Development and Pilot Testing of Evidence-Based Interventions to Improve Adherence after Receiving a Genetic Result

Population genomic screening, defined as using genomic sequencing to identify those at increased risk of a disease or condition but without previous genetic disease risk, is being used in more research settings. Due to this increase in population genomic screening, it is necessary to examine post-result utilization of recommended risk management to demonstrate clinical utility and determine public health benefit. Early results indicate that genomic screening can improve ascertainment of at-risk individuals and provides clinical benefit for individuals who act on the results, but there is still widespread underutilization of current risk management recommendations after receiving a result [1‒5].

Closing this gap in adherence to recommended risk management can improve equitable population benefit related to health outcomes of morbidity and mortality [6‒9]. Surgery in females with hereditary breast and ovarian cancer syndrome (HBOC) is associated with reduction in breast and ovarian cancer risk of at least 90% [10] and 80% [11, 12], respectively, and uptake of oophorectomy among females with HBOC has been associated with a 77% reduction in all-cause mortality [13]. Further, magnetic resonance imaging surveillance leads to earlier detection at more treatable stage for these individuals [14, 15]. More frequent and earlier colonoscopies among individuals with Lynch syndrome (LS) have been associated with a 62% reduction in colorectal cancer risk [6] and a 70% reduction in colorectal cancer mortality [2, 16‒18]. Hysterectomy and oophorectomy in females with LS significantly reduce endometrial and ovarian cancer incidence [19].

Based on the effectiveness of risk-reducing surgery, surveillance, and pharmacologic intervention, professional organizations have established clinical guidelines for risk management [20‒25]. However, a substantial proportion of individuals do not receive recommended risk management after genomic screening [2, 26], with the absence of coordinated and/or longitudinal follow-up care and the clinician’s lack of knowledge about their genetic condition reported as significant barriers that are potential targets for intervention [27].

Adherence relies on multiple factors, including patients’ knowledge and understanding of recommendations, effective communication between the provider and patient, and trust in the therapeutic relationship [28]. A statement by the American College of Medical Genetics and Genomics emphasizes the importance of utilizing evidenced-based strategies to ensure clinicians, patients, and their families understand results and actions that should be taken [29]. This statement emphasizes the use of these strategies by multiple provider types to ensure patients can take appropriate steps [29].

Two evidence-based practices (EBPs) to address adherence issues, problem solving (PS) and motivational interviewing (MI), have been tested in a variety of health care settings, with many different health behavior change goals, and have been provided through different modalities with a variety of clinicians, including genetic counselors (GCs) [30‒41]. These EBPs are particularly useful for targeting individual-level barriers to improve patient self-efficacy and promote adherence to health behaviors. Identifying individuals’ perceived barriers, understanding, thoughts, and feelings about genomic screening results may help them make decisions and take appropriate action based on their risks. PS is useful for addressing “lower level” barriers (e.g., Can I do it?) and can increase self-efficacy, while MI addresses “higher level” barriers (e.g., Is it worth doing?), in particular ambivalence, which can decrease motivation to act.

This manuscript describes the use of multiple methods from implementation like science, behavioral science, and human-centered design to develop and pilot a program to identify individuals who have not followed up on guidelines for care after a genomic result and provide EBPs for improving adherence. For this investigation, we identified barriers experienced by individuals in a population genomic screening program and determined acceptability and feasibility of the proposed EBPs. We then collaboratively developed a program to conduct PS or MI EBPs to motivate adherence to recommended medical follow-up. Finally, we piloted the program and assessed fit, acceptability, and feasibility in clinical practice through post-pilot engagement groups with clinicians who delivered the program.

This study was conducted at Geisinger, an integrated healthcare delivery system in rural Pennsylvania, which provides care for over one million individuals annually and insures about one-third of those individuals through the Geisinger Health Plan. Much of the Geisinger service area is designated as rural, with an average household income 15% lower than the USA’s average. Geisinger’s MyCode^® Community Health Initiative (MyCode) is a precision medicine program that pairs exome sequence data with clinical data for discovery and clinical research. Within MyCode, the Genomic Screening and Counseling (GSC) program is a population-based genomic screening program that began disclosing clinically actionable results to participants in 2015 [42‒44]. As of February 2024, MyCode had consented over 340,000 patients, received almost 245,000 samples, and returned clinical genomic results to over 4,600 participants [45]. The MyCode GSC program confirms pathogenic or likely pathogenic (P/LP) variants, deposits clinically confirmed results into the participants’ electronic health record, notifies participants’ primary care physician, mails a packet with the results to the participant, and offers a complimentary genetic counseling appointment [44].

