Agreement between self-reported and health insurance claims on utilization of health care: A population study

To investigate the agreement between self-reported and insurance claim of the utilization of health care.

Data were from the Taiwan National Health Interview Survey, conducted in 2001. There were 22,121 (94.2%) individuals who completed the interview, 18,143 of them aged 12 and above. Among them, 87% signed a consent form to allow us linking their interview with data with the National Health Insurance database. After deleting those with illegible identification code or date of birth, 15,660 subjects were included in the study. Kappa (kappa) was used to describe the visit concordance, whereas the intraclass correlation coefficient was used to describe frequency consistency. Logistic regression was used to investigate factors associated with the agreement.

All had reasonable visit and frequency agreements. Logistic regression showed that males and those living in rural areas tended to have a lower likelihood of agreement in reporting outpatient visits compared with others. In reporting emergency room visits, males and those with a higher education level, self-perceived health of "not good," chronic disease, or at least one difficulty with activities of daily living (ADL) had higher chances of agreement.

Agreement between self-reported utilization and insurance claims were fairly good in general population.