Study design and methods for the physical activity index (PAI) feasibility pilot trial for breast and colon cancer survivors in North Carolina.

1
Introduction and background
Maintaining an active lifestyle is an integral part of clinical recommendations for cancer survivors, now numbering more than 18 million in the US [1], especially after completion of cancer treatment. However, “survivors” (i.e., individuals diagnosed with cancer) [2] often do not meet recommended physical activity (PA) requirements following treatment, which include moderate-to-vigorous intensity physical activity (MVPA) for ≥150 min of moderate or 75 min of vigorous intensity activity (i.e., two and a half hours in a week), 3–5 times per week, and ≥2 strength-training sessions per week [3,4]. This is especially true for middle-aged and older breast cancer survivors (BCS) and colon cancer survivors (CCS) to achieve optimal recovery, as they may be concurrently managing age-related chronic illnesses and cancer symptoms [[5], [6], [7]]. BCS and CCS often face similar treatment-related issues, such as inflammation, fatigue, and metabolic changes., for which they may specifically benefit from a tailored exercise program compared to other cancer survivors.
To reduce physical dysfunction associated with cancer treatment, survivors (especially BCS and CCS), are specifically advised to avoid sedentary behavior (i.e., excessive sitting) [8]. Although all cancer survivors are advised to do physical activity throughout the cancer continuum, BCS and CCS are at particular risk of insufficient physical activity and excessive sedentary behavior, and this may negatively impact health outcomes [9]. Although a physician's recommendation promotes adoption of PA and other healthy behaviors [10,11], there are few clinical tools available to support patients and facilitate lasting success in achieving these goals.
The completion of primary cancer treatment (i.e., surgery, radiation and/or chemotherapy) has been suggested as a ‘teachable moment’ to influence positive health behaviors in the cancer recovery process [12] The physical activity index (PAI) was developed by the Principal Investigator (SMB) as a screener using population-based data as a proof of concept to show that collecting multiple physical performance measures (e.g., MVPA, strength, sedentary behavior) provides a more precise estimate of PA and its relationship to key health outcomes (e.g., multimorbidity, self-rated health) than basic tools currently used in clinical practice [13,14]. By asking simple, validated questions on MVPA, strength training, and sedentary behavior at each follow up visit and encouraging discussion on how those results relate to recommendations, it provides a real-time snapshot, or ‘report card,’ of how survivors are doing relative to recommendations.
To understand the potential clinical utility of the PAI screener, we conducted preliminary interviews with providers (both nurses and physicians) and survivors to introduce the PAI concept to these clinical stakeholders for potential use in counseling survivors about lifestyle. Survivors and providers generally viewed the PAI concept favorably [15], but questions emerged about how such a tool would fit in the clinical work flow. Providers also expressed concerns about having the time and necessary expertise to counsel patients to makes changes in physical activity and reduce sedentary behavior. To wit, we addressed concerns about lack of knowledge concerning exercise counseling by: providing educational resources for providers and participants about recommendations for PA and sedentary behavior; offered Garmin activity trackers to all participants so they could self-monitor their (in)activity, and; provided referrals to certified exercise physiologists once they had received the recommendation to exercise from their providers.
In this paper, we describe our plans to test the feasibility of a multicomponent PAI intervention that includes the PAI screener in a clinical setting, also providing education, self-monitoring tools and coaching sessions with an exercise physiologist to help breast and colon cancer survivors to increase their physical activity and reduce their sedentary behavior. In this pilot study, we will determine feasibility by examining recruitment, retention, acceptability, and PAI intervention adherence goals. Secondarily, we will measure changes/variability in achievement of behavioral outcomes for PA and sedentary behaviors to help inform power calculations for a future efficacy trial.

2
Methods
2.1
Overview
The primary goal of this two-arm 1:1 randomized pilot trial with a waitlist control (n = 20, 10 BCS and 10 CCS) is to establish the feasibility of intervening with the PAI at the “teachable moment”, using the PAI screener, coaching and self-monitoring tools (e.g., an activity tracker plus PAI ‘report card’) to increase PA and reduce sedentary behavior. The pilot will span 12 months, with a six-month active PAI intervention period. Table 2 lists the standardized study measures and the timing of measures at baseline, three months, six months, and for nine and 12-month follow-up visits. For the waitlist control group, the PAI intervention visit will start at six months (after receiving two assessment visits at baseline and three months) and proceed to nine and 12-month visits. All participants will have five total research assessments in a one-year period. A depiction of the study schema is provided in the appendix.
All study procedures were approved by the Institutional Review Board (IRB) at Atrium Health Wake Forest Baptist Comprehensive Cancer Center (AHWFBCCC). The study is registered at www.clinicaltrials.gov (NCT06340503).