Multiple data collection methods were used as appropriate for each purpose within the study: surveys, focus groups, and interviews with MyCode patient-participants; a design-thinking workshop with GSC program clinicians to co-develop the program; weekly case discussion and clinical supervision during implementation; and a post-pilot group engagement session with clinicians after piloting the program. Table 1 provides an overview of the purpose, participants, and data collection methods iteratively utilized throughout the study. This study was reviewed and approved by the Geisinger Institutional Review Board (IRB# 00008345).

Participants who had received P/LP results associated with cardiac disease (Cardio) (long QT syndrome, hypertrophic cardiomyopathy, or arrhythmogenic right ventricular cardiomyopathy), LS, or HBOC through the MyCode GSC program were eligible to participate in focus groups planned for spring of 2020. For the in-person focus groups, individuals residing within 25 miles of the facility were invited. One focus group each for the LS and HBOC cohorts was conducted in February 2020; the cardio cohort focus group was cancelled prior to scheduling in March 2020 due to COVID-19. Due to a high no-show rate and COVID-19 in-person meeting constraints, the recruitment methodology was revised to conduct surveys followed by virtual one-on-one interviews. Focus groups and interviews were recorded and transcribed verbatim. Detailed memos and a summary of each group/interview were also created. All surveys were delivered and collected via REDCap [46, 47]. Survey invitations were sent to a random sample of eligible participants from each cohort. At the end of the survey, participants were offered the opportunity to be contacted for an interview. Gift cards were given as incentives for completing the survey (USD 10) and for participating in the interview (USD 25). Two scenario skits were created to demonstrate a PS interaction and an MI interaction for the participants. For the focus groups, the skits were acted out by the facilitators and study staff. For the surveys, the skits were video-recorded and embedded into the REDCap survey, and a checklist of barriers to risk management adherence was created from literature review. The semi-structured focus group interview guide and survey contained items developed by the investigators included demographic questions, thoughts, and feelings about receiving a genetic test result and questions about whether participants experienced specific barriers that have been described in the literature related to acting on the result. After each skit (or skit video), questions were asked about whether the PS or MI interaction would or would not help address the barriers (online suppl. Appendix 1; for all online suppl. material, see https://doi.org/10.1159/000541745), with the list for of barriers presented again for questions after each skit, as well as additional questions such as how helpful the intervention would be for a similar issue with response options ranging from not at all to a great deal.

Design thinking workshops were conducted with clinical stakeholders using the Double Diamond model of the design process [48] for the end goal of co-designing with clinicians a pilot program to facilitate adherence to recommendations that included the EBPs of PS or MI to address identified barriers. The first clinical stakeholder session presented an introduction to MI and PS and shared the findings from the interviews and surveys as well as the barriers checklist used in the patient surveys. Clinical stakeholders were also presented with the current state of adherence to risk management guidelines among individuals with a P/LP identified through MyCode [2]. Participants were then instructed to each sketch ideas for a program workflow to be piloted. A moderator (K.M.R.) then led a discussion of each workflow design, barriers and facilitators to implementation, including other funded studies or MyCode GSC program activities that might overlap or could be leveraged. Clinical stakeholders then voted on the different workflow designs that they thought would be feasible and effective for delivering the intervention. In a second session, the project leads (A.B., A.K.R.) presented a final draft workflow for the pilot program, and the moderator led the discussion for feasibility, risks to success, and logistics (time, resources, roles, and responsibilities) necessary for a successful pilot.

The final workflow consisting of a barriers checklist followed by PS/MI was piloted with MyCode participants identified as non-adherent (not actively engaged in any recommended risk management related to their condition) after receiving a P/LP result. A REDCap database was created for clinician tracking of individuals contacted as well as for recording patient-reported barriers and other observations about each patient after an appointment. A single post-pilot engagement session was conducted through a one-hour online session with the clinicians delivering the program.

