Rationale and design of extended cancer education for longer term survivors (EXCELS): a randomized control trial of ‘high touch’ vs. ‘high tech’ cancer survivorship self-management tools in primary care

EXCELS is a two-phase research study to develop and evaluate the comprehensive intervention in a RCT. Phase 1 used a mixed-method, multi-step iterative process to develop intervention components. The development of the EXCELS intervention components was conducted with 45 patients and 10 clinicians. The objective of the Phase 2 is to determine the efficacy of the EXCELS intervention to increase engagement in preventive health services and monitoring for recurrence and late effects. The 18-month EXCELS trial implements a 2 × 2 factorial RCT design comprised of three intervention conditions and a comparison condition (see Fig. 1). Four hundred and eighty Breast, colorectal, and prostate cancer survivors are randomized to receive either EXCELS-website (high tech), EXCELS-health coaching (high touch), EXCELS- website and health coaching (integrated tech-touch), or a comparison condition (National Cancer Institute Facing Forward booklet) [26]. It is hypothesized that compared to the comparison condition, participants randomized to EXCELS-high tech, EXCELS- high touch, and EXCELS- integrated tech-touch conditions will demonstrate increased levels of engagement in preventive health services and monitoring for cancer recurrence and late effects via self-report (6 months, 12 months, and 18 months) and medical chart review for a subset of survivors (20 months). The secondary outcomes include the examination of mediators (i.e., patient activation, recurrence, and perceived intervention utility) and moderators (i.e., patient perceived cancer vulnerability and use of primary care) on the intervention’s impact via self-report.

EXCEL phase 2 ethics approval was received from the Rutgers University Institutional Review Board and the CentraState Healthcare System Institutional Review Board. Written informed consent will be obtained from all trial participants. The EXCELS trial is registered at www.​clinicaltrials.​gov, trial # NCT03233555.

The EXCELS trial is guided by the Cognitive-Social Health Information Processing Model (C-SHIP), an integrative theoretical framework which addresses how individuals process information about cancer threats and prevention options based on their perceived vulnerability, disease and self-efficacy expectations, goal and values and coping strategies [27]. The key constructs of the C-SHIP model operationalized in EXCELS include: 1) cancer and health relevant interpretations; 2) beliefs and expectancies about cancer treatment and disease outcomes; 3) cancer and health relevant goals and values; 4) cancer and health-relevant affective and emotional states; and 5) self-regulatory competencies. EXCELS supports processing of tailored cancer survivorship information using traditional formats (e.g., print materials), new technologies, and health coaching techniques to test their impact on survivor’s adoption of cancer control and health monitoring behaviors. The objectives of the EXCELS trial are to: 1) promote accurate processing of recurrence risk and risk of secondary cancers; 2) identify new or lingering treatment-related symptoms; 3) reframe beliefs and expectations about need for risk-based preventive care for existing sequelae; 4) articulate goals to encourage consistency between patient behavior and desire; and 5) identify barriers to goal-directed decision making. To foster accurate cancer and health-related interpretations, American Cancer Society guidelines for primary care providers on breast, colorectal and prostate cancer follow-up were converted into printed health coaching manuals and web-based education materials [9‐11].

EXCELS seeks to recruit 480 cancer survivors who are long-term survivors whom may seek follow-up care in primary care settings for management of late and long-term effects of cancer and its treatment. Two recruitment strategies are utilized to optimize patient recruitment into EXCELS, the New Jersey Primary Care Research Network (NJPCRN) and the New Jersey State Cancer Registry (NJSCR). To be eligible for EXCELS, survivors must be: 1) diagnosed with breast, colorectal, or prostate cancer (stages 1–3); 2) have completed active treatment for their cancer diagnosis (except hormonal treatment); 3) have access to a computer, a smartphone or an I-Pad/tablet equivalent with internet access; 4) have access to a phone for contact with health coach; 5) able to read and speak English; and 6) be competent to provide written informed consent. Survivors are excluded or removed from the trial if they experience a cancer recurrence or a new cancer diagnosis during their participation in the EXCELS trial.

