CHAT-S Study Protocol: A randomized controlled trial of a health insurance literacy education program for young adult cancer survivors

Abstract

Background:

Health insurance education could mitigate financial toxicity experienced by young adult (YA) cancer survivors by increasing confidence when navigating cancer care costs. This paper describes the protocol in a randomized controlled trial (RCT) to test a virtual patient navigation program designed to help YA cancer survivors understand their health insurance.

Methods:

This is a two-arm, multi-site (Huntsman Cancer Institute, Intermountain Health) RCT wherein intervention participants receive four sessions with a patient navigator (PN) and a booklet on insurance; usual care receives the booklet. We will enroll 300 YA cancer survivors (n=200 intervention; n=100 usual care) diagnosed with breast, testicular, lymphoma, sarcoma, colorectal, melanoma, or thyroid cancer between the ages of 26 and 39, who have completed treatment in the past two years. All participants will complete three surveys: enrollment, 6 months, and 12 months; medical records/insurance claims data will be collected out to 18-month follow-up. Recruitment began in the fall of 2023 and is expected to last approximately 2.5 years. The primary efficacy outcomes include improvement in health insurance literacy and financial toxicity at 6 months. Secondary outcomes include adherence to cancer surveillance guidelines at 18 months. We will also conduct cost-effectiveness and budget impact analyses.

Discussion:

Anticipated results from this trial could identify key information that YA cancer survivors need to improve health insurance literacy and survivorship care.

Introduction

Cancer survivors often face financial toxicity, the adverse financial effects from medical expenses that impact their health-related quality of life (Di Giuseppe et al., 2023). In the United States (US), financial stress due to medical treatment of cancer can be especially impactful for young adult (YA) cancer survivors diagnosed in their 20s and 30s, as the insurance system is largely employer based; at diagnosis YAs are often at the beginning of their careers, and may be transitioning off their parent’s insurance or employed at a position without insurance coverage (Kirchhoff et al., 2024). Others are experiencing other life changes that are costly such as having children (Di Giuseppe et al., 2023; Salsman et al., 2019). YA cancer survivors are less likely to have insurance coverage and are underinsured at higher levels than older cancer survivors and report difficulty covering medical costs (Corrigan et al., 2022; Kirchhoff et al., 2013; Thom et al., 2023; Zheng et al., 2019). Over 70% of US adults with medical debt are insured (Pollitz et al., 2014). While having quality health insurance can be an important resource for cost mitigation, navigating health insurance can be stressful and confusing for YA survivors given many have little experience managing insurance due to their young age and health status (Waters et al., 2022).

Health insurance literacy, a patient’s ability to make informed decisions about choosing and using health insurance, can be an important factor in navigating one’s health care costs (Paez et al., 2014). A lack of understanding of health insurance can result in unexpected cost to patients or can lead to patients choosing to skip treatments or appointments due to cost concerns (Waters et al., 2022). Yet, there have been few interventions designed or tested for addressing the link between low health insurance literacy and lessening the financial hardship experienced by cancer survivors (G. L. Smith et al., 2022). One promising strategy for addressing these needs is patient navigation. Patient navigators are individuals that provide support for the logistic and administrative part of cancer treatment. Patient navigators as a means to address patient financial concerns and improve insurance literacy are a promising strategy as they have successfully assisted in cancer settings forhelping patients access health insurance, complete paperwork for financial assistance applications, and reduce costs by connecting patients with financial assistance programs (Chan et al., 2023). Patient navigation can improve quality of life, patient satisfaction during survivorship, and reduce hospital readmission during active treatment and survivorship (Chan et al., 2023).

We are now enrolling YA survivors in Let’s Chat about Health Insurance for Survivors (CHAT-S), which is a virtual patient navigation health insurance literacy intervention efficacy trial (Kirchhoff et al., 2024; Park et al., 2024) (Table 1). The CHAT program was adapted from the Health Insurance Navigation Tools (HINT) program, a navigation program for childhood cancer survivors; both have demonstrated proof of concept in improving health insurance literacy with two pilot trials listed in Table 1. CHAT and HINT program development are described elsewhere (Vaca Lopez et al., 2023; Warner et al., 2024).