The rapid assessment process [49] was used to analyze the patient qualitative data and to expediently identify the themes. This technique utilizes Microsoft Word and/or Excel [50] or other general-purpose word processing and spreadsheet software, to facilitate organizing, reducing, and analyzing data from qualitative research. It is well suited for mixed-method studies that require rapid turnaround, which was necessary in this case to identify topics for consideration during the design thinking workshop. Data analysis occurred using a model of coding the transcript data and memoing, displaying the data, and drawing conclusions [49]. A templated summary table was created to facilitate collection of data and to display the data. The table contained neutral domain names as column headers derived from using the interview guides and episodic summaries. Domains were piloted using one participant interview to complete the templated summaries to assess the accuracy and completeness of the domain template. Once this was optimized, the remainder of the transcripts was summarized using the templated summary document. Two of four main ideas were listed for each domain per participant and transferred into a matrix containing all participant data to facilitate identification of themes. Members of the study team met weekly throughout the data collection and analysis period to discuss the data and emerging themes.

Both sessions of the design thinking workshop were conducted remotely, using Microsoft Teams video conference software and Miro, a digital whiteboard and collaboration tool [51]. A summary of the workshops and suggestions were created by K.M.R. Study PIs (A.K.R. and A.M.B.) reviewed the workflow designs and votes with the summary information.

The pilot program was offered by clinicians as video/telemedicine visits. All PS/MI sessions were recorded and transcribed for training and evaluation purposes. The post-pilot clinical stakeholder engagement group was conducted over Microsoft Teams, recorded, and transcribed. Data were summarized and reviewed by A.B. and A.K.R. and presented to the entire team for feedback. Participant totals for each data collection method and cohort are summarized in Table 2.

A list of barriers was compiled based on the previous literature on barriers to risk management that patients experience [52‒57] and expert opinion on what specific factors may be relevant for additional barriers and on phrasing of barriers.

Demographics of individuals who answered the survey are presented in Table 3. Data were examined for completeness and summarized to describe results. A total of 408 individuals were invited to complete the survey. Forty-six individuals completed the survey. No surveys were started but not submitted. Questions were analyzed individually such that each question was analyzed by the number of respondents who completed that question. When presented with the comprehensive list of potential barriers to following medical recommendations for their condition, most barriers were endorsed by at least one survey participant, despite survey response rate being low (see Table 4). In addition, all barriers were endorsed at least once as a suitable target addressable with PS or MI. Each skit video demonstrating MI and PS techniques presented a clinician assisting a patient with a typical barrier addressable by each method (PS = worry about cost of screenings; MI = worry about communication with family). Over half of respondents (10 HBOC; 7 Lynch; 5 Cardio) reported that a PS conversation with a GC would be a good or great deal helpful. Almost half of respondents (9 HBOC; 3 Lynch; 6 Cardio) also felt that the MI conversation would be a good or great deal helpful for helping patients with similar barriers.

Patient interviews (7 from each group for a total of 21 interviews) about the acceptability for PS/MI conversations with a clinical provider to address adherence to medical follow-up provided useful information for developing a clinical program. Thematic areas discussed by the interview participants included context of receiving their result, adherence intention and barriers, and access to GCs and current resources. Feedback on the PS and MI communication styles were examined individually.

Context of receiving their result: In the cardio and LS cohorts, the presence or absence of a personal or family history of disease or symptoms related to the condition impacted the participants’ level of concern or surprise. The absence of disease history was associated with more concern/surprise than the presence of disease history. Individuals from the HBOC cohort either did not comment on their immediate reaction to receiving results or expressed only lack of surprise when results were returned to them. For example, one participant shared that “I was kind of taken back at first, but I was not surprised. My mother passed away from ovarian cancer, so I already suspected that I had BRCA2 or could potentially have it, so I was not surprised.” [HBOC06].

