Theory-based behavior change interventions.

The theoretical basis of an intervention and the integration of behavior change techniques (BCTs) are characteristics that can influence intervention efficacy [9]. Theories help to identify intervention targets and outcomes because they offer explanations of complex phenomena, include key drivers of behavior change, and identify constructs to assess [10]. Self-determination theory (SDT; [11]) is a macro theory of human motivation widely used to develop behavior change interventions due to its focus on psychological factors and processes that determine motivated behavior in diverse health contexts and populations [12–14]. Within SDT, satisfying one’s psychological needs for competence, autonomy, and relatedness is theorized to influence behavioral outcomes by promoting autonomous motivation (i.e., motivation reflecting self-endorsed reasons for engaging in a behavior; [11]). Systematic reviews and empirical studies support the use of SDT to develop PA behavior change interventions for adults with and without cancer [13, 15–18]. Furthermore, within SDT, it is theorized that the extent to which one’s psychological needs are satisfied is largely influenced by their social environment [11]. Thus, the integration of behaviors and communicative techniques that social agents (e.g., PA counsellors) can use to support, or actively satisfy, participants’ psychological needs to promote autonomous motivation and behavior change are central to SDT-based behavior change interventions.

SDT-informed (i.e., those vaguely describing SDT use) and SDT-driven (i.e., those integrating SDT throughout intervention planning, design, and evaluation) behavior change interventions (hereinafter collectively referred to as SDT-based or theory-based when nonspecific to SDT) have comprised several behavioral and communicative techniques designed to support participants’ psychological needs, and in turn promote autonomous motivation and behavior change. For instance, autonomy supportive techniques such as cultivating autonomy support, providing structure, and establishing strong interpersonal relationships with participants have been used and found to promote behavior change via psychological needs satisfaction and increased autonomous motivation [19–21]. Owing to the growing use of various techniques across different SDT-based interventions, recent work has sought to identify, describe, and organize the essential techniques, referred to as either behavior change techniques (BCTs) or motivation and behavior change techniques (MBCTs), used to promote sustained behavior change, and thus critical to intervention success [22, 23]. BCTs and MBCTs are replicable components of an intervention designed to facilitate behavior change by either augmenting factors that facilitate behavior change or by mitigating factors that inhibit behavior change. This work has revealed a large range of potential BCTs/MBCTs that could be selected for SDT-based behavior change interventions [22, 23]. Whilst it remains unclear which techniques or set of techniques are most efficacious for promoting autonomous motivation and behavior change via psychological needs satisfaction, studies show PA interventions integrating BCTs/MBCTs are not only effective for adults after cancer treatment ends, but are more effective than those without BCTs/MBCTs [24, 25]. There is, however, a lack of evidence for the feasibility, acceptability, and effects of interventions based on theory (SDT or otherwise) that also integrate evidence-based BCTs/MBCTs that may effectively promote PA in young adults after cancer treatment ends [26, 27]. Because not all interventions may be feasible, acceptable, and/or effective in promoting PA in young adults who have completed cancer treatment, there is some uncertainty about the value of such interventions.

Real-time videoconferencing interventions.

To enhance public health impact, it is important that behavior change interventions not only target theoretical factors and integrate BCTs/MBCTs, but that they be readily available to the target population. Many theory-based behavior change interventions developed to promote PA in persons diagnosed with cancer have been delivered face-to-face, which can be inconvenient for participants and reach only a small proportion of the target population. It is therefore pertinent to consider alternative means of delivering theory-based behavior change interventions, investigate the extent to which these are feasible and acceptable, and assess their effects and costs. Synchronous (or real-time) videoconferencing provides an alternative means to deliver theory-based behavior change interventions to technology-adept young adults who report heavy reliance on the Internet for access to health information and who desire interactive and trustworthy content [28–31]. The use of videoconferencing technology has the potential to reach a large proportion of young adults and can address participation barriers (e.g., pandemic restrictions, travel restrictions for those living in remote locations, scheduling difficulties; [18, 32]). Moreover, videoconferencing can reduce physician burden by shifting the responsibility of promoting PA to trained PA counsellors [33, 34], which is critical as treatment and follow-up care centres have restricted infrastructure and resources, and in some cases limited expertise, to address young adults’ specific PA needs [35]. As behavior change interventions delivered via videoconferencing technology have been shown to promote health behavior change (including PA) in clinical and non-clinical populations [36–38], the World Health Organization (among other organizations) now advocates for enhanced technology use in healthcare settings [39]. On this basis, research investigating whether a novel, SDT-driven behavior change intervention that is delivered using real-time videoconferencing and integrates empirically based BCTs/MBCTs to promote PA in young adults who have completed cancer treatment, and in turn improve patient-reported outcomes, is feasible, acceptable, effective, and cost-effective is warranted.

