Something to chat about: An analysis of genetic counseling via asynchronous messaging following direct‐to‐consumer genetic testing

Advances in technology, decreasing cost of genetic testing, and growing public interest in genetics marked by an increased uptake of genetic testing, particularly direct‐to‐consumer genetic testing (DTC‐GT), have led to an overwhelming demand for genetic counseling services. As such, various alternative service delivery models have been proposed to increase access to genetic counseling. Some service delivery models, such as asynchronous messaging, remain unexplored in the genetic counseling literature. The purpose of this study was to assess communication during genetic counseling for DTC‐GT through asynchronous messaging. A thematic analysis was conducted on 34 de‐identified chat transcripts between genetic counselors and clients who underwent DTC‐GT. Six categories of communication were identified and were grouped based on communication sources from either the client or the genetic counselor. Categories observed in client communication were motivations for seeking DTC testing and/or genetic counseling services, questions posed to the genetic counselor, responses provided during the session, and psychosocial aspects of the session related to the clients' mental, emotional, social, and spiritual needs. Categories of communication that emerged from the genetic counselors' communications were educational aspects of the session and counseling strategies to address concerns that are not related to educational or informational needs. Most clients had specific questions about variants detected or specific conditions. Many clients asked about appropriate subsequent steps related to additional testing or medical management. Genetic counselors discussed the limitations of DTC‐GT and recommendations for clinical grade testing in almost all chat transcripts. In several chats, the genetic counselor provided advice to the client related to minimizing time sorting through likely benign results and refraining from altering medical management. Results suggest that genetic counselors are able to provide genetic information to clients and respond to their mental and emotional needs through asynchronous chat following DTC‐GT. Findings from this study provide initial insight into a unique genetic counseling delivery model and reveal the informational and counseling needs of clients following DTC‐GT.


| INTRODUC TI ON
Advances in technology, decreased genetic testing costs, and increased public interest have driven the creation of genetic services including genetic testing and genetic counseling that are marketed directly to the public outside the traditional structure of a referring provider or medical system. Direct-to-consumer genetic testing (DTC-GT) gives individuals access to personalized genomic information without requiring the involvement of a healthcare provider.
Various motivations for pursuing DTC-GT have been reported, such as a desire for information about one's ancestry, as well as an interest in genetic traits and disease risk .
Some individuals pursuing DTC-GT elect to use third-party interpretation tools to review their raw genomic data. One study reported that around 40% of participants wished that their DTC genetic test report provided more information , potentially driving the use of third-party interpretation services. A majority of individuals stated that having access to their raw data was an important aspect in deciding to undergo DTC-GT and that they were interested in using a third-party interpretation tool to explore their ancestral information, understand personal and family health implications, or satisfy their curiosity about new technology (Nelson et al., 2019;Wang et al., 2018). Most consumers of DTC-GT that have entered raw data into third-party interpretation tools have reported feeling satisfied with their results; however, over 40% described feeling confused or upset by the interpretation (Nelson et al., 2019). One reason why some individuals felt dissatisfied with their results interpretation was because they did not understand the information in the way it was presented to them (Nelson et al., 2019).
Information gained from reports using third-party interpretation tools is often shared widely with family and friends as well as healthcare providers . The importance of clinical confirmation of results from DTC-GT or third-party interpretation tools has been demonstrated, as one study revealed that 40% of positive findings using raw data from DTC genetic test results were false positives (Tandy-Connor et al., 2018). It is therefore recommended that those seeking to use their DTC-GT results to guide clinical management undergo clinical testing for confirmation of their result (ACMG Board of Directors, 2016; National Society of Genetic Counselors, 2019). Some individuals reported that when discussing DTC-generated reports with a healthcare provider, they perceived the provider was disinterested or discounted their results . Genetic counselors have been able to provide individuals with information about how follow-up clinical testing could impact medical management and/or provide individuals with reassurance .  (Koeller et al., 2017). Motivation to seek genetic counseling after DTC-GT may be driven by feeling concerned or confused by results and interest in learning about genetics-related risks for their children or other family members, understanding how to plan for their future, understanding information about drug response, and improving their health (Brett et al., 2012;Koeller et al., 2017;Marzulla et al., 2021). Individuals who pursued genetic counseling after receiving DTC-GT results did so to address concerns about how their results could affect insurance, uncertainty about what results meant for personal risk, questions about appropriate prevention strategies, and difficulties discussing results with others, such as family members or healthcare providers (Koeller et al., 2017;Marzulla et al., 2021). Individuals who undergo DTC-GT may have lower confidence in their genetics knowledge following testing (Carere et al., 2016), prompting them to seek follow-up from a content expert. While many individuals were interested in clarifying risk, fewer individuals were motivated to seek genetic counseling to have additional genetic testing or to confirm the validity of DTC-GT results (Brett et al., 2012).
Although genetic counseling services can benefit individuals who have undergone DTC-GT or used third-party genomic interpretation K E Y W O R D S communication, direct-to-consumer genetic testing, service delivery models

What is known about this topic
Growing public interest in how genetics affects our health has contributed to increased uptake of direct-to-consumer genetic testing. Various service delivery models have been introduced to improve access to genetic services.

