Identifying the Content of Care Delivered to Control Groups

Abstract

Background: Much attention has been directed at behavior change techniques (BCTs) in interventions. Little attention, however, is being paid to the BCTs in control groups receiving usual care, yet studies suggest usual care content can vary between study sites and impact outcomes. This study explored a method of assessing the active content of usual care provided to control groups. Methods: HIV-nurses (n=21) completed an open-ended questionnaire about nine domains in adherence counseling. The content was coded by two coders based on a medication adherence BCT taxonomy. We examined intercoder reliability (kappa), internal consistency of a standard care quality (SCQ) scale, and variability in SCQ scores. Findings: Responses could be reliably coded with the taxonomy. The mean kappa was .79. 19 of the 22 BCTs coded produced a kappa >.6. The SCQ scores displayed substantial variability (range=7-19 BCTs), and showed good internal consistency (?=.74). Discussion: The BCT taxonomy could be reliably applied to responses to open questions on adherence counseling. Moreover, adherence counseling can be compiled in one quality score. The tool may be useful for exploring control group content.