2.2
Eligibility
Table 1 describes the inclusion and exclusion criteria for the PAI study. Potential participants who are referred will be screened using a standardized telephone screening script. These individuals are offered a baseline appointment for further screening and enrollment if they report being age ≥50 years, have access to the Internet, have a prior diagnosis of breast or colon cancer, no high-risk symptoms for cardiac events during exercise based on the Physical Activity Readiness Questionnaire items included in the phone screening script [16] or have provider permission if they report any high-risk symptoms. Interested individuals must also demonstrate no major cognitive deficits based on EORTC cognitive screening items [17] also included in the phone screening script.
As we are seeking to understand feasibility from both the patient and provider perspectives, we have two sets of eligibility criteria.

2.3
Recruitment and screening procedures
Patient electronic medical records will be accessed by a member of the study team to identify potentially eligible patient participants. Some potential participants will be chart screened first from a list of potentially eligible patients, provided by the translational data warehouse offered through the Wake Forest Clinical and Translational Institute (CTSI). These will be noted on a study log. For those that qualify, we will contact the patient and a phone screening form will be completed to determine interest. Using information obtained from screening, we will use de-identified patient information including demographic (age, race, ethnicity, gender assigned at birth), cancer type (breast or colon), eligibility, and potential reasons for declining to participate (if applicable) for all potential participants to understand the current patient population. In addition, patients will be recruited through referrals from medical oncologists and Advanced Practice Providers (APPs) in each oncology clinic (breast and colon). Study fliers, approved by the IRB, will also be posted in patient waiting areas and shared with community partners, including the Office of Cancer Health Equity.
Patients will receive modest compensation after completion of each research visit ($25 per visit, for a total of $125). Signed informed consent will be obtained from each subject by the study team at their baseline visit for all study activities at AHWFBCCC.

2.4
Randomization
Patient participants enrolled will be randomized to two groups in the PAI intervention: the immediate exercise group (n = 10) or waitlist control group (n = 10). Allocation will be computer-generated, stratified by cancer type (i.e., breast, colon) by Ms. Evans (biostatistician) and/or staff from the AHWFBCCC biostats core.

2.5
PAI intervention procedures
2.5.1
PAI intervention overview
The Physical Activity Index (PAI) intervention is designed to engage cancer survivors during the “teachable moment”—a critical window early in recovery when individuals may be more receptive to adopting lasting health behaviors [12,18,19]. Grounded in behavioral theory and mapped to constructs aligned with the teachable moment model (Fig. 1, Table 3), the intervention aims to promote increased physical activity and reduced sedentary behavior, consistent with survivorship care guidelines from the American Cancer Society and American Heart Association [[20], [21], [22]].
Participants are recruited shortly after completing cancer treatment, leveraging this transitional phase as a behavioral cue to action [23]. At the initial study visit, they complete the PAI screener on paper, a format favored in preliminary interviews. They then receive educational materials that highlight the role of physical activity in recovery, helping establish foundational knowledge.
To support self-monitoring, participants are provided with both paper-based exercise logs and a wrist-worn activity tracker. The intervention also includes provider endorsement—delivered during a routine follow-up visit, via phone, or through secure messaging—where oncology providers reinforce the importance of physical activity and reduced sedentary behavior. Endorsements typically occur within two to three months of intervention initiation.

2.5.2
PAI screener and provider feedback
During the initial study visit, participants complete a brief worksheet adapted from validated National Health and Nutrition Examination Survey (NHANES) measures [24] to assess current exercise habits. Research staff score the worksheet and generate a one-page “report card” that categorizes physical activity levels into four groups: no activity, needs improvement, meets recommendations, and exceeds recommendations. Each category is scored from 0 to 3, resulting in a total score ranging from 0 to 9.
The report card is shared with both the participant and their clinical team. It includes motivational messages and curated resources—such as websites and suggested activities—to encourage increased physical activity and reduced sedentary behavior. Providers are encouraged to review the report card with patients during a surveillance visit or through secure messaging. This tool complements standard care by incorporating physical activity into survivorship care planning (Fig. 2).

