Design: This study adopted a complementary mixed-methods study design including a qualitative approach using focus groups and a prospective cohort study. Both study components were approved by the National Health Service (NHS) Health Research Authority, NRES Committee South West – Exeter, UK [Ref: 14/SW/1183]. All participants provided written informed consent and the study was undertaken in accordance with the Declaration of Helsinki.
Participants: Patients were recruited into the prospective cohort study from referrals into a NHS-run Inflammatory Arthritis Exercise Programme (IAEP) across Greater Glasgow & Clyde (GG&C) Health Board. The NHS is a nation-wide universal health care system in Britain which is free at the point of provision. GG&C Health Board is the largest Health Board in Scotland serving 1.2 million people with wide and variable socioeconomic characteristics. The IAEP is a 12-week exercise programme run by rheumatology physiotherapists across GG&C Health Board. Any adult within the Health Board who has a clinician confirmed IJD and is under the care of the Rheumatology Department can be referred into the programme. The prospective cohort study involved collection of data at 3 key time points: prior to commencing the IAEP (baseline) – assessment 1 (A1), post completion of the IAEP (12 weeks from baseline) – assessment 2 (A2), and 9 months from baseline – assessment 3 (A3). Participants were recruited into the focus groups from the prospective cohort study between A2 and A3 – see Fig. 1.
Inclusion criteria: Patients referred into the IAEP were included in the study if they met all of the following inclusion criteria: 1) physician-confirmed diagnosis of IJD such as Rheumatoid Arthritis, Psoriatic Arthritis, Ankylosing Spondylitis or any other type of inflammatory arthritis/polyarthritis, 2) were aged 18 years or over.
Exclusion criteria: Patients were excluded from the study if they met any of the following criteria: 1) did not provide informed consent to be part of the study, 2) were unable to complete the study within the designated data collection period, 3) the presence of co-morbidity severely limiting the patient’s ability to participate in an exercise programme such as unstable angina, heart failure, uncontrolled heart arrhythmias, uncontrolled hypertension, severe respiratory condition, uncontrolled epilepsy, uncontrolled diabetes, recent medical instability such as a stroke, wheelchair user and pregnancy.
Recruitment Strategy: Sampling was undertaken by convenience through identification of eligible participants from consecutive referrals. The sampling frame was limited to the study population of interest which comprised of patients who were under the care of the Rheumatology Department across GG&C Health Board and who were referred into the IAEP between March 2015 to July 2017. Referrals into this programme were made by rheumatology consultants, rheumatology nurse specialists, rheumatology allied health professionals and patients via self-referral. Every patient who was referred into this programme and met the inclusion/exclusion criteria for the study was informed in writing and verbally of the research project by their rheumatology specialist physiotherapist at a screening appointment prior to attending the programme. If the patient was interested in being part of the study, they were subsequently contacted by the researcher for further information. Once willingness was confirmed, participants were booked in for their baseline session where written informed consent was obtained.
Prospective Cohort Study Data collection: Data was collected by the researcher (KB) at each study time point (see Fig. 1). Health related quality of life (HRQoL) was measured using the Short Form – 36 (SF36) and Hospital Assessment Questionnaire – Disability Index (HAQ-DI); self-perceived levels of control were measured using the Arthritis Self-efficacy Scale (ASES); attitudes and beliefs towards physical activity were measured using the Exercise Attitudes and Beliefs Questionnaire for patients with RA (RA-EAQ); and mental health was measured using the Hospital Anxiety and Depression Scale (HADS), all of which have good psychometric properties which have been verified in populations with IJD [22–26]. The Scottish Index of Multiple Deprivation (SIMD) is a composite measure of social deprivation which has seven domains: current income, employment, health, education, skills and training, housing, geographic access and crime. These seven domains are calculated and weighted for small areas, called 'data zones', with roughly equal population and can be obtained using participant postcodes .
The Disease Activity Score (DAS-28) was recorded as a marker of disease activity by the researcher who was trained in undertaking the DAS-28. Acute phase reactants from blood test results (within 3 months of each data collection session) were obtained from the patient’s medical records to complete the DAS-28 score. Disease duration was measured from the date of physician-confirmed diagnosis which was obtained from the participant’s medical records. Drug therapy was obtained from the patient’s medical records and clarified with the patient in case of any recent changes; the level of pain on average over the past week was measured using a pain visual analogue scale (VAS) and the level of fatigue was measured using a 100mm fatigue VAS [8, 28].
To evaluate whether there are any physical-condition-related and/or environmental factors that could determine physical activity levels and sedentary behaviour the following measurements were undertaken. Body Mass Index (BMI); 6-minute walk test [29–32]; grip strength using a JAMAR grip dynamometer and the Southampton protocol [29, 32, 33]; and a custom-made environmental questionnaire to elicit information concerning attendance to an exercise facility or exercising independently, cost, affordability, transportation to/from and the variety of activities on offer at the community exercise facilities. A successful transition could be determined from this questionnaire which was defined as still exercising in the community 6-months post discharge from the NHS-run IAEP. This questionnaire was developed with assistance from the study Advisory Board which consisted of rheumatology clinicians, NHS health improvement officers, patients and academics.
Focus Group Methods:
Topics for discussion were developed with assistance from the study Advisory Board which consisted of rheumatology clinicians, NHS health improvement officers, patients and academics. Topics were attitudes towards exercise, beliefs about the impact of exercise on their disease, other personal factors that can act as barriers or facilitators towards sustained healthy exercise behaviour and environmental factors. They discussed how these attitudes and beliefs have changed by participating in the IAEP and how they are self-managing in the community. Three focus groups were conducted with patients who were recruited using purposive convenience sampling from the prospective cohort study. The researcher (KB) lead the semi-structured focus groups with an assistant (MC) who recorded level of consensus using a focus group consensus matrix  (supplementary material). Both researchers were physiotherapists who had undertaken training in qualitative research. All focus groups were recorded using a digital voice recorder and recordings were transcribed verbatim.
Analysis: Descriptive statistics were used to summarise the cohort study data. All variables were then assessed for normality of distribution using the Sharpiro-Wilk test. Paired t-tests or Wilcoxon tests were undertaken for two consecutive assessment timepoints; one-way repeated measures ANOVAs or Friedman’s tests for three consecutive assessment timepoints. Data analysis was undertaken using IBM SPSS version 26 and statistical significance level was p < 0.05. Three focus groups were undertaken. The same themes ran through all the focus groups suggesting theoretical data saturation was reached. Micro-interlocutor analysis  was used to analyse the focus group data which included thematic analysis of the focus group transcriptions with additional analysis of the matrix for assessing the level of consensus within the focus groups. This enabled group dynamics to be included in the data analysis which increases scientific rigour of focus group analysis . To further enhance scientific rigour two researchers (KB, MC) independently analysed the transcripts using thematic analysis to confirm emerging themes [35, 36]. A final discussion was conducted between the researchers where the data from the three focus groups were integrated, discussed and clarified using Micro-interlocutor analysis . After completing the quantitative and qualitative analyses independently, data from both sets were linked for a more robust understanding of findings. Areas of congruence and incongruence were explored by confirming the statistical results against the qualitative results. The adapted ecological model of the determinants of physical activity  was then used as a framework to describe the findings from both the quantitative and qualitative analysis. The framework has 5 main categories: individual, interpersonal, environment, regional or national policy and global. This framework has been used to describe the determinants of physical activity in adults and children across the world .