What this paper adds to this topic
Asynchronous messaging is a service delivery model that allows clients to provide information about their motivations for DTC-GT and ask questions about genetic testing via a HIPAA-compliant online platform. This novel delivery method also allows genetic counselors to provide education and discussion of the limitations of DTC-GT. tools, genetic counselors have reported several challenges when counseling these clients in the clinic. Preparing for sessions with third-party interpretation reports from DTC-generated raw data can be more time-consuming due to the length and degree of detail commonly found in reports. Some genetic counselors do not feel adequately prepared for conducting counseling sessions regarding raw data or extensive information from third-party interpretation tools, and some clinical centers have a policy against taking referrals related to third-party interpretation (Allen et al., 2018). Furthermore, the demand for genetic counseling services, in general, continues to grow and a shortage of genetic counselors in direct patient care is projected over the next several years (Hoskovec et al., 2018).
This combination of factors has resulted in few genetic counselors who are available and willing to counsel those that have undergone DTC-GT.
The challenges of counseling clients who have undergone DTC-GT in combination with the shortage of genetic counselors working in direct patient care settings highlight the need for innovative methods to increase the efficiency and productivity of those providing genetic counseling services (Hoskovec et al., 2018).
Alternative service delivery methods may increase the efficiency of genetic counseling and meet the needs of more clients, including those undergoing DTC-GT. According to the NSGC Service Delivery Model Task Force, classes of service delivery include in-person, telephone, group, and telegenetics. Telegenetics is defined as utilizing audio-visual technology and is therefore distinct from telephone alone (Cohen et al., 2012). Strengths, limitations, and barriers for many service delivery models have been well described (Cohen et al., 2016;Otten et al., 2016). The benefits of alternative service delivery models, such as telegenetics, include reduced travel times for both genetic counselors and clients, cost-effectiveness, reduced wait times, and enhanced financial and geographic access to genetics services (Cohen et al., 2016;Otten et al., 2016). Some genetic counselors have reported a desire to implement additional service delivery methods and that the benefits of offering telegenetics outweigh limitations, such as additional considerations that may be required for arranging testing or equipment setup (Cohen et al., 2016;Greenberg et al., 2020).
Additional delivery models have emerged since the publication of the findings of the NSGC Service Delivery Model Task Force. A growing body of literature describes the use of chatbots to increase the uptake of genetics services and participation in research (Ireland et al., 2021;Nazareth et al., 2021;Schmidlen et al., 2022;Siglen et al., 2022), and their use has been supported by patients (Schmidlen et al., 2019). The utilization of electronic health records and other web-based platforms enabling automated processes of sharing genetic information and test results have also been described (Arjunan et al., 2020;Kalejta et al., 2019). However, as automated delivery models, they may not be personalized to the informational and psychosocial needs of the client. Some genetic counseling telehealth companies are beginning to use alternative web-based platforms like live chat or asynchronous messaging to increase access to genetic services. Live chat and asynchronous messaging services have not been studied extensively in the genetic counseling context but have been used in other fields, such as mental health counseling (Stubbings et al., 2015). Although live chat sessions require the client and genetic counselor to remain active simultaneously, asynchronous messaging is a communication platform in which both parties do not need to be concurrently active in the conversation. This method allows more flexibility and convenience in consumer and clinician communication in a nonurgent manner via a secure platform.
Although several studies have described individuals' motivations for seeking genetic counseling after DTC-GT or using third-party interpretation tools (Koeller et al., 2017;Nelson et al., 2019;Roberts et al., 2017;Wang et al., 2018), there is little research focused on the actual topics and concerns addressed during a genetic counseling session with a client who has undergone DTC-GT. Furthermore, since genetic counseling services using messaging (live chat or asynchronous) have not been used extensively in the past, there is little-to-no information about counseling for DTC-GT through those platforms. The purpose of this study is to reveal themes surrounding the needs of clients who have undergone DTC-GC and communication patterns between those clients and genetic counselors, which may aid genetic counselors in communicating with clients that have undergone DTC-GT.

