Nvivo v 12

The audio-recordings were transcribed verbatim into separate Word documents and then checked by the author for accuracy against the original recording. The author assigned a unique identifier to each transcript denoting the service type: regional (ReHS) or rural (RuHS); the participant type: target AH staff (TAHS) or key informants (KI); and the interview number for that service and group type (e.g., ReHS-KI-6). The author then conducted a thematic analysis of the data using NVivo v12 software (QSR International) [17 (link)]. The WoP-RIF’s three domains—workplace/organisational, role/career and community/place—provided structure for the first level of analysis—coding and categorisation. Identification of emergent themes drew on the key influences on staff job and personal satisfaction under each WoP-RIF domain (see Table 1). These identified themes underpinned the development of a set of recommendations for each service to support a sustainable allied health workforce. Fourteen recommendations were made for the ReHS and 13 for the RuHS; 10 of these recommendations were common to both services. This paper focuses on the data underpinning these 10 shared recommendations because these are most likely to resonate with, and have utility for, other rural health services.

Audio-recordings were transcribed by a professional transcriber and then checked for accuracy against the original recordings. Each interview was assigned a unique identifier. Three authors conducted a thematic analysis of the data taking an inductive approach. NVivo v12 software (QSR international) was used. Six initial interviews with participants who had different roles were read by two members of the research team who independently created initial codes. One author then coded all remaining transcripts to identify patterns in the codes and links between the codes thereby organising the data into meaningful themes [16 (link)]. To validate the emergent themes, three further interviews from each of the ACCHSs were independently coded by two of the research team members. Any disparities identified were discussed and resolved.

Written informed consent was obtained, including consent for audio recording and transcription. Patient focus groups were facilitated by an experienced qualitative researcher (R.M.) and observed by an additional investigator (M.J.B.). HCP focus groups were facilitated by two investigators (M.J.B. and C.L.H. or R.J.B.). Participants in all groups were invited to discuss perspectives on their care priorities when attending outpatient services (patient groups) and providing care (HCP groups). This included personal experiences of clinic attendance, opinions on patient-centred care and aspirations for future improvement. Groups followed a semi-structured format, with standardized prompt questions developed by the authors (Fig. 2) to promote discussion around care domains that could be addressed by a PREM, such as how care occurs, care content and impact on individuals. The question guide was informed by prior work on rheumatology patient experiences in published literature [24 , 25 (link)].
Focus groups were audio-recorded, transcribed verbatim using a professional transcription service, and imported for coding into NVivo v.12 software (QSR International).

Pseudonymised demographic data will be collected from all participants to provide a descriptive and contextual picture of the participants working in each of the six hospitals, for example demographic details such a surgeon years in post and additional qualifications will be sought. Personally identifiable data used in the data pseudonymisation process will be collated in a password protected database which will be stored to enable member checking later in the course of analysis. In total, we anticipate conducting approximately 120–140 interviews over the course of the 5-year study. During the longitudinal data collection, we will seek to obtain a full dataset from each individual participant, hospital and year. During the cross-sectional national surgeon leadership interviews, it is possible that data saturation will be reached before the maximum number of planned interviews are conducted (n = 10). All interviews will be audio-recorded, professionally transcribed and processed in NVivo v12 software (QSR International 1999) [58 ].

Written informed consent was obtained, including consent for audio recording and transcription. Patient focus groups were facilitated by an experienced qualitative researcher (R.M.) and observed by an additional investigator (M.J.B.). HCP focus groups were facilitated by two investigators (M.J.B. and C.L.H. or R.J.B.). Participants in all groups were invited to discuss perspectives on their care priorities when attending outpatient services (patient groups) and providing care (HCP groups). This included personal experiences of clinic attendance, opinions on patient-centred care and aspirations for future improvement. Groups followed a semi-structured format, with standardized prompt questions developed by the authors (Fig. 2) to promote discussion around care domains that could be addressed by a PREM, such as how care occurs, care content and impact on individuals. The question guide was informed by prior work on rheumatology patient experiences in published literature [24 , 25 (link)].
Focus groups were audio-recorded, transcribed verbatim using a professional transcription service, and imported for coding into NVivo v.12 software (QSR International).