Eligibility criteria.

Young adults are eligible to participate if they: (1) are currently between the ages of 18–39 years; (2) received a first diagnosis of invasive, non-metastatic cancer between the ages of 18–39 years; (3) completed primary treatment for cancer within the past 5 years; (4) are able to read and speak English; and (5) have access to videoconferencing technology (e.g., Zoom). Exclusion criteria include: (1) evidence of current cancer (i.e., recurrent, secondary, or relapse); (2) physical impairments precluding participation in PA; (3) self-report engaging in ≥150 minutes of moderate-to-vigorous intensity aerobic PA per week in the month prior to screening; and (4) non-ambulatory status. Young adults enrolled in other studies are potential candidates for the ACCESS trial.

The ACCESS intervention (intervention group).

Participants randomized to the intervention group receive six real-time, one-on-one 60–minute videoconference sessions with the same PA counsellor over a 12-week period. See Table 2 for an overview of the session topics.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. ACCESS intervention session topics and activities/worksheets.

https://doi.org/10.1371/journal.pone.0273045.t002

The ACCESS intervention is multicomponent and includes essential elements derived from previous studies with young adults who have completed cancer treatment [18], as well as from published interventions yielding significant changes in PA [48]. The intervention targets SDT constructs, focusing specifically on: (1) providing autonomy support, structure, and interpersonal involvement; (2) increasing perceptions of autonomy (i.e., perceived control over one’s actions by providing a rationale, structure, and emphasising responsibility), competence (i.e., perceived mastery of tasks and skills by providing support and encouragement, information feedback, and support barrier identification), and relatedness (i.e., perceived belonging and connection to others by encouraging social support seeking); and (3) increasing autonomous motivation (i.e., acting through self-endorsement and volition because the activity holds inherent interest and/or personal value). Accordingly, the intervention includes several content and relational techniques to augment these SDT constructs and facilitate PA behavior change. Techniques were selected based on the theoretical construct they are proposed to target, wherein we drew on published literature [10, 22, 23, 49] to draw out which techniques target specific SDT constructs. BCTs used to guide the content of the intervention and inform the development of activities/information sheets are presented in Table 2. Moreover, PA counsellors use MBCT and motivational interviewing (MI) techniques that align with SDT [22] for the relational component of the intervention to support the delivery of the content; these are presented in Table 3. As content and relational techniques are interrelated, there is overlap between some techniques presented in Tables 2 and 3. By using MI techniques, PA counsellors aim to help participants explore and resolve their ambivalence toward PA behavior change, and thus they tailor the content of the intervention to match participants’ readiness to change, perceived advantages/disadvantages for change, and self-efficacy beliefs [50]. To this end, PA counsellors avoid argumentation and “roll with resistance,” acting as an engaged problem-solving partner [51]. MI principles and techniques used by PA counsellors, which align with SDT [49], include empathic/reflective listening, asking open-ended questions, prompting participants to ask questions, prompting participants to offer solutions, seeking participants’ permission to provide information and advice, shifting focus, supporting change/persistence, and showing unconditional regard.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Overview of ACCESS intervention components organized by primary SDT construct.

https://doi.org/10.1371/journal.pone.0273045.t003

Training and ongoing supervision of PA counsellors. Owing to the nature of the intervention, PA counsellors who are current or past graduate students in kinesiology or a related field are hired. PA counsellors are trained, supervised, and supported by the first author (who holds a PhD in PA and health psychology) and the second author (who holds a masters in Human Kinetics). PA counsellors complete a 2–hour training session (conducted by the first and second authors), are observed performing a mock counselling encounter, and take part in a bracketing interview to examine how their own backgrounds, perceptions, and interests could impact them in their role. Ongoing supervision consists of 30–minute team meetings every 2 weeks focused on updates on participants’ progress, discussion of challenging sessions, deviations or modifications to sessions, and feedback on session notes. During these meetings, PA counsellors are invited to reflect on and discuss their experiences conducting the sessions. Of note, these meetings also serve to monitor data on outcomes, accrual, and adverse events, and oversee participants’ safety as a data monitoring committee is not needed for the ACCESS trial because the intervention is non-invasive with minimal risk of harm.

