A positive working relationship between college counselor and client, or the therapeutic working alliance (TWA), is one of the most important factors in successful counseling (Bachelor & Horvath, 1999; Falkenström et al., 2014; Flückiger et al., 2020; Norcross & Goldfried, 2019). Hubble et al. (1999) reported that the therapeutic relationship contributes to about 30% of the total effect of treatment, regardless of the theoretical orientation of the counselor. In fact, many psychotherapy researchers have shifted their focus from various theoretical strategies to focusing on the process aspect of change (Flückiger et al., 2018; Eldridge et al., 2008), which includes the TWA.

Examining how the use of technology impacts TWA continues to be a complex question. First, systematic reviews from Greenwood et al. (2022), Berryhill et al. (2019), and Norwood et al. (2018) indicate that telemental health is an effective method of delivering counseling services. The National Institute of Mental Health (NIMH) defines telemental health as “the use of telecommunications or videoconferencing technology to provide mental health services” (NIMH, n.d.). Many of the aforementioned reviews are not always clear about the exact delivery method, from asynchronous websites and apps to synchronous video counseling sessions. This variability greatly impacts the connection between counselor (if there is any counselor involvement) and client. Second, although there have been systematic reviews of research looking at client and therapist satisfaction with TWA, many of the studies have not been outcome-based or conducted with controls. Therefore, treatment effects may have been overestimated. In addition, few studies have examined working alliance differences with college students specifically, and this population has the potential to benefit from access to telemental health (Ollio et al., 2022).

The connection between TWA and counselor perceptions of telemental health has been examined, albeit with various findings. For example, researchers in the early 2000’s reported a concern among psychologists for the effect of videoconferencing on the TWA (Rees & Stone, 2005; Wray & Rees, 2003). Rees and Stone’s (2005) study focused on psychologists’ perception of their clients’ sense of warmth and empathy in treatment. These researchers noted that the lack of experience using videoconferencing among the psychologist participants as well as the impact of negative expectations on their perceptions may have been possible limitations of their study. Relatedly, Simms et al. (2011) showed that counselors who had been in the field longer and had more training on the use of technology were more likely to use online counseling with some frequency, suggesting that the mental health provider’s comfort with the medium may be a significant issue. Lopez’s (2019) recent systematic review of the literature shows that the comfort of the counselors in the use of technology has a significant impact on their perceptions of the alliance in this delivery method.

Studies have also focused on the impact of TWA on treatment outcomes. For example, Klein et al. (2010), in an experimental design study, addressed the outcome of online treatment for posttraumatic stress disorder (PTSD), with a focus on symptomology, treatment satisfaction, and the TWA. This online treatment protocol for PTSD appeared to be effective, considering that 69.2% of participants (n = 16) no longer met the criteria for PTSD immediately post-treatment, and 77% no longer met criteria three months post-treatment. In addition, 69% of participants reported satisfaction with treatment, and 87.5% rated the TWA highly based on the Traumatic Antecedents Questionnaire. However, the small number of participants was a limitation.

Preschl et al. (2011) conducted a randomized controlled study (N = 53) looking at cognitive-behavioral therapy for depression delivered online (with direct therapist communication) and in-person. The online and in-person participants had similar scores on the Working Alliance Inventory (WAI), with the exception of therapists’ ratings for the Task subscale, which were significantly higher for the online group. Though the TWA was not as impactful on symptom reduction in the online group, treatment outcome was similar across groups. Other similar research postulated that the TWA might be an “additional indirect measure of outcome rather than a predictor of treatment outcome” (Knaevelsrud & Maercker, 2006, p. 8). These studies suggest that the TWA may have a different relationship to outcome when conducting counseling using online technologies.

In continuing efforts to examine the impact of telemental health delivery on the TWA as well as treatment outcomes, Watts et al. (2020) showed an equivalent working alliance using synchronous video delivery of counseling compared to in-person delivery, specifically for those with generalized anxiety disorder (GAD). In fact, overall clients with GAD preferred telemental health, indicating that videoconferencing was “less intrusive and intimidating” (p. 218). This is important given that anxiety is the most frequent reason given for college students to attend appointments at their campus counseling center (AUCCCD, 2021).

