Abstract
OBJECTIVE: The aim of this study was to evaluate the validity of self-reported medication use for hypertension, diabetes, and dyslipidemia by comparison with health insurance claims among employees of large-sized companies in Japan. METHODS: Participants were 61 676 participants of 13 large-sized companies in Japan. Self-reports on medication use were obtained through web- or paper-based questionnaires conducted at the annual health checkup in fiscal year 2016. Health insurance claims for medication were obtained from corporate health insurance associations from April 1, 2016, to March 31, 2017. Agreement rate, sensitivity, specificity, positive and negative predictive values (PPV and NPV), and kappa statistics of self-reporting were examined for different reference periods (1-, 2-, and 3- months, and 1-year). Subgroup analysis was conducted stratified by sex, age, body mass index, smoking, alcohol drinking, blood pressure, hemoglobin A1c, and low-density lipoprotein cholesterol. RESULTS: Agreement, sensitivity, specificity, PPV, and NPV were 0.98, 0.90, 0.98, 0.87, and 0.99 for hypertension, 0.99, 0.89, 1.00, 0.89, and 1.00 for diabetes, and 0.98, 0.86, 0.99, 0.83, and 0.99 for dyslipidemia, respectively, between self-reports and claims data for 3 months. Kappa statistics were highest with the 3-month reference period of claims data for hypertension, diabetes, and dyslipidemia. No major concordance was observed between the subgroups. CONCLUSION: This validation of self-reported medication use for hypertension, diabetes, and dyslipidemia showed almost perfect reliability among employees of large-sized companies in Japan.