Adherence intention and barriers: Regardless of initial reactions to the result, adherence intention was high among participants across all cohorts. The pandemic, which occurred during the time of the interviews, was not reported as impacting intention. A typical example of intention was reported as “Obviously, I felt that I would go through with all the recommendations that they had given me and I am. I am following up with both an endoscopy and the colonoscopy” [Lynch07]. However, even when intention was high, some participants expressed caveats such as the participant who shared that “I am happy to do follow-ups. They did stress tests and I have done two of those so far...He did mention that they were going to prescribe a medication for me....and I don’t like to take medications and I think I told him that I am not going to take medication and he said that’s fine, optional, so I decided that was one of the recommendations that I didn’t follow.” [Cardio05]. Participants also reported difficulties following up with adherence intentions due to lack of informed clinicians or conflicting information. For example, “So, I’m okay with that [routine screenings]. But like I said, if you go see a doctor and he doesn’t know anything about it, is not educated, and then tells you, well we didn’t find anything, so you’re going to have to come back in 2 years. But somebody at Genetics who should know something about it says, … they want you to come in every year to do it. Then, I’m saying obviously maybe the physicians need to be educated more on this…because it seems like not everybody is educated well enough.” [Lynch03]. Other reasons for not following through with intention to adhere to guidelines for care included concern about insurance coverage and more pressing healthcare priorities.

Access to GCs and current resources: Other topics important to following risk management guidelines included access to GCs and up-to-date resources. The ability to have ongoing communication with a GC was a prevalent theme, whether for PS or wanting additional information on follow-up care. This is especially important if there is any additional surveillance, prevention, or treatment options that become available [HBOC040]. Lynch02 mentioned having the GC provide patients with an action plan and timeline would be very helpful. Participants had varying opinions on optimal communication methods. Some felt instant messaging, texting, and email were optimal, while others preferred a phone call, video conference, or a face-to-face encounter for discussions. Participants also expressed that increased telemedicine exposure during the pandemic was increasing their comfort levels with telemedicine visits. Finally, many interview participants felt simply having a phone number to call to ask questions or having a GC regularly check in would be helpful to further understand what recommendations are and remind them when or how to follow-up on them.

Feedback on PS: After viewing again and discussing the PS skit, most participants across all cohorts reported the interaction as valuable. A few mentioned the interaction was helpful because it empowered the patient to come to his/her own solution. Cardio06 indicated that the GC in this intervention did well at facilitating the patient to brainstorm solutions. “By brainstorming you tend to find solutions more often than excuses.” [Cardio06]. Some participants had difficulty relating to the PS example scenario because it was not applicable to them (e.g., they did not experience insurance concerns related to healthcare recommendations [Cardio04, Cardio02, HBOC02]). However, most of these participants went on to say that if they were experiencing a problem like the example, PS would be helpful. Areas where participants reported PS would be beneficial included discussing language to use when calling the insurance company about coverage or preexisting conditions or when speaking with a PCP about the result and recommendations, obtaining healthcare or finding a PCP if they do not have one, finding reliable healthcare resources, identifying other clinicians if a PCP is not willing or able to be involved with follow-up care, identifying the best places to follow through with the given recommendations, safely obtaining testing during the pandemic, and locating transportation for follow-up appointments. Only Lynch03 felt they would not speak with a GC to problem-solve unless it was absolutely necessary, saying an internet search would be most comfortable and useful and they would only consider consulting a GC to help resolve conflicting information.

Feedback on MI: Regarding the MI discussion, most participants across cohorts indicated it was very useful. HBOC01 explained this interaction would make them feel like they had been truly heard and their feelings were validated, whereas HBOC07 mentioned this type of interaction takes the feeling of fault off the patient and focuses on how to constructively speak with family. One participant indicated that she thought PS and MI went hand-in-hand, but she really liked the reflective listening that was used in the MI scenario. Lynch04 corroborated the usefulness of MI for helping with personal health decisions and understanding the different options available. However, one participant shared that during the family communication scenario that despite good intentions, MI may not be useful in some situations as “You can only do so much...Some family members just don’t want to hear about it. That’s just like my brother, they don’t want to, there’s nothing you can do...You send a letter, you do so much.” [Lynch03]. HBOC06 felt MI should be carried out by the most appropriate medical provider to be effective, which may be a PCP who has an existing relationship with a family.

