration, averaged over three years, with compliance based on the monitor with the highest concentrations. Numerous epidemiological studies have evaluated associations between ozone and health effects, but how the ozone concentrations derived from those studies can be compared to the ozone NAAQS is not clear, because of the complexity of the standard. The purpose of the present work was to determine how ozone summary metrics used in key epidemiology studies compare to the metrics that comprise the ozone regulatory value. Evaluation of epidemiology studies used for quantitative risk assessment in the 2015 ozone NAAQS review demonstrated that the most commonly used summary metrics that differed from the NAAQS were: 1-h maximum or 24-h average concentrations; multiple-day averages from 2 to 30 days; and averaging of ozone concentrations across all monitors in an area and over different months of the year. Using different ozone summary metrics to calculate the ozone regulatory value in twelve US cities for 2000-2002 or 2013-2015 generated alternative ozone regulatory values that were often substantively different and that may or may not vary commensurate with the regulatory standard. Comparison of epidemiology study metrics to other countries' ozone standards or guideline levels produces similar challenges as described here for the NAAQS. In conclusion, many of the ozone concentration metrics used in epidemiology studies cannot be directly compared to the ozone NAAQS, and using simple conversion ratios adds substantial uncertainty to concentration estimates. These summary metrics must be reconciled to the regulatory value before any judgements are made as to the protectiveness of current and alternative standards based on epidemiology study results.