OBJECTIVES: The Adelphi Adherence Questionnaire (ADAQ©) is a patient-reported 13-item generic medication adherence instrument, developed with qualitative patient/clinician input. Measurement properties of the ADAQ© were evaluated across multiple disease areas, modes of administration, and treatment regimens.

METHODS: Data were drawn from the Adelphi Real World Disease Specific Programmes (cross sectional surveys conducted in the USA) in Atopic Dermatitis (N=180), Heart Failure (N=544), Human Immunodeficiency Virus (HIV; N=247), Multiple Sclerosis (N=1337), and Osteoarthritis (OA; N=723). Item response distributions and inter-item polychoric correlations were examined. Latent variable modelling was used to evaluate the ADAQ© measurement model across disease areas (including exploratory and confirmatory factor analysis, bifactor modelling, exploratory graph analysis, Item Response Theory, and Mokken scaling). Reliability was estimated using coefficients alpha and omega. Convergent validity was evaluated using Spearman’s and polyserial correlations with the Adherence to Refills and Medications Scale (ARMS) and clinician-reported adherence/compliance where available.

RESULTS: Consistent with previous research, potential item-level ceiling effects were observed across disease areas (i.e., large proportions of patients selecting the most adherent response for an item). Latent variable modelling consistently indicated that an 11-item scoring of the ADAQ© provided the best model fit across studies, with bifactor modelling suggesting the appropriateness of an essentially unidimensional measurement model. A scoring algorithm (the unweighted average of items 1 to 11) represents the overall ADAQ score, with items 12 and 13 providing additional information. Reliability coefficients omega hierarchical and alpha indicated the ADAQ© has excellent reliability (>0.9) across all disease areas. Convergent validity was demonstrated with the ADAQ© scores showing moderate-to-strong correlations with ARMS scores (0.591 in HIV to 0.774 in OA), and moderate correlations with clinician-reported adherence/compliance (all >|0.3|).

CONCLUSIONS: Evidence across multiple disease areas supports use of an 11-item ADAQ© score as a unidimensional, reliable, and valid measure of patient-reported adherence to medication.