Table 1:

Health Insurance Literacy (HIL) Patient Navigation Interventions for Cancer Survivors

HINT: Health Insurance Navigation Tools Program
Intervention Name	Implementation Phase	Participants	Findings
HINT 1:  NCT04520061	Preliminary testing: Pilot Trial	Adult survivors of childhood cancer	Program is feasible, acceptable, and improves HIL (Park et al., 2024)
HINT 2:  NCT05527392	Type I hybrid effectiveness-implementation trial	Adult survivors of childhood cancer	Recruitment ongoing; Protocol is reported elsewhere (Park et al., in review)
CHAT: Let’s Chat about Health Insurance Program
Intervention Name	Implementation Phase	Participants	Findings
Huntsman-Intermountain Adolescent and Young Adult (HIAYA) CHAT NCT04448678	Preliminary testing: Pilot Trial	Adolescents and young adults diagnosed with cancer in the past year	Program is feasible, acceptable, and improves HIL (Kirchhoff et al., 2024)
CHAT-S NCT05829070	Efficacy Trial	Young adults finishing cancer treatment within the past 2 years	Recruitment ongoing

Both the HINT and CHAT health insurance literacy interventions are informed by the Andersen and Aday behavioral model of health services use and Levy and Meltzer’s proposed relationships between health and health insurance (Figure 1) (Andersen & Aday, 1974; Levy & Meltzer, 2001). Andersen and Aday describe the relationships among pre-disposing characteristics (e.g., demographics, health beliefs), enabling resources (e.g., personal and community resources), need (perceived and evaluated) and the eventual use of health services. This complements the Andersen and Aday framework by proposing interrelationships among these resources and behaviors. Thus, by intervening on health insurance literacy (as an enabling characteristic), CHAT-S aims to empower YA cancer survivors to maximize the utility of their health insurance with the hope that it will positively impact both their use of health services and their overall health, as well as reduce their financial toxicity.

Figure 1.

CHAT-S Theoretical Framework

As the original Huntsman-Intermountain Adolescent and Young Adult (HIAYA) CHAT pilot trial demonstrated that younger patients (18–25) and patients currently undergoing intensive cancer treatment were often unable to participate in multiple intervention sessions, the current CHAT-S protocol focuses on patients ages 26 to 39 at diagnosis who have finished treatment. Herein we describe the trial protocol of the National Cancer Institute (NCI) funded CHAT-S efficacy trial, which began recruitment in November 2023.

Objectives/Aims

The purpose of the CHAT-S intervention trial is to determine the efficacy of a patient navigation intervention for improving health insurance literacy and financial toxicity among YA cancer survivors compared to participants in the usual care arm (see Intervention below for description of the two study arms). The following aims will guide this trial:

1. Determine the efficacy of CHAT-S in improving health insurance literacy and financial toxicity outcomes (primary outcomes).

2. Investigate whether CHAT-S improves surveillance care (i.e., care to identify potential recurrences; secondary outcome).

3. Conduct a cost-effectiveness and budget impact analysis of CHAT-S for improving health insurance literacy, financial toxicity, and surveillance care to inform sustainability, as well as future dissemination and implementation efforts.

Trial design

The study design is a randomized controlled trial with parallel groups investigating whether the intervention group has superior outcomes to the usual care group (Figure 2).

Figure 2.

CHAT-S Trial design

Methods: Participants, interventions, and outcomes

Participants

Study setting

Participants will be recruited from Huntsman Cancer Institute (HCI) and Intermountain Health (IH). HCI is the only NCI-designated Comprehensive Cancer Center in the Mountain West region, serving all of Utah, Idaho, Montana, Nevada, and Wyoming. IH is comprised of 23 hospitals across Utah and Idaho. Together, both locations provide the majority of oncology services for Utah. HCI and IH are also home to the HIAYA Adolescent and Young Adult Cancer Care Program, which provides navigation and programmatic support for AYAs with cancer in the region. Participants in the trial have access to the HIAYA Program activities, which can include navigation at their treating institution for health education, fertility services, support for concerns such as work, school, financial stress, and other psychosocial needs. They can also participate in HIAYA Program activities, which includes speakers and group activities.

HCI and IH have a unique and integrated data infrastructure through the Utah Population Database resource, which allows us to capture electronic health record (EHR) data for these two systems, statewide insurance claims data, and statewide information on health care encounters and medical procedures from hospital facilities databases, to identify use of medical care among our sample. All survey data for the trial is collected and managed using REDCap, an electronic data capture tool hosted at HCI (Harris et al., 2009).

Eligibility criteria

Eligible individuals will include YA cancer survivors who were originally diagnosed with cancer between the ages of 26 to 39 and have completed initial treatment for their cancer (i.e., radiation, chemotherapy, or surgery) within the past two years. We will include breast, testicular, lymphoma, sarcoma, colorectal, melanoma, and thyroid cancer diagnoses as these are common cancers for young adults (American Cancer Society, 2019). Potential participants who do not have health insurance, do not speak English, or have a developmental delay will be excluded. Also, if potential participants cannot participate through a phone or video-capable computer, they will be excluded. As needed, we will allow participation over the phone by mailing/emailing the materials ahead of each session—this should be minimal, as 90% of YAs have internet access (Pew Research Center, 2021). Following baseline survey completion, participants will be randomized to either the usual care or intervention arm.

