© 2004 by Oxford University Press
© 2004 Oxford University Press
CORRESPONDENCE |
Re: Extended Follow-up of a Cohort of British Chemical Workers Exposed to Formaldehyde
Correspondence to: Morris Greenberg, MB, FRCP, FFOM, 74 North End Road, London NW11 7SY, U.K. (e-mail: gmgreenberg{at}macunlimited.net)
Recently, Coggon et al. presented data from an extended follow-up of a cohort of British chemical workers exposed to formaldehyde (1). In 1984, Donald Acheson and Martin Gardner initiated the mortality study and circulated a prepublication draft of their report (2) to a peer group of epidemiologists and invited them to discuss it at the Medical Research Council’s Environmental Epidemiology Unit in Southampton. During the discussion, the important question was raised as to the most suitable population to use for comparison.
Although it may seem obvious to compare the rates derived from the group under investigation with those obtained from the local population for cancer mortality and morbidity studies, members of the audience queried whether it was appropriate for this study (2). By then, members of the audience were aware of a number of studies in which a comparison using local rates had been erroneous. One type of error arose in the study of mortality in a group of workers at a factory where antimony smelting took place (3). The study reported an excess of lung cancer deaths when national rates were used for comparison, which persisted when regional rates were used. A "shadow" study comparing the rate of cancer mortality among the study group with the rates derived from an even more restricted geographical unit effectively eradicated this excess (unpublished data) because the overlap between the study and the control populations was so great that, in effect, the subjects were being compared with themselves, and in consequence no excess mortality was to be expected.
Another complication in interpretation arises when most workers in the catchment area have been exposed to the same permutations and combinations of carcinogenic agents, such as when, for historic or economic reasons, noxious industries are juxtaposed. By comparing the standardized lung cancer rate by age and sex of the study group with the local rates, the magnitude of the cancer excess calculated will be an underestimate of the true risk.
For the epidemiologist studying the burden of common tumors such as bronchial carcinoma in a working-class population, there is often a need to use imperfect historic data, which were neither collected nor preserved for retrieval and analysis, and an awareness that there might be important confounding factors to consider and adjust for. Consequently, the epidemiologist has the option of making comparisons with national rates adjusted for social class. Coggon and his colleagues might care to discuss the merits of their comparing the study group with local geographic variations rather than social class-adjusted national rates.
Exposures that led to an excess mortality of less than 2% (for example, in the case of ionizing radiation) were once considered acceptable. With such institutions as the Royal Society categorizing "acceptability" of risk as an excess mortality of less than 1 in 100 000, researchers and epidemiologists should declare when the ability to detect such levels with confidence for bronchial carcinoma is beyond the limits of methodology. Although such frankness would put severe constraints on the "selling" of epidemiologic studies to workforces or their representatives and on the research ethics committees to whom study proposals are made, it is important that individuals involved in formulating public health policy be informed of the limitations of study population sizes and of methodologies to measure such low-order risks. Coggon et al. (1) concluded that further follow-up of industrial cohorts, particularly those with relatively high levels of exposure to formaldehyde, may help resolve the outstanding uncertainties. What industry population size did they have in mind that would permit the calculation with confidence of the orders of risk at the exposure levels experienced occupationally and domestically?
REFERENCES
1 Coggon D, Harris EC, Poole J, Palmer KT. Extended follow-up of a cohort of British chemical workers exposed to formaldehyde. J Natl Cancer Inst 2003;95:1608–15.
2 Acheson ED, Barnes HR, Gardner MJ, Osmond C, Pannett B, Taylor CP. Formaldehyde in the British chemical industry. An occupational cohort study. Lancet 1984;1:611–6.[Web of Science][Medline]
3 Jones RD. Survey of antimony workers; mortality 1961-1992. Occup Environ Med 1994;51:772–6.
Correspondence about this Article
CiteULike 
Connotea 
Del.icio.us What's this?
J Natl Cancer Inst 2004 96: 1037-1038.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||