Optimizing measurement of self-reported adherence with the ACTG Adherence Questionnaire: a cross-protocol analysis

Nancy R Reynolds, Junfeng Sun, Haikady N Nagaraja, Allen L Gifford, Albert W Wu, Margaret A Chesney
2007-12-01
Abstract:Background/Objective: The AIDS Clinical Trials Group (ACTG) Adherence Questionnaire is used extensively, but investigators frequently only use the first item of the questionnaire (4-day recall).Design/Methods: A secondary analysis was conducted to (1) estimate the validity and reliability of each of the 5 scale items and (2) compare the approach commonly used to summarize adherence data collected with the instrument (average 4-day recall) with alternate approaches derived using principal component (PC) analysis and the full questionnaire. We hypothesized that an estimate of adherence taking all items of the questionnaire into account would provide a stronger measure of adherence.Results: Logistic regression analyses showed that the first PC identified (PC1) was significantly correlated with plasma HIV RNA outcome (P< 0.0001 for ACTG 370 data and P= 0.006 for ACTG 398 data) and correlated with …
What problem does this paper attempt to address?