Table 1

Study design using video-reflexive ethnography

Hospital A (ICU)Hospital B (2 surgical wards)
ParticipantsNinety participants were recruited, comprising 49 nurses, 8 nursing students, 20 doctors, 3 medical students, 5 allied health practitioners and 5 administrative or cleaning staffEighty-seven participants were recruited, comprising 52 nurses, 2 nursing students, 20 doctors, 1 medical student, 2 allied health practitioners, 6 administrative or cleaning staff and 4 patients
Phase 1
Interviews and observations
Thirty semistructured interviews with a range of staff were conducted, including 18 nurses, 7 doctors, 2 cleaners, 1 medical student, 1 nursing student, 1 wards-person and 1 dietitian.
Field observations were carried out and recorded in field notes by the researcher (S-yH), an experienced hospital ethnographer.
Sixteen semistructured interviews were conducted, with 12 nursing staff, 1 doctor, 1 cleaner and 2 senior general services staff.
Phase 1 was abbreviated in hospital B, due to the limited availability of staff for interviews. Instead, more informal field interviews were conducted, and more observations and field notes were undertaken.
Interviews and observations focused on identifying the routine activities of the ward and on how IPC was (or was not) part of them, according to participants. At each site, an initial content analysis of field notes and interview transcripts was undertaken to guide phase 2 of the study.
Phase 2
Videoing and reflexive sessions
Videoing focused on routine activities of the ward, guided by interview data. Footage was then edited into clips, 2–10 min long, to be shown to stimulate discussion during reflexive sessions. Footage was chosen for feedback if it showed a routine activity requiring attention to IPC and/or if participants stated an interest in viewing particular footage.
Reflexive sessions were facilitated to encourage discussion around IPC in the footage. Where cross-contamination risks were identified, participants were encouraged to consider how they might practically overcome these risks. Reflexive sessions were video-recorded and/or audio-recorded and transcribed for analysis. Content and thematic analyses were carried out, to describe the main features of the reflexive discussions, as well as the themes that emerged in how clinicians discussed IPC issues. Both the content (problems identified, solutions offered, etc) and a selection of thematic findings from these analyses were presented to participants during feedback sessions in phase 3.
We facilitated a total of five reflexive sessions (three with nurses, and two interdisciplinary sessions with both doctors and nurses).We facilitated a total of 18 reflexive sessions (seven with nurses, three with infection control practitioners and three with doctors).
Phase 3
Feedback sessions
One feedback session was conducted with the senior ICU doctors.Two feedback sessions were conducted for the nurses, one for the doctors and one for the infection control practitioners.
Postintervention analysisFollowing the completion of fieldwork, a posthoc analysis using a modified constructivist grounded theory approach23 was undertaken with the data created from all three phases and wards, to produce the findings presented in this paper.
Additional thematic coding was undertaken of all interview and reflexive session transcripts, during which we developed initial codes relating to IPC practices such as ‘crossing clean/dirty boundaries’ and ‘making clean/dirty distinctions'. Instead of collecting more data through theoretical sampling, we performed a process of abduction24 through iterative comparison of our codes with each example of IPC work that was found in our data (especially drawing on the video data).
Correspondingly, our reading of the literature16 25 on the spatial dimension of IPC contributed the terms ‘boundary work’ and ‘buffers’, which we have adopted and extended in this paper to account for our data.
  • ICU, intensive care unit; IPC, infection prevention and control.