Study design

The study design was a feasibility randomised trial with the published protocol (version 1) [19] and the final trial protocol registered on ClinicalTrials.gov NCT03644160. Ethical approval was granted by the University Hospital Limerick (UHL) Research Ethics Board (approval number 064/19). Reporting followed the relevant extensions of the Consolidated Standards of Reporting Trials [20].

Participants and setting

Inclusion criteria for the intervention part of the study were: adults (over 18 years) with a diagnosis of RA based on the American College Rheumatology 2010 criteria and low PA levels using the Godin−Shephard Leisure−Time Physical Activity Questionnaire (GSLPAQ) [21]. Patients identified from clinic charts as having a diagnosis of RA were approached by the study research assistant who introduced the study and completed the GSLPA) with each patient. If the patient had a score of less than 14 on the GSLPAQ they were given an information leaflet on the study and any questions were answered. A consent form was given and a time was arranged for a follow-up call within the next week to determine the person’s interest or not in participating the in the study. Informed consent was then obtained prior to formal study enrolment at the arranged baseline meeting with the study team. Patients continued with their usual care throughout the study.

For the interview part of the study patients in both intervention and control arms were invited to participate following completion of the main study. Health care professionals involved in rheumatology clinics the patients were recruited from (rheumatologists, medical teams and nursing staff) as well as the sessional physiotherapists who delivered the intervention were invited to participate on completion of the study. Informed written consent was recorded for each participant in the interview part of the study.

The study was conducted at University Hospital Limerick (UHL), Ireland in the outpatient rheumatology department (recruitment) and the hospital’s Clinical Education and Research Centre (CERC) (intervention delivery and assessments). Participants were recruited from weekly outpatient rheumatology clinics in UHL across two periods (due to interruption from Covid-19) between October 2019 and March 2020 and November 2020 to May 2021 (9 months total).

Patient safety was monitored throughout the study by the study team. Discontinuation criteria were determined at the outset of the study as follows: a) immediate discontinuation will occur, regardless of the stage of the study, if participants develop any severe adverse events during the study which may or may not be related to the intervention—such as chest pain; b) Participants could also choose to discontinue the intervention at any time for any reason. The type and frequency of adverse events were recorded and reported to the study Data and Safety Monitoring Committee (DSMC) if they occurred.

Randomisation, allocation concealment and blinding

The CERC Unit provided a computer randomisation system. Randomisation was performed by computer generated random numbers with a 1:1 allocation ratio to the intervention or control arms. The allocation sequence was generated independent of the study team. Allocations were stored in a locked cabinet. Participants were informed of allocation by the research assistant after completion of their baseline assessments. Each participant was assigned a code number, which was used on all outcome assessments in place of their names. The assessor was blinded to group allocation. The person assigning the allocations was blinded to allocation until after the participant was assigned. The statistician who completed the data analysis was blinded to group allocation.

Intervention and control group

The PIPPRA trial is a theoretically underpinned physiotherapist led BC intervention to promote PA delivered over 4 1:1 sessions by a sessional physiotherapist in person or virtually over an 8 week period. The intervention was tailored to each individual depending on the discussion with the sessional physiotherapist on the type of PA they were undertaking. Each session was a maximum of one hour. The original design was in-person intervention delivery only; however due to the timing of the study during the Covid-19 pandemic, virtual sessions were necessitated to ensure the completion of the study. Virtual sessions were delivered using WhatsApp video calls. The in-person sessions were delivered in a meeting room with minimal infrastructure and no exercise equipment. This was deliberate to ensure the transfer of the intervention delivery to any setting and eliminate the reliance on specialist settings or equipment.

PIPPRA comprises individual sessions based on the specific BC techniques, as described by the Behaviour Change Taxonomy [11]. A detailed description of the final programme is included in supplementary file 1 and in the protocol paper [19]. Briefly, across the 4 sessions participants were guided using a range of ‘education’ techniques’ (e.g. instruction on how to perform a behaviour, discussion on health consequences), ‘enablement’ techniques (e.g. goal setting, problem solving, action planning, feedback, self-monitoring, social comparison) and ‘modelling’ techniques. A PA behaviour plan was also prepared by each participant and reviewed at each session and adjusted as needed and agreed by the participant and therapist. Both control and intervention group received an information booklet about PA based on the existing guidelines for people with RA (Supplementary file 2).