Four GCs from the MyCode GSC and two GSC program leaders participated in the design thinking workshop. All participating clinical stakeholders agreed that assessing barriers in non-adherent individuals through PS/MI EBPs was acceptable and feasible. They also determined a program could fit within the current recontact framework of the GSC, and each created an example workflow they thought might be best (for a total of 6 potential clinical workflows). However, due to ongoing grant-funded studies in cardiovascular genetics, it was determined the program could only be piloted with MyCode participants with HBOC or LS results. After reviewing the 6 different workflows and subsequent discussion in the first design thinking workshop, it was determined the best workflow would be for GSC GCs to contact non-adherent individuals, utilize the barriers checklist to facilitate the discussion, and then conduct PS/MI as appropriate.

The program was piloted from September 22, 2022, to November 1, 2022. Three GCs were trained by A.B., a clinical psychologist, on both PS and MI over two training sessions. A convenience sample of 76 individuals in the MyCode patient population previously identified as nonadherent because they had not yet completed at least one guideline-based risk management behavior for either HBOC or LS from an analysis of 255 MyCode participants eligible for risk management reported elsewhere [2] was selected for this pilot program. Prior to starting the program, chart reviews were conducted by a genetic counseling assistant, and 67 of the 76 were identified as still nonadherent to recommendations. These 67 individuals were sent a letter that a MyCode GC would be contacting them and offered them the opportunity to call in and opt-out of this conversation. Individuals were then contacted 2 weeks later by one of three MyCode GCs. GCs attempted to contact each patient three times. In total, 14 MyCode participants (21%, ages 32–78 years) were reached through this pilot program. A.B. conducted weekly supervision sessions with the GCs to answer questions and give feedback on fidelity to PS and MI techniques as well as the overall pilot workflow.

All 3 GCs participated in the post-pilot engagement group on November 28, 2022. Overall, they felt the program was acceptable to them as clinicians and to the patients they contacted, fit well with the clinical workflow, and was feasible for GCs and patients. They felt MI and PS both were worthwhile EBPs for patients with barriers. For improving adherence, they felt recontacting the patients and addressing issues using both PS and MI as determined by their clinical judgment was useful. GCs also reported doing more PS than MI overall. The GCs liked the flexibility to utilize the PS or MI technique based on patient needs and expressed intent to continue to use these techniques in other appropriate patient interactions. One GC noted participation in this pilot made patient-level barriers more salient overall and they were now more likely to address barriers in their patient interactions.

The recontact of patients through the pilot was deemed acceptable by clinicians for reasons beyond the offering of the MI and PS. For example, GCs noted some patients did not remember getting a result; the recontact therefore allowed the GCs to remind them of the result and to provide the information about recommended risk management. GCs also noted that the recontacting was helpful for individuals who did not ascribe importance to discussing the result with their PCPs. Some patients were focused on more pertinent health concerns that took precedence over the risk management recommendations for the genomic screening result. For some of those patients, the recommendations for the genetic result were contraindicated or they decided not to engage in the recommended behavior because of their focus on the more pressing medical need. One patient simply needed assistance scheduling the follow-up care. Others indicated needing more information or had misunderstandings about their result; recontacting these individuals afforded the GC an opportunity to provide necessary information.

Suggested changes to the program after the pilot involved framing of the check-in for the patients, adjusting the workflow, and altering the barriers checklist and/or its use. Framing the check-ins as continued engagement and care facilitation versus addressing nonadherence specifically was postulated by the GCs to be more helpful for patients, especially if it were to be a scheduled part of routine follow-up care. GCs felt touching base and discussing the MyCode result would be useful even for patients receiving most of their care outside of the system. In terms of workflow, the GCs felt that experienced genetic counseling assistants (GCAs) [58] could take on the work of contacting patients for follow-up, give appropriate information (like scheduling phone numbers) when applicable and easily addressed, and transfer the patient to the GC for in-depth follow-up and MI/PS as appropriate. Since patients contacted for the pilot had varying time periods since receiving their MyCode result (4.5 years–6.5 years), the GCs also suggested that a structured check-in program using MI/PS at designated timepoints after result disclosure could facilitate adherence, even for individuals who do not express a specific barrier to adherence using the checklist. They felt teaming up with PCPs and other clinicians could also facilitate follow-up. GCs were cognizant of and concerned about addressing medical advice of other clinicians, such that they felt this teaming up was necessary for patients with a specific medical reason for nonadherence, or where the recommendations relayed by the patient were inconsistent with guidelines. Finally, GCs felt the barriers checklist they used as part of the pilot could either be shortened or simplified (broader barriers vs. specific) or be given electronically in advance to allow the GC to spend more time in conversation addressing issues using MI/PS techniques.