| ME THODS
Thematic analysis of online conversations between genetic counselors and clients that underwent DTC-GT was conducted to identify common questions and counseling issues. The IRB at the University of North Carolina at Greensboro determined that this project did not constitute human subjects research as defined under federal regulations. No informed consent was required from subjects as data were anonymously extracted from DNAvisit. All procedures followed were in accordance with US Federal Policy for the Protection of Human Subjects.

| Access to chat transcripts
DNAvisit is a telehealth company that provides genetic consultations via telephone, video, and an asynchronous messaging/ chat platform. Indications for counseling vary and include risk assessments for Mendelian disease in specialties such as cancer and cardiovascular genetics, as well as referrals pertaining to direct-toconsumer testing, raw data interpretation, and nonmonogenic diseases, such as polygenic risk assessment and pharmacogenomics. Of the over 20 genetic counselors, medical geneticists, and pharmacists providing consultations through DNAvisit across these indications, five licensed and certified genetic counselors and one medical geneticist manage referrals from the chat service. The chat service is more heavily utilized by individuals and families seeking the expertise of a genetic counselor to provide information regarding the use and interpretation of raw genomic data acquired from DTC-GT.
Most individuals are self-referred to DNAvisit and enroll in genetic counseling services online. Some of the self-referrals received by DNAvisit are related to specific partnerships with other companies, whereas others have found DNAvisit on their own. When an individual has chosen the chat interface, they are assigned to a genetic counselor to begin the process. In most instances, clients have uploaded raw genetic data acquired from DTC-GT to generate health reports through a third-party service, though some clients have questions about the use and interpretation of their raw data and specific areas of interest/concern or both. If applicable, clients may upload third-party reports or screenshots of their data to the chat interface for use in the session. DNAvisit provides clients with the option of either 30-or 60-day access to the chat interface to message a genetic counselor, as well as the option to schedule a telephone call at any time. Clients may also extend their chat sessions beyond 30 and 60 days. Within the chat interface, clients are guaranteed a response from a genetic counselor within two business days. Authors H.W. and A.K. are genetic counselors who are part of the chat service team at DNAvisit.
Although each client interaction is unique, there are common topics that the genetic counselor reviews during the chat sessions with the majority of clients. Each chat session begins with an introductory message in which the genetic counselor reviews the chat process, sets expectations, and invites the client to ask their questions via the online chat portal. Once the client begins the conversation in the chat portal, the genetic counselor typically requires access to the client's reports of interest if the client has more specific questions about a test they have completed through another service. Prior to the initiation of this research study, anecdotal recollection suggested that chat conversation typically consists of the genetic counselor answering specific questions posed by the client and providing education about the differences between DTC-GT and other types of genetic tests. Explaining the difference between consumer-initiated genetic testing and physician-ordered genetic testing typical of all chats. If a genetic test or report identified a variant of concern, the chat conversation may mimic other types of genetic counseling sessions in which a genetic disease may be reviewed along with considerations for testing of other family members. However, in most instances, the focus of the chat session concentrated on educating the client about the tests performed, including an in-depth review of the limitations of such testing. The review and assessment of the chats for this research would indicate if these anecdotal observations were typical across a majority of chat sessions.
One hundred and thirty-two chat conversations that occurred between May 2019 and December 2020 were related to DTC-GT and able to be included in the study. An initial sample of 34 chat transcripts pertaining to DTC-GT was randomly selected by DNAvisit and de-identified for review by the research team, with the option to select and identify additional chats if needed. The number of exchanges and the duration of each chat were calculated. The number of exchanges was defined as the number of times the client or genetic counselor communicated through the chat interface. Multiple messages sent before the other party responded were counted as one exchange. Duration is defined as the amount of time in days in which messages were exchanged between the genetic counselor and the client. Messages that were both directly related and not related to DTC-GT were included when determining the number of exchanges between clients and genetic counselors in order to indicate the total length of the chat.