Intervention

Description of the usual care group

Individuals in the usual care group, following completion of the baseline survey and randomization, will receive an email that links to a website with the digital health insurance booklet, which includes resources regarding health insurance, financial support, and survivorship care. The digital booklet provides the same information that the patient navigator (PN) will cover in the intervention sessions.

Description of the intervention group

The CHAT-S intervention is a series of four one-on-one PN-led videoconference sessions that will consist of guided conversation between the PN and the participant about important information relating to health insurance (Table 2). Each session will have a focus on a different aspect of health insurance navigation and is guided by the digital booklet. After the baseline survey, intervention participants will be contacted by the PN to schedule their virtual education sessions. CHAT-S sessions last for 35–45 minutes and will be held once a week for four weeks. The first session will focus primarily on survivorship care resources, important insurance terms and concepts (e.g., coinsurance), and usual costs. The second session will cover different insurance plan types, interpreting and accessing information about what goods and services are covered, and how to interpret medical bills. The third session will describe relevant and important health insurance legislation and the appeals process. Lastly, the fourth session includes a conversation on managing care costs including coaching on how to have cost of care conversations with health care team members and how to budget. While the content of each session is fixed, the PN will tailor sessions to the individual participant based on their knowledge of the participant’s diagnosis, information gained through their conversations during sessions, and any specific questions the participant may ask.

Table 2.

CHAT-S Intervention Learning Content

	Session 1: Introduction to health insurance	Session 2: Deep dive into your health insurance	Session 3: Health insurance legislation and appeals	Session 4: Cost conversations and budgeting coaching
Fixed	• Basic insurance terms and concepts • Phases of survivorship • Survivorship tools including NCCN guidelines	• Types of insurance coverage • Benefits of individual’s chosen insurance plan • Breakdown of bills and explanation of benefits • Deductibles and other cost-sharing mechanisms	• ACA, FMLA, Hospital Transparency, No Surprises Act, and other laws and protections • Relevance of laws for cancer survivors • Appeals process	• Assertiveness training • Budgeting examples and further resources
Tailored	Specific questions, responses to anecdotes, adjusting time spent on each topic (i.e., covering a topic less if participant is already familiar with it so they can spend more time on a less familiar topic)

Strategies to improve adherence to the intervention

Navigator Training: Prior to launch of the trial, the study-specific PN will complete trainings to enhance skills in motivational interviewing, health insurance and financial needs of cancer patients, and interacting with participants of different cultures/backgrounds. PN with nursing, social work, public health, health education, or other related disciplines can be trained to deliver the content. The current PN was hired for the CHAT-S study and has background in public health interventions and community health education.

The PN will complete two trainings developed by the George Washington Cancer Center: 1) “Oncology Patient Navigator Training: The Fundamentals” and 2) “The Executive Training on Navigation and Survivorship”. The PN will also complete several trainings on insurance and finances, including the Triage Cancer Finance and Insurance Intensive and the Association of Community Cancer Centers Financial Advocacy Bootcamp, and receive guidance from oncologists on our team regarding survivorship care. Other trainings include the Utah Pride Center’s Cultural Competency Training for health care professionals, and the Racial Equity Institute’s virtual workshop.

In addition, we will train the PN in the theoretical basis and rationale for the study design, as well as developing standardized responses about the study to tell participants who may ask common questions (Andersen & Aday, 1974; Bellg et al., 2004). There will also weekly supervision meetings between the navigator and a study co-investigator with expertise in navigation interventions, to debrief on: strategies to direct the conversation during difficult conversations, answering challenging participant questions, and identifying ongoing insurance policies and legislation changes that may affect the intervention. Together, these trainings and supervision will provide a foundation to ensure the intervention happens consistently (e.g., sessions are roughly the same length, all participants receive knowledge regardless of varying demographic characteristics) and accurately (e.g., PN is trained to know how to answer unforeseen follow-up questions related to insurance or other circumstances that may arise).

Intervention Fidelity: To monitor intervention fidelity, we will audio record and evaluate the PN sessions, using checklists regarding session content and quality (Bellg et al., 2004). Following the PN Research Program’s navigation performance checklist (Freund et al., 2008) that specifies two quality indicators of care: 1) participant interaction (e.g., established rapport) and 2) intervention delivery (e.g., relevant information provided and understood by participant), we will review 15% of sessions to confirm that appropriate information was shared and that the intervention modules were completed. To accomplish this, the PN will complete a fidelity checklist after each session to ensure the material is covered as intended and complete a post-session form to document delivered session content. Two co-investigators trained to assess intervention fidelity will also complete fidelity forms in REDCap as they listen to the randomly selected 15% of sessions. The PN and co-investigators meet weekly to discuss fidelity-related questions or concerns, which can include challenging sessions, specific health insurance policies, and guidance for difficult conversations.