Therapist intervention training

The sessional physiotherapist was trained in BC techniques and motivational interviewing by the study’s postdoctoral researcher and co-author (LL). The postdoctoral researcher was a qualified physiotherapist with a PhD in the area of behaviour change for PA in rheumatology and also co-applicant on the study and lead author on many of the underpinning studies. The sessional physiotherapists were newly qualified therapists with no experience in specialist rheumatology settings. The rationale for involving newly qualified therapists was to explore if a BC intervention could be delivered with minimal specialist training. A detailed session log of each session was maintained by the sessional physiotherapists (supplementary file 3a and 3b).

Modifications to study design due to COVID-19

COVID-19 restrictions in Ireland came into effect in March 2020 and necessitated a transition, after a 5 month pause of the study, from in-person to remote, virtual assessments and intervention delivery [22]. The COVID-19 pandemic also resulted in lost data as we were unable to follow-up some participants who withdrew from the study for reasons linked to Covid-19.

Data collection

The data collection took place at baseline (Time 0 (T0), 12 weeks (Time 1 (T1)) and 24 weeks (Time 2 (T2)). Questionnaires were completed via phone call with participants ActivPal devices were sent to and returned via post. The data were downloaded from the ActivPal to the study laptop and the output file was uploaded to the e-Case report form database (Clindox Limited, hosted by the University of Limerick encrypted server). Participant’s out of pocket expenses were reimbursed using a shopping voucher.

Primary outcomes

A number of feasibility targets were calculated:

Recruitment and retention

Feasible eligibility—the total number of eligible participants from UHL group rheumatology clinics

Recruitment rate—the number (%) of participants recruited and the rate of participants recruited to the study.

Refusal numbers—the number (%) of eligible participants that refused to participate and reasons why.

Protocol adherence

Minimum average attendance—the number of participants who attended over 80% of the intervention sessions.

Minimum outcome assessment target—retention of at least 80% of recruited participants with valid 12 (T1) and 24 (T2)—week primary outcome data, i.e. less than 20% attrition at outcome assessments at 12 and 24-week follow-up

Intervention acceptability

Semi-structured interviews were undertaken on completion of recruitment and the intervention with patients and the HCPs (rheumatology clinic medical team and nursing staff and sessional physiotherapists). Patients and HCPs were provided with verbal and written information about the study. Each participant provided informed consent prior to participation. The interviews aimed to determine patient and HCP acceptability of the intervention, understand some of the trial design and processes and to consider the acceptability of the secondary outcomes used. Interview questions for the interview guide (Supplementary file 4) were developed from an extensive literature review on behaviour change interventions to promote physical activity behaviour in people who have RA [9, 10] and from the previous qualitative research in this area [17, 18]. Interviews were audio-recorded and conducted by co-authors LL and S McK, chartered physiotherapists and researchers who have experience of undertaking and analysing qualitative interviews. Audio recordings from the semi-structured interviews were transcribed, anonymised and saved to a password protected laptop from University of Limerick. Each interviewee was offered a copy of their transcript to review and advised to amend the transcript if necessary. Once interviewees were satisfied that the transcript reflected their views and opinions accurately the finalised transcript was included in data analysis. Preliminary data analysis was conducted concurrently with data collection, to enhance understanding about the questions being asked and facilitated minor revisions of the questions. The analytical approach for this data was content analysis. NVivo (version 14 QSR International) was used to support the qualitative analysis.

Secondary outcomesPhysical activity

The ActivPAL™ (PAL Technologies Ltd, Glasgow, UK) measured time spent sitting, standing, lying and time spent in moderate and vigorous intensity aerobic physical activity (MVPA) [16]. It has been validated in RA samples [23]. The ActivPAL™ is a small, lightweight activity monitor that uses proprietary algorithms. It is a tri-axial accelerometer that produces a signal related to thigh inclination and needs no calibration before use. Recordings were processed for daily minutes of moderate to vigorous PA. The ActivPAL™ was worn for 8 days beginning week 1 before the start of intervention (T0, for 8 days 1-week post-intervention (T1) and at the 24-week assessment (T2). The first 24 h of recording were not included in the analysis to minimise the effects of reactivity. A minimum recording duration of 3 days from the 7-day period, including at least one weekend day was required for data processing; samples of lower than 3 days were not included.

