Semi-structured Interviews with Clinicians

METHODS

Study Design

We conducted a qualitative study using semi-structured interviews with clinicians at a large academic medical center in the Northeast. Approval to conduct this study was obtained from the Dartmouth- Hitchcock's Medical Center's (DHMC) Committee for the Protection of Human Subjects.

Participants and Recruitment

We used a criterion-based sample of clinicians (n=58), including attending surgeons, house officers, clinical nurse leaders, respiratory therapists, registered nurses, and nurse assistants. The units where interviewees worked varied by several measurable characteristics, including primary surgical specialty, number of inpatient beds, and acuity of patient population. We purposefully sampled by clinician type, years of experience, and patient population to maximize diversity and generalize findings to a wider demographic. The hospital's site coordinators and nurse managers on the unit helped identify potential participants for the study based on interdisciplinary team composition. The study team contacted individuals directly for participation. We obtained consent from participants to record and transcribe their interview.

Data Collection

Interviews were conducted between February and June 2017. A registered nurse trained in qualitative interview skills conducted all interviews, each lasting between 30 minutes to 1 hour on average. A semi-structured interview guide was developed based on a conceptual model developed by A.G., grounded in concepts of high reliability organizations and adapted from a 2014 study by Wakeam et. al..(8). The goal of the interview guide was to understand not only provider perceptions around rescue, but organizational influences on the rescue process and how that is conveyed to staff. The interview guide explored positive and negative influences of rescue (see Appendix A). All interviews were audio recorded and transcribed verbatim.

Data Analysis

Interview transcripts were managed and analyzed using NVivo 11 qualitative software (QSR International). We conducted thematic analysis and consensus coding of verbatim transcripts. Analysis began with independently reading, highlighting, and annotating transcripts to identify patterns. We refined these patterns into a systematic list of codes (codebook) to describe the content of each data excerpt. After this initial development, the codebook was applied to a new transcript, and again refined by author consensus. After three iterations of codebook refinements, we reached thematic saturation as no new themes or patterns emerged. All transcripts were coded using the final codebook. We used a Framework Matrix approach to synthesize our coding and identify overarching themes that facilitate or impede the rescue process (9).

Cite this page