We gathered information and feedback from individuals who have received a P/LP result through a genomic screening program, as well as clinicians experienced in disclosing those results and educating on recommended risk management. Using these data, we co-designed with stakeholders and piloted a program to facilitate adherence to follow-up medical activities based on the EBPs of PS and MI. Through first engaging with individuals who received a result through the MyCode program, we found barriers experienced after genomic screening were consistent with barriers to adherence in other settings [52‒57, 59, 60]. As all barriers were endorsed by at least one participant during the assessment phase, and participants specifically mentioned components of PS and MI strategies that could be helpful to address specific barriers. A draft checklist was presented in the design thinking workshop for consideration as part of the PS/MI pilot program and including this barriers checklist was endorsed and included in the final pilot program workflow. These barriers are endorsed across the literature broadly in many different patient groups [52‒57, 59]; however, our numbers were too small to look for any patterns between genetic conditions. Through our program, we did not see a reason to believe these barriers were condition-specific as many barriers were experienced by multiple participants across conditions and leading us to believe MI/PS interventions should be applicable in addressing barriers regardless of genetic result.

Within the MyCode GSC program, GCs and GCAs disclose the population genomic screening results to participants, educate on recommended risk management, and conduct follow-up outreach with participants [42]. PS and MI EBPs are well suited for use in genetic counseling and have been previously utilized [30‒41]. GCs are well positioned to implement this type of intervention effectively in the future as it fits well within their scope of practice and enhances feasibility [34]. This type of communication can fit within current return of results workflows or by scheduled by follow-up. It also provides the opportunity for GCs to continue to practice at the top of their scope, focusing their role in posttest population management, while pre-test education shifts to other clinicians. A key barrier noted in patients with LS is insufficient provider knowledge [27]. As experts in genetic conditions, GCs are accustomed to staying current with rapidly changing information related to actionable genetic results. The MyCode GCs also reported they received some training in these techniques through their graduate education and continuing education at conferences and reported finding these techniques useful for genetic counseling. Therefore, through the design thinking workshop, it was determined appropriate and feasible that the GSC GCs would pilot the PS/MI program and use of the checklist to guide a conversation on recommended medical follow-up care.

Piloting the PS/MI program determined such a program is feasible to deliver, acceptable to patients who were reached by the pilot program, and appeared to be valuable to the patient from the viewpoint of the GCs who contacted them. We confirmed patients are interested in long-term follow-up contact with GCs, are open to and even requesting recontact, and that having this information and reminders were important to them. The pilot also identified workflow changes that could improve program reach and efficiency, such as setting a specific timetable for adherence conversations, utilizing GCAs for outreach and review of the barriers checklist to immediately address logistical or resource-related barriers (PS strategies), and coordinating with other clinicians to assist in creating unified communication about personalized recommendations to each patient – particularly those with other pressing medical conditions.

