The PhysioDirect service investigated in the randomised trial is an example of a complex intervention and, in line with MRC guidance,98 a qualitative study was nested alongside the main trial. The aims of the qualitative investigation were to explore the acceptability of PhysioDirect, to determine the key elements in how the service was perceived, and to understand the key barriers and facilitators to the successful implementation of the service. The perspectives of patients, physiotherapists, physiotherapy managers, GPs and PCT commissioners were sought through one-to-one, semistructured interviews in person or by telephone. Each of these interviews was directed by a topic guide informed by available literature. Appendix 3 shows an example of a topic guide. Patients, managers, GPs and commissioners were interviewed at one point in time, while longitudinal qualitative data were collected with a sample of physiotherapists (before and after the trial) to investigate if and how their views changed over time in the light of their experience of providing the PhysioDirect service. This chapter focuses on the beliefs and experiences of only patients and the physiotherapists who provided the PhysioDirect service and their managers.
Data collection and sampling
Patient interviews
Interviews were carried out after the patient's discharge from their physiotherapy episode of care in the trial. The aims were to explore the accessibility and acceptability of the physiotherapy service, the influence of PhysioDirect and usual physiotherapy care assessment and treatment on patients' perceptions of their care and its impact on their MSK problem, and to explore whether and how patients perceived that services could be improved.
Purposive sampling based on age, sex, presenting clinical problem and participating PCT was used to generate a diverse range of patient perceptions and experiences. In total, 388 patients were invited to take part in the qualitative interview study between August 2009 and April 2010, with 82 agreeing to be interviewed and 57 interviews being conducted. The reasons why the number agreeing to be interviewed and the final number of interviews were different were due to the difficulties in arranging convenient times for interviews and saturation of the data, i.e. no new themes emerged after the first 57 interviews.
provides a summary of the characteristics of patients who took part in interviews, showing their key characteristics according to each of the sampling criteria. Slightly more women were interviewed than men, and although the average age was 58 years, there was a large range of ages, with the youngest person interviewed being 19 years and the oldest being 87 years. More patients were interviewed in the group that was randomised to the new PhysioDirect service than in the usual-care group, including those who received some or all of the components of the new service, as well as those who were randomised to the new service but who never telephoned or contacted the service. This was deliberate in order to fully explore patients' perceptions of the new service. To facilitate comparisons between the new PhysioDirect service and the usual physiotherapy service, nine interviews were conducted with patients randomised to usual care. Interviews were conducted with patients from each of the four participating PCTs, and patients were sampled to ensure a breadth of MSK problems affecting different bodily regions.
Summary of patient characteristics according to the interview sampling criteria
Interviews with physiotherapists and their managers
The aim of the interviews with physiotherapists and their managers was to identify the salient issues involved with providing and managing the new PhysioDirect service and to explore its acceptability and implementability. A longitudinal design was used with the participating physiotherapists so that each was interviewed twice: once before treating patients in the PhysioDirect trial, and again when the trial had finished. The aim of the longitudinal design was to obtain insight into each physiotherapist's views, hopes and concerns prior to using the new service with patients and then in the second interviews to explore their experiences of using PhysioDirect with patients, following up on the key issues arising from the first interviews, while also exploring if there were any changes over time. The physiotherapy service managers were interviewed once after the trial had finished and these interviews explored the perceived impact of the PhysioDirect service, how it had been set up in each PCT, what it was like to manage the new service and other issues of importance to service managers which were perceived to facilitate or hinder its implementation or use.
