Beyond point estimates in Air Quality Health Indices: A severity-informed analysis of pollutant risk weights

Abstract

Air Quality Health Indices (AQHIs) translate epidemiological concentration–response functions into composite risk metrics for public health communication. Recent proposals grounded in meta-analytic evidence have advanced the construction of globally harmonized indices by deriving pollutant weights from excess mortality risks at World Health Organization guideline concentrations. However, these approaches implicitly treat point estimates as exact quantities, overlooking the evidential uncertainty surrounding concentration–response parameters. In this commentary, we examine the implications of this assumption using the severity framework of Mayo and Spanos. We show that conventional point-estimate–based weighting structures embed a form of false precision that systematically exaggerates inter-pollutant risk differentials. By deriving severity-informed minimum warranted excess risks, we construct alternative weight ratios that reflect the magnitude of differences supported by the data rather than their nominal estimates. Numerical application to WHO guideline concentrations demonstrates that gaseous pollutant contributions are consistently underweighted when uncertainty is ignored. Importantly, this critique does not challenge the underlying epidemiological associations but distinguishes statistical significance from evidentially warranted magnitude. Incorporating severity-based reasoning into AQHI construction yields weighting schemes that are more conservative, epistemically defensible, and better aligned with the inferential limits of meta-analytic evidence. These findings suggest that future composite health indices should explicitly propagate parameter uncertainty to avoid overstating pollutant risk contrasts in public-facing tools.