It is important to contextualize the findings of this pilot by addressing some limitations as well as what is generalizable. The low survey response rate could limit the generalizability of our findings; the survey was sent out at the height of the pandemic. As is typical with adherence research, there is also ascertainment bias that those who participate are likely to have higher adherence and fewer barriers. While we did seek out participants for the pilot who were already identified as nonadherent, this bias could be present in the results from the surveys and interviews conducted prior to the pilot program. In addition, both survey and pilot participants were also older in age, many who likely had disease progression or comorbid health concerns, making it more difficult to participate. While our population may not be diverse in race, ethnicity, or education, Geisinger’s coverage areas represent a medically underserved population with an average household income 15% lower than the USA’s average and designated as rural. Recruitment of family members to interviews and surveys was extremely low, and the design-thinking workshop determined that including cascade testing discussions and contacting family members as part of the initial pilot was too burdensome. Due to a shortage of GCs, some healthcare systems may not have the resources for GCs to implement this intervention. However, the utilization of GCAs and other trainees and clinicians is likely feasible as other studies have shown PS and MI are EBPs known to improve adherence behavior in a variety of settings [31, 33, 36, 37, 40, 41, 61, 62]; this pilot was designed to identify the feasibility and acceptability of PS/MI for this purpose and focused on co-developing and testing a clinical workflow to be implemented. Evaluating patient behavior change in regard to adherence to follow-up was outside the scope of this pilot. Further studies using hybrid implementation-effectiveness designs [63] are needed to determine how and for whom PS and MI work to improve adherence to guideline recommended care after genomic screening results and how to optimize healthcare resources and implementation strategies for maximal patient and population impact. This pilot was only conducted on a sample of individuals in MyCode [2] who were confirmed as non-adherent to care through records review. This may not be a feasible or sustainable component of an adherence program as population genomic screening becomes more prevalent or as focus broadens from diagnostic activities. Even in this small pilot, we identified patients with competing health conditions as well as system-level barriers to adherence. Therefore, it is important to acknowledge the many barriers patients face within the process of following medical recommendations after receiving genomic results, as well as competing health and life priorities, patient choice, and quality of life. It is therefore likely that multiple EBPs will be needed for a substantial population impact on adherence to occur.

Finally, MI/PS interventions are likely to be feasible and acceptable for multiple contexts within genetic counseling (e.g., facilitating test uptake, facilitating cascade testing). The GCs who co-developed and delivered this pilot reported intention to utilize the PS and MI approaches throughout other clinical interactions with patients, not just related to adherence conversations.

In conclusion, we were able to demonstrate that a program to facilitate adherence to medical recommendations after population genomic screening based on the EBPs of PS and MI were seen by patients as helpful interventions to barriers and can fit within GC scope. Such follow-up and continued contact with patients appear important for addressing barriers and may aid in improving adherence to guideline-based care and positively impact patient outcomes. These results will help inform further trials to study the use of these interventions to address gaps in adherence in a population of patients who have received actionable genomic results through population screening programs.

We would like to acknowledge the GCs that participated in the design thinking workshop and the three GCs that piloted the intervention – Miranda Hallquist, Cassidi Kalejta, Nicole Deckard, Cassandra Pisieczko, Rachel Schwiter, and Alyson Evans.

This study protocol was reviewed and approved by Geisinger Institutional Review Board (IRB# 00008345). Opt-out informed consent protocol was used for a portion of these research purposes. This consent procedure was reviewed and approved by the Geisinger Institutional Review Board (IRB# 00008345) on 24 March 2020.

Written informed consent was obtained from participants (or their parent/legal guardian/next of kin) to participate in the study during the initial MyCode study, and letters were sent for opt-out of recontact. Participants gave verbal consent for interviews, focus groups, and the pilot intervention and consented to participate in the survey.

Mr. Buchanan has equity in MeTree and You, Inc. He has received grant funding from the NIH, Exact Sciences, and Freenome Holdings. All the other authors have no conflict of interests to disclose.

The authors received funding from the Bucknell-Geisinger Research Fund in the amount of USD 20,000. The MyCode GSC program is funded by internal Geisinger funds and by the Horace W Goldsmith Foundation.

Anna Baker and Alanna Kulchak Rahm co-led study conceptualization, design, data collection, and analyses and did initial writing as well as editing of the majority of the manuscript. Dr. Baker trained and supervised GCs on the intervention and led the qualitative feedback session. Jessica Goehringer assisted in IRB submission, led qualitative and quantitative data collection and analyses for the participant portion of the study, and assisted in editing the rest of the manuscript. Makenzie Woltz aided in data collection and analyses for the participant portion of the study and reviewed the manuscript. Katrina Romagnoli led the design thinking workshop portion of the study and wrote the methodology and results, along with editing the remainder of the manuscript. Gemme Campbell-Salome aided in the design of the study as well as data collection and assisted in editing the manuscript. Amy Sturm, Adam H., and Marc S. Williams Buchanan consulted on the study design and edited the manuscript.

Deidentified survey data are available upon request. Data supporting the study findings are not publicly available due to risk of identification of participants. Reasonable requests for de-identified survey data may be made by contacting Adam Buchanan ([email protected]).