Data Generation and Handling Data collection from discussions

The professional discussion within the consensus meeting created an opportunity to reflect and explore perspectives emerging from the results within a wider health economy, actively involved in care delivery. The discussion also generated data that highlighted the dynamic interactions between professionals. This facilitated a greater understanding of the collective meaning of what and how triggers were perceived and their feasibility within COPD care delivery, in contrast to their individual, professional grouping, organisational contexts or constructs (Frey and Fontana, 1991). The aim was to explore care professionals’ perspectives in as

many settings as possible (e.g. hospice physiotherapist and a hospital physiotherapist), whilst gaining a view collectively from a certain professional grouping (e.g. physiotherapists).

Discussions from the meeting were recorded on a flipchart and the author clarified that the comments captured reflected what participants’ perspectives were (Denzin, 1989). The discussions were not digitally-recorded and therefore it was not possible to assign an individual identifier for said participant and their comments, nor listen to the verbatim dialogue. Despite several rehearsals to explore if an audio-recording was feasible, the challenge of the room layout and ensuring all voices were captured using only a single digital recorder, proved too difficult and was thus abandoned. Therefore, the process of capturing the discussions was facilitated by a colleague, unknown to the participants, who recorded key phrases and discussion points.

4.2.4.1 Initial presentation of background information

Within the professional consensus group, an initial presentation highlighted the supportive and palliative care needs of individuals with COPD and their carers. This identified the current service provision, with particular emphasis given to the supporting literature for the content of the workshop. This included important national and international policy documents and an appraisal of the literature to date (as discussed in Chapter 2).

4.2.4.2 The consensus process

Table 7 illustrated how the different rounds took place within the consensus method. There were three in total.

Table 7: The Consensus Process qualitative, secondary analysis and potential eight candidate events as triggers: drawing on their individual clinical experience and

3: Further over-arching discussion of the concept of triggers

Round 1

This round started with an explanation of what the attributes of a trigger should comprise within the context of this piece of work and discussion about the eight potential events as triggers. Participants had an agenda to help orientate and navigate them through the format of the meeting (Appendix D:

Agenda for the consensus meeting) and each round had a scoring sheet to help maintain confidentiality and anonymity within the consensus process (Appendix E: Scoring sheets).

The discussions were time-limited so that first impressions of the identified events as potential triggers could be captured. Professionals were asked to rate from 1-10 (10= highest utility, 1= lowest) all eight of the suggested candidate events in their ability to act as triggers for a holistic needs assessment.

Round 2

After the initial rating (round 1), the participants had the opportunity to discuss the merits of each candidate event in turn, seeking clarification and exploring the feasibility of this event within their sphere of clinical practice. Further rating of the candidate events’ utility occurred after discussion of each trigger. The ratings aimed to quantify the level of consensus, whilst the discussions that ensued gave a very rich, qualitative component to the process. In this round again, professionals were asked to rate from 1-10 (10= highest utility, 1=

lowest) all eight of the suggested candidate events in their ability to act as triggers.

Round 3

Once all candidate events were discussed in turn, a final rating of all eight triggers occurred. This facilitated discussion of the feasibility and applicability of triggers as a concept throughout the disease trajectory of severe COPD.

4.2.5 Data Analysis

Data were entered onto Excel (Microsoft, 2008) spreadsheets and the medians for Rounds 1, 2 and 3 were calculated during the course of the meeting, for feedback in the subsequent rounds.

Definition of consensus

Consensus was defined a priori as 75% of the group rating the candidate event with a score of eight or above, using the median scores (Murphy et al., 1998).

4.2.5.1 Process of the analysis

Consensus methodology depends on all participants contributing to every stage in order to reflect the process as they reach consensus. However, five hours is a substantial time commitment in the middle of a working day, and the research team recognised that not all professionals would be able to attend all phases. In view of this, the decision was taken to exclude ratings from participants unable to contribute to every stage, but to include their comments in the discussion.

4.2.5.2 Thematic content analysis from group discussions

This qualitative approach is a method of textual investigation of the content of the data, to categorise recurrent or common themes (Silverman, 2005). The approach is commonly used in health-related, qualitative research and aids data interpretation when views from particular groups, and in this case professionals (health and social care), are required. Given the limitations for capturing the group discussions on a flip chart, the data could not be coded sufficiently to categorise respondents’ accounts, limiting the ability to classify them into themes that were common and recurring in the data sources (Green and Thorogood, 2004, pp. 176-177).

4.2.6 Reporting

Addressing the potential for bias

To ensure transparency in reporting, the Guidance on Conducting and Reporting Delphi studies (CREDES) in palliative care has been used as the reporting checklist (Junger et al., 2017). The principles of reporting and assurances suggested by this approach, whilst addressing the nuances of reporting within palliative care, identify it as a suitable and transparent reporting standard for this chapter of the thesis. In the absence of any published checklists specific to the consensus methodology, namely nominal group technique used in this study, the CREDES recommendations were used as a helpful checklist (Appendix F: CREDES checklist).

4.2.7 Ethical Considerations