A grounded theory approach was adopted,11
using interviews to explore how GPs, patients, and informal supporters of depressed individuals understand depression and its management.
Participants were recruited in and around Southampton, apart from two GPs in Leicester. A total of 135 GP study information packs and 1078 patient and supporter packs were mailed to potential participants at seven primary care trusts, and posters were displayed in a students' union, public library, and a hospital. Each pack contained a response slip and freepost envelope for individuals to submit contact details. Participating GPs came from 28 practices. Most of the patients and supporters were recruited through 10 of these practices via mail or face-to-face distribution of information packs by practice staff. Thirteen were recruited through other routes through mental health support groups, a carers' group, a youth service, poster advertising, snowballing (participant identification of further participants), and word of mouth.
The patient and supporter information sheets stated that ‘The purpose of this study is to explore the different ways people experience and understand depression. To do this, we will be asking you about your thoughts and/or experiences of depression’. The GP information sheet stated that ‘The purpose of this study is to clarify how medical practitioners, their patients, and their patients' significant others understand depression in relation to its management’.
A total of 147 participants agreed to participate. As recruitment criteria and data gathering were refined through theoretical sampling and analytic saturation, 23 offers of participation to patients and supporters were declined; a further seven could not be contacted and six withdrew. Sampling was carried out purposively to obtain a diverse range of participants, summarised in . A total of 111 participants were interviewed: 61 patients (28 who were experiencing an episode of depression at the time of the interview, 18 with a past history of depression, and 15 who had never been depressed), 18 supporters, and 32 GPs. Some of the supporters and GPs had experienced depression themselves. Exploring GP experiences of depression had not been an aim of the study, and so participating GPs had not given informed consent for these issues to be explored during the research interview. Therefore in-depth information was not gathered on these experiences, and they are not included in the analysis. Most participants with experience of depression had suffered recurrent or persistent, rather than acute, depression. The group was heterogeneous as to severity of depression, which was self-defined rather than based on standard psychiatric classification.
How this fits in
Guidelines for depression management have been developed but little is known about GP and patient goals, which are likely to have an impact on offers of treatment, uptake, and adherence. While GPs take patient preferences for treatment into account, their perceptions of patient attitudes are only moderately related to patients' self-reports. GPs, patients, and their supporters describe a wide range of different concepts of depression and goals for its management. GP goals for managing depression may be perceived as irrelevant or unachievable by some patients, and GP responses may be considered as unhelpful; therefore findings emphasise the value of listening to patients, and sensitivity to alternative perspectives.
Interviews were conducted between May 2002 and March 2004 and usually took around an hour, ranging from 30 to 150 minutes. Data collection ceased when saturation of categories was achieved. Data were collected and analysed iteratively, starting with a semi-structured topic guide (Box 1
), but allowing the interviewers to follow participants' responses, gradually focusing on emerging themes (for example, in order to further explore the emergent theme of goals for the management of depression, later participants were asked about their main goals and how they viewed their GP's role in achieving these). All team members engaged in the analysis, which followed grounded theory procedures,11
and assumed the principles of the ‘critical realist perspective’,12
or ‘subtle realism’.13
Each transcript was analysed independently by the interviewer and another team member. For each group of participants (for example, GPs, supporters), one researcher reviewed the whole dataset.
Transcripts were divided into meaning units (the smallest self-explanatory piece of information), which were grouped into categories. Constant comparison of units and categories stimulated thinking on the properties of, and relations between categories which developed iteratively from the descriptive towards higher-order, more abstract categories. Theoretical sampling (recruitment of participants guided by the emerging themes so as to develop the analysis by obtaining crucial new information) and comparison of data from different participant groups constituted a form of triangulation;14
the goal in this context was completeness rather than convergence or consensus.15
The interdisciplinary team procedures helped to develop both depth and range of meaning in the data, and to assure the ‘trustworthiness’ of the analysis.16–18
- an audit trail for transparency of methodology and analysis development (for example through transcribing analysis meetings);
- all team members keeping reflective diaries, discussed regularly during meetings, which tracked developing ideas and explored the links between the researcher and the research; and
- agreement on ‘waving a red flag’,17 should any researcher be concerned that unwarranted assumptions or beliefs were beginning to drive the analysis.