Data analysis

Interviews were audio-recorded using two digital recorders, transcribed verbatim by a professional transcription service, and ranged in length from 13 to 46 minutes. Given the study’s interest in understanding provider training needs, participant responses to the following questions were analyzed and further detailed: (a) What do you think are the key prerequisites for providers to be able to implement the prescription of PrEP successfully?, (b) If you have an interest in prescribing PrEP in the future, what skills do you think you need to be able to successfully implement PrEP in your practice/clinic?, and (c) What should you know about PrEP, or what do you need to prescribe it?

An inductive thematic analysis was conducted to report semantic-level themes describing the PrEP training needs for our participants (Braun & Clarke, 2012). The analytic process commenced in Phase 1, with three researchers with experience in qualitative data analysis (1 doctoral student, 1 research coordinator, and the principal investigator) familiarizing themselves with the data. In Phase 2, two researchers (1 doctoral student and 1 research coordinator) read five transcripts to generate initial codes and develop a codebook. In Phase 3, the researchers independently coded the remaining transcripts and updated the codebook when new codes emerged. Coding discrepancies were discussed throughout all phases in consultation with the principal investigator to establish consensus. In the final phase of the analysis, researchers identified themes by collating relevant codes. In an effort to establish consistency of themes throughout the data analytic process, the three researchers met on a bi-weekly basis to discuss findings. All overarching themes and subthemes were discussed in detail, and representative quotes were selected for each theme that emerged. Microsoft™ Excel was used for data management in all phases.

The rigor and trustworthiness of the data were determined by criteria established by Guba (1981). To assess credibility, interviewers familiarized themselves with the semi-structured interview guide prior to engaging in data collection. Interviewers also completed post-interview assessments to appraise the authenticity of participant responses. Additionally, interviewers met weekly throughout the data collection phase for peer debriefing of completed interviews and to provide ongoing feedback. Dependability and confirmability were assessed through the stringent adherence to the approved study protocol and with thorough documentation during the entire data collection phase. Criteria were also assessed during the analysis phase, as the researchers met on a bi-weekly basis to discuss the accuracy and reliability of codes between investigators to build consensus.