Methods
A qualitative approach was adopted, guided by a critical realist perspective. The phenomenon of interest (i.e. the smoking-related behaviour of people with MHPs) was explored through the lens of social cognitive theory (SCT). SCT can be used as a framework to explain how and why people acquire and maintain behavioural patterns; it can also inform the design, implementation and evaluation of behaviour-change interventions. It was, therefore, considered useful for our purposes. Social cognitive theory proposes that a person's behaviour both influences and is influenced by environmental factors (including other people's behaviour and social norms), personal factors (including motivation) and attributes of the behaviour itself. Central tenets of SCT include self-efficacy and the value placed on the perceived outcomes or consequences of a change in behaviour.
Participants were recruited by one of the authors (CW) from three Health Boards that serve a large geographical area in the west and south of Scotland. The target populations were smokers who had an enduring mental health problem and members of the multi-disciplinary health and social care team who had regular contact with this client group. Participants were recruited purposively from urban and rural areas with varying levels of socio-economic deprivation.
People with MHPs were recruited with the assistance of key workers in community-based mental health teams. Eligibility criteria included: being a current smoker; presence of an enduring mental health problem i.e. schizophrenia/delusional disorder, an affective disorder or a neurotic/stress-related disorder; being 18 years or older; living at home in the community at the time of recruitment. People were excluded if they had florid or very active symptoms of mental illness. The aim was to recruit 25 people with MHPs.
Health and social care professionals (HSCPs) were recruited with the assistance of senior managers in Community Mental Health Teams and the Scottish Primary Care Research Network. Participants included psychiatrists, mental health nurses, occupational therapists, psychologists, general practitioners and social workers. To be eligible to participate, the professionals had to have regular contact with people with MHPs. The aim was to recruit 50 HSCPs.
Ethical approval was granted by the Multi-Site Research Ethics Committee (MREC) in Scotland. The principles of the UK Data Protection Act (2003) were observed. All participants provided informed consent.
The data were collected by CW during one-to-one interviews (outlines of the interview guides are presented in Table 1). Participants with MHPs were normally interviewed in their own homes, with HSCPs being inter-viewed on the telephone. The interviews were digitally-recorded and lasted between 30 and 60 minutes.
Following verbatim transcription, the data were analysed with the assistance of qualitative analysis software i.e. NVivo v8. Drawing on the principles of 'framework analysis', a method commonly used in the applied health field, the data were analysed both deductively and inductively. Key issues and concepts were initially identified by drawing on a priori reasoning and linked to questions in the interview guide. The data were then indexed thematically (and inductively) based on the participants' discussion of key barriers and facilitators to smoking/smoking cessation. In the final stage, and with the aim of understanding the mechanisms involved, the findings were viewed through the lens of social cognitive theory. Data analysis was undertaken by SK and discussed and agreed with CK.