Table 2.

Exposure assessment bias

	Exposure Metric	Risk of exposure assessment bias
A	Exposure as a surrogate for air pollution (e.g. distance to road, personal recall)	High
A	Use of single fixed monitoring system	High
A	Use of multiple fixed monitoring systems	≥3 Moderate
		< 3 High
B	Personal exposure	Low
B	Atmospheric dispersion models	Moderate
B	Land use regression models	Moderate
B	Satellite/remote sensing	Moderate

Combination of > 2 of the above (providing at least one B measure) = low risk of bias