Damian Sendler: Having a strong therapeutic relationship is considered essential to delivering psychotherapy in an ethical manner. Many questions remain unanswered about the alliance between text-based digital psychotherapy and its importance in delivering mental health care via the internet. In order to use text-based digital therapies responsibly, more knowledge about the type and strength of the therapeutic alliance online is required.
Damian Jacob Sendler: Anxiety and/or depression were the most common diagnoses in the research on text-based digital interventions that looked at the therapeutic alliance. ICBT and tailored platforms for specific client groups were the primary treatment modalities used in this study (e.g., PTSD). To provide feedback on tasks, nearly all treatments employed some form of text-based, asynchronous communication, such as e-mails and built-in messaging functions. As a standard measurement tool, most studies employed a modified WAI (Working Alliance Inventory). A weighted mean score of 5.66 (on a scale of 1 to 7) and a weighted standard deviation of 0.84 were found using the WAI or WAI-short form, indicating a good therapeutic alliance. Significant associations were found between the therapeutic alliance and treatment outcome in many studies (n = 8 out of 10) as well as significant therapeutic alliance-related outcomes (n = 5 out of 6).
Dr. Sendler: Psychiatric treatment can be delivered via the internet, which has been studied for a long time. The advantages include broad and easy access, as well as lower costs (1). Online psychotherapy, in particular, has the potential to expand mental health services to underserved populations and during times of crisis, and thus serves as a viable alternative or supplement to face-to–face therapy. Research shows that online therapy has moderate to large effects and comparable results in randomized controlled trials (RCT’s) for a variety of psychopathological symptoms (3–5). Despite the overwhelming evidence that both approaches are equally effective, many people still believe that the therapeutic relationship between the therapist and the patient is critical to the process of successful psychotherapy. The importance of this relationship in online therapy is being debated by researchers, practitioners, and even some clients (6).
In general, therapists believe that digital psychotherapy offers a wide range of benefits, including new treatment options and even a more intimate therapeutic relationship (7). In times of crisis, the lack of physical proximity may prevent them from performing the proper caretaking actions, but the increased accessibility of digital psychotherapy gives them a greater sense of moral responsibility. Clients’ trust and openness in their therapists could be crucial in these situations, so that the therapists can respond quickly to their concerns (8). A lack of knowledge and trust are also major obstacles for therapists when it comes to using digital interventions (7, 9). It is important to know how much of a good therapeutic alliance can be established in various digital treatment modalities in order to use digital therapy responsibly.
Text-based digital psychotherapy is one type of online psychotherapy where the role of the therapeutic relationship is still ambiguous. There are many forms of digital psychotherapy, but the most common form is e-mail or chat-based, such as internet cognitive-behavioral therapy (ICBT) (10). Using digital psychotherapy in this way can help the therapeutic relationship in novel ways, such as by giving clients more time to reflect on the written expression of their feelings and thoughts without being observed. Writing allows for more time to think before responding, which can benefit therapists, as well (2, 11). To further encourage alliances and self-disclosure, there is a lack of social cues, making it easier for people to open up and share their thoughts and feelings without fear of reprisal (12). Unlike traditional face-to-face or videoconferencing therapy, this could foster a close and trusting relationship between clients and therapists (13).
Online text-based therapy, on the other hand, can lead to concerns from therapists about the lack of nonverbal cues and whether a therapeutic alliance can be formed through text-based online counseling (14). A further question arises: Is it beneficial to have a certain type of alliance in the therapeutic process? (10). If you want to use text-based digital psychotherapy in a responsible way, this scoping review will help you understand the therapeutic alliance and how it relates to therapy outcomes.
The term “working alliance” is frequently used to operationalize the therapeutic alliance. The therapist-client working alliance encompasses a variety of cooperative aspects (15). It is worth noting that Bordin (16) developed a model of the working alliance that is still widely used today. This model focuses on the therapeutic partnership between therapist and client from a variety of theoretical perspectives. Both the therapist and the client must agree on therapeutic goals in order for therapy to succeed. Processes and behaviors in therapy sessions that have a direct bearing on the actual therapeutic work are referred to as “tasks.” Interpersonal attachment between the therapist and the patient is referred to as the bond, and it should be characterized by mutual trust as well as confidence, acceptance, and respect (16).