| Instrumentation
The team developed an initial codebook based on a review of the

| Procedures and data analysis
The initial set of 34 chat transcripts had the potential to have enough informational power (Malterud et al., 2016) to reveal themes related to the needs of clients who have undergone DTC-GC and communication patterns between those clients and genetic counselors given the density of chats, the specificity of the sample, and the application of previously known themes related to genetic counseling for DTC-GT. Recognizing that reaching thematic saturation may not be feasible given the varied indications, personal goals, questions, and experiences of clients (Wainstein et al., 2022), transcripts were reviewed until the research team identified the general essence of the client and genetic counselor experience with asynchronous chat.
Using a step-wise process, chats were reviewed and analyzed approximately 10 chats at a time until no new themes emerged that were generalizable to a majority of transcripts. Chat transcripts were analyzed and coded independently by C.K., a genetic counseling student at the time, and R.M., a certified genetic counselor who is not affiliated with DNAvisit. R.M. has previous experience counseling patients with DTC-GT results via a telehealth platform as well as providing education about DTC testing and counseling for DTC testing to genetic counseling students, practicing genetic counselors, and other practicing clinicians. The previously described codebook was used to analyze data; new codes were added to the codebook using an iterative process; and data were re-analyzed for consistency following the addition of new codes. Discrepancies in coding between investigators were discussed until a consensus was reached. Qualitative coding was managed and analyzed using ATLAS.ti. Review and analysis of 34 transcripts revealed a common pattern of communication, reaching theoretical saturation related to the general experience of clients and genetic counselors with asynchronous chat.

| RE SULTS
Thirty-four de-identified chat transcripts between genetic counselors and clients who previously completed DTC-GT were analyzed.
The number of exchanges in the analyzed transcripts ranged from 2 to 55. Most chat transcripts had fewer than 10 exchanges, with the median being eight exchanges (Figure 1). The range of time elapsed between the start and end of the chat ranged from 1 to 288 days.
The median time elapsed for the analyzed chats was 6 days. All but one chat lasted less than 60 days. The chat that spanned 288 days had a period of inactivity lasting 208 days before the service was renewed with an extension.
All chats were initiated by the client. In approximately 91% of chats (31 out of 34 chats), the client initiated the conversation by asking about test results. Three clients (9%) initiated the conversation by asking about the platform or the process of counseling through asynchronous chat. For 32% of chats (11 out of 34 chats), an automated response was sent by the support team which detailed the process of using asynchronous chat and included information about timelines responses from a genetic counselor. Approximately 24% of chats (eight out of 34) ended with a closing from the client like "goodbye" or "thank you," whereas most chats (26 out of 34 chats) ended without a closing farewell.
Components of the transcripts were categorized into four types of client communication and two types of counselor communication.
"Motivations for genetic counseling and/or DTC-GT" were communications from the client that suggested why they pursued DTC-GT and/or why they sought genetic counseling to understand their results. Questions asked by the client were categorized as "Questions posed to genetic counselor." Comments and responses from the client were categorized as "Responses from clients during the session." Psychosocial needs of the clients were categorized as "Psychosocial aspects of the session." The techniques used by counselors, including those that responded to the psychosocial needs of the clients, were included in "Counseling strategies." "Educational aspects of the session" included information about results, testing, and/or conditions that the genetic counselor provided to the client. Subthemes and example quotations can be found in Table S1.

| Motivations for genetic counseling and/or DTC testing
Twenty-five clients (74%) were interested in understanding the risk of developing a specific condition or group of conditions. Examples of specific conditions or groups of conditions included mental health conditions, bleeding disorders, myotonic dystrophy type II, and Ehlers Danlos syndrome. Clients also sought more information about specific variants or results identified themselves or thirdparty interpretation tools. Most clients wished to place information found within a report in the context of symptoms or family history.
For example, Client #11 said, "I have some questions about my DNA variants. It appears I have several pathogenic markers for Lynch syndrome and I do have significant family history of cancer." Often, they were looking to use their reports to provide information to their primary care provider (PCP) to help explain symptoms. Some clients indicated that they were using testing as "baseline" information, and others indicated that they were using testing as a last resort because their PCP had no more evaluations to offer them.
While most clients had specific questions about their report and/ or raw data, several clients had nonspecific questions or were seeking general risk information. For example, Client #5 asked "What are the biggest risks you are concerned about with my DNA?" Clients also sought information about raw data analysis. This included clients who wanted information about how to search through or "mine" raw data more effectively. One client expressed that they were seeking to make reproductive decisions. This client was using information about carrier status to determine whether they should use a sperm donor.
Two clients indicated issues with accessibility to clinical genetics services as the motivator for seeking DTC-GT over clinical testing. Inaccessibility was related to clinic wait time and cost of clinical testing. Four clients asked directly about familial risks or testing. Clients who requested resources appeared to be people who had a medical background and were looking for primary literature. In two cases, individuals asked about protections related to insurance.
In five chats, clients asked about reports or data other than their own. In some cases, the other individual's data were the focus of the session. In one chat, a client used their family member's data as an example for a specific question. In 10 cases, the categorical types of questions asked changed throughout the chat as the genetic counselor provided more information, such as a more detailed inquiry about the limitations of DTC-GT.

