The impact of measurement error in modeled ambient particles exposures on health effect estimates in multilevel analysis: A simulation study

Various spatiotemporal models have been proposed for predicting ambient particulate exposure for inclusion in epidemiological analyses. We investigated the effect of measurement error in the prediction of particulate matter with diameter <10 µm (PM(10)) and <2.5 µm (PM(2.5)) concentrations on the estimation of health effects. METHODS: We sampled 1,000 small administrative areas in London, United Kingdom, and simulated the “true” underlying daily exposure surfaces for PM(10) and PM(2.5) for 2009–2013 incorporating temporal variation and spatial covariance informed by the extensive London monitoring network. We added measurement error assessed by comparing measurements at fixed sites and predictions from spatiotemporal land-use regression (LUR) models; dispersion models; models using satellite data and applying machine learning algorithms; and combinations of these methods through generalized additive models. Two health outcomes were simulated to assess whether the bias varies with the effect size. We applied multilevel Poisson regression to simultaneously model the effect of long- and short-term pollutant exposure. For each scenario, we ran 1,000 simulations to assess measurement error impact on health effect estimation. RESULTS: For long-term exposure to particles, we observed bias toward the null, except for traffic PM(2.5) for which only LUR underestimated the effect. For short-term exposure, results were variable between exposure models and bias ranged from −11% (underestimate) to 20% (overestimate) for PM(10) and of −20% to 17% for PM(2.5). Integration of models performed best in almost all cases. CONCLUSIONS: No single exposure model performed optimally across scenarios. In most cases, measurement error resulted in attenuation of the effect estimate.