Methodologicalconsiderations

102 et al. 2009, Slettebø 2009). This is part of the core of clinical ethics (de Vries and Gordijn 2009, Leget, Borry et al. 2009).

When situations become complicated, utilising the National Guidelines for Decision-making Processes (Norwegian Directorate of Health 2009) can ensure that important facts, values, interests and assessments are not omitted, for example demands from the next of kin for treatment that is not in line with the physician’s professional assessment.

Operative clinical ethics committees have been set up in some municipalities, and the committee at the local hospital can also be used in some places for questions concerning difficult ethical dilemmas on life-prolonging treatment in nursing homes. Relevant dilemmas in nursing homes often concern an assessment of whether elderly, dying patients can benefit from hospital treatment.

103 6.5.1 Limitations of the study

Weakness of the sample includes the exclusion of informants from other national backgrounds than ethnic Norwegian. A critical inspection would also point out that the exclusion of Northern Norway fails to allow for maximum variation in the sample. The inclusion of only physicians and nurses from among the health workers in the nursing home was based on the nature of the research questions, which were associated with issues pertaining to life-prolonging treatment, and which made this choice the most relevant solution. Physicians are legally responsible for the provision of treatment (Norwegian Ministry of Health Care Services (HOD) The Patients` Right Act of July 1999). The nurses are delegated the task of providing most of the medical treatment prescribed by the physicians. In a nursing-home context, the physicians and the nurses tend to cooperate on these matters. One weakness of this study is that all the physicians in the selected nursing homes were men, and all the nurses were women.

6.5.2 Rigour and trustworthiness

The standards used for assessing the quality of research must be relevant to the research method applied. The knowledge arising from this study has been sourced from in-depth interviews and the data have been manually transcribed and interpreted; consequently, the study shows clear signs of subjectivity. From the outset till the very end, it has therefore been important to focus on validation. I have tried to provide clear reasons for my various choices of method at all stages in order to ensure transparency (Malterud 2001). Kvale and

Brinkmann (2009) use the concepts of reliability and validity when dealing with the validity assessment of qualitative studies. Reliability is the standard against which the study’s

handling of data is gauged, whereas a study’s validity tells us whether it investigates what it is meant to investigate. Furthermore, transferability is an understandable concept which

measures a study’s applicability within similar contexts; in my opinion, this is a more appropriate concept than generalisation in a qualitative setting (Reiter-Theil 2004).

6.5.2.1 Process validation

Critical awareness of the researcher’s own role is important throughout all stages outlined above, also because the choices made are associated with the role of the researcher. Tranøy (1986) writes that if we assume the moral calibre of researchers to be on a par with that of everybody else, then openness to opposing views and criticism is imperative. Self-criticism plays a key role in this respect.

104 Whilst the continuous practical process of this study is based on Kvale and Brinkmann‘s seven stages of an interview investigation, I present a summary process assessment of the consistency (reliability) of the results and whether the study has investigated the issues it was intended to investigate (validity). This process validation is inspired by Kvale’s sixth phase (Kvale & Brinkmann 2009).

Textbox 4.7: The seven stages of validation – a summary

1. Thematisation. At this stage, validity is linked to the correlation between the topic studied and the theoretical basis, and also to the derivation of the research questions. Medical ethics and the autonomy principle formed the theoretical points of departure. However, although this was a relevant starting point, the data collection process introduced a need to draw on other aspects of medical ethics in order to enable discussion of the findings in the light of theory beyond the autonomy principle. Norwegian legislation also represented a key framework for these discussions, as well as literature on frameworks and organisational factors.

2. Planning. My choice of method and design was based on the research question. In order to gain knowledge about the decision-making processes experienced by physicians, nurses and relatives while mapping what was happening, qualitative in-depth interviews emerged as a natural choice. The application of emergent design, which was developed in pace with the data collection process and the textual analysis, proved to be important for my ability to uncover new knowledge and to continue probing in further depth.

3. Interviewing. At each of the nursing homes, participants were selected through “purposive sampling”

by a nurse who was familiar with potential participants with experience of the research question. By personally transcribing the interviews, I was in a position to listen to my own interviews and the responses I received. Combined with interview guidance, this process was useful in allowing me to probe sufficiently deeply, thus improving the quality of my interviews.

4. Transcribing. All interviews were transcribed verbatim. This was in order to capture all the content as accurately as possible. When reviewing the material it was easy to read and detect nuances in a holistic way. Nothing was added or subtracted. The transcripts were further validated by a participant reading through them and confirming their accuracy. My supervisor also checked my transcripts against the audio files.

5. Analysing the data. Large chunks of my work with the data involved making them accessible from within the context of the research questions and their textual presentation. Three separate analyses resulted in a descriptive presentation of the findings. My primary task in interpreting the findings was to “sort” the participants’ various statements to ensure they were put into the correct category.

Furthermore, textual presentation will always be coloured by interpretation. This was checked and discussed with my supervisors.

6. Validation. I have sought to question the study’s reliability and validity at every stage. I have constantly stopped to focus on the core questions in order not to lose the “thread” throughout the study.

Because the interviews provided a rich source of information, it was tempting to include other associated topics. I noted these associated topics as proposals for further research (Appendix 6). This was helpful in putting associated topics to one side in order to maintain my focus on “end-of-life decision-making processes in questions pertaining to life-prolonging treatment in nursing homes”.

7. Reporting. I have chosen to present my findings in the form of multiple papers in the hope that this will facilitate the communication of knowledge about the study to other interested parties in a nursing home context, nationally and internationally. The presentation of my findings is kept in a descriptive format, interspersed with selected quotes which I find to be particularly illustrative of the various findings.

105 Textbox 4.7 presents a brief validation summary for all stages (Kvale and Brinkmann 2009).

Please refer to the descriptions of the various stages of my work for a more detailed assessment of the quality of the study.

6.5.2.2 Pragmatic validation and transferability

My epistemological position and my declared view that the development of knowledge is a social construction (Rose and Webb 1998), will inevitably affect my attitude with respect to quality assessment of this study in a wider context. When I set out to map decision-making processes at end-of-life in nursing homes, my presentation of the findings was never meant to constitute a correct reflection of an isolated situation within the ten specific nursing homes.

The objective was for the descriptions provided by representative participants (purposive sampling) with experience of this particular situation to shed light on as many factors as possible associated with end-of-life decision-making processes in nursing homes. From there on, the aim is for audiences in other nursing homes to consider the findings in the light of their own setting. My critical and normative discussion of the findings can be further assessed and discussed and may stimulate improved practice. Kvale and Brinkmann (2009) refer to this process as pragmatic validation, a method by which the social sciences provide input to an ongoing dialogue about quality in society, rather than isolated questions about verified knowledge. This makes it possible for the present qualitative study to be considered by physicians and nurses in similar settings, allowing the findings to be used as potential working hypotheses for further work rather than absolute truths (Malterud 2001). One objective is to draw attention to subjective descriptions of practice based on relevant experience. Another objective is to stimulate further research into the issues raised (Angen 2000). In this way the outcome becomes a continual validation of research beyond the individual studies.

106