| Responses from clients during the session
In 13 cases (38%), clients shared personal or family health information after being prompted by the genetic counselor to share this information. Personal and family history information ranged from brief provision of information to more detailed descriptions of personal health information and personal diagnostic odyssey. Four clients expressed concerns about the cost of clinical testing. Some individuals mentioned a lack of access to clinical testing at the time of the discussion due to financial barriers while others inquired about appropriate labs to use that might accept insurance. Several clients expressed that they understood the limitations of DTC-GT with some indicating that they knew the limitations prior to undergoing testing. Others stated that the limitations of DTC-GT were new information. Client #21 stated that "Wow thanks, didn't realize [DTC-GT company, name redacted] could be inaccurate at times."

| Psychosocial aspects of the session
Clients in several sessions provided responses related to their mental, emotional, social, and spiritual experiences, well-being, or needs.
Gratitude was expressed most frequently, which occurred in 19 out of 34 sessions. When gratitude was expressed, it was most often in relation to the information received during the chat session rather than gratitude toward the data they received through DTC-GT. This was most frequently communicated when the client understood something that they did not prior to the interaction.
Frustration was expressed by nine clients (26%). When clients expressed frustration, it was most often over their results or the limitations of DTC-GT. For example, client #2 expressed frustration that their data could not detect triplet repeat expansion conditions by saying, "I had no idea, figured DNA was dna and that my 30x sequencing could get me where I needed to get." Some clients expressed frustration about the lack of availability of informational resources, such as tools to help them mine their data. Others stated frustration with technical aspects of companies offering DTC-GT or generating reports using previously obtained raw data. One client was frustrated with the interpretation by the genetic counselor who suggested that carriers of autosomal recessive conditions are typically asymptomatic, which contradicted information the client learned elsewhere.
A few clients communicated worry or fear about results while others expressed hope. Clients who expressed fear or worry related to a result did so at the beginning of the session and retrospectively, after information was provided by a genetic counselor. In cases where fear or worry was retrospectively communicated, humor was often used. For example, Client #30 said, "I have been an emotional wreck for days LOL," after the genetic counselors responded to their immediate concerns. Four clients expressed hope over the present analysis or over potential future analyses. A sense of urgency was conveyed in one instance, which related to the use of DTC-GT information for reproductive decision-making.
In nine chats (26%), previous experiences with providers were mentioned. Most of these sentiments had a neutral tone. Clients explained which clinical tests their provider previously ordered and provided insight into the individual's personal health journey, as several clients described how their provider already utilized all other potential evaluations and tests.
Furthermore, several clients suggested that while they knew some information prior to the chat interaction, they desired to have more, particularly in the form of personal knowledge or knowledge of a family member. Client #3 expressed some of their genetics knowledge by stating, "I'm going to be upfront and admit that I know enough about genetics to be dangerous, but not enough to like, ACTUALLY know what I'm doing."

| Counseling strategies
Attempts at establishing rapport through introductions and responses to client statements occurred in every chat. For example, the genetic counselor stated, "Good to hear you have normal lab results!" in Chat #13 after the client shared information about their medical history. In several cases, the genetic counselor explained their training or qualifications. This typically occurred after the client requested information about to whom they were speaking prior to asking questions or explaining personal/family history information.
Contracting took place in all but one session. For example, the genetic counselor asked about the client's main concerns at the beginning of the chat. Often, there were multiple instances of contracting throughout the session. In the session where contracting did not occur, the patient sent a long list of questions and the counselor responded by answering these questions directly. This chat was limited to three exchanges.
In 26 cases (76%), genetic counselors attempted to manage expectations, including information related to the services offered by DNAvisit. Most often, it was explained that the genetic counselor would not be able to analyze every piece of data available in a report.
Management of expectations also related to testing and specific conditions. Genetic counselors used open-and closed-ended questions to understand patient concerns and to elicit personal/family history information. Normalization of both concerns and results occurred in several sessions as well. For example, when discussing a client's variants of uncertain significance (VUSs), the genetic counselor in chat #28 explained, "when we test over 100 genes, it is typical to have several VUSs identified." One case involved the facilitation of decision-making. In another case, DTC-GT was framed as one step in a journey for someone seeking more actionable information. These discussions included the availability of clinical genetic testing for personal or family history as well as recommending clinical genetic testing or confirmatory testing based on results. In one instance, the process of undergoing clinical genetic testing was described in response to questions from the client. In 28 cases, genetic counselors interpreted results to explain the clinical significance of results. For example, a genetic counselor explained to Client #4, "As for the colon cancer risk, it is likely due to you having a few of these low-risk variants. Independently they aren't significant, but the more people have, the higher their GENETIC risk gets." Genetic counselors also directed the client to a clinic or provider, both genetics and nongenetics, who could address their needs.