The alliance is more important in face-to-face psychotherapy than in online psychotherapy, according to therapists (17). Internet-based psychotherapy has been found to have a lower level of trust in the ability of therapists to form a therapeutic alliance (17). Internet-based cognitive behavior therapy studies, however, show that the therapeutic alliance is at least as strong in face-to-face therapy and also highlight the association of the alliance with online treatment outcome, whereas this is the case in face-to-face therapy (15, 18, 19).
According to meta-analyses of the alliance in both traditional and online psychotherapy for adults, there was a strong correlation between the alliance and treatment outcomes in both settings (r = 0.278 and r = 0.275), suggesting that the alliance matters when it comes to getting the best results (15). According to the findings, the relationship between the alliance and outcome appeared to be consistent across various alliance and outcome measures, treatment approaches (such as Cognitive Behavioral Therapy and Psychodynamic Therapy), and client characteristics. The overall correlation between alliance and outcome was nearly identical to the correlation found in an earlier meta-analysis (20).
It has been found that there is a strong correlation between the level of trust between the therapist and the patient and the level of improvement in the patient’s mental health. The effectiveness and alliance-outcome associations examined in many recent reviews and meta-analyses fail to distinguish between various types of psychological online interventions. Many different types of electronic mental health were included in the meta-analysis by Flückiger and colleagues (15), for example (via internet, e-mail, videoconferencing and phone). This restricts our ability to make informed decisions about the various ways in which these interventions can be implemented online.
Berger and Andersson (21), who make a distinction between online psychological interventions’ modes of communication, offer a way to classify them. Text-based and asynchronous (e.g., e-mail) or near-real-time or synchronous (e.g., chat) communication is possible, as is audio- or video-based synchronous communication (e.g., video-conferencing). Berger’s (10) review was the first to use Berger and Andersson’s (9) categorization of the working alliance to examine ratings of the alliance (21). However, Berger’s (10) review was narrative, and did not provide an overview of all the alliance measurements and their relation to outcomes that studies reported in the literature. More research is needed into text-based digital psychotherapy, which Berger (10) found to have a limited database on the alliance compared to other forms of online interventions. A therapeutic alliance in text-based internet psychotherapy, which is defined by Berger and Andersson (21) and involves therapist-client communication via the internet and text, is examined here specifically in this review (e.g., e-mail, chat). To focus on the therapeutic relationship between the client and the therapist, studies on self-help text-based digital interventions are not included in this section.
This communication method, which included feedback on self-help tasks and written assignments as well as emailing and using an integrated text or chat function within the treatment platform (n = 26), was the most common in the reviewed interventions (Table 1). When clients sent messages, asked questions, or completed writing assignments, therapists always responded with text-based responses within 24–48 hours. One intervention did not specify the method of communication, while another simply stated “written exchange” without further explanation. There were only two interventions in which the client and the therapist communicated in real time via text in a chat room.
There were 16 instances in which the client and therapist had weekly contact for interventions. 2–3 times a week (n = 2) and every 10–14 days (n = 1) were also used. On-demand access to free online counseling was made possible through one such intervention. There was no mention of how often the participants were contacted during nine interventions. Time spent by therapists on each client was reported for eleven of the interventions studied. A weekly average of 15 minutes was the most common (n = 5). When the therapist responded to writing assignments, the average time per client per text increased from 20 to 50 minutes to 45–50 minutes. A 45-minute weekly session duration was reported for one of the two interventions that utilized synchronous chat sessions. The interventions’ treatment durations ranged from five to sixteen weeks, six to ten modules, or nine to fourteen sessions. This review’s treatment lengths were typically 8 (n = 6) or 10-weeks-long (n = 6) for the vast majority of the interventions it included. Using instant messaging as a form of free online counseling, some participants accessed the intervention two to five times or more.
