Design
The study was a single-blind RCT designed and reported in line with the CONSORT statement and checklist (Figure 1) [23], and the TIDieR checklist [24]. The completed CONSORT and TIDieR checklists are included in Additional files 1 and 2. The trial was registered with the Australian and New Zealand Clinical Trials Registry (ACTRN12619000036112) prior to participant recruitment.
Participants
Participants were recruited from an ambulatory, secondary care clinic in a major tertiary hospital in regional Victoria between January 2019 and September 2019. Recruitment involved consulting orthopaedic and general surgeons providing, during the normal course of the consultation, a verbal recommendation to engage in PA coaching and a sequentially numbered research flier (Additional file 3) to adult patients who, in their view, would benefit from increased PA. Potential participants contacted the research team of their own volition using information on the flyer. Sequential numbering allowed the research team to quantify both the number of individuals referred by surgeons and the number who subsequently acted on this referral.
Participants were included if they were between 18 and 69 years, and did not meet the recommended PA guidelines [25]. A single item question “As a rule, do you do at least half an hour of moderate or vigorous exercise (such as walking or a sport) on five or more days of the week?” was used to identify insufficiently physically active individuals [26]. The question has demonstrated good sensitivity (76.7%), high specificity (81.1%), and a high positive predictive value (86.7%) for identifying those not achieving the recommended 150 minutes of MVPA per week [26]. Exclusion criteria included: sufficiently physically active; an existing medical condition that contraindicated PA (indicated by the Physical Activity Readiness Questionnaire); deaf/hearing impaired; poor comprehension of English language; disabling neurological disorder; severe mental illness such as psychosis; learning disability; dementia and cognitive impairment; registered blind; housebound or resident in nursing home; non-ambulant; pregnancy; advanced cancer; and scheduled for surgery with 30 days of the clinic presentation.
Procedure
Participants who were eligible and consented to take part were randomly allocated to either the intervention or the control group based using block randomization generated by a computer generated random number sequence (randomizer.org). The allocation sequence was generated by a research assistant who was not involved in the data collection process. Assignments were prepared and sealed in sequentially numbered opaque envelopes. Intervention assignment was made by opening the next envelope in the sequence, after: (1) the recruiter had determined eligibility for the study; (2) participants had consented to take part; (3) attendance at an education session was confirmed; and (4) baseline measurements were completed.
Intervention
All enrolled participants attended an education session prior to group allocation. The education session was a facilitated learning session based around PA self-management and was carried out using a self-determination theory (SDT) framework [27]. SDT is a motivational theory commonly used to illicit changes in PA [28]. SDT was used in this group setting to support, educate and motivate participants around PA changes [27].
The intervention group completed a telephone-based coaching intervention that comprised integrated motivational interviewing and cognitive behaviour therapy (MI-CBT). The coaching was delivered in five by 20-min sessions over 12 weeks. The intervention schedule, theories and techniques are displayed in Table 1. The intervention was delivered using a motivational interviewing (MI) framework, with MI used as the underpinning approach to influence motivation, ambivalence and self-efficacy to be physically active [29]. MI microskills (open-ended questions, affirmations, reflections and summaries) were utilised in all sessions [29]. The CBT coaching focused on six theory-derived determinants of PA: PA outcome expectations, PA outcome experiences, PA values, PA barrier self-efficacy, social support and relapse prevention [30]. The cognitive behaviour therapy (CBT) strategies were incorporated within a MI framework to support PA change and maintenance [31]. The intervention was delivered by an experienced Australian Health Practitioner Regulating Authority registered physiotherapist. The clinician was trained in MI-CBT, including workshop attendances, and one-on-one coaching from an experienced practicing psychologist. The clinician previously delivered 145 hours of MI-CBT in the H4U study [20]. Fidelity was assessed using the MI-CBT fidelity scale [32] using audio-recorded sessions.
Outcome Measures
Participants’ outcome measures were recorded at baseline, after 3 months of intervention (post-intervention) and at 9 months (follow-up) by assessors blinded to the study group assignment. Information on treatment allocation was not provided to data collectors. The data collection forms only contained unique identifier codes assigned to each participant.
Primary outcome
The primary outcome of interest was the change in MVPA (minutes/day) over time. MVPA was objectively assessed using a tri-axial accelerometer (wGT3X-BT; Actigraph, USA). Participants were instructed to wear the accelerometer on their hip at all times over 7 consecutive days, excluding sleep and water-based activities. PA was calculated using the manufacturers software (Actilife; Actigraph, USA) with cut points by Freedson Adult (1998) used to provide daily measures of MVPA (> 1951 cpm) [33]. Accelerometer wear time was based on activity counts per minute (CPM). Non-wear time was defined as 60 minutes or more of consecutive activity counts of zero, with a spike tolerance of 2 min and 100 cpm. Participants used logbooks to report activities and periods of accelerometer non-wear. Non-wear time was compared to participants’ notes on their logbook. A minimum of 10 h per day was used as the cut-off for a valid day of measurement and a minimum of 5 days of data were required including at least 1 weekend day [34]. Weekly MVPA was computed based on the average of all valid days per person. A daily average of 21 mins of MVPA over a 7 day period is sufficient to meet the recommended PA guidelines [25].
Secondary outcomes
Change over time for the following secondary outcomes was also assessed. Waist circumference was measured to the nearest 0.1 cm using a rigid anthropometric measuring tape (Lufkin, US). Body mass was recorded to the nearest 0.1 kg using a calibrated scale (model 813; Seca, Germany). Free standing stature was recorded to the nearest 0.1 cm using a calibrated equipment with the participant barefoot (Portable stadiometer; Seca, Germany). Body mass index (BMI) was calculated by dividing body mass by the square of height. Self-efficacy to be physically active was measured using a validated PA self-efficacy survey [35]. Health-related quality of life (HrQoL) was measured using the Medical Outcomes Study Short Form 12 Health Survey (SF-12) which is a reliable tool with published psychometric support [36].
Study size
Utilising data from the H4U study [20], a sample size of 50 participants per arm was calculated to be sufficient to detect a between group difference of 30 ± 53 (mean ± SD) mins/week MVPA, with the alpha set at 0.05, and the power set at 0.80. Protecting against a drop-out rate of 20% over the 9-month study duration, 60 participants were recruited and randomised into each arm.
Data analyses
Analyses were carried out using IBM SPSS Statistics for Windows (Version 26.0; IBM Corp., USA) and statistical significance was set at an alpha of 0.05. Data were assessed for normal distribution by Shapiro-Wilk tests. Homogeneity of variances and covariances were assessed by Levene’s test and Box’s M test, respectively. Grouped data are presented as mean ± standard deviation. For the main analyses, a series of mixed-model ANOVAs (within: time; between: intervention) were used to assess the effects of the PA coaching intervention on each of the outcome variables separately. Mauchly’s test was consulted and Greenhouse–Geisser correction was applied if the assumption of sphericity was violated. A significant interaction effect was interpreted to demonstrate that the change in dependent variables was influenced by intervention. Where statistically significant interactions were observed simple main effects analyses were carried out. Where data were in breach of Shapiro-Wilks test of normality, sensitivity analyses were performed. Data were explored for significant outliers and repeat sensitivity analyses were undertaken on data with outliers removed. Repeated sensitivity analyses provided no indication that the outliers had a significant effect on the outcome; therefore, all data were included in analyses. Intention-to-treat analysis was used for missing data using the last observation carried forward approach [37]. Repeat sensitivity analyses were undertaken on data with and without imputing the last-observation-carried forward value and provided no indication that the imputed values had a significant effect on the outcome.