Scientific Evidence Portal
Reanalysis of epidemiological evidence on lung cancer and passive smoking | J B Copas, J Q Shi
Article Published: 2000/02/12
Funding Source: Economic and Social Research Council. Competing interest: None declared.
Published By: BMJ Vol. 320
Further Information A re-examination of the “mountain of evidence” that proves nothing, that on passive smoking, indicates that a strong publication bias has been used in the studies that examined the effects of passive smoking on human health. What does that mean? In simple terms, it means that the belief of the researchers (that passive smoking must be harmful to health) got in the way of the research, and thus of its results.
This study (not financed by the tobacco industry) has recomputed the figures of the “mountain of evidence”, and the results are predictable to all those who understand antismoking ideology: the already contradictory, uncertain and tiny risk elevations, calculated on the basis of methodologically defective interviews on vague memory recalls of unverifiable exposures — products of studies that do not meet any basic epidemiological standard — simply vanish into a virtual zero.
These are the key points of this analysis:
A systematic review of epidemiological studies on passive smoking estimated the increased risk of lung cancer as 24%. (Please note that 24% is already orders of magnitude below the
200% risk elevation which constitutes the smallest risk level which can potentially be suggested with this type of study. Even a 200% elevation indication is highly dubious, as are all results from such statistical studies on lifestyle factors, which must always be reviewed with scientific scepticism. Pretension to decimal accuracy is cosmically ludicrous, but as noted by this analysis, the meaningless 24% becomes a " strong epidemiological evidence for an association between lung cancer and passive smoking". This is ideology, an enemy of science, talking. This analysis is itself insufficiently sceptical but shows the foolishness of pretension to impossible precision in shining but a single ray from a standpoint of critical perspective. There is clear evidence of publication bias in these studies.
Reanalysis of the data allowing for the possibility of publication bias substantially lowers the estimate of relative risk.