Eligible participants will be recruited from the Mary Pack Arthritis Program (Vancouver Coastal Health Authority) and Fraser Health Authority in British Columbia, Canada. Study information will also be posted on social media (Facebook, Twitter, Kajiji, Craigslist) and distributed by our patient pratner's organizations (Arthritis Research Canada, and Arthritis Consumer Experts). Individuals are eligible if they have a diagnosis of RA or SLE, have an email address and daily access to internet, and are able to attend the 1.5-h education session. We will exclude people who have previously used any physical activity wearables or are unsafe to be physically active without health professional supervision, as identified by the Physical Activity Readiness Questionnaire (PAR-Q) .
After completing the baseline measures, participants will be randomly assigned to the Immediate Group or the Delayed Group (i.e. control) in 1:1 allocation ratio. Randomization, stratified by diagnosis (RA or SLE), will be performed using numbers generated by SAS v9.4 (SAS Institute, Cary, North Carolina, USA) in variable block sizes to ensure adequate allocation concealment.
Wearable and online technology
The intervention will include a Fitbit Flex 2™ wristband. Fitbit® is a commercial wearable device which tracks and displays steps walked, gross level of physical exertion, and the time spent being active. Fitbit® has an open source platform that permits customization of a new app, FitViz, to enhance the use of the data as part of our activity coaching strategy. To ensure user friendliness, FitViz was co-developed with 3 patient research partners from Arthritis Research Canada and Arthritis Consumer Experts. Using FitViz, the participant can share information with a study PT who will coach them to set activity goals by phone and adjust the activity parameters in the app remotely. These parameters include: 1) the upper and lower bound of intensity and duration of MVPA, 2) the duration when a sedentary behaviour should be interrupted, and 3) the rest time in between vigorous activities (i.e., pacing). By combining the use of a wearable, an app and coaching from a PT, we maximize the use of behavioural change techniques for supporting people with arthritis to engage in an active lifestyle .
The Immediate Group will receive the 8-week intervention immediately after randomization. Participants will attend a 1.5-h session where they receive: 1) 20 min of standardized education about physical activity, 2) a Fitbit Flex 2™ and a FitViz app account, and 3) individual coaching by a study PT trained in motivational interviewing . The coaching will follow the Brief Action Planning approach , whereby the PT guides individuals to set goals, develop an action plan, and identify barriers and solutions. The PT will then adjust the activity parameters on the app based on the participants’ goals.
Participants’ physical activity will be captured continuously by the Fitbit® and wirelessly synchronized with FitViz 150 times/day. During Weeks 1–8, a study PT will review participants’ progress and coach them to modify their physical activity goals via 4 biweekly phone calls. A counselling guide will be used and the discussion will be documented by the PT. Participants may also contact the PT via email with questions. At the end of the intervention, participants may keep their Fitbit® and FitViz account, but will have no contact with a PT.
The Delayed Group will receive the intervention in Week 10. During the waiting period (Delayed Group only) or post-intervention period, participants will receive monthly emails of arthritis news, which are unrelated to physical activity.
To better understand the reasons people do or do not adopt and maintain recommended levels of physical activity, we will interview 20 participants with RA and 20 with SLE for 1 h by phone after the intervention. Interviews will focus on 1) goals set, strategies used, barriers/facilitators to being active, 2) their experience with the intervention, 3) the nature of activities they engage in, and 4) their experience of being a research participant. These data will enrich the RCT data, and inform the design of the future implementation strategy, if the intervention is found to be effective.
Participants will be assessed at baseline (T0), and Weeks 9 (T1), 18 (T2) and 27 (T3). Our primary outcome measure will be mean daily MVPA time measure with SenseWear Mini, a multi-sensor monitor that is worn on the upper arm over the triceps. It integrates tri-axial accelerometer data, physiological sensor data and personal demographic information to provide estimates of steps and energy expenditure. Tierney et al.  has showed that SenseWear is a valid tool for estimating energy expenditure during activities of daily living in people with RA (ICC = 0.72). A strong relationship was also found between SenseWear and indirect calorimetry measures of energy expenditure for activities of daily living (Pearson’s r = 0.85) . SenseWear can be worn 24 h a day. Hence, it can capture a full picture of physical activity and the off-body time throughout the day [28, 29]. An important feature of SenseWear is its ability to differentiate between sedentary and light physical activities , making it an ideal instrument to assess both active and sedentary behaviours. Participants will wear a SenseWear Mini for 7 days at each assessment. Almeida et al.  determined that a minimum of 4 days of wear is required to reliably assess energy expenditure from different levels of physical activity in people with RA (ICC > .80).
We will calculate the average daily MVPA accumulated in bouts per day. A bout is defined as ≥ 10 consecutive minutes at the level of ≥ 3 METs (i.e., the lower bound of MVPA), with allowance for interruption of up to 2 min below the threshold . Additional analysis will be performed with a cut-off at ≥ 4 METs which reflects purposeful activities .