Damian Jacob Markiewicz Sendler: Psychotherapists in training (n = 6), licensed but unregistered psychologists, psychotherapists, and psychiatrists were among those who offered or guided treatment via the internet, according to the number of interventions that did so (n = 6). For additional interventions, students in the final stages of their master’s degree in psychology or social work were hired (n = 7). One study did not specify who was in charge of communicating with clients during the treatment program (n = 7) while others used the term “online therapist” or “therapist” to refer to their practitioners.
The cognitive behavioral therapy approach was used in the majority of the studies reviewed (n = 18). Cognitive behavioral therapy and narrative exposure therapy were used together in one study, while mindfulness-based cognitive therapy and psychodynamic treatment were used in another. Another three studies did not identify which psychotherapeutic approach was used in their treatments.
The Working Alliance Inventory (WAI) (n = 19) was used in the majority of the studies reviewed (24). Using the WAI’s client and therapist versions, one can arrive at a composite score that incorporates the two. However, the majority of the studies included here (n = 17) used only the client version of the WAI. Only one study utilized the original WAI, which consisted of 36 items graded on a 7-point Likert scale. The WAI-S, a 12-item shortened version of the WAI, was used in fourteen studies. For internet-guided interventions (n = 2) or online support specifically for women (n = 1), a modified version of the WAI-S was used in a few studies. Four studies also utilized the WAI-SR, a revised short form with 12 items but a 5-point Likert scale (51). When it comes to internet interventions, a new scale has been created called the WAI-I, which was derived from the WAI-SR (36). Additionally, there are 12 items on the WAI-I, each with a Likert scale of 1 to 5.
For the WAI and WAI-S, a score between 1 and 7 is commonly calculated (50). The average working alliance with the WAI or the WAI-S was measured in 12 studies examining 15 interventions and ranged from 4.30 (SD: 1.27) to 6.3 (SD: 0.54). Overall, there was a positive working alliance with a weighted average score of 5.66, and the standard deviation was 0.84. A weighted average of 3.23 was found in only three of the four studies that used the revised short form (with a maximum score of 5) of the WAI (weighted SD: 0.8). Only one had a mean score of 2.34 (SD: 0.98) out of the others (38).
The therapeutic alliance was also assessed using a variety of scales in a few studies. TAQ (52) was used in two studies that focused on three different interventions, and it only assesses clients’ perspectives. The maximum TAQ score is 102, and a score above 80 indicates a strong working alliance (37). Based on the mid-treatment scores when they were included, the two studies here showed an average weighted TAQ-score of 85,26 with an average weighted standard deviation of 12,44, indicating an effective working relationship.
It was found that one study used a Scale to Assess the Therapeutic Relationship [STAR (53)], which has a separate patient (STAR-P) and clinician (STAR-C) scale, both with 12 items and scores that range from 0–48, with higher scores indicating a better therapeutic relationship. There was an average STAR-P score of 37.41 and an average STAR-C score of 30.54 in the study included here (standard deviation: 1.5).
Finally, a short version of the Agnew Relationship Measure (ARM) was used in one study. Clients and therapists can use the same 12-item Likert scale and 7-point Likert scale for the ARM’s parallel versions. Client and therapist subscale scores were only reported in this study, not a composite score. Patients’ sub scores ranged from 5.27 to 6.19 (out of 7) and therapists’ sub scores ranged from 4.73 to 5.76 (out of 7), which suggests a positive therapeutic relationship.
There were 28 interventions in the 23 articles that were reviewed. For 22 of these interventions, studies found statistical correlations between measures of the therapeutic alliance and outcomes. There were significant correlations between the therapeutic alliance measures and the change-scores on one or more primary outcome measures in 8 out of 13 interventions (8 out of 10 studies), ranging from small (n = 1) to moderate (n = 7) and strong (n = 1) in magnitude for the 8 interventions. Moreover, the higher the therapeutic alliance, the better the treatment outcomes were predicted for six of the seven interventions (5 of the six studies) that were examined on the predictive value of the alliance on subsequent outcomes.
