Do risk assessments justify banning chrysotile or not?

An asbestos ban today implies replacing high-density chrysotile products by technically equivalent substitutes of unknown but suspected carcinogenicit y. Deciding to ban asbestos should thus rely on a comparison of the risks o f chrysotile and the risks of alternative products. No asbestos ban to date has been based on an effective comparative risk assessment. If deciders ov erestimate the risks of chrysotile while downplaying the risks of its subst itutes, a ban may not improve public health safety. Past quantitative risk assessments (QRAs) of asbestos have targeted risks a ssociated with asbestos exposures to extant products "containing small amou nts of crocidolite and substantial quantities of amosite" (cf. EPA, 1986). So past QRAs pooled studies of workers exposed to mixed asbestos fibers (av eraging 25-30% amphiboles), resulting in a tendency to overestimate the ris ks of pure chrysotile exposures. Over the last 15 years, updates of these s tudies and new occupational and environmental studies have corroborated the hypothesis that chrysotile is significantly less potent than amphiboles. T his new evidence has not been integrated in recent QRAs. The implied extra safety margins for chrysotile are protective when establishing hygiene stan dards. Substitution is a different ball game. Substitution requires more realism or "impartiality" to ensure that existin g risks will not be replaced inadvertently by larger ones. Only a formal co mparative QRA can support the substitution of chrysotile. It is proposed th at asbestos QRAs be updated methodologically and data-wise, that they focus on today's chrysotile products, that uncertainty be formally accounted for , and that the risks of chrysotile be compared impartially with the risks o f its substitutes, applying equal precaution to both parts of the compariso n. A comparative assessment should incorporate qualitative factors formally , perhaps with a bayesian approach.