From a total sample of 32 physiotherapists trained in the delivery of the PhysioDirect service, 16 physiotherapists were identified to be interviewed – four from each PCT. In each of the four PCTs, the key physiotherapy service manager overseeing the operational issues of the physiotherapy service was also invited for interview. The purposive sampling of physiotherapists was based on sex, clinical experience, whether they worked exclusively in the NHS or also in private practice, their experience of using telephone assessment systems previously and the circumstances under which they had became involved in the trial. All physiotherapists and managers who were invited for interview agreed and were interviewed.
provides a summary of the physiotherapists and their managers who took part in the interviews, showing their key characteristics according to sampling criteria. Of the 16 physiotherapists interviewed across the four PCTs, 75% were female. They had a range of MSK outpatients experience ranging from 1 to 30 years, with an average MSK experience of 12 years. The majority (81%) worked exclusively in the NHS and 62.5% of the physiotherapists had previous experience of using telephone assessment within physiotherapy. Half of the physiotherapists reported that they had been invited to participate in the PhysioDirect trial by their managers, while the other half reported that they had volunteered to deliver the new PhysioDirect service in the trial. Of the four managers interviewed, 75% were female and had a range of experience of managing MSK services from 1 to 8 years, with an average of 4 years.
Summary of physiotherapists' and managers' characteristics according to the interview sampling criteria
Interviews with GPs and commissioners
The aim of the GP and commissioner interviews was to explore some of the key organisational and contextual issues that might influence the acceptability and implementation of the new PhysioDirect service. In each PCT, GPs and commissioners of MSK services were interviewed after the trial had finished but before the results of the trial were known. The sampling criterion for the commissioners was based upon whether they had a clinical or non-clinical background. A total of eight commissioners, two from each PCT, were identified and invited to take part in the interview study, of whom four (one from each PCT) responded and agreed to participate. GPs based in practices that had participated in the trial were invited to take part in the trial, based upon two key criteria: the PCT and their referral patterns to physiotherapy (high vs low referrers to physiotherapy). Eighty GPs were sent postal invites of whom 26 replied, 15 agreed to be interviewed and eight GPs in total were eventually interviewed (two from each PCT area). The GPs and commissioners were interviewed at their place of work.
provides a summary of the GPs and commissioners who took part in the interviews, showing their key characteristics according to the sampling criteria. Of the eight GPs interviewed across the four PCTs, five were male. They had a range of general practice experience ranging from 10 to 30 years with an average experience of 20 years. Of the four commissioners interviewed, one was female, three had a non-clinical background and overall they had a range of experience of managing MSK services from 1 to 5 years with an average of 2 years.
Summary of GP and commissioner characteristics according to the interview sampling criteria
Recording and management of data
All of the interviews were digitally recorded. The audio recordings were stored on the researchers' password-protected drive and the audio files on the Dictaphone were then erased. The audio files were transcribed in full and saved in Microsoft Word (Microsoft Corporation, Redmond WA, USA) format and then anonymised. The transcripts were saved to both a password protected computer and the Framework software programme. Each case was given an identification number.
Analysis
A Framework approach was used to analyse the qualitative data from each of the three participant groups.99 Framework is an analytical method that is pragmatic and is commonly used in health services research.100 The Framework approach is a thematic, cross-sectional analysis that allows the researcher to simultaneously analyse across themes and cases. It enables qualitative findings and interpretations to build from the original data, allowing the analysis to maintain a clear auditable trail. The approach is a matrix-based method for analysing qualitative data that includes familiarisation with the data, the creation of a theoretical framework, indexing the data according to the Framework and the creation of summaries from the indexed data. These summaries are then finally mapped in charts and descriptive explanatory accounts are created. The audio files of the patients, the physiotherapists and their managers as informants were listened to several times and transcripts reread to identify key themes and concepts. The data were sorted and reduced to a manageable form, and a theoretical index was refined to summarise the essence of the transcripts. In order to ensure confirmability and trustworthiness, a sample of transcripts was double coded and the thematic framework was reviewed by the research team before it was applied to all of the data. A data summary was attached to each of the data labels on the index. Large charts of the index headings and attached summaries were created and descriptions that captured the essence of the summarised data across the cases were made with key themes identified. Finally, the themes were mapped and interpretation of those themes was reviewed by several study team members in order to construct overall explanations of the data.
