A method of estimating risk for occupational factors using multiple data sources: the Newfoundland lip cancer study

This paper describes how national census data, a provincial cancer registry, hospital patient charts, and the results of a household questionnaire survey in turn were used in attempting to estimate the risk of cancer for a specific occupation. A case control study was done in the Province of Newfoundland (Canada) during the summer and fall of 1973 to assess the role of the occupation of fishing as a risk factor in cancer of the lip. The following information on occupations must be available in order to adequately assess an occupation as a risk factor of cancer: 1. The occupations of the respondent over his lifetime; 2. The length of time the respondent was employed in each occupation; 3. The respondent's occupation at the time of the interview. Cancer registries and hospital charts should be target areas for improving the recording of occupation data. Clinicians should be encouraged to record at least the most usual occupation of their patient with additional information recorded for patients who work in areas known to be potentially hazardous to health. After defining the 'cases' and appropriate controls, interviewing subjects themselves with a standardized questionnaire should be seriously considered when specific occupational activities are suspected as carcinogenic. It is possible to design questionnaires which elicit detailed occupational information which have internal validity as the interobserver reliability test results have indicated. The household questionnaire data on work activities or working conditions of fishermen as possible risk factors failed to show evidence of positive associations which suggest any independent contribution to malignancy. The questionnaire data also revealed that using the mouth as a 'third hand' in the handling of nets protected fishermen, rendering them less than half as likely to acquire the disease from those in the occupation who employed other techniques.