The therapeutic alliance was found to be a significant predictor of both compliance and client satisfaction in three of the four studies. Both studies found that the alliance had a significant impact on customer satisfaction. There was a significant (n = 1) or significant and moderately correlated (n = 2) relationship between subscales of the WAI, such as agreement on tasks, in all three studies.
The therapeutic alliance was examined in 28 text-based interventions in 23 articles. These articles were analyzed for a variety of factors, including the type of participants, study design, and intervention, as well as the relationship between the therapeutic alliance and treatment outcomes.
Damian Sendler
Research on this topic has been conducted with a wide range of clients and treatment approaches, as our findings demonstrate. But the majority of studies used internet-based cognitive behavioral therapy (ICBT) to treat patients with anxiety and/or depression symptoms. Asynchronous communication methods like emails and delayed chats, as well as text-based features built into websites and platforms were the norm for most digital treatments based on text.
Damien Sendler: It is crucial to know whether or not a therapeutic alliance can be formed in this therapy format and whether or not the strength and type of a therapeutic alliance that is formed through text is comparable to those found in face-to-face treatment if digital text psychotherapy is to be used responsibly. This suggests that positive alliances can be established in digital interventions even if only text-based communication modalities are used between clients and therapists, as the therapeutic alliance scores in the reviewed articles on digital text-based psychotherapy were generally high Therapeutic alliances have been found to have a significant impact on treatment outcomes in most studies.
Only a few studies have been done on the therapeutic alliance in text-based digital therapy in previous reviews of this topic (6, 10). For the first time in a while, a review of the therapeutic alliance in this therapy format has been updated. In Berger’s narrative review (10) there were a total of 23 included studies, seven of which were published in 2017 or later. This suggests that the body of evidence supporting text-based digital treatment alliances is growing. 14 studies in this review focused on therapeutic alliances as a primary objective, compared to six studies in a 2012 review that focused on this topic (6). According to the current review, the therapeutic alliance in digital therapy and face-to-face therapy has been shown to have a mixed to positive relationship with treatment outcomes based on the larger evidence base (6, 10).
Despite the fact that the studies examined covered a wide range of treatments for different client groups, the majority of them focused on (a combination of) anxiety and depression symptoms. Considering that anxiety and depression are the most common mental illnesses, this is not surprising (56). It is also clear that, in line with previous findings on internet psychotherapy, cognitive behavioral therapy dominates the treatment options online because the vast majority of the studies reviewed here use this approach or framework (6, 10). CBT’s short-term interventions and techniques fit well in an online format (57), and may also be more easily integrated into internet-based psychotherapy given the length of the treatments reviewed here (five to sixteen weeks), which is consistent with the findings in this study. Psychodynamic approaches and third-wave cognitive behavioral therapy (CBT) were not adequately represented in this review. Even as internet-based psychodynamic treatments and third wave approaches are becoming more common, the working alliance or other concepts relating to a therapeutic relationship have not yet been studied (58, 59).
It is worth noting that the authors of the studies examined in this review often come from the same research groups in Germany and Sweden, which may have an impact on the results. There appears to be a lack of diversity in researchers studying the therapeutic alliance online, which suggests a lack of interest in this topic and indicates that research into this area has not yet become widespread. Furthermore, caution should be exercised when interpreting the results of studies on (digital) psychotherapy because the country in which they were conducted is likely to influence the generalizability of the findings. This study’s findings may not be applicable to other countries or cultures because of the North European context and culture.
Most of the studies examined (n = 10) used the Working Alliance Inventory [WAI (24)] or its short form [WAI-S (50)] to gauge the therapeutic alliance. The average combined WAI and WAI-S score was 5.66, which was generally high (out of 7). According to the results of a study on the Working Alliance Inventory in face-to-face treatment, this indicates a good working alliance that is comparable to ratings in face-to-face treatment (50). Additional measures of the working alliance were used in four studies, as well. According to all of these measures, the working alliance had (far) above the midpoints of the various scales high scores.