Another non-inferiority study conducted in Canada found that the TWA was comparable regardless of delivery method for those with panic disorder and agoraphobia (Bouchard, 2020). Outcomes also were similar in terms of reduced anxiety and agoraphobia systems for both telemental health and in-person groups. The TWA was strongly correlated with a reduction in dysfunctional beliefs which were then predictive of treatment outcome. One consideration for further exploration is whether clients with anxiety symptoms have a natural preference for telemental health, given possible fears inhibiting social situations.

Solution-Focused Brief Therapy

Solution-Focused Brief Therapy (SFBT) was developed by de Shazer (1985), Insoo Kim Berg, and Peter De Jong (1996). The question of efficacy of SFBT has been examined through systematic reviews. Gingerich and Peterson (2013) conducted a systematic review of 43 controlled studies. A high percentage (74%) of these studies showed significant positive benefit for clients, especially those with depression symptoms. This review also found that fewer sessions were used to achieve this benefit than in other models of treatment. Kim (2008) examined 22 studies using hierarchical linear modeling software in order to calculate an overall effect size. SFBT showed positive treatment effects, although most were small. There was a statistically significant result (p < .05) on internalizing behavior problems. Finally, Kim et al. (2019) provided an update on the evidence base for SFBT and indicated that there has been growth in the number of experimental design studies that give evidence for efficacy of SFBT, including with diverse populations.

In terms of the college population, there are few studies on SFBT. An Indonesian study examined the effect of Solution Focused-Brief Counseling (SFBC) on the stress levels of students in college in a one-group pre- and post-test design (Sitindaon & Widyana, 2020). There was a drastic decrease in stress levels after the SFBC intervention in this population, although the limited number of participants (N = 6) is a concern. Wettersten et al. (2005) compared SFBT sessions at a university-based counseling center with archival data from clients engaged in Brief Interpersonal Therapy (BIT). Twenty-six therapist–client dyads were included in the SFBT sample. Clients completing SFBT were 73% women and 89% Caucasian, while clients in the BIT group were 79% women and 87% Caucasian. The researchers used the WAI, the Brief Symptom Checklist (DeRogatis, 1993), and the Counseling Center Follow-up Questionnaire (Gelso & Johnson, 1983). This study found that the relationship between the TWA and treatment outcome was not correlated significantly when using SFBT. In contrast, the archival BIT outcome ratings were correlated with TWA. The researchers suggested that in SFBT, the counseling relationship may be a means to an end rather than the end in itself (Wettersten et al., 2005).

Although the literature listed above on the TWA using SFBT provides some useful information, there are significant gaps. For example, the TWA in SFBT has not been explored in an online environment using randomized, controlled methods, including treatment for college students with anxiety symptoms. Although emerging adults may have greater comfort with online delivery of services (Hanley, 2012), few randomized controlled studies using online treatment with college students have been conducted. Information in this area will inform the theory and practice of telemental health as well as SFBT for anxiety.

Purpose

The purpose of this study was to conduct a randomized, controlled study to provide data on the TWA using SFBT in college students with anxiety and to compare an online delivery system with treatment-as-usual (in-person). There is a clear gap in the literature and a need for more experimental design studies comparing the TWA in online delivery systems to treatment-as-usual (in-person) using brief treatment protocols.

The predictive value of the TWA on the efficacy of anxiety treatment for both modalities was examined with the goal of informing the use of online counseling in college counseling settings. This study addresses the following research questions:

  1. 1.

    Is the TWA (as measured by the client version of the WAI) in telemental health comparable to the TWA in in-person counseling after the initial session using SFBT to treat emerging adults at college with mild to moderate anxiety?

    H0: The TWA scores reported by clients using telemental health will not be statistically inferior to the TWA using in-person counseling, as determined by this non-inferiority study.

  1. 2.

    Is the TWA (as measured by the client version of the WAI) in telemental health comparable to the TWA in in-person counseling after the third session using SFBT to treat emerging adults at college with mild to moderate anxiety?

    H0: The TWA scores reported by clients using telemental health will not be statistically inferior to the TWA using in-person counseling, as determined by this non-inferiority study.

  2. 3.

    Is the TWA (as measured by the client version of the WAI) after the first session of SFBT predictive of the level of anxiety post-treatment (as measured by the Beck's Anxiety Inventory (BAI)) after controlling for the pre-treatment level of anxiety (as measured by the BAI)?

    H1: The TWA explains significantly the variance of post-treatment anxiety.

