Measures of adherence based on self-report exhibited poor agreement with those based on pharmacy records

J Clin Epidemiol. 2005 Sep;58(9):924-33. doi: 10.1016/j.jclinepi.2005.02.002.


Objective: To assess the level of agreement between a self-reported measure of adherence with prescribed drug and a measure of adherence based on pharmacy data.

Methods: During an in-home interview, people aged 65+ were asked to report all prescription drugs they had taken in the preceding month. For each drug, a four-item questionnaire was used to determine self-reported adherence. In the pharmacy records, each drug that had been filled at least four times was analyzed, and the percentage of days with the drug available was calculated. Two types of adherence were studied: (1) adherence by individual, and (2) adherence by drug. The level of agreement was assessed using kappa (kappa) statistics and proportions of agreement.

Results: We compared the adherence measures among 189 individuals (880 drugs). Among all, 90 individuals (48%) self-reported adherence, whereas 95 individuals (50%) were adherent according to the records. The level of agreement between these two measures was slight (kappa=0.16 [95% CI: 0.02-0.30]). Individuals self-reported to be adherent for 81% of the drugs, while pharmacy records showed adherence for 83% of them (kappa=0.13 [95% CI: 0.05-0.20]).

Conclusion: Self-reported measures of adherence exhibited poor agreement with those based on pharmacy records.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Age Factors
  • Aged
  • Cross-Sectional Studies
  • Drug Prescriptions
  • Drug Therapy*
  • Female
  • Humans
  • Male
  • Medical Records*
  • Pharmacies
  • Self-Assessment*