However, only one study, out of the 38 examined, reported a mean score of 2.34 for the WAI’s revised short form (the WAI-SR, with a maximum score of 5). By week 9, however, the researchers’ additional scores did show an improvement in the working alliance ratings. The study’s focus on patients with chronic tinnitus and the majority of their participants being male made it stand out. Because of the lack of research on online treatment for this particular client group, it is unclear whether the relatively low score and late improvement in alliance quality are due to client characteristics, the specific treatment or other factors.
Damian Jacob Sendler
Some of the results were inconsistent when it came to the statistical correlation between the working alliance and treatment outcomes. Many studies have found moderate correlations between total working alliance scores or subscale scores and outcomes measures [e.g., (31)]. While a few studies didn’t find significant correlations with primary outcome measures [e.g., (28)], one study found small correlations (8) and another found strong correlations (47).
Because the therapeutic alliance has a direct correlation to treatment outcomes, text-based digital psychotherapy appears to be a good fit for this type of therapy. A previous meta-analysis of the alliance-outcome association in digital psychotherapy found a significant overall correlation of r = 0.28 (15), as well as a more recent meta-analysis on the same topic that found a significant overall correlation of r = 0.20, both of which support the finding that many studies found a significant and moderate relationship between the alliance and treatment outcomes (19). Due to the lack of research on this topic until recently (10), our review provides the necessary update on the working alliance in this evolving and growing field.. [Read more…] about Our Review. In the hope that our findings will help therapists feel more confident in their ability to form a functional working alliance in internet-based psychotherapy, prior research has found this confidence to be lacking in therapists (7, 17).
Our review shows that good therapeutic alliances can be established in text-based psychotherapies, especially for clients with anxiety, depression, or PTSD symptoms and CBT-based digital text-based treatments. Due to the lack of face-to-face interaction, the alliance does not appear to be jeopardized here. Digital text-based psychotherapy research is still in its infancy when compared to other client groups and treatment modalities. We must therefore exercise care when applying the results of this review to other client populations or treatment modalities. The working alliance in text-based digital treatment is expected to work similarly across the board, in line with findings on the internet in general, because we found no obvious differences for different client groups or treatment forms (15).
A quantitative assessment of therapeutic alliance quality, such as the WAI, may provide an overly narrow view of the therapeutic alliance when determining whether or not digital text-based treatments are a responsible option. Unlike the working alliance, which was typically measured at one point in time in most studies, the therapeutic alliance is a dynamic construct that changes over time and should be closely guarded by the therapist throughout the therapy process (20). As a result, experience sampling (60) or some other new method of studying the therapeutic alliance might be a good fit. In the therapeutic relationship, it is possible that experience sampling could provide more information about the natural ruptures that occur and the repair work that is required to restore the relationship to its original state. This is despite the fact that the WAI does not take these into account, despite the fact that ruptures are likely to occur in relationships based on reduced communication cues and responsiveness (10). There’s a chance that other, more nuanced aspects of the relationship, like self-disclosure or empathy, are currently being overlooked (61).
As an example, compassion between the therapist and the patient is an essential value in mental health care (62, 63). Recognizing suffering, understanding its universality, emotional resonance, coping with discomfort, and the desire to alleviate suffering are all components of compassion (64). The therapeutic alliance can be strengthened by the presence of compassion in treatment (65). (66). Digital treatments and therapeutic (text-based) relationships have not yet been studied for their ability to express and transfer compassion, and no scale exists to measure compassion in digital treatment forms.
The WAI, for example, has three subscales that assess the degree of agreement on goals and tasks between the client and the therapist as well as the therapeutic relationship (16, 24). Although the bond subscale is the closest to capturing the essence of compassion in therapy, its items are restricted to feelings of mutual liking, respect and appreciation. They don’t represent the full scope of compassion. As previously mentioned, the WAI dates back to 1989 (24), before digital treatment options were even considered. As a result, a new scale to assess whether or not a fundamental value such as compassion is lost in digital, text-based treatment when compared to face-to-face treatment is required.