Method

Participants

Participants were recruited from the undergraduate student population at a private, Catholic university in the Northeast. There were 52 participants recruited, and, due to self-attrition, 49 remained in the study. In terms of demographics, 4 participants identified as male, 45 as female, none as non-binary or genderqueer. There were 5 participants who identified as African-American, 2 as Asian-American, 4 as Latine, and 38 as White with no international students participating. These demographic identities (gender identity and ethnicity) are in line with the demographics of the typical clients attending this university’s counseling center as available through Titanium Software. Twenty-four participants reported having had previous counseling while 25 had not. Of the participants, there were 19 first-year students, 10 sophomores, 11 juniors, and 9 seniors. Participants’ ages ranged from 18 to 22 (M = 19.29, SD = 1.2).

Participants were included in the study based on scores on the Counseling Center Assessment of Psychological Symptoms (CCAPS-62; Locke et al., 2011). For generalized anxiety, the range of scores for inclusion was between 40 and 90 and for social anxiety between 50 and 95. There were several other exclusion criteria. Students who scored in the clinical range on the substance use subscale (above 80) were not eligible for the study because of the possible need for more substance-specific treatment. These students were referred to the alcohol/drug counselor for further evaluation. Finally, on the depression subscale, there are questions about suicidality. If a student answered the question “I have thoughts of ending my life” with very much like me or extremely like me, they were not eligible for the study, but instead were immediately seen for further suicide risk assessment and possible treatment. Although no students began taking psychotropic medication during the course of the study, students who already were taking medications were included in the study. They still needed to score in the appropriate range on the generalized anxiety and/or social anxiety subscale to qualify.

Treatment

Participants engaged in three SFBT counseling sessions post-screening. Based on reports generated by the university counseling center’s electronic records system, the average number of sessions attended for this counseling center is three; therefore, this protocol represented average attendance at this site. Furthermore, SFBT is most effective in the 3- to 5-session range (Beyebach et al., 2000).

The first author and another licensed clinical social worker from the counseling center implemented the treatment protocol. Both clinicians have expertise in college counseling work, having been at this counseling center for over 20 years. In preparation for delivery of the SFBT protocol, the second author, who had training in SFBT and had experience using it, provided a 2-day intensive training for the research clinicians. Before the research project began, sample online sessions were recorded by the counselors as part of training feedback, and the textbook Interviewing for Solutions (4th ed.; DeJong & Berg, 2013) was used as an instructional text. Supervision by the second author continued throughout the period of data collection.

Synchronous video sessions were provided through Doxy.me software, a HIPAA-secure platform (see Platform section).

Procedure

Sample

The majority of participants were recruited by screening counselors (other than first author and research clinician) in the university counseling center. Additional participants were recruited by flyers that were distributed in residence halls, in high-traffic areas on campus, through global e-mails, and to faculty. Participants were offered a $30 Amazon gift card for participating.

The study was approved by the participating institution’s IRB. If students expressed interest in the study either by responding to one of the recruitment flyers or by entering the counseling center for an appointment, they were given an appointment for an in-person screening, as is the procedure for all students. Students completed all intake forms through the Titanium system, which includes the CCAPS-62 (see Measures). Titanium Schedule is a college counseling electronic medical record program used by numerous college and university counseling centers around the world. This symptom checklist was used by the screening counselor to determine initial eligibility. Then, the screener’s brief clinical interview allowed for final determination of eligibility. Screening counselors did not begin the therapeutic relationship at this point. If the student met research inclusion criteria, the screener described the study and reviewed the written informed consent. Informed consent was obtained for all those who chose to participate. Those students who were not interested or who did not qualify went to the administrative assistant for a follow-up appointment in accordance with the counseling center’s established procedures. These initial screening appointments were conducted similarly, whether the student was randomly assigned to an online or in-person counseling treatment group.

The screening counselor handled the review and signing of the informed consent with the participant. After this process, the student was directed to the administrative assistant’s desk to schedule future appointments and receive a randomized assignment. The administrative assistant used a 6-sided die to assign them to the online or in-person counseling group (1–3 = in-person, 4–6 = online). She then scheduled the participant with one of the participating counselors based on a general balance of sessions between counselors. In addition, each participant was given an identification (ID) number for anonymity when filling out all Qualtrics questionnaires. Participants were asked to provide basic demographic information.

