Derivation and validation of the ASK-12 adherence barrier survey.

BACKGROUND The ASK-20 survey is a previously validated patient-report measure of barriers to medication adherence and adherence-related behavior. OBJECTIVE To derive and validate a shorter version of the ASK-20 scale. METHODS Patients with asthma, diabetes, and congestive heart failure were recruited from a university medical center. Participants completed the ASK-20 survey and other questionnaires. Approximately one-third of participants were randomized to a 2-week retest administration. Item performance and results of an exploratory factor analysis were examined for item reduction and subscale identification. Subsequent analyses examined reliability and validity of the shorter version of the ASK. RESULTS A total of 112 patients participated (75.9% female; mean age 46.7 y; 53.6% African American). Eight items were dropped from the ASK-20 based on factor loadings, floor effects, Cronbach's alpha, and the ability of each item to discriminate between groups of patients differing in self-reported adherence. The new total score (ASK-12) had good internal consistency reliability (Cronbach's alpha 0.75) and test-retest reliability (intraclass correlation 0.79). Convergent validity was demonstrated through correlations with the Morisky Medication Adherence Scale (r -0.74; p < 0.001), condition-specific measures, the SF-12 Mental Component Score (r -0.32; p < 0.01), and proportion of days covered by filled medication prescriptions in the past 6 months as indicated by pharmacy claims data (r -0.20; p = 0.059). The ASK-12 total score also discriminated among groups of patients who differed in self-reported adherence indicators, including whether a dose was missed in the past week, the number of days medication was not taken as directed, and treatment satisfaction. Three subscales were identified (adherence behavior, health beliefs, inconvenience/forgetfulness), and results provided initial support for their validity. CONCLUSIONS The ASK-12 demonstrated adequate reliability and validity, and it may be a useful brief measure of adherence behavior and barriers to treatment adherence.