Semi-structured interviews of the physiotherapists were completed by a researcher (NK) experienced in qualitative descriptive methodology. Questions for these interviews are presented in Box 2. These questions sought to explore the physiotherapists’ perspectives of what worked well and provided additional value, what didn’t work well and potential challenges to delivering the approach from their own perspective, and their perceptions

of the patients’ perspectives. Patient interviews were conducted by a physiotherapist academic or research assistant experienced in qualitative interviews, who was not involved in providing the activity coaching intervention to the patient. For these interviews, questions explored what worked well, any added value of the program to their health selleck chemical and wellbeing, and anything they didn’t like or did not work well. Interviews lasted between 20 and 40 min, were audio recorded, and a denaturalised transcription Navitoclax concentration was used (Oliver et al 2005). What was your

overall impression of the activity coaching process? How have the activity coaching sessions affected your health and well-being? Has the programme affected other areas of your life? What have you liked about the activity from coaching process? What has worked well for

you? • Prompt to clarify what factors were most motivating and how these were identified if not already identified What has not worked well for you? What have you not liked about the process? Is there anything else you would like to tell us about the programme or how it has affected you that you would like to talk about? Do you have any suggestions for improvement? During the data preparation phase, each transcript was read through several times by two researchers (CS, SM) to first get an idea of the whole of each interview and notes were taken of impressions and thoughts (Sandelowski 1995). A data reduction framework based on the interview guide was used to prepare data for analysis (Sandelowski 1995). Data were analysed using conventional content analysis not only to identify themes of importance within and across the two participant groups, but also to look for any differences between experiences (Hsieh and Shannon 2005). Clusters of codes and categories were grouped to form core themes. A third researcher (NK) independently reviewed the codes as a form of member checking to ensure consistency of interpretation with identified themes and to ensure theme names adequately captured the data coded to that theme.