Secondary outcomes will measure 1) mean daily time in sedentary behaviour, 2) average daily step count, 3) McGill Pain Questionnaire Short Form (MPQ-SF), 4) Fatigue Severity Scale, and 5) Partners in Health Scale. Sedentary behaviour and step count will be measured with SenseWear. For sedentary behaviour, we will calculate the mean daily time spent with an energy expenditure of ≤ 1.5 METs, occurring in bouts of > 20 min during waking hours [34,35,36,37]. The MPQ-SF contains 15 pain-related words, which can be rated from 0 to 3 (higher = more severe) . The Fatigue Severity Scale, which consists of 9 questions measuring the impact of fatigue, has demonstrated excellent internal consistency (Cronbach’s α = 0.89) . Construct validity was demonstrated by a moderate correlation with pain (r = 0.68) and depression (r = 0.46) . The Partners in Health Scale is a 12-item measure designed to assess self-efficacy, knowledge of health conditions and treatment, and self-management behaviour such as adopting a healthy lifestyle (Cronbach’s α = 0.82) .
Tertiary outcome will include Patient Health Questionnaire-9 (PHQ-9)  and Self-Reported Habit Index . The PHQ-9 consists of 9 questions (rated from 0 to 3) that correspond to the diagnostic criteria for major depressive disorder. A total score of greater than 11 indicates a major depressive disorder . A difference of at least 5 points indicates clinical change over time . The Self-Reported Habit Index is a 12-item scale, rated on a 7-point Likert scale, that measures characteristics of habitual behavior (reliability minimum α = 0.81). We will ask participants to rate their strength of habit for 3 specific activity-related behaviors: sitting during leisure time at home, sitting during usual occupational activities, and walking outside for 10 min. A higher score indicates a stronger habit or behaviour that is done frequently and automatically [43, 44].
Data analysis and monitoring
Our collaboration with health authorities and patient groups will allow the study to recruit 130 eligible participants within 24 months. In one of our proof-of-concept studies on a similar physical activity counselling intervention involving 61 people with osteoarthritis, we estimated a standard deviation (SD) of 52.0 min of bouted MVPA performed in sessions of ≥ 10 min (unpublished data). Assuming an attrition rate of approximately 15%, we anticipate 110 of the 130 participants will complete the study (55 per group). With a sample size of 110 and α-level of 0.05, we will have 80.5% power to detect a between-group difference of at least 25 min post intervention (via one-sided test).
Intervention Fidelity and adverse event monitoring
We will monitor intervention fidelity by tracking participants’ Fitbit/FitViz app usage statistics (frequency & duration of use) during the evaluation periods. Further, we will analyze PTs’ physical activity counselling records to ensure the discussions follow the brief action planning approach. Participants will report any serious adverse events (falls, cardiovascular and musculoskeletal events)  to the study coordinator at any time during the study period. In addition, we will ask participants to record all adverse events related to their physical activity in the follow-up questionnaire at Weeks 9, 18 and 27.
An intention-to-treat analysis will be performed by a biostatistician who is blinded to the group assignment. For the main comparison, we will use the Shapiro-Wilk test to assess normality of the outcome variables. If normality assumption is rejected a suitable transformation will be selected to achieve an approximately normal distribution . Analysis of covariance (ANCOVA) will be used to evaluate the effect of the intervention on the outcome measures, adjusting for 2 strata and blocking. If blocking is found to play no role, then it will be removed from the subsequent analyses.
Since we expect the randomization schedule to be implemented as planned, any differences between groups at baseline should be due to chance. Hence, the main analysis will not adjust for baseline differences [47, 48]. We will perform sensitivity analyses to adjust for baseline differences that appear to be clinically important to determine if they affect the conclusion from the main analysis. The first contrast will compare T0-T1 between the 2 groups to determine if the intervention is superior to the control. The second contrast will compare T0-T1 with T1-T2 in the Delayed Group. Unlike the first contrast which provides between-subjects treatment effect estimate, this second contrast uses within-subject pre-post comparison for treatment effect estimates. We will use linear mixed-effects longitudinal models to combine the first and second contrast for an overall treatment effect estimation. This combined estimate has the potential to substantially improve the precision of treatment effect estimates as compared with using either one alone. The third contrast will compare T0–T1 in the Immediate Group against T1-T2 in the Delayed Group. The forth contrast will compare T0-T1 in the Immediate Group against T1-T3 in the Delayed Group. The last two models will assess if the 10-week delay had an impact on the efficacy of the intervention. We will use descriptive analysis to summarize participant characteristics, comorbid conditions and adverse events, which will be adjudicated by the first author.
For the qualitative interviews, we will conduct an iterative content analysis, whereby codes will be identified and revised as interviews are analyzed. Initial open coding (i.e., assigning conceptual labels to the content) will be followed by clustering the labels into thematic categories. Quotes representative of the thematic categories will be identified to illustrate participants’ perspectives on physical activity, nature of activities, and their experiences as research participants. These data will inform the interpretation of statistical analyses and the design of future studies and implementation strategies, for example, ways for PTs to provide feedback about physical activity to people with inflammatory arthritis.