Reliability of self-reported household pesticide use

Household pesticide exposure has been associated with cancer risk in both adults and children. We investigated the reliability of reported lifetime household pesticide exposure through repeated administration of a standardized questionnaire. A questionnaire including detailed questions about lifetime frequency and duration of pesticide use in nonoccupational circumstances was administered on two occasions to 163 cutaneous melanoma cases and 113 controls. We investigated the agreement between the two measurements taken on average 12 months apart and studied the association between differences in the two measurements and a set of explanatory variables. According to the results of the reliability analysis, we also corrected the odds ratio estimates from the main study. The agreement for duration and frequency of use of pesticides outdoors was 89.5% (Cohen's kappa = 0.48) and 92.0% (Cohen's kappa = 0.40), respectively, whereas duration and frequency of use of pesticides indoors agreement was 75.4% (Cohen's kappa = 0.32) and 77.4% (Cohen's kappa = 0.28), respectively. The agreement was higher for duration (97.4%; Cohen's kappa = 0.72) and use of pesticides on domestic animals (86.4%; Cohen's kappa = 0.68). The corrected odds ratio showed a moderate increase with a reinforcement of the effect of pesticides. Overall, there was a good reproducibility in self-reported exposure to pesticides. This findings may reinforce earlier studies that showed that residential pesticides may cause cancer.