Usual care group.

Usual care as a comparison group is necessary as the aim of a future large, full-scale RCT will be to compare the ACCESS intervention with de facto usual care (or treatment as usual). Without it, it would not be possible to claim that the ACCESS intervention be the new standard of care. Participants randomized to the usual care group are asked to continue their usual activities of daily living. No restrictions are placed on their PA behavior. Once participants complete follow-up assessments, study staff email them a copy of the intervention materials to review on their own in hopes that it could possibly lead to increases in PA even though they do not have an opportunity to engage with a PA counsellor. Another reason to provide participants randomized to the usual care group with a copy of the intervention materials once they complete follow-up assessments is to promote retention.

Pilot RCT primary outcomes.

Feasibility [throughout the trial]. Drawing on the CONSORT 2010 statement for pilot and feasibility trials [42] and CONSORT flow diagram, the following feasibility aspects are tracked by study staff: (1) number of months needed to reach target sample size; (2) number of young adults assessed for eligibility (with referral source) and reasons for exclusion (i.e., not meeting inclusion criteria, declined to participate, other reasons); (3) number of young adults who consented and reasons for refusal; (4) number of participants randomized and reasons for refusal; (5) number of participants allocated to the intervention group and the usual care group; (6) number of intervention sessions attended and proportion of participants who adhered to the intervention (i.e., attended all sessions); (7) number of participants who discontinued the intervention and reasons; (8) number and type of adverse events related to the intervention and/or trial procedures; (9) number of participants lost to follow-up (i.e., who do not complete post-intervention and follow-up assessments) and reasons; (10) number of participants analyzed, number excluded from analysis, and reasons; and (11) percentage of missing data on quantitative measures and number of participants who completed an interview. Study staff conduct a biweekly audit of these metrics (along with adverse events, if any); the first author/principal investigator reviews these during biweekly meetings.

Criteria were set a priori based on relevant literature and the authors’ experiences to determine feasibility of the intervention and trial protocol. The intervention and trial protocol will be considered feasible if: (1) target sample size is reached within 20 months; (2) 2–3 young adults are referred or self-refer per month and are assessed for eligibility; (3) ≥70% of eligible young adults consent and are randomized, with 50% being allocated to the intervention group and 50% to the usual care group, and ≤10% discontinuing the intervention; (4) ≥75% of participants in the intervention group complete all six videoconference sessions for an average of ≥50 minutes; (5) ≤25% of participants are lost to follow-up (i.e., ≥75% complete post-intervention and follow-up assessments) allowing for ≥22 participants (≥11 per group) to be analyzed; (6) <10% missing data on quantitative measures and ≥75% of participants are interviewed; and (7) no adverse events related to the intervention are reported. These data will serve to determine if/what changes/modifications are needed to improve the intervention and/or trial protocol, and will inform the timeline and budget for a large, full-scale RCT.

Acceptability–participants [post-intervention period]. All participants are invited to take part in an audio-recorded, semi-structured interview by phone or videoconferencing technology to share their opinions about the trial protocol. Additionally, participants in the intervention group are asked questions related to the intervention (i.e., content, delivery mode, length, duration, PA counsellor), SDT constructs, and BCTs/MBCTs. Their feedback will serve to determine if/what changes/modifications are needed to improve the intervention and/or trial protocol. It will also offer insight into the impact of the various intervention components in achieving the overall outcome of increased PA.

Acceptability–PA counsellors [at trial cessation or contract termination]. Each PA counsellor will be invited to take part in an audio-recorded, semi-structured interview by phone or videoconferencing technology to explore their experiences delivering the ACCESS intervention, as well as their thoughts about the content of the intervention, required training, ongoing supervision, using an intervention manual, conducting sessions using videoconferencing technology, maintaining fidelity to the manual, and their motivation and confidence to conduct the sessions. Their feedback will serve to determine if/what changes/modifications are needed to improve the ACCESS intervention and help to highlight factors that may facilitate or hinder the effectiveness of the intervention and/or its implementation in the future.