All online participants were given instruction on (a) how to access the Doxy.me account, (b) use of audio and video components with the device they were using, and (c) how to upload documents. The online participants were given a technology check appointment per best practices in telemental health (Langarizadeh et al., 2017) to ensure a smooth first session.

The technology check reviewed the procedures for continuing the session if there were a technology failure (i.e., phone contact). In addition, emergency procedures were reviewed with the participants at this appointment per the American Counseling Association (2014) Code of Ethics for distance counseling, as well as confidentiality limitations. An online informed consent form separate from the main research informed consent was reviewed and signed through e-mail by all online participants.

Intervention

The SFBT protocol was a 3-session treatment for both online and in-person deliveries, based on DeJong and Berg (2013). Session 1 included: (1) brief exploration of the problem with scaling of anxiety (10–15 min); (2) asking the miracle question and brainstorming client’s solutions (20–30 min); (3) break for feedback and homework to be written (5 min); and (4) client given feedback and homework (2 min). Session 2 protocol was: (1) scaling of level of anxiety in the past week (1 min); (2) ideas of what would move number on scale to reduce anxiety (15 min); (3\2) review of homework and what worked/what did not (15 min); (4) break for feedback/homework to be written (5 min); and (5) client given feedback and homework (2 min). Session 3 protocol included: (1) scaling of level of anxiety (1 min); (2) review of homework (15 min); (3) discussion of what was achieved given final session (strengths of client) (10 min); and (4) client sets goals going forward (5 min). The client version of the WAI was given after the first SFBT session and after the third and final SFBT session in both delivery systems. Students used only their ID numbers when filling out these inventories in Qualtrics, a web-based survey site. At no time did the counselors have access to identifying information on responses to the WAI. The informed consent document explained to participants that the counselors would not access or review the data collected in Qualtrics while the research was still ongoing. Data were only reviewed by the researcher upon completion of the research project, with research IDs serving as the only identifying information.

For treatment fidelity, all sessions were audio-recorded, whether in-person or through Doxy.me. The online recordings were recorded through an external recording device since Doxy.me does not allow for internal recording. Treatment fidelity was assessed by the second author and another counselor who participated in the initial SFBT training together with the two treatment counselors. All first sessions were reviewed for treatment fidelity as well as 10 randomly selected subsequent sessions. Fidelity checklists were developed for each session to establish SFBT fidelity, and treatment fidelity rate was determined to be 95%.

Measures

Counseling Center Assessment of Psychological Symptoms-62

The CCAPS-62 (Locke et al., 2011) is a psychological symptom checklist designed for use with college students. The eight subscales are: (a) Depression, (b) Generalized Anxiety, (c) Social Anxiety, (d) Academic Distress, (e) Eating Concerns, (f) Family Distress, (g) Hostility, and (h) Substance Use. The initial iteration of the scale was released in June 2009, and an update was released in 2012. The subscales have a minimum factor loading of 0.32 for each item (Tabachnick & Fidell, 1996) and an item total correlation of 0.30 or above (Nunnally, 1978). The CCAPS-62 is commonly completed at the first intake appointment in college counseling centers while the shorter version, the CCAPS-34, is used for regular assessments of treatment efficacy (Locke et al., 2011). The total time to complete the longer version is 7 to 10 min.

The test–retest reliability for the Generalized Anxiety subscale was 0.782 after one week and 0.842 for a two-week period. The internal consistencies of the subscales were high, with all being above 0.80. Locke et al. (2011) looked at reliabilities among diverse student populations and found a reliability index greater than 0.75.

McAleavey et al. (2012) found good convergent validity when comparing the CCAPS-62 to established psychological symptom checklists used by clinicians. The correlation results of 0.692 and 0.643 (Generalized Anxiety) and 0.747 and 0.733 (Social Anxiety) were found for the anxiety subscales used in McAleavey et al.’s study.

Working Alliance Inventory

The WAI (Horvath & Greenberg, 1986) was constructed using the three central factors in Bordin’s (1979) theory of alliance: goal, task, and bond. This is a 36-item questionnaire anchored to a 7-point Likert scale (1 = never to 7 = always). Eaton et al. (1988) showed that the TWA could be established within the first three sessions, and most research on the TWA recommends having these measures completed after the third session. This makes the WAI ideal for this study using SFBT.