2.5.3
PAI intervention procedures for patient participants
At enrollment, all participants—regardless of group assignment—receive a research-grade accelerometer (ActiGraph wGT3X-BT, Pensacola, FL). The device is worn for seven consecutive days to objectively assess baseline physical activity prior to any provider-suggested lifestyle recommendations.
Participants randomized to the immediate exercise group receive multiple resources to support behavior change.•Educational materials from the NCCN® guidelines on physical activity in survivorship [21].

•A set of resistance bands (valued at $25)

•A Garmin VivoFit activity tracker (valued at $100)

•Personal exercise logs for daily tracking, retained by participants post-trial

They also complete.•An initial intake consultation

•Five monthly video sessions with an ACSM-certified exercise coach affiliated with the Wake Forest University Healthy Exercise and Lifestyles Program

Coaching is guided by the Moving Through Cancer: A Guide to Exercise for Cancer Survivors framework [25]. This approach emphasizes meeting survivors “where they are”(i.e., physically and emotionally); tailoring recommendations based on individual goals, preferences, physical limitations, and treatment-related effects and; repeatedly updating personalized exercise prescriptions to reflect progress throughout the intervention.
Participants assigned to the waitlist control group receive the same materials, procedures, and coaching beginning at their six-month follow-up visit.

2.5.4
PAI engagement with clinical team members
To facilitate engagement with patient participants, clinicians will be provided with a trial orientation/educational session, including receiving: 1) A copy of NCCN survivorship care physical activity recommendations; 2) A copy of the PAI scoring criteria and example talking points for how to coach patients based on results, (e.g., “Looks like you are spending too much time on screens. Let's try to reduce your screen time to 2 h per day. If you can't do that, at least try to stand once per hour.”); 3) short video explaining PAI report card that they can reference at any time and follow up with study team with any questions [22].
Clinical team members will also participate in qualitative activities to address our questions about how the PAI intervention would fit in the clinical workflow. Leaders within the care team will identify and recruit relevant healthcare staff members for interviews and focus groups based on their roles in the practice setting and their expected involvement in referring patients to the intervention. If they do not end up referring any patients to the study but could be involved in the future implementation of the PAI intervention, they may still be asked to participate in qualitative activities to inform our approach. We plan to enroll at least 2 physicians, 2 APPs, and 2 other staff from the clinical oncology team (including patient navigators, social workers, physical or occupational therapists, etc.) that are interested and available to participate in qualitative research activities to understand how the PAI would fit in the clinical workflow. These activities will include one focus group and three interviews with key clinical stakeholders. Clinical team members will receive modest compensation ($40) after sharing their perspectives. A waiver of written consent will be requested for clinical team members. All clinical team members will be provided with an information sheet outlining the study requirements.
Select clinical team leaders [3] will each participate in a single interview at one timepoint and other clinical team members [5,6] will participate in one focus group discussion at one timepoint. All interviewees and focus group participants will receive information about the trial as needed, and receive access to the planned tools for clinicians to provide feedback.
Interview: An interview guide will be developed with open-ended questions and probes to facilitate discussion. The interviews will be conducted by a trained Q-PRO interviewer and audio recorded. Interviews are expected to last approximately 30 min. Discussion will include acceptability of the PAI intervention, related tools, and clinical experiences and will help shape/inform refinements needed for a larger efficacy trial (see Table 3). Interview procedures will be submitted to the AHWFB IRB for approval.
Focus group: The Focus Group guide will draw from concepts specified in the RE-AIM Quest mixed-methods evaluation framework (Table 3). The guide will be developed with open-ended questions and probes to facilitate group discussion. The focus groups will be facilitated by a trained Q-PRO moderator, audio recorded, and transcribed verbatim. A second trained Q-PRO staff member will attend the focus groups and take detailed notes. Focus groups are expected to last approximately 45–60 min. Discussion will include perceptions about the value of physical activity to patient health and will help shape/inform barriers and facilitators for implementing physical activity counseling in the clinical setting. Focus group procedures will be submitted to the AHWFB Institutional Review Board for approval.

