World health organization (WHO) consensus questionnaire on validity of methods for assessment of carcinogenicity of man-made fibers.

In May 1992, the WHO European Office held an expert consultation on the assessment of carcinogenicity of man-made fibers. WHO decided to determine whether the conclusions of such an expert consultation would be supported by scientific opinion internationally. For this experiment, WHO sent the 28 conclusions of the consultation, as a Consensus Questionnaire, to scientists worldwide seeking agreement, or otherwise, with the conclusions. Two hundred scientists responded. For every conclusion, at least 70% of respondents agreed, and for over half the conclusions more than 90% agreed. For some conclusions there were significant differences by country of respondent and/or by work environment. Respondents from the U.S.A., or in Government employ, tended to agree with the questions less often, and those from Germany, or from Industry, most often. The disagreements were most often related to questions concerned with the models for assessing carcinogenicity, but in every case the disagreements were expressed by fewer than one in three of respondents. There was minority opinion that the inhalation model was not adequately sensitive for assessing human hazards and that the intraperitoneal test could give validated answers on the carcinogenicity of fibers. Three questions demonstrated significant variability in response rate both from country to country and by work environment group. Each of the original WHO conclusions was concerned with shortcomings of the intraperitoneal test, when used for carcinogenicity classification. For these questions, the overall agreement rates were 85 to 91%. Respondents from U.S.A. and from Government disagreed with the WHO conclusions more often than others. Those from Germany and in Research agreed most often. This experiment has shown that the Consensus Questionnaire approach can be used to seek international scientific opinion. It has also shown that the conclusions of future expert consultations need to be formulated in a manner suitable for the Consensus Questionnaire approach, to avoid the ambiguities that can arise if too many concepts are included within a single conclusion.