Validity assessment of self-reported medication use by comparing to pharmacy insurance claims.

Self-reported questionnaire data on medication use had sufficiently high validity for further analyses. Item responses showed close agreement with actual prescriptions, particularly those covering 3 months.