Kokotovic and Tracey (1990) found subscale reliability on the client version to range from 0.88 to 0.91. Cook and Doyle (2002) found that reliability coefficients “were slightly lower, with a composite alpha of 0.86 and subscale scores of 0.59, 0.70, and 0.76 for tasks, bonds, and goals respectively” (p. 99). Construct validity has been evidenced based on high correlations between the WAI and other working alliance instruments (Tichenor & Hill, 1989). In addition, Horvath and Greenberg (1986) reported high internal consistency on the client form (0.93) and construct validity on the client form. The reliability coefficient of the WAI for this study was 0.904.

Apparatus

Doxy.me is videoconferencing telemedicine software that is HIPAA-secure and provides the clinician with a business associates’ agreement. It is marketed for use with patient care in telemedicine and telemental health fields. It is accessible on computers, tablets, and mobile phones. In this study, we used the synchronous video and audio functions of Doxy.me to deliver SFBT for mild to moderate anxiety. The software allows counselors to create a waiting room, with check-in capabilities for clients, as well as screenshare capability between counselor and client.

Survey data was collected on Qualtrics, an online research tool that was used for participants to complete the WAI after the first and third session of treatment. Qualtrics was licensed by Oregon State University.

Data Analysis

SPSS Version 25.0 was used for data analysis. Dataset was saved on Dryad (DOI: https://doi.org/10.5061/dryad.b2rbnzspt). All 49 participants completed more than 90% of the WAI after Sessions 1 and 3, and missing data were addressed by using the mean of completed items for that participant per best practices with psychometric scales (Siddiqui, 2015). Outliers were assessed through evaluations of z scores: less than − 2.68 or greater than + 2.68 were considered outliers (Hawkins, 1980). To test the hypothesis that there would be no difference in WAI scores between the online and in-person groups after the 1st and 3rd session, an independent samples Welch’s t test was conducted. This test performs better for unequal variances and offers the same result as Levene’s test for equal variances. For Research Question 3, a hierarchical regression analysis was employed. The conventional method in clinical trials of answering the question of non-inferiority was used. The null hypothesis that the TWA scores reported by clients using online counseling would not be statistically inferior to the TWA using in-person counseling would be rejected if, and only if, the lower limit of the confidence interval (because higher scores are better on the WAI) divided by the mean of the independent variable group were greater than − 0.2 (Chen et al., 2006).

Results

Group differences were examined between the two groups of participants: SFBT in-person (treatment-as-usual) versus online. An examination of group differences revealed no significant difference in WAI between the two delivery methods (see Table 1). The results of the Welch’s t test were as follows: 1st session, F(1) = 0.196, p = .660; 3rd session, F(1) = 0.026, p = .872, showing that the WAI scores were not statistically significantly different between in-person and online delivery methods between the sessions.

Table 1 WAI mean scores after first and third sessions
Full size table

The non-inferiority score for Session 1 was − 0.06. Therefore, the data does not support the rejection of the null hypothesis; this indicates that online counseling was not statistically inferior to in-person counseling in terms of WAI. For Session 3, this score was − 0.03, again indicating non-inferiority.

A hierarchical analysis also was conducted which used two models: (a) Model 1 predicted the post-BAI score from the pre-BAI score; and (b) Model 2 added the WAI 1st session score. The model predicting post-BAI score from pre-BAI score accounted for a significant amount of variance, F(1,34) = 16.202, p < .001. When controlling for the pre-BAI score, the addition of the WAI 1st session significantly increased the variance accounted for in the post-BAI outcome F(2,33) = 11.993, p < .0001. The overall regression model predicted approximately 43% of the variance in the post intervention BAI score (R2 = 0.43, F(2,33) = 11.993, p < .001). First session WAI accounted for about 10% in the variance in post-BAI scores (Table 2) after controlling for pre-BAI scores.

Table 2 Regression analysis showing BAI Pre and WAI 1st session as predictors of BAI Post
Full size table

Discussion and Implications

In this study, we explored the impact of the TWA on anxiety treatment outcomes, comparing online to in-person delivery of counseling in a college counseling center setting. Concerns about TWA are a central roadblock to mental health providers using an online delivery method, and there has been a question as to the non-inferiority of the TWA using telemental health. Results of the present study showed that there was no statistically significant difference in TWA for telemental health compared to treatment-as-usual (in-person) in the first and third sessions of SFBT, as measured by the WAI. This supports the original hypotheses, as well as previous research in this area (Knaevelsrud & Maercker, 2006). Blackwelder’s (1982) non-inferiority statistical analysis showed that, for this small clinical trial, telemental health was not inferior to in-person counseling in terms of the WAI. In addition, the random assignment of clients to online vs. in-person delivery of counseling supports the findings that goals, tasks and bonds (Bordin, 1979) had been established in both delivery methods.