Outcomes

Primary Outcome Measures

The primary outcomes to be collected from the surveys are 1) health insurance literacy and 2) financial toxicity at 6 months (Table 3). The measure used to assess health insurance literacy will be the Health Insurance Literacy Measure (HILM) (Paez et al., 2014); HILM is a 21-item validated tool designed to capture health insurance and cost literacy, scored from 0–84 with higher scores indicating greater literacy. Financial toxicity will be measured using the COmprehensive Score for financial Toxicity (COST) (de Souza et al., 2014; de Souza et al., 2017), which ranges from 0–44 with lower scores representing greater hardship.

Table 3.

Study Measures of Interest

Outcome Measures
Primary Outcomes: Survey
Health Insurance Literacy (Paez et al., 2014)	1. Understanding of terms (e.g., premium, deductible, co-payments, co-insurance, out-ofpocket spending limits, annual limits) 2. Confidence in choosing, comparing, and using insurance plans
Financial Toxicity (de Souza et al., 2014; de Souza et al., 2017)	Perceptions and behaviors regarding costs of cancer care
Secondary Outcomes: EHR/claims/facilities data
Surveillance Care for Recurrence	Determinations about the appropriateness of surveillance for each participant will follow National Comprehensive Cancer Network (NCCN) for each disease type. These guidelines incorporate factors such as cancer histology and stage. We will use procedure and diagnosis codes from the medical records, facilities data, and claims data to identify the types of care received.
Other Medical Care	Types of providers seen (oncologist, primary care, social workers, health educators, etc); Number of visits/types in the past year. We will use procedure and diagnosis codes from the medical records, facilities data, and claims data to identify the types of care received.
Secondary Outcomes: Survey
Survivorship Care	Did they discuss with a provider the need for regular follow-up care and monitoring even after completing your treatment? Do they have a primary care provider?
Familiarity with important Health Care Legislation	Familiarity with Affordable Care Act policies (Park et al., 2015) such as appeals, preventive care coverage, and summary of benefits/coverage; No Surprises Act (e.g., appeals on out of network cost sharing)
Care access	Trouble finding a provider who accepts insurance/getting an appointment as needed; Unmet health care need due to cost; Out of pocket medical costs>10% of income; Worry about medical costs
Cost Related Literacy	Comfort/confidence talking about medical costs with providers (Bestvina et al., 2014) understanding of financial concepts of care
Satisfaction with navigation sessions	Acceptability of the intervention
Other Measures
Measures from Session Transcripts
Navigator Rapport	Linguistic Inquiry Word Count documenting the proportion of positive and affirming language exchanged between the navigator and participants
Demographic Measures: Survey
Individual	Age, gender, sexual orientation, education, race, ethnicity, marital status, rurality, employment status
Household	Household & personal income, and housing stability, food insecurity, transportation insecurity
Perceived Stress	Stress levels from Perceived Stress Scale
Insurance characteristics	Insurance status; Policy holder (e.g., self, spouse, parent) and source Plan source: Employer-sponsored, direct purchase (exchange or outside), Medicaid, Medicare Type of plan: High-deductible plan; Narrow network plan Being forced to switch plans because of cancellation (in past year) Satisfaction with current insurance plan or previous insurance plan if uninsured
Clinical Measures: EHR/claims/facilities data
Cancer information	Cancer diagnosis, age at and years since diagnosis, treatment, health status

Secondary Outcome Measures

Secondary outcome measures include 1) adherence to cancer surveillance guidelines, 2) receipt of survivorship care, 3) access to care and health insurance changes, 4) aspects of cost-related literacy, and 5) familiarity with relevant health care legislation.

For adherence to surveillance guidelines and receipt of survivorship care, we will investigate whether participants met (partially or fully) or failed to meet cancer surveillance screening recommendations for recurrence and whether they received other survivorship care or not. Survivorship care will be evaluated through the 12-month follow-up survey and medical records using current procedural terminology (CPT) codes to designate whether participants have visited with both their primary care providers and oncologists in the past year, similar to other studies (Hahn et al., 2019; Ramsay et al., 2022).