The first recruitment strategy is within primary care practices affiliated with NJPCRN, a practice-based research network that has well established collaborative research linkages. To identify survivors, practices request a report from their billing system using 57 ICD-10 codes for breast, colorectal and prostate cancer diagnosis or history. Identified patients receive a mailed invitation from their primary care practice that also includes a letter from the study principal investigator (PI). The letter from the PI highlights trial eligibility criteria, offer three methods to ‘opt-in’ (postage-paid postcard, telephone, or email) to receive a telephone call to assess eligibility, and include the study telephone number and e-mail address. Research assistants send up to two repeat mailings after the initial mailing. Patients that meet eligibility criteria and volunteer to take part in the trial are emailed a link to complete an electronic informed consent. Participants recruited though NJPCRN will have their medical charts reviewed to determine receipt of preventive health service use and cancer related follow up care. The chart audit data will be compared to patient self-report to verify self-management behaviors. In addition, patients recruited through NJPCRN practices are asked to provide written release for cancer-related information (cancer site, stage, tumor characteristics e.g., hormonal status, Gleason score) obtained from the NJSCR. Patient cancer-related data will verify patient self-reported data at baseline and in the educational tailoring process to access potential implementation issues.

Participants are also recruited from the NJSCR, a resource established by the New Jersey Legislature to evaluate the impact of cancer at the state level (N.J.S.A. 26:2–104 et seq.). The NJSCR can only release identifiable information about individuals diagnosed with cancer for research with patient consent. Therefore, initial patient contact to determine interest and trial eligibility is through the NJSCR staff. NJSCR staff identifies patients from the registry with diagnoses of breast, colorectal and prostate cancers and contact the patient’s physician of record. The physician of record receives a mailed informational letter describing the trial and NJSCR’s intention to contact the patient for screening and trial participation. The physician of record is given a two-week window to notify the NJSCR staff of any objections to contacting the patient, (e.g., death, mental illness, severe illness, etc.) otherwise, passive physician consent is assumed. After physician notification, EXCELS recruitment packets are mailed from NJSCR to potentially eligible survivors. The contents of the packets include: 1) an invitation letter that highlights eligibility criteria for the EXCELS trial; 2) EXCELS information sheet with instructions on how to give agreement to contact (e.g., over telephone, email or by returning signed postage-paid release); 3) NJSCR frequently asked questions brochure; and 4) EXCELS trial recruitment flyer. One week after the recruitment packet is mailed a NJSCR team member calls potential participants to confirm receipt of the recruitment packet, answer questions about EXCELS, and screen for trial eligibility. Potential participants are called once per week for up to three weeks. If participants are not reached during this period, a final recruitment packet is mailed. Participants who are reached, screened, and eligible then have a consent discussion by telephone and e-mailed a link to complete an electronic informed consent. Once consented, the participant’s contact information is released to the EXCELS research team. Participants may also return a written release of information to NJSCR via postage paid envelope from the recruitment packet. In this case, participant information is securely forwarded from NJSCR to the EXCELS research team. The EXCELS research team proceeds with eligibility screening and subsequent electronic informed consent.

Upon completion of electronic informed consent, participants complete an electronic baseline questionnaire (see Fig. 1). After completion of baseline questionnaire, participants are randomized into one of four intervention conditions using a computer-based algorithm that will equalize groups across conditions through the DatStat software system: 1) EXCELS-high tech; 2) EXCELS- high touch; 3) EXCELS- integrated tech-touch and 4) comparison-booklet. The randomization algorithm equalizes the groups across experimental conditions, 120 participants per condition for 480 participants. The breakdown of cancer disease site is 200 breast cancer survivors, 200 prostate cancer survivors, and 80 colorectal cancer survivors.

Participants complete self-administered questionnaires at baseline, 6-months, 12-months, and 18-months post-baseline. Each questionnaire is administered as an online survey through DatStat software that takes approximately 30 min to complete. Reminder emails are sent to enrolled participants one week before the scheduled follow-up assessment date. Demographic characteristics (e.g., gender, race/ethnicity, age, income, smoking status, etc.) are assessed at baseline only. Participants receive a $20 gift card for each of the four questionnaires completed. Variable selection was guided by the C-SHIP model constructs. Measure selection was guided by empirical evidence of instrument reliability, validity, and sensitivity to change. A list of primary outcomes, mediators, and moderators, along with their associated data source and time interval are provided in Table 2. These data are captured through patient self-report and a confirmatory medical chart review in the subset of participants recruited from the NJPCRN.