Large, full-scale RCT effectiveness outcomes.

Primary–PA behavior [baseline, post-intervention, follow-up]. Objective PA behavior is assessed using an accelerometer (ActiGraph GT9X-BT [also known as ActiGraph GT9X Link]). Total volume of PA (i.e., overall number of minutes spent engaging in light, moderate, and vigorous intensity PA per week) will be considered rather than only moderate-to-vigorous intensity PA (MVPA) for several reasons: (1) studies have highlighted concerns with generic application of intensity cut points as most were developed from young healthy volunteers, whereas young adults who have completed cancer treatment may have ongoing cancer-related side effects that impact their functional capacity; (2) MVPA may not be appropriate or desirable for certain participants with functional/mobility issues that preclude intensive PA; (3) data show that light intensity PA (not just MVPA) confers health benefits across the cancer spectrum [52]; and (4) using total volume of PA will facilitate comparisons across studies as researchers who have summarized accelerometer data into discrete variables (e.g., light intensity PA, MVPA) cite different cut point algorithms. The primary endpoint is total volume of PA at post-intervention (week 13); the secondary endpoint is total volume of PA at follow-up (week 24). Data are managed with ActiLife v6.13.4 software using established wear time criteria [53] and data processing items will be reported (e.g., definition of non-wear time, filters applied during processing, valid day definition and minimum number of valid days, epoch length, vectors used during processing) in forthcoming publication of trial results. Compliance with accelerometer wear will also be reported. Accelerometer based PA measurements taken in free living conditions have been shown to be reliable and valid [54–56]. These data will be used to produce standard deviation estimates to inform the sample size calculation for a large, full-scale RCT.

Secondary–patient reported outcomes [baseline, post-intervention, follow-up]. Patient reported outcomes are assessed using the following questionnaires: RAND 36-Item Short Form Health Survey [57, 58], Positive and Negative Affect Schedule Short Form (PANAS-SF; [59]), Patient Health Questionnaire Depression Scale (PHQ-9; [60]), Patient Health Questionnaire Generalized Anxiety Disorder Scale (GAD-7; [61]), and the Impact of Cancer Instrument–Adolescent and Young Adult Module [62]. These data are collected because they represent secondary outcomes for a large, full-scale RCT, and thus the suitability of the measures and the methods for data collection must be tested.

Secondary–putative mechanisms of PA change [baseline, post-intervention, follow-up]. Testing mechanisms of behavior change is central for a large, full-scale RCT to assess how the intervention works [63, 64]. Thus, SDT constructs are assessed at each timepoint using the following measures to ascertain feasibility of data collection: Psychological Need Satisfaction in Exercise Scale [65, 66] modified to assess basic psychological needs satisfaction in relation to PA, Behavioural Regulation in Exercise Questionnaire [67, 68] modified to assess autonomous motivation for PA, and the Health Care Climate Questionnaire [69] modified to assess autonomy support, structure, and interpersonal involvement in the context of the intervention sessions. The Health Care Climate Questionnaire is only included in the post-intervention questionnaire and is only be administered to participants in the intervention group as the items are irrelevant for participants in the usual care group who do not have a PA counsellor (e.g., “My PA counsellor conveys confidence in my ability to make changes regarding my PA behavior”). Also, a questionnaire covering BCTs/MBCTs was developed by the research team and is administered to assess which specific techniques and strategies participants used to change their PA behavior.

Additional measures.

Personal and medical factors [baseline unless otherwise specified]. At baseline, participants self-report a range of personal (age, biological sex, gender identity [using categories currently employed by Statistics Canada in the Census that are based on wording from the Employment Equity Act], race, civil status, work status, education attainment, income) and medical information (height, body mass, cancer type and stage, cancer treatments received, current medication use, comorbid conditions using the Cumulative Illness Rating Scale; [70]). At post-intervention and follow-up assessments, they are asked to self-report potentially changing personal and medical information (i.e., civil status, work status, education attainment, income, comorbid conditions, body mass, current medication use), as well as use of other resources (e.g., personal trainer, psychosocial support). These personal and medical data are collected to describe the sample and assess if such factors impact feasibility outcomes.