The remaining secondary outcomes will be elicited through the 6-month and 12-month follow-up surveys. These surveys have questions identifying any changes in insurance coverage or care access, cost-related literacy, and familiarity with important health insurance legislation. Changes in health insurance will be evaluated through a series of items that identify the participant’s current health coverage. We will ask for changes regarding policy holder status, benefit coverage, and satisfaction with their health insurance coverage, as well as items relating to access to care (e.g., ability to find a provider in network, receiving treatment that was not covered by insurance). A component of cost-related literacy includes ability to discuss financial burden with health care providers and will be evaluated in follow-up survey questions asking if and with whom on their care team participants spoke with about health care costs (Park et al., 2017; Zafar et al., 2015). We will assess familiarity with Affordable Care Act (ACA) protections (e.g., appeals process) and familiarity with other important insurance legislation (e.g., FMLA, ADA, No Surprises Act) using a series of Likert scale items ranging from very familiar to not at all familiar (Park et al., 2015).

Other Study Outcomes

We will assess intervention feasibility and acceptability among intervention arm participants. Feasibility will be assessed by examining the participation rate, as well as comparing the percentage of participants enrolled to the percentage of participants that complete all four sessions. Acceptability will be assessed via participant satisfaction with the intervention, as measured on a 5-point Likert scale of satisfaction with CHAT-S (To what extent did CHAT-S meet your needs? Did you get the kind of insurance assistance that you wanted? How helpful has this program been for you for accessing survivorship care?). We will also conduct a linguistic analysis using the Linguistic Inquiry Word Count (LIWC) tool to identify and describe rapport building between the participant and the PN. Specifically, we will focus on the proportion of language that is positive and affirming exchanged throughout the four navigation sessions.

We will conduct cost-effectiveness analyses (CEAs) to determine the impact of CHAT-S on cost outcomes relative to health outcomes. These CEAs will include both a short-term horizon perspective and long-term perspective. In our short-term time horizon CEA, we will combine an assessment of costs from the societal perspective with the results of the impact of CHAT-S on our primary outcomes of interest over the 12-month period. The long-term time horizon CEA will use a Markov simulation model to extrapolate the results from our trial into transitions between cancer remission, recurrence, and death health states over a survivor’s lifetime with costs captured from a societal perspective. To complement the CEA estimates, we will conduct a budget impact analysis (BIA) to calculate the total financial outlay from a health system’s perspective for implementing the CHAT-S intervention.

Participant timeline

Study participation lasts 12 months, beginning on the day the consent form and baseline survey are completed (Figure 3). Once a completed baseline survey is submitted, REDCap will automatically randomize the participant based on a pre-programmed randomization schema and triggers an email to the participant explaining which arm they have been assigned to and what the next steps are. Each participant will then engage in the intervention or usual care arm and complete additional surveys at 6 months and 12 months post enrollment. Each survey is estimated to take 15–20 minutes to complete. Participants will be given $30 Amazon gift cards for each completed survey, for a total of $90 if all surveys are completed.

Figure 3.

Participant Timeline

Those assigned to the usual care arm will be emailed a link to a REDCap landing page where they can download a PDF copy of the digital health insurance education booklet. For tracking purposes, each participant who downloads the PDF will be asked for their name. Participants assigned to the intervention arm will receive an email explaining the components of the intervention and a link to self-schedule their first session. Study staff follow-up at this point and then the PN will reach out to the participant if they do not self-schedule within one week. Ideally, sessions will be scheduled every week or every other week until they are completed, but the timing between sessions can be flexible depending on participant schedule. The PN will send the participant a follow-up email after each session, which includes the corresponding section of the booklet that was covered in the session. The intervention arm will receive five additional questions in their 6-month follow-up survey regarding navigation session satisfaction, while the usual care arm will receive two questions regarding booklet access. Otherwise, the two follow-up surveys for the intervention and usual care arms are the same.

Sample Size

To ensure power of at least 80% for the primary outcomes, we will enroll 300 participants and randomize 200 to the intervention arm and 100 to the control arm. This 2:1 randomization schema allows for more YA cancer survivors to access the intervention while still maintaining the statistical power needed to assess effect size. This sample size also accounts for an attrition rate of 20%, which is based on prior studies (Kirchhoff et al., 2024), and subsequently results in an anticipated final sample of n=240 (160 randomized to intervention arm, 80 randomized to control arm).

Recruitment

The study teams at HCI and IH will receive periodic data pulls from their respective EHR systems containing potentially eligible patients. Data analysts will generate queries at each institution using eligibility criteria. Referrals will also be gathered from oncologists and patient navigators at HCI and IH. The study team will use the EHR to verify age, diagnosis, and whether the potential participant has finished treatment. Potentially eligible patients will be added to a screening log and contacted approximately 8 times over a 4-week period for recruitment into the study, unless they consent or decline to participate. These contact attempts include phone calls, patient portal messages, and emails; HCI patients are also contacted via text. The type of contact, date of contact, and any other relevant information will be noted in the screening log. Additional recruitment will occur on social media channels run by groups at HCI and IH. We plan to enroll 10–12 participants a month to reach our goal of 300 participants over 2.5 years (30 months; recruitment started in November 2023 and is anticipated to go through May 2026). We anticipate at least 1,026 eligible patients will be treated at HCI or IH during our recruitment period.