| Educational aspects of the session
Basic genetic concepts were reviewed in chats with 18 clients (53%). Some examples of concepts discussed include inheritance, somatic versus germline testing, mitochondrial DNA, intronic DNA, and differences in testing strategies. Genetic counselors also discussed how variants were identified and/or classified. Management for common conditions, such as cardiovascular disease and thrombosis, as well as less common conditions, such as Mendelian conditions or inherited cancer predisposition syndromes, were discussed.
In four cases, genetic counselors presented unprompted risk information that was identified by the DTC-GT but not requested by the client. These included risks of familial hypercholesterolemia, GJB2-related nonsyndromic hearing loss, and hereditary breast and ovarian cancer. In a case involving a variant of uncertain significance in BRCA2, a client asked follow-up questions about the specific variant and requested primary literature. In other cases, clients did not ask additional questions about the finding acknowledged by the genetic counselor and returned to their initial inquiry.
In 10 cases (29%), the genetic counselor provided advice and recommendations related to DTC-GT. Advice included recommendations to seek evaluation by a provider, obtain clinical testing, and refrain from making any significant changes to lifestyle or medications until speaking with a provider. In Chat #9, the genetic counselor explained, "I wouldn't condone making any drastic changes [to diet and exercise based on results alone]." Genetic counselors also advised on how to conduct additional research on the clients' DTC-GT results.

| DISCUSS ION
Genetic counselors have reported several challenges with counseling clients who have undergone DTC-GT and raw data interpretation (Allen et al., 2018). Furthermore, the shortage of genetic counselors working in direct patient care settings presents a need for alternative service delivery models (Hoskovec et al., 2018), such as asynchronous messaging/chat. The results of this study provide insight into the interactions between clients who have undergone DTC-GT and/or raw data interpretation and genetic counselors on an asynchronous messaging/chat service.
The motivations for genetic counseling uncovered in this study are consistent with other studies involving DTC-GT (Brett et al., 2012;Marzulla et al., 2021) and include feeling confused about results as well as seeking information about results in the context of personal and family histories. However, compared with previous studies, the present study indicated that clients were less interested in risk information for family members or reproductive risk information (Brett et al., 2012). Instead, clients desired information regarding "next steps" including information about additional testing and sharing results with other healthcare providers. Clients also asked genetic counselors to explain or clarify test reports and information they had discovered through their own information-seeking prior to genetic counseling.
Genetic counselors have previously reported that they perceived DTC-GT as anxiety-provoking for clients (Brett et al., 2012).
Conversely, in the present study, a minority of individuals expressed fear or anxiety over results. There are a multitude of possible explanations for this finding: this finding may be due to selection bias, where individuals who choose to utilize an asynchronous chat platform are collectively less anxious, use of the platform may have also promoted lower rates of anxiety, or a greater portion of the clients had feelings of anxiety than those who expressed them. The observed difference between our data and that previously reported may also suggest that genetic counselors interpret anxiety and fear in clients more frequently than clients would self-identify as anxious or worried. Due to the exploratory nature and recreational use of many DTC-GT services as well as findings that are often nonurgent, it is possible that clients' information-seeking may be rooted in curiosity rather than anxiety.
In a study of DTC-GT customers, around 88% felt satisfied with the information gained from the third-party interpretation tools (Nelson et al., 2019). While clients in the present study suggested gratitude toward the counselor that they were speaking with, no clients directly expressed gratitude toward the test itself. Conversely, many expressed frustrations with the limitations of testing and raw data analysis.
Differences between the findings of our research and those of previous studies may be explained by distinction in service delivery method. With a chat-based service delivery method used in the present study, clients were self-referred rather than physician-referred.
Furthermore, the data set of chat transcripts used in our study may not accurately reflect clients' perspectives or feelings which may be otherwise captured in a survey or interview.

