Methods
We conducted in-depth interviews with in fifty patients (28 who accepted HIV testing and 22 who declined HIV testing) three northern California EDs. The ED settings have been described in detail elsewhere. All sites were recipients of CDC funding to increase HIV testing among disadvantaged urban populations. As previously described, two sites in Oakland offered non-targeted oral swab screening at registration or triage while a third site in San Francisco utilized clinician-initiated diagnostic venipuncture testing and targeted testing of admitted patients.
All patients offered an HIV test in the ED were eligible to participate in this study, although recruitment of patients was limited to those who had already received their test results. Patients with reactive test results were excluded from recruitment given the sensitive nature of these test results and the need for immediate follow-up counseling. In addition, to gain a range of perspectives, study investigators decided a priori to attempt to sample ten accepters and ten decliners at each site. Emergency department staff helped identify patients for recruitment, and flyers advertising the study were posted throughout the EDs. Once patients were recruited, researchers explained the goals of the study and obtained verbal consent for an interview. Interviews were conducted in a private setting in the ED and lasted approximately twenty to thirty minutes. Information on demographics and health care utilization, including age, gender, race/ethnicity, insurance/care status, and reason for ED visit, was collected at the beginning of each interview. Participants were each reimbursed thirty dollars for their participation. Study data were collected from April to June 2009. The institutional review boards of the University of California San Francisco and all participating sites approved this study.
A semi-structured interview guide was developed to cover participant experiences with HIV testing in the ED, reasons to accept or decline HIV testing, prior HIV testing history, relationship status and perceptions of HIV risk behavior, and participant attitudes towards HIV infection. Interviews were transcribed verbatim and entered into Atlas.ti for organization and easy retrieval of text elements. Three analysts employed a framework analysis approach to the data, which included two distinct analytic tasks: managing and interpreting the data. The analysts began managing the data by reading a subset of the interviews to gain an initial familiarity with the dataset and to produce a preliminary list of coding categories. Subsequent steps included refinement of codes, as well as sharpening the shared understanding of how to apply codes. Each interview was coded by a primary analyst and reviewed by a secondary analyst. The analytic team then selected text associated with key codes across cases to be read and summarized as a group in six three-hour interpretative analysis sessions. The goal during the interpretative phase was to produce an exploratory analysis to uncover overarching attitudes towards HIV testing in the context of a visit to the ED.