Methods: Assignment of interventions

Sequence generation

Once a participant signs the consent form, they will be automatically randomized to one of the two arms in the REDCap system by a randomization schema. This randomization position is held until the participant finishes the survey and clicks “Submit” on the last page. Once participants submit their survey, they will be emailed information about what arm they have been assigned to and relevant instructions.

Blinding

The study principal investigators and study team will not be blinded to randomization. The lead statistician and senior data manager (who oversees the study data team) will be blinded to treatment allocation until the analysis is complete. To ensure the data team remains blinded, direct access to the survey and randomization information in REDCap is restricted.

Methods: Data collection, management, and analysis

Data collection

All baseline and follow-up survey data will be stored in REDCap until the completion of the trial. EHR/claims/facilities data will be obtained for all participants out to 18 months after enrollment into the study from the Utah Population Database, which provides crosswalks between the Utah Cancer Registry, HCI EHR, IH EHR, insurance claims, and other facilities data for the state (K. Smith et al., 2022).

Strategies for retention

In our pilot studies, our most successful strategies to promote retention were 1) delivering intervention sessions virtually through videoconference with maximal flexibility, 2) conducting follow-up assessment reminders via multiple modality options, 3) promoting familiarity with the study team via sending participants study updates halfway through the trial linking them to research team social media pages, and 4) providing remuneration for surveys completed (Kirchhoff et al., 2024).

We will apply current best practices for retention for randomized controlled trials and gathering survey data on participants at HCI/IH clinics. We will obtain phone numbers, email addresses, and mailing addresses to limit loss to follow-up. In addition, we have designed the CHAT-S intervention to be delivered through videoconference, meaning that the sessions will occur at the convenience of the participant. If needed, we will also allow participation in the intervention sessions and surveys via telephone to promote retention.

Data Management

All surveys will be completed and stored in REDCap. Study coordinators will monitor REDCap weekly to check for any irregularities in data collection forms. Participant records in REDCap will include consent forms and surveys. Audio recordings of navigation sessions for intervention arm participants will be stored on the research team’s secure, password-protected computers and encrypted HCI servers for the planned LIWC analyses to investigate navigator rapport. Participant audio recordings will be restricted to only the study staff directly assessing intervention fidelity or providing conceptual feedback on this process. No audio recordings or identifiable participant details will be used in public presentation of results. No physical data will likely be collected for this study, but if any participants were to request paper surveys, they would be held in locked research team offices for 3 years post-trial and then destroyed.

Statistical methods

Aim 1 Analyses

We will use descriptive statistics to measure differences in the primary outcome measures (HILM and COST) between the baseline and follow-up timepoints (6 months and 12 months). Imbalanced covariates will be included as adjustment variables. Changes in our primary outcome measures will be compared between the intervention and usual care arms using multivariable linear mixed effects models, adjusted for stratification factors. In the context of these multivariable mixed effects models, the test of the CHAT-S intervention effect on change in primary outcomes will be the Wald test of the corresponding study arm by follow-up time interaction.

In addition, in exploratory analyses to identify potential subpopulations that may experience differential intervention effects (e.g., differences between female and male survivors), we will examine differences in the primary outcomes by key groups (age, gender, and cancer diagnosis) at 6 months; if we identify statistically significant differences, we will run stratified models of the HILM (health insurance literacy) and COST (financial toxicity) analyses to explore differences in intervention effects among subgroups. We will also explore how rapport differences measured by LIWC affect intervention outcomes.

All randomized participants will be analyzed according to intention-to-treat for the primary comparisons. Further, all randomized patients will have complete baseline data on cancer type, rurality, and ethnicity as well as study endpoints (financial toxicity and health insurance literacy). Participants with missing follow-up data on study endpoints at 6 and 12 months will be included in analyses, but only with their baseline measures. This analysis will ensure validity of comparisons between study arms under missingness-at-random (data values and missingness pattern independent conditional on the observed data), the most general data-centric missing data assumption (Rubin & Schenker, 1991). Additional exploratory analysis aimed at data-driven patient cluster identification will be conducted, where the trial population will first undergo a data-driven clustering (via a spectrum of candidate techniques including k-means, hierarchical, and mixture model clustering), then clusters will be examined for differential treatment effects (effect moderation) in clusters.