| Asynchronous messaging compared with other service delivery models
Although our research did not directly compare asynchronous messaging to other delivery models, some differences were noted as we consider genetic counseling literature and practice. Analysis of transcripts revealed genetic counselors offering advice or directive guidance via the chat. Historically, advice-giving or directiveness in genetic counseling has been discouraged to support client autonomy; as such, a principle of nondirectiveness has been encouraged.
Directiveness has been defined as a genetic counselor attempting to persuade a course of action of a patient (Kessler, 1992). Advicegiving has been critiqued as a failure to remember that clients are capable of sorting through information and reaching their own conclusions (Kessler, 1997). It has also been suggested that if advice is provided, it is best done in an indirect way, by asking patients to relate an approach to their own situation (Butler et al., 2010). However, more recently, many have questioned the counseling principle of nondirectiveness, highlighting its insufficiency as a model of ethics for genetic counseling (Austin, 2021;Rehmann-Sutter, 2009;Wolff & Jung, 1995). Further, practicing genetic counselors have expressed a desire for flexibility in approach depending on the needs and values of the clients as well as clinical circumstances (Weil et al., 2006).
Providing advice is increasingly accepted in some cases, especially when advice is based on standards of care, the genetic counseling process, and patient behavior (Veach et al., 2018). The present study suggests that genetic counselors using asynchronous messaging provided advice to the patient, including advice related to refraining from changing medical management and spending time sorting through likely benign results. This directive guidance likely occurred for several different reasons, including the nature of the chat compared with a telephone, video, or in-person visit. With chat, it is possible for the client to leave and no longer check or respond to messages at any time. Therefore, it may be more critical to communicate messages directly in a chat-based session. However, because our study did not specifically compare chat to other service delivery methods, it cannot be ruled out that the advice provided in the present study was in part due to a trend away from nondirectiveness in the genetic counseling field.
The asynchronous chat for genetic counseling also appeared different from other genetic counseling delivery models as it relates to building rapport and facilitating decision-making. Building rapport during a genetic counseling session can be accomplished both verbally and nonverbally (Accreditation Council for Genetic Counseling, 2019). Chat transcripts revealed genetic counselors attempting to build rapport by introducing themselves, sharing their background and credentials when asked, and reflecting information provided by the clients. While primary empathy and physical attending are often discussed as methods to build rapport with clients (Veach et al., 2018), discussing credentials with a patient is less often considered a form of rapport building. In several chats, the client inquired about the credentials of the genetic counselor prior to sharing more extensive or personal information, suggesting that genetic counselors may have shared credentials as a means of establishing credibility. Requests from clients regarding credentials may be due to the online nature of the chat and a need to know more information about with whom they are speaking so information and questions could be appropriately tailored.
Genetic counselors are trained to facilitate decision-making and counsel individuals about genetic topics not only relevant to them but also to their family members (Accreditation Council for Genetic Counseling, 2019). However, few chats reviewed as part of this study involved seeking information related to family members, including familial risks or familial testing. Furthermore, decision-making was only facilitated in one chat. Conversations about risks to family members or facilitating decision-making may be premature given that results from DTC-GT generally require confirmatory testing.
While several distinctions from in-person sessions have been noted, there also appear to be several similarities. Contracting is the process by which a mutually agreed upon agenda is established with the client and is done throughout a session as needed (Accreditation Council for Genetic Counseling, 2019; Veach et al., 2018). Genetic counselors in the present study used contracting throughout the chat to manage expectations both about the service and about topics discussed. In addition to contracting, building rapport, normalizing results and concerns, and implementation of interview skills were counseling strategies used in this format which are also frequently used during in-person sessions. (Veach et al., 2018).

| Limitations
While this study was a novel exploration of chat transcripts from both clients that had undergone DTC-GT and genetic counselors, a few limitations should be noted. We did not have access to the demographic information of clients who participated in these chats. It has been reported that non-Hispanic Black individuals and Hispanic individuals are equally likely to have DTC-GT compared with non-Hispanic White individuals. However, non-Hispanic White individuals are more likely to seek clinical genetic services when they receive an abnormal result (Carroll et al., 2020). Without demographic information, we do not know if these clients are representative of a population that seeks DTC-GT and/or genetic counseling. Without demographic information related to age, gender, or other sociodemographic characteristics, it is unclear if the thematic findings are due to the characteristics of clients or whether they are more generalizable. Additionally, this study utilized data from a single company, DNAvisit; therefore, findings may not be generalizable to other organizations that offer asynchronous genetic counseling via chat.

