2002 Volume 68 Issue 2 Pages 43-53
To assess the usefulness of the recall-based self-report on cataract diagnosis or extraction as a means of case definition in a cohort study, the answers for an originally developed questionnaire on cataracts were compared with the medical records among 56 residents living in Iwate, Nagano, and Okinawa in Japan who had reported past history of cataract diagnosis. The difference in date of diagnosis between self-reports and medical diagnosis was very widely distributed, while the positive predictive value of self reports on past history of cataracts diagnosis exceeded 90%. As regards the history of cataract extraction, no discrepancy was found between self-reports and medical records. Although the self reported dates of extraction were more accurate than that of diagnosis, there were still a few largely discordant reports on extraction date and a tendency for reporting earlier date than real extraction date. From the comparisons on causes for cataracts, it was considered that the number of senile cataracts could be underestimated by self report. These results suggest that the recall-based self-report on cataract extraction may be a practical method for case definition in a cohort study, if it is followed by the confirmation of extraction date and cause for cataracts through the review of medical records, as compared with the use of self-reported cataract diagnosis.