2.6
Measures
Table 3 lists the standardized study measures and the timing of measures at baseline, three months, six months, and for nine and 12-month follow-up visits. For the waitlist control group, the baseline visit will start at six months.
The primary objective of the study is feasibility, which will be measured by participation, retention, and PAI intervention adherence goals, in addition to acceptability, as follows.•Participation will be measured by the number of participants (i.e., the first twenty who meet eligibility requirements) who agree to participate.

•Adherence will be measured by the percent of participants who complete at least three out of five exercise coaching sessions, plus consultation visits (>70 % of sessions).

•Retention will be defined as the number of participants who complete study questionnaires at baseline as well as at least one follow-up timepoint after six months (target >70 %).

•Acceptability: will be defined as individuals who rate satisfaction with the PAI intervention as 7 or more on a 10-point Likert scale.

2.7
Analysis of primary outcome
We will provide quantitative data on participation rate, adherence, and retention rates to guide future study planning. We will calculate 95 % CI for each of the feasibility measures to determine the range of estimates that are consistent with our data. We will also use one-sample tests of binomial proportions to compare the participation, adherence, and retention rates to the hypothesized values of 40 %, 70 % and 70 %, respectively. In exploratory analyses, we will compare participants who are non-adherent or those who drop out by demographic characteristics, intervention group, and baseline scores of the measures. We will also investigate any differences in participant participation, adherence, and retention rates or acceptability by cancer type to identify an optimal patient population for a future study. Quantitative acceptability measures will be summarized using descriptive statistics and 95 % confidence intervals and confirmed through comparison with qualitative data.

2.8
Analysis of exploratory and secondary outcomes
Qualitative and quantitative analyses will be evaluated using the RE-AIM QuEST mixed-methods translational framework [26] for consistency and discrepancies to refine the protocol in future studies (see Table 4). For this step, we will collaborate with the Wake CTSI Qualitative and Mixed Methods Patient Reported Outcomes (QPRO) shared resource for expertise in program evaluation and collection of process and qualitative measures. Acceptability will be evaluated by conducting semi-structured interviews and using the RE-AIM Qualitative Evaluation framework to understand barriers and facilitators for implementing the PAI intervention. We will also ask participants to rate their satisfaction with the intervention on a scale of 1–10 (low to high for satisfaction). We expect that a rating of 7 or higher will affirm PAI intervention acceptability.
Analyses of behavioral outcomes will include descriptive statistics of MVPA, sedentary behavior, background characteristics (i.e., demographic) and other possible confounding variables (e.g., type of cancer, type of treatment, medications) overall and by treatment assignment group. The primary goal of the statistical analysis of these measures will be to estimate standard deviations (SD) for use in future studies.
Additional analyses will include fitting mixed ANCOVA models (adjustment for baseline) to model the trajectory of MVPA and sedentary behavior by time and group accounting for the repeated measures on a subject; if any meaningful differences between the groups are found at baseline, we will include them in these models. The purpose of these models will be to obtain estimates of the SD of change and the within-person correlation of the repeated measures, not to perform formal hypothesis testing.

2.9
Sample size calculations
The sample size is calculated to estimate the rates of the feasibility using measures from similar studies (e.g., enrollment of 40 %) [27] to inform the design of a larger efficacy trial. With n = 20, we will be able to estimate these rates within ± 22 % using a two-sided 95 % Cl. If the recruitment rate is below 40 % and the adherence and retention rates are below 70 %, a larger study may not be feasible. Assuming a negative binomial distribution and true rate of 40 %, the probability that we would have to approach 65 or more people to recruit 20 is < 0.05. Therefore, if we approach ≥65 participants to enroll 20, it is unlikely the true probability is above 40 %, and we will conclude the study may not be feasible.
Similarly, if the true rate of intervention completion and study retention is 70 %, we anticipate that ≥11 of 20 will participate/be retained. If 10 or fewer do not participate/are not retained the study may not be feasible; we have 80 % power to test that the rate is 70 % compared to the alternative that is 40 % or less.
This current pilot study is not designed to assess the effect of the PAI intervention on any outcome, but to characterize variability in the study population for the future outcome measures. With n = 20 participants we can estimate the lower limit of the estimate of the SD within 15 % of the true SD with 80 % confidence. Accordingly, we will accrue a total of 20 participants (10/group), because this number provides reasonably tight estimates of our parameters of interest. The number of interview participants will relate to sufficiency of input from key stakeholders (clinical staff and patient participants) to achieve data saturation [28].

