Reliability and validity of a telephone questionnaire for estimating lifetime personal sun exposure in epidemiologic studies

Our Australia-wide case-control study of ocular melanoma diagnosed in 1996 to 1997 needed a short telephone interview on sun exposure. We constructed one by examining data from 700 controls ages 40 to 64 years in the Geraldton Skin Cancer Survey in 1988; they had answered a "whole-of-life" questionnaire in a face-to-face interview. Sun exposure in their first 4 decade years of age best predicted their lifetime annual average sun exposure, so the shortened questionnaire asked about sun exposure in these 4 decade years only. Retesting 60 participants 1 year later with the whole-of-life questionnaire gave an intraclass correlation coefficient of 0.65 (95% confidence interval, 0.48-0.78) for ranked total sun exposure between the two interviews; the intraclass correlation coefficient was higher in men (0.73) than in women (0.54). Correlations were also high between parallel measurements of sun exposure on working days in the decade years and in outdoor occupations throughout life in the telephone interview of the ocular melanoma study (Spearman's R = 0.75) and in another study of non-Hodgkin's lymphoma (R = 0.71). Agreement between simultaneous parallel measures of total exposure (nonworking + working day and recreational + occupational exposure) was slightly weaker and of nonworking day and recreational exposure much weaker. Occupational exposure in women was much less strongly correlated with total exposure than it was in men possibly because of their frequently combined work and family roles, which the questionnaires did not try to separate. Research is needed into how this might be done to improve sun exposure measurement in women.