Aim 2 Analyses

Binary endpoints of surveillance (e.g., meet/failed to meet recommended surveillance care to screen for cancer recurrence by 18-month follow-up) will be compared via analogous multivariable logistic mixed effects models. We will also examine potential moderators of interest including demographic factors; we will examine correlation among the moderators of interest and use that to inform moderator selection for formal testing. Pre-specified tests for effect modification driven by each of the potential moderators will be conducted by including potential moderator and study arm by follow-up time by potential moderator interaction terms and conducting Wald tests of the interaction in the corresponding multivariable models. Corresponding subgroup analyses will also be conducted. Notably, the moderator and subgroup analyses are exploratory, with several candidate effect modifiers examined, thereby increasing the risk of false positive as well as false negative results. These analyses will require cautious interpretation and follow-up validations. We will compare baseline/end-of-treatment endpoints within groups as supportive analyses, via paired t-tests for quantitative outcomes and McNemar’s tests for categorical outcomes. We will ascertain whether rurality, ethnicity, and other factors, such as insurance changes throughout follow-up, may be associated with the intervention effects. Finally, from the medical records data we will identify whether recurrences or new primary cancers occur during the study, as this could disrupt the patient’s survivorship care plan and thus impact our Aim 2 analysis.

Aim 3 Analyses

The CEAs and BIA analysis will be led by a co-investigator who is a health economist, who will utilize EHR/claims/facilities data for health care utilization, study team time tracking data for staff time and intervention costs, and literature about current medical costs. We will assess cost differences between each treatment group using ordinary least squares, generalized linear models, or two-part models chosen based on the distributional characteristics of the outcome variables. We will use the estimates of the CHAT-S intervention and health care utilization to conduct a BIA of a health system implementing the intervention over varying time periods.

For our long-term time horizon CEA, we will construct Markov simulation models to extrapolate the results from our trial. Estimated differences in cost and effectiveness outcomes derived from regression and simulation models will be combined to construct incremental cost-effectiveness ratios (ICERs) which will convey the cost per additional unit of health insurance literacy associated with the CHAT-S intervention. Results from our model will be presented as ICERs comparing the CHAT-S intervention to usual care.

Methods: Monitoring

Data Monitoring

The data and safety monitoring plan includes conducting a twice-yearly review of data safety and enrollment for this trial. The lead statistician will compile and lead analysis of data safety and enrollment information with support from the data manager. As part of the safety plan, the principal investigator will review each participant’s record to ensure that appropriate mechanisms to protect the safety of study participants are being followed, that protocol requirements are adhered to, and that data are accurate, complete, and secure. The principal investigator’s safety monitoring will be supported by study coordinators. If adverse events such as emotional discomfort occur as a result of participating in the study or if a participant expresses significant psychological distress during data collection, appropriate referrals for supportive counseling or psychotherapy will be made with the support of the PN.

Protocol Deviations

Any protocol deviations will be communicated to the principal investigator, who will decide appropriate next steps and file a protocol deviation form with the institutional review board (IRB).

Ethics and Dissemination

Research ethics approval

This trial was approved by the University of Utah IRB (IRB_00164028) using a single IRB model that includes Intermountain Health.

Protocol amendments

Any amendments to the intervention or changes in eligibility will be reported to the University of Utah IRB and funding institution (National Cancer Institute).

Informed Consent

Informed consent will be obtained by research team members trained in responsible conduct of research principles. Patients will read the consent form and sign it electronically in REDCap prior to starting the first survey. All patients will be told that study participation is voluntary and that declining to participate will not affect their medical care in any way. Patients will also be informed about the purpose of the study, potential risks and benefits, the need to access medical/claims records, and compensation for participation (up to $90). There are no biological specimens collected for this, thus eliminating the need for additional consent provisions.

Declaration of Interests

There are no competing interests for any investigators included in this trial.

Access to Data

Only investigators and study staff directly involved in this trial will have access to participant data. Data analysts will only have access to blinded data until the analysis is complete.

Dissemination policy

At the completion of the trial, we expect to have evaluated the efficacy of a virtual patient navigation intervention to improve health insurance literacy, reduce financial toxicity, and improve surveillance for recurrence among YA cancer survivors. Thus, wewill be well positioned for future work to identify implementation strategies to ensure widespread dissemination into YA programs nationally. The study team has considerable experience in implementing and disseminating interventions across large-scale health care-systems and throughout community cancer centers nationally. We will use the information on efficacy and implementation costs of CHAT-S to integrate the intervention into YA cancer programs, hospitals, primary care and oncology care clinics nationally.

In addition to communicating our findings to health systems, our team is dedicated to dissemination of research findings to patients, providers, and policymakers. Plans for dissemination of findings from the proposed project include 1) local cancer organizations, including the HIAYA Cancer Care Program patient newsletter and HCI/IH social media, 2) national advocacy groups, 3) planned conferences and manuscripts; and 4) submission of summary results to clinicaltrials.gov within the first year of all outcome and administrative data collection completion.

