Participant selection

My commitment and desire to cover as many points of view on delirium determined that the data collected would contain multiple and diverse perspectives. Originally I thought that interviewing approximately five people over the age of 65 years who had been delirious, five family members who were associated with someone who had been delirious, five sets of client hospital notes and selected scientific, medical and nursing texts available on delirium would be sufficient data sources for this study.

However, once I had begun the interview process and had completed interviews with five people who had been delirious and five family members I realised that a further and wider data source was needed. At the time the delirium team, comprising of a medical consultant and two registered nurses made themselves available to be interviewed. The interviewing of this additional group added another dimension to the object of delirium. In addition, the opportunity to interview another person who had been delirious

presented itself and so the final sources of data comprised of six people who had been delirious, their clinical notes, five family members, two registered nurses and one medical consultant, making a total of twenty sets of texts.

It was clear that the exact number of participants interviewed would depend on the number of volunteers, the range of experiences they represented and the amount of data collected in each of the interviews. The relatively small sample of participants is indicative of the labour intensive nature of utilising discourse analysis as a means of data analysis. This is supported by Potter and Wetherell who claim:

because one is interested in language use rather than the people generating the language and because a large number of linguistic patterns are likely to emerge from a few people, small samples or a few interviews are generally quite adequate for investigating an interesting and practically important range of phenomena (1994, p.161).

The age criterion for selecting the people who had been delirious was influenced by the array of gerontological texts and New Zealand public documents available identifying that in order to be considered an older person you needed to be over the age of 65 years (Ministry of Health, 2002; O’Neill, 2002). The method of obtaining participants for this study was to approach people over the age of 65 years who had been delirious or became delirious as part of their illness experience. Family members who had been associated with an older person who had experienced a delirious event were also approached to be interviewed.

The sampling method was a combination of purposive and convenience. Purposive sampling was utilised for the group who had been delirious and family members. Older people were approached after being nominated by the clinical charge nurses or the medical consultants at a large urban acute care hospital or at a medium sized assessment, treatment and rehabilitation hospital for older people. The clinical charge nurses and medical consultants were asked to identify clients who, in their assessment, were over the age of 65 years, were no longer delirious and did not have an associated dementing illness. Family members who had a significant other who had been delirious were also approached. On the other hand convenience sampling was utilised for the two registered nurses and the medical consultant. This group of people were known to me

and were available to be interviewed so in essence this approach is what Ingleton (1998) terms opportunistic.

As mentioned above, potential participants were initially approached by either the clinical charge nurses or medical consultants. Information about the study was communicated verbally and in written form through the provision of an information sheet (see appendices 1, 2 and 3). Potential participants needed to be deemed, by the clinical charge nurse or the medical consultant, no longer in the acute phase of their delirium and had to have the cognitive capacity to consent to participating in the research. If people met these two criteria and were interested in being interviewed they made contact with the clinical charge nurse or medical consultant, providing them with their contact details. These were then passed onto me. This approach was approved as the contact plan by the relevant ethics committees.

I tried, where at all possible, to make contact with people who had expressed an interest in participating in the study within 24 hours of receiving their contact details. Once contact was made and agreement was reached we identified a mutually convenient time and place for the interview to take place. If people were still in hospital then I would go in and personally meet them before organising an interview. If they had been discharged I made telephone contact with them in order to arrange an interview. If people preferred to be interviewed within the hospital setting then I arranged for a place for this to occur, otherwise interviews occurred in people’s homes. The same process was used for family members.

On the day before each scheduled interview I contacted each person again to check that they still wished to have a meeting. This gave potential participants the opportunity to decline to be part of the study without me being physically present and time to decide whether they were comfortable sharing their stories with someone they knew little about. Making contact with potential participants several times before the interview took place is consistent with viewing informed consent as a process (see Reid, Ryan & Enderby, 2001; Usher & Arthur, 1998). Process consent ensures vulnerable populations, for example older people who have been delirious, have plenty of opportunities to be reminded of and understand what they are consenting to be involved in. This was important for the older adults in the present research who may still have been

experiencing some residual cognitive difficulties, even though those difficulties were not sufficient for them to be categorised as still being delirious by nursing and medical staff. By utilising this form of consent I was attempting to ensure the participation of people with delirium was voluntary. All participants contacted wanted to continue with the interview. The fourteen people interviewed were all pakeha New Zealanders. This is not surprising given the population make up of the Canterbury area. Of the people who had been delirious three were male and three were female. All five family members were female. The two registered nurses were also female and the medical consultant male.

At each interview I confirmed that each participant had read the information sheet. I reiterated verbally what the study was about by saying “so you know that this study is about …” Doing this provided me with the opportunity to use my professional judgement to ensure all participants understood what the study was about and that they were able to voluntarily give their consent. In addition, this was also an opportunity for people to clarify any further questions that they may have had. Several people did have questions and I answered these as they arose. All people participating in the research were informed of their right to turn the tape off at any point and/or terminate their involvement in the research at any time. A consent form was signed before the interview began. A copy of the consent form can be found in appendix 4. Consent was also obtained to access the client’s hospital records. The purpose of this exercise was to identify for analysis as many representations of delirium as possible.