Abstract
When assessing epidemiological evidence for the purposes of classifying carcinogenicity, the International Agency for Research on Cancer (IARC, World Health Organisation, Lyon, France) adopt high scientific standards. The classification of carcinogenicity is perhaps open to question for a few agents evaluated in monographs 39–63, but clearly aligns with the evidence summarised for the great majority. Since monograph 38 in 1986, when the evidence on environmental tobacco smoke (ETS) and lung cancer was considered equivocal, IARC staff and associated scientists have published numerous papers which led them to their current view that the carcinogenicity of ETS is ‘well established.’ The content of these papers is critically examined and is found to contain numerous limitations, including failure to report results using standardised indices of ETS exposure, failure to show weakening of the association over time, failure to make it clear the association with lung cancer is only for spousal smoking and does not apply for workplace exposure, failure to investigate sources of between-study heterogeneity, failure to consider study quality adequately, failure to consider histological type, seriously inadequate consideration of sources of bias, overstatement of biologic plausibility, and inadequate consideration of proof of causation. The strength of the epidemiological evidence relating ETS to lung cancer appears to be less than that for all other agents classified by IARC as having ‘sufficient’ evidence of carcinogenicity, and is no stronger than that for various agents with a ‘limited’ classification. ‘Limited’ evidence of carcinogenicity would appear a more appropriate classification for ETS.