Discussion

Health care costs continue to escalate with cost sharing increasingly being shifted to patients, requiring patients to be savvy insurance users to avoid high cancer-related financial toxicity (Harnett, 2019). In this randomized controlled trial that focuses on improving health insurance literacy through patient navigation, our goal is to demonstrate the efficacy of the CHAT-S intervention on improving health insurance literacy, mitigating financial toxicity, and improving health care utilization among YAs with cancer. The health insurance education sessions are led by a PN who is trained on available resources supporting this population. Benefits of this intervention include tailored sessions and accessibility (e.g., sessions occur via videoconferencing and are scheduled by the participant to accommodate their schedule), that may improve the efficacy of the patient navigation.

This intervention is innovative in that it specifically tailors the health insurance literacy intervention concept originating with the HINT trial (Warner et al., 2024) and modifies it to fit a population with significant needs regarding health insurance and costs. YA cancer survivors may experience financial hardship due to lost income due to cancer treatment while also experiencing strain on their emerging careers (Di Giuseppe et al., 2023). These phenomena may also impact the survivor’s partner if they act as a caregiver and have experienced employment disruptions due to this role (Reuvers et al., 2023). The employment impacts coupled with not having adequate time to acquire savings given their relatively new introduction to the workforce compounds the financial hardship that can be experienced by these cancer survivors. By situating this intervention while YAs are in survivorship and have completed treatment, we believe it will impact and improve the efficacy of this intervention on our primary outcomes of health literacy and financial toxicity. Additionally, we anticipate an improvement in YAs engagement with their medical team – both primary care and oncology – past their initial treatment, and throughout their survivorship by mitigating the barrier of avoidance due to cost.

It is possible that recruitment of underrepresented populations may be a challenge we encounter in this intervention trial. To address this, we have incorporated many different avenues of recruitment and retention including contacting eligible patients found through EHR records via multiple modalities, posting study advertisements on social media channels and bulletin boards (physical and virtual) of community organizations that serve young cancer survivors, and informing oncologists at HCI/IH so they can share with potentially eligible patients. Additionally, we anticipate there will be instances where individuals start the intervention sessions but do not remain enrolled to finish all four sessions. To combat this, participants will be told at the end of each session what content will be covered in the subsequent sessions, will receive multiple follow-up attempts to continue completing sessions, and will be told they can work with the PN to schedule at a time that’s best for their schedule. By doing this, we hope that individuals will remain in the study and complete all four sessions to receive information specifically tailored to their situation that they can practically use. Lastly, we anticipate more long-term challenges will be sustainability and incorporating the services provided by our study specific PN into practice settings. To address this, we have incorporated a CEA and a BIA. We are hopeful that these two economic evaluations will provide compelling evidence to health systems that these roles and services provided are critical to not only the financial health of patients, but also the financial health of the health system providing care to these patients.

Trial Status

The CHAT-S Study was registered in the ClinicalTrials.gov system in April 2023 (NCT05829070). Recruitment for this trial began in November 2023. We expect recruitment to be completed by late spring 2026.

Limitations

This study has potential limitations. Recruitment is limited to English-speaking participants from HCI and IH. HCI serves a large, five state catchment areas, whereas IH has hospitals throughout Utah. Thus, while our remote intervention will allow participants throughout these systems to participate, our sample demographics will largely reflect that of HCI and IH. For example, Utah’s Hispanic and Latino population is undergoing huge growth; however, the majority of YAs are either English dominant or bilingual (Lopez et al., 2018). Thus, future iterations of CHAT-S should be translated into other languages, such as Spanish. Another limitation of the study includes not enrolling participants who are uninsured and could also benefit from health insurance education; however, future scale-ups of this intervention will explore ways to include this population. Additionally, while health insurance literacy is likely a critical component of accessing care for this population (Waters et al., 2022), other factors such as provider recommendation and care coordination play a role as well; while we are unable to capture all these factors, our combined survey and medical records data provide a unique way to understand how patterns of care (e.g., use of both primary and specialty care) may further influence access to surveillance care. Finally, in this current trial we are not collecting information on other critical health outcomes such as health-related quality of life; future efforts should evaluate the efficacy of CHAT-S on such outcomes.

Conclusion

The CHAT-S intervention is a remote-based program that is anticipated to provide needed assistance to YA cancer survivors in navigating their health insurance, optimizing their use of available financial resources, which could reduce health care costs and barriers to surveillance care. Patient navigation programs that support financial and insurance needs, directed both to patients in active treatment and off-treatment survivors, are likely needed to address the evolving social and economic complexities that YAs face during both cancer care and survivorship.