Printed in Great Britain Repeatability of Self-Reported Data on Occupational Exposure to Particular Compounds

The repeatability of serf-reported occupational exposure to eight particular compounds was investigated in 209 males aged 49-87 years. The subjects completed an initial mailed questionnaire and were interviewed by telephone 3-5 weeks later. The study was carried out as part of a case-referent study on the relation between occupation and prostate cancer. The repeatability of answers was found to be better in the case of some agents than in that of others: kappas calculated as a measure of concordance range from 0.36 to 0.70. No substantial influence was found to be exercised by age or socioeconomic status, or by case or referent status. Although the repeatability observed was not flawless, it was concluded that self-reported exposure data would appear to be a sound basis for epidemiological studies on the aetiology of disease, whenever objective information on occupational exposure is not available. In occupational epidemiology information is often re-quired about the work histories and the occupational exposures of the study subjects. Whenever possible, objective sources of information should be used to gather these data. Unfortunately the subjects are often the sole sources of the required data. The only way then to obtain information is from self-administere