3
Discussion
In this pilot trial, we will evaluate the feasibility of this multicomponent intervention that combines traditional behavioral change techniques (e.g., education, goal setting) with complementary elements specifically developed for the clinical environment. These include serial assessments of behaviors through the PAI screener, personalized exercise coaching, self-monitoring and obtaining a provider recommendation for the participant to exercise. It will also include educational training for the clinical team. We hope to establish that the PAI intervention is feasible and move on to test the refined PAI intervention in a larger efficacy trial. The long-term goal is to see if our multicomponent approach could increase MVPA and reduce sedentary behavior in breast and colon cancer survivors, leading to a path for achieving a sustainable lifestyle in post-treatment recovery.
Increasing PA is an important and well recognized strategy for improving health outcomes among cancer survivors. Despite the benefits associated with PA, many BCS and CCS do not meet PA recommendations [29,30]. Moreover, most existing PA interventions currently only demonstrate short-term increases in MVPA and there is poor integration in the clinical setting. We also note that poor long-term maintenance of improved behavior is still a developing area of research. The PAI Trial is designed to provide the skills to live a more active lifestyle with a period of follow -up to determine if those skills were successfully developed. We also considered that the potential benefits of decreasing sedentary behavior are not as broadly acknowledged as may be most helpful to cancer survivors. Reducing sedentary behavior is likely a behavioral change that may be more readily achieved by survivors who may find increasing MVPA challenging [31], but more evidence on the specific applicability for BCS and CCS is still needed.
The PAI Trial may have limitations. Our sample may not be representative of all BCS or CCS residing in our catchment area. By nature of our decision to make this a remotely delivered intervention, we may have less participation from lower income patients or those with access challenges. However, we have included several elements to make the PAI intervention address personal needs, including offering one-on-one tech consultations to those with lower tech literacy. We are also recruiting from larger rural areas as well as large urban areas (both Winston-Salem and Charlotte), which will be very informative about the needs and preferences of the patients served by our health system. However, following COVID, we have determined that many middle-aged and older adults are far more comfortable with remote access for information than may have been true prior to that global pandemic. Indeed, we believe providing a remotely-delivered intervention will facilitate recruitment and participation, as well as scalability. One potential challenge may be that we are planning to test the feasibility of providing referrals to exercise physiologists, with knowledge that this is not a standard resource available everywhere. We hope this study will generate more data to support the value of such a service for cancer survivors, and model a proficient strategy for this service in cancer centers in the future.

CRediT authorship contribution statement
Shirley M. Bluethmann: Writing – review & editing, Writing – original draft, Supervision, Software, Resources, Methodology, Investigation, Funding acquisition, Data curation, Conceptualization. Janet Tooze: Writing – original draft, Resources, Methodology, Conceptualization. Joni K. Evans: Writing – original draft, Methodology. Jeffrey Katula: Writing – original draft, Resources, Methodology, Investigation, Conceptualization. Kristy Wood: Writing – original draft, Project administration, Methodology, Investigation. Lesley Hitariansingh: Writing – original draft, Project administration. Charlotte Crotts: Writing – original draft, Project administration, Methodology, Investigation. Heidi D. Klepin: Writing – original draft, Investigation. Ravi Paluri: Writing – original draft, Methodology, Investigation, Conceptualization. Kunal Kadakia: Writing – original draft, Resources, Methodology, Investigation. Katherine Ansley: Writing – original draft, Resources, Methodology, Investigation, Data curation, Conceptualization.

Funding acknowledgements

•Dr. Bluethmann was supported by a Mentored Research Scholar Grant in Applied and Clinical Research, MSRG-18-136-01-CPPB, from the American Cancer Society. Dr. Bluethmann is supported by a pilot in Cancer Prevention and Control Program from the Atrium Health Wake Forest Baptist Comprehensive Cancer Center: and a supplement funded through the National Cancer Institute
3P30CA012197-49S2, associated with Cancer Center Support Grant award number P30CA012197.

•Dr. Paluri has a private grant from Pfizer pharmaceuticals.

•Drs. Klepin and Bluethmann are both associated with the Cancer and Aging Research Group, currectly supported by the National Institute on Aging/National Institutes of Health under award number. R33AG059206.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.