The Yale Physical Activity Survey (YPAS) [24] measured participants self-report PA and has been used in the previous arthritis studies and older adults [25, 26]. The 2-part YPAS measures PA over a time period of a typical recent week (part 1) and from the past month (part 2). The total energy expenditure per week was calculated from part 1 of the questionnaire for each participant at each time point.

Psychological beliefs

The theory of Planned Behaviour Questionnaire (TPBQ) consists of three elements; beliefs about the likely consequences of the behaviour (behavioural beliefs), beliefs about the normative expectations of others (normative beliefs), and beliefs about the presence of factors that may facilitate or impede performance of the behaviour (control beliefs) [27]. Behavioural beliefs produce a favourable or unfavourable attitude toward the behaviour; normative beliefs result in perceived social pressure or subjective norm; and control beliefs give rise to perceived behavioural control. Thus, attitude toward the behaviour, subjective norm, and perception of behavioural control led to the formation of a behavioural intention. The TPBQ aims to capture all three elements which contribute to behavioural intention.

Disease activity

The Disease Activity Score 28 (DAS28) [28] was used to combine single measures into an overall, continuous measure of RA disease activity. The DAS28 includes a 28 tender joint count, a 28 swollen joint count, erythrocyte sedimentation rate, and a general health assessment on a visual analogue scale. After study interruption due to COVID-19 this measure was not recorded as participants were not able to attend the unit physically for blood draws.

Pain

Pain was recorded using a 10 cm visual analogue scale (VAS), which is sensitive to detecting changes in pain in inflammatory conditions and has good reliability and validity [29].

Fatigue

The Bristol Rheumatoid Arthritis Fatigue Multi-Dimensional Questionnaire (BRAF-MDQ) [30] measured the impact of fatigue for people with RA and disease specific. It has acceptable to good convergent validity [31] and consists of 20 items combined to create 5 scores, where a higher score reflects worse fatigue. The total score range is 0–70.

Sleep

Sleep was assessed using the Pittsburgh Sleep Quality Index (PSQI) [32] and has been used in previous arthritis studies [33, 34]. It is a 19-item self-rated questionnaire for evaluating subjective sleep quality over the previous month. The 19 questions are combined into 7 clinically-derived component scores, each weighted equally from 0 to 3. The clinical and psychometric properties of the PSQI have been formally evaluated by several research groups [33].

Quality of life

The Rheumatoid Arthritis Quality of Life Questionnaire (RAQoL) measured disease related quality of life [35] on ADLs, social interaction, emotional well-being, and relationships. The questionnaire consists of 30 statements that have a yes/no response. Items are scored one for yes and zero for no. Scores for each item are summed to give an overall quality of life score with a higher score indicating a poorer quality of life.

Sample size and data analysis

The target sample size for the pilot study of 40 participants, with 20 participants in both the control and intervention groups was determined pragmatically. This sample size was expected to provide sufficient data to meet the primary outcomes and was in the appropriate range for pilot studies [36]. Descriptive statistics were used to determine the recruitment and retentions rates and for the secondary outcomes. Data analysis was undertaken in SPSS version 27 (IBM Corporation, New York, USA). Given the small sample size it was not deemed appropriate to present differences between arms.

Progression criteria

The following criteria were agreed for progression to a full trial.

1.

In order for a definitive trial to be feasible we project that we need to recruit participants at a minimum recruitment rate of 4 per month; i.e. 48 per year. The primary feasibility target is a minimum recruitment rate of 4 participants per month.

2.

Following recruitment of participants, a minimum attendance at 100% of the intervention sessions is expected. A second feasibility target is thus a minimum average attendance by recruited participants of 100% of the intervention sessions.

3.

Outcome assessment at 12-week follow-up—retention of at least 80% of recruited participants with valid outcome assessments, i.e. less than 20% attrition. Thus a third feasibility target is a minimum retention of 80% of recruited participants providing valid 12-week outcome data