Influence of the mode of administration on the results of medication adherence questionnaires

J Eval Clin Pract. 2017 Dec;23(6):1252-1257. doi: 10.1111/jep.12773. Epub 2017 Jul 17.


Introduction and objectives: Adherence to medication regimen is commonly assessed through questionnaires, some of which are validated via self-administration. The inadequate health literacy of elderly people pushes researchers to the use of interviews as a method of administration. The aims of this study were to compare the results obtained with an interviewer-administered and a self-administered medication adherence questionnaire and to evaluate the consequences of the adherence status classification of individuals.

Methods: A cross-sectional study was performed in which the Medida de Adesão aos Tratamentos adherence questionnaire was administered to adult patients who were taking at least 1 antihypertensive drug. The data were collected in 7 community pharmacies in central Portugal between March 2014 and September 2015 in 2 different phases: in the first phase, the questionnaire was applied during a healthcare professional interview, and the second phase involved a self-report administration. A confirmatory factor analysis was conducted, and the measurement and structural invariances across the application methods were examined.

Results: A sample of 425 patients with a mean age of 68.21 ± 10.56 years participated in the study. The confirmatory factor analysis revealed that both the interview and self-report had a good fit with the original model, although the self-report results exhibited a better fit. In the interview administration, we obtained lower values for skewness and higher levels of kurtosis. The patients subjected to the interview administration presented with a 9.7% higher tendency to answer "never" when compared with the self-administered application, which overestimated adherence.

Conclusions: The interview administration method induced bias that led to a higher percentage of "never" answers and a subsequent overestimation of adherence levels. Self-report administration should be preferred in the application of medication adherence questionnaires.

Keywords: bias (epidemiology); medication adherence; patient outcome assessment; patient preference; reproducibility of results; surveys and questionnaires.

MeSH terms

  • Aged
  • Aged, 80 and over
  • Bias
  • Community Pharmacy Services / statistics & numerical data
  • Cross-Sectional Studies
  • Female
  • Humans
  • Interviews as Topic / standards*
  • Male
  • Medication Adherence / statistics & numerical data*
  • Middle Aged
  • Portugal
  • Self Report / standards*
  • Surveys and Questionnaires / standards*