Evaluating the sensitivity of mortality attributable to pollution to modeling Choices: A case study for Colorado

Abstract

We evaluated the sensitivity of estimated PM_2.5 and NO₂ health impacts to varying key input parameters and assumptions including: 1) the spatial scale at which impacts are estimated, 2) using either a single concentration-response function (CRF) or using racial/ethnic group specific CRFs from the same epidemiologic study, 3) assigning exposure to residents based on home, instead of home and work locations for the state of Colorado. We found that the spatial scale of the analysis influences the magnitude of NO₂, but not PM_2.5, attributable deaths. Using county-level predictions instead of 1 km² predictions of NO₂ resulted in a lower estimate of mortality attributable to NO₂ by ∼ 50 % for all of Colorado for each year between 2000 and 2020. Using an all-population CRF instead of racial/ethnic group specific CRFs results in a 130 % higher estimate of annual mortality attributable for the white population and a 40 % and 80 % lower estimate of mortality attributable to PM_2.5 for Black and Hispanic residents, respectively. Using racial/ethnic group specific CRFs did not result in a different estimation of NO₂ attributable mortality for white residents, but led to ∼ 50 % lower estimates of mortality for Black residents, and 290 % lower estimate for Hispanic residents. Using NO₂ based on home instead of home and workplace locations results in a smaller estimate of annual mortality attributable to NO₂ for all of Colorado by 2 % each year and 0.3 % for PM_2.5. Our results should be interpreted as an exercise to make methodological recommendations for future health impact assessments of pollution.