Preventive services use is measured as adherence to evidence based guideline care based on author adapted measures [28] updated using criteria from the Physician Quality Reporting System (PQRS) measures [29]. Eligibility for guideline-related services is determined based on the medical assessment, including patient/family history, laboratory and physical findings in the medical records and patient surveys. Measures of preventive health services include influenza vaccination for patients 50 or older, pneumonia vaccination for patients 65 and older, receipt of annual physical exams, breast cancer screening with mammography, and colorectal cancer screening. Adherence is computed as a percentage of the eligible guideline-related services each participant has received.

Cancer recurrence is assessed via patient self-report “At any time since you were diagnosed with your primary cancer, did a doctor or other health professional tell you that your cancer had come back (did you have a recurrence)?” The validated Charlson Comorbidity Index (CCI) will assess late effects of cancer treatment and comorbid conditions [30]. The CCI, based on 19 chronic conditions, is a weighted index that takes into account the number and seriousness of comorbid diseases. We will also assess additional chronic diseases via participant self-assessment.

The statistical power calculation of the trial is based on the primary outcome of increased engagement in preventive health services and monitoring for recurrence and late effects. The key assumptions for the power calculations stem from the literature and preliminary data. Given the trial’s hypothesis participants randomized to EXCELS intervention conditions will demonstrate greater effects over those in the comparison condition. Assuming a Pearson chi-square test for comparing rates of adherence to recommendations between two arms, we calculated the power attained using 120 individuals per condition arm. EXCELS has four condition arms, thus six pairwise comparisons of conditions and each comparison can be tested at the 0.00833 level (Bonferroni correction 0.05/6) to attain a 0.05 family-wise error rate. We conservatively assume that across conditions, the rate of adherence is 50%. With a rate ratio of 2.0 (0.66 versus 0.33), the power to detect an effect of treatment is 85.9%. Power would be greater if rates of adherence for the comparison group are less than 0.33 or at 0.50 or greater.

The primary outcomes are preventive health services engagement, monitoring for recurrence and late effects. Descriptive and exploratory analyses will be conducted for all baseline measures to characterize the sample and stratify by treatment arm to describe the inferential population. Primary analyses will be conducted with analysis of variance (ANOVA) to compare the proportions of preventive and surveillance services used and proportions of potential late effects tracked for which the participant was eligible, between participants in intervention conditions and the comparison condition. To examine mediators of primary outcomes, ANOVA models will be expanded to include a main effect of the mediator as well as interaction between the conditions and the potential moderator. F-tests of the interaction will assess any moderating effects. If significant, linear combinations of regression coefficients will estimate the effect of the intervention at varying levels of the moderator. For each potential mediator, the four component steps of the mediation model will be examined via regression models. Bootstrap procedures will formally test for significance and estimate the percent contribution of the mediating pathway to the main effect of the intervention. All participants will be informed by e-mail about the final results of the trial. Scientific publications will also be planned.

Participants using a secure website will complete online surveys using a secure website (hosted on DatStat servers). The software allows for research study personnel to be assigned data access and privileges specific to their role on the study. Databases for participant recruitment, tracking, medical records review data, and survey data are maintained by the Rutgers Cancer Institute of New Jersey, Population Science Research Support Core using HIPAA-compliant DatStat software.

Rutgers University maintains specific formal policies and infrastructure that have been developed to assist investigators with ensuring that research participants and research data are adequately protected from harm. In accordance with these policies, for this study we will use standard procedures and infrastructure for data and safety monitoring. We will not use a protocol-specific Data Safety Monitoring Board since the study involves minimal risk. Thus, the oversight committees is adequate to ensure participant safety and data integrity. Study protocol will be submitted for annual review to the Rutgers University Institutional Review Board.