Adverse events [throughout the trial]. Becoming more physically active is very safe for most young adults after cancer treatment ends and can yield many health benefits [71, 72]. Though the risk of injury is estimated to be very low in this trial, PA counsellors educate participants on the warning signs that may indicate a problem as per Canadian Society for Exercise Physiology guidelines (e.g., chest discomfort, unusual shortness of breath, dizziness or light-headedness, heart rhythm abnormalities) and tell participants to seek immediate medical attention should one of these signs occur. Adverse events (i.e., any unfavourable and unintended sign, symptoms, or disease) definitely, probably, or possibly related to engaging in PA or trial procedures (e.g., wearing an accelerometers) will be tracked. To this end, all adverse events occurring during trial participation are documented by PA counsellors. Each session, PA counsellors record any adverse event(s) participants share (if any) in their notes and report them immediately to the principal investigator via email. At the time of reporting, the intervention is paused for the participant reporting the adverse event until clearance from an appropriate healthcare provider to resume the intervention is obtained. All adverse events reported will be shared with Institutional Review Boards and reported in forthcoming publication of trial results. Additionally, there is a risk that participants experience distress in response to certain questions included in the questionnaire packages and/or posed during the interviews; participants receive a list of free resources to consult in case this occurs during the consent process (e.g., Canadian Cancer Society–Peer support / cancer information specialist: 1-888-939-3333).

Intervention fidelity [throughout the trial]. The assessment of fidelity (i.e., the extent to which PA counsellors were able to follow the intervention protocol as intended) is critical when counsellors use an intervention manual so as to ensure the intervention delivery is standardized with limited variation [73]. However, the intervention is delivered by persons (i.e., not automated), and their personal backgrounds, perceptions, and interests may influence their interactions with participants. Therefore, as mentioned above, each PA counsellor participates in a bracketing interview with the second author. The goal of the bracketing interview is to help them explore and address any biases that may inadvertently influence how they deliver the intervention content and interact with participants. In addition, following each session with participants, PA counsellors take detailed notes regarding required deviations, implementation facilitators and challenges (including strategies used to overcome challenges), participants’ responses, external influences on the intervention, and any factors they believe may have influenced their ability to follow the intervention protocol (i.e., intervention fidelity). These notes are reviewed during the biweekly meetings to facilitate fidelity monitoring, and will be synthesized to provide a broad assessment of the intervention implementation process.

Main analyses.

Quality checks and content validation will be utilized to ensure high quality data (i.e., identify sources of inconsistency, checking for errors). Quantitative and qualitative data analyses will be conducted. SPSS and SAS will be used for quantitative data management and analysis; NVivo will be used for qualitative data management and analysis. Descriptive statistics will be computed to describe the sample based on baseline personal and medical factors, and report on feasibility and acceptability outcomes described above. Specifically, categorical variables will be described using proportions and frequencies, and continuous variables with means and standard deviations (or with medians and interquartile ranges for data not normally distributed). Content analysis [74] of the transcribed interviews will be conducted to identify themes and see if/what changes to the intervention and/or trial protocol are warranted.

Additional analyses.

This pilot RCT is not powered to determine the effectiveness of the ACCESS intervention, which is why the main analyses will focus on feasibility and acceptability outcomes. Nonetheless, effectiveness outcomes will be analyzed descriptively to inform the design of the future large, full-scale RCT. The primary effectiveness outcome (i.e., objectively measured PA behavior assessed at baseline, post-intervention, and follow-up) will be analyzed by fitting a linear regression model accounting for the correlation in repeated measures on the same participant over time explicitly modeling a covariance matrix that best fits the data based on the likelihood ratio tests and information criteria. The model will include the stratification factor (i.e., biological sex), PA counsellor ID, and characteristics commonly associated with attrition (e.g., length of chemotherapy) as covariates under the assumption of missing at random. The model allows all available observations on each participant to be used without having to use an imputation procedure. The model will be estimated by means of Restricted Maximum Likelihood (REML) and used to produce estimates of usual care group means and event rates, and variance and correlation estimates to inform the sample size calculation for a large, full-scale RCT. Degrees of freedom will be computed using the Kenward-Roger approach [75].

Additionally, the amount and potential richness of data collected facilitates supplementary exploratory analyses. Therefore, after the main analyses are conducted to address the pilot trial objectives, it is envisaged that exploratory analyses may be undertaken to address subsequent research objectives that will make unique theoretical and empirical contributions (e.g., changes in secondary outcomes).