| Practice implications
Themes identified in the chats revealed questions from clients who underwent DTC-GT and/or raw data interpretation. These include questions related to appropriate next steps, genetic counseling services, and technical questions about testing. Furthermore, this study uncovered counseling strategies that were often utilized during sessions, including managing expectations and giving advice. In any setting or with any counseling delivery model, genetic counselors may consider the informational and counseling needs of clients revealed by this research to prepare for sessions with clients that have had DTC-GT.
This study also explored a service delivery model that has not been described in the genetic counseling literature. In the context of this study, asynchronous chat as a service delivery model for clients who have undergone DTC-GT enabled genetic counselor response to clients' questions through conversations tailored to their needs.
As indicated by the number of exchanges, length of time the service was used, and the information clients sought, the needs of each client were drastically different. This study demonstrated that psychosocial aspects, such as gratitude, frustration, and hope were able to be communicated through this format while also providing answers to questions and factual information such as the limitations of DTC-GT. However, there are several shortcomings in using chat for clients that have undergone DTC-GT. These include the inability to fully assess client affect and, in some cases, the inability to ensure full communication of the limitations of DTC-GT results before the client leaves the chat session or ceases responding.

| Future research
While this study suggests several similarities and differences between in-person and asynchronous messaging-based sessions, it was not designed to directly compare service delivery formats.
Future research may focus on evaluating the similarities and differences between these two service delivery models. Topics discussed, psychosocial needs, and counseling strategies may be compared, which could, in part, be used to determine which patients may benefit from one format over another. Additional research related to the feasibility of implementation of this novel delivery model could also be investigated further. Additional topics that may be explored are comparisons of genetic counselors' preparedness and comfort with this delivery model, time and resource utilization differences between asynchronous chat and other delivery models, and patient characteristics that may impact updates and satisfaction with asynchronous chat.
Furthermore, there have not been any studies related to directiveness and providing advice during genetic counseling sessions to our knowledge. Our study demonstrated genetic counselors were direct in several instances including recommending to not altering medical management based on DTC-GT results or spending time researching likely benign results. Studies that involve understanding how and in what context genetic counselors provide advice during sessions that involve DTC-GT and how clients react to this information may be helpful for counselors.

| CON CLUS ION
The need for genetic counseling for DTC-GT remains high, particularly as access to these tests continues to increase. However, the number of genetic counselors working in direct patient care is limited, especially those who are able and willing to see patients that have undergone DTC-GT. This study explored the use of one service delivery model, asynchronous messaging, to address the needs of clients that have undergone DTC-GT. Results suggest that chat-based genetic counseling can address the vastly different informational and psychosocial needs of clients following DTC-GT.
Chat-based genetic counseling may be distinct from other service delivery models by necessitating an increase in directiveness due to clients' flexibility to abandon the session, a need to identify credentials, and a lack of ability to use visual cues to assess client affect.
Future research may focus on comparing asynchronous messaging to other service delivery models as well as exploring directiveness and how the use of directiveness may differ in chat-based service compared to other models of genetic counseling.

AUTH O R CO NTR I B UTI O N S
CK contributed to study conceptualization, investigation, methodology, formal analysis, project administration, writing of the original draft, and writing as a reviewer and editor. HW contributed to study conceptualization, data curation, formal analysis, writing of the original draft, and writing as a reviewer and an editor. AK contributed to formal analysis, writing of the original draft, and writing as a reviewer and an editor. LED contributed to writing as a reviewer and an editor.
RM contributed to study conceptualization, investigation, methodology, formal analysis, project administration, supervision, writing of the original draft, and writing as a reviewer and an editor. Authors CK and RM confirm that they had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. All the authors gave final approval of this version to be published and agreed to be accountable for all aspects of the work and ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

ACK N OWLED G M ENTS
This research was conducted as part of the Capstone Experience submitted to the Faculty of the Genetic Counseling Program at The University of North Carolina at Greensboro in partial fulfillment of the requirements for the degree Master of Science in Genetic Counseling for the first author (C.K.).

CO N FLI C T O F I NTE R E S T
Heather Wetzel and Andrie Klass were contractors for DNAvisit at the time of this study. Cari Koerner, Lauren Doyle, and Rachel Mills declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Due to the personalized nature of the data as transcripts of counseling sessions, complete data cannot be made available. A subset of coded data similar to what is presented in Table S1 is available from the corresponding author upon reasonable request.

H U M A N S TU D I E S A N D I N FO R M E D CO N S E NT
The IRB at the University of North Carolina at Greensboro determined that this project did not constitute human subjects research as defined under federal regulations and did not require IRB oversight. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000.

A N I M A L S TU D I E S
No non-human animal studies were carried out by the authors for this article.