College and university counseling centers reverted to some form of online counseling during the COVID-19 pandemic, as did most of the mental health field. As this pandemic has subsided, many college students continue to request counseling through some kind of technology (Field, 2021). This study adds to the literature that supports the development of the TWA online, which adds to the efficacy of this delivery system.

In addition, given the prevalence of brief therapies in the current mental health marketplace, it is important to address the speed with which the TWA can be established, whether online or in-person. These results support a central tenet of SFBT, that one session can be enough for noticeable reduction in symptoms using this strengths-based approach (DeJong & Berg, 2013). This finding helps to lessen the concern some counselors may have about the development of the TWA, a critical predictor of treatment outcomes, using brief therapeutic modalities with either form of delivery.

Limitations and Recommendations

Despite the findings, it is important to acknowledge the limitations of this study. The sample size was small, and therefore generalizability is limited. The participants were students at one university in the northeastern United States and therefore are not necessarily representative of national or international students. Also, although the percentage of male-identifying students in this study was small, it is in line with the overall percentage of males seeking support at this counseling center. Replication studies with more diverse client samples at other universities should be conducted to verify the findings in this small, initial study.

This study focused on college students with mild to moderate anxiety, as determined by the BAI. Although this primary diagnosis of anxiety is a common one in college students, especially post-COVID-19 (Ding et al., 2022), it is possible that the particular symptoms and behaviors of a student with anxiety created a tendency for preferring telemental health, i.e. discomfort in social situations, navigating a counseling center in-person. The only caveat to this is that most participants were in the “mild to moderate” range per the BAI and had no panic or agoraphobia symptoms. However, further study focusing on non-anxiety symptomology using telemental health is warranted.

Additionally, given that this was a “real world” counseling center setting rather than a laboratory, there were real-world variations in counseling styles and other details regarding protocol that were difficult to control. For instance, with some online participants, the feedback sheet was e-mailed to them after the session ended; for others, a break was taken, with both client and counselor remaining online and then reviewing the feedback together. Also, some sessions were significantly shorter than others, depending on the counselor and style of delivery. A laboratory setting result would be helpful in order to control for some of these factors, and results could be combined with these to get a more complete picture. Finally, the number of participants made it difficult to get a medium effect size using the t test.

Non-inferiority studies are more challenging and typically are undertaken with clinical medical trials to determine if one treatment is “as good as” the standard treatment. That being said, this study used this statistical method with fewer participants. Increasing the number and diversity of participants would improve the generalizability of these non-inferiority results.

Additional research could compare the therapist’s rating of the WAI with the client’s rating when random assignment to the delivery method is employed. As described in some of the previous research on therapist-rated WAI in telemental health, this comparison would elucidate the impact of therapist perception of alliance on the client’s sense of connection when technology is being used. Also, given previous research indicating a lack of correlation between the TWA and outcome when using online counseling (Knaevelsrud & Maercker, 2006), the use of an inventory like the WAI, which was designed for in-person counseling interactions, comes into question. Does online delivery require an inventory designed to assess the unique ways that clients connect through technology? Perhaps, the balance of goal, task, bond developed by Bordin (1979) can be affected by an online intervention. Researchers may also consider using qualitative inquiry to examine if there are qualitative differences in clients’ perceptions and experiences of the TWA between in-person and online delivery.

Conclusion

This non-inferiority study examined the differences in the TWA using the WAI in online versus in-person SFBT for undergraduate college students with mild to moderate anxiety, as well as the predictive relationship of WAI on outcome. The findings indicated that telemental health was not inferior to treatment-as-usual (in-person) when it comes to TWA. These findings are strengthened by the fact that participants were randomly assigned to delivery method and did not choose the method with which they felt more comfortable. In addition, the strength of the TWA accounted for roughly 10% of the overall variance in treatment outcomes after the first session. Since TWA is often the greatest concern for counselors in providing counseling through technology, this study strengthens the telemental health literature. This is particularly useful given the dramatic increase in counseling through technology that occurred during the COVID-19 pandemic.