Psychotherapy

Clinical Impact Statement Question: Does therapeutic alliance mediate the relationship between antisocial personality traits and treatment outcome? Findings: We found that Personality Assessment Inventory scales, antisocial traits (ANT), treatment rejection, and patient-rated alliance were significantly related to patient-rated outcome (n = 65). Regression analysis indicated that both ANT (β = −.29, p = .014; f² = 0.09) and treatment rejection (β = −.28, p = .019; f² = 0.09) were unique predictors of outcome, explaining 18% of the variance. Adding patient-rated alliance to the model (β = .34, p = .003; f² = 0.13) increased the explained variance to 25%. Subsequent analyses revealed that the negative impacts of treatment rejection on outcomes occur separate from both ANT and alliance, while the impact of ANT is partially mediated by the alliance (indirect effect = −.10, lower level confidence interval = −.208; upper level confidence interval = −.003). Meaning: In terms of ANT endorsement on the Personality Assessment Inventory, “a little goes a long way” in its impact on decreasing treatment outcomes in the presence of a therapist who has more difficulty facilitating and maintaining strong therapeutic alliance. In our study, patients most at risk for poor outcomes had higher levels of both ANT and treatment rejection. When working with such individuals, therapists may need to direct efforts early in treatment to improve readiness by exploring patient expectations, collaborative goal setting, providing psychoeducation on the nature of therapy, enhancing problem recognition, and focusing patients on personal motives for change. Next Steps: Future research should focus on rupture and resolution as well as session-by-session interventions and subsequent symptom changes that occur with these patients to garner a clearer clinical picture of what variations throughout treatment may produce better overall outcomes for these patients.

Longitudinal evidence indicates struggles with spirituality/religion (S/R; e.g., feeling distant or abandoned by God, guilty for not attaining standards of moral perfection, judged by one’s family or community) may cause distress and hinder recovery from mental health challenges. However, research has not examined temporal associations between spiritual struggles and psychological distress in spiritually integrated psychotherapies (SIPs). In total, 175 practitioners of SIPs from 38 clinics and other settings in a practice-research network assessed these outcomes with 1,404 clients over the first 6 weeks of treatment (Time 1 [Weeks 1–2], Time 2 [Weeks 3–4], and Time 3 [Weeks 5–6]). In keeping with findings from other samples (e.g., Cowden et al., 2022, 2024; Currier et al., 2015, 2018), cross-lagged panel analyses across the three time points revealed spiritual struggles predicted clients’ psychological distress at the next interval rather than vice versa. Specifically, whereas psychological distress was not predictive of later spiritual struggles, clients who were struggling with their S/R at Time 2 were generally more psychologically distressed at Time 3. Further, the positive prospective association between spiritual struggles at Time 1 and psychological distress at Time 3 was mediated by the severity of spiritual struggles at Time 2. In combination, these findings affirm spiritual struggles drive many clients’ depression, anxiety, and other symptoms over the course of SIPs. As such, clinicians will ideally cultivate foundational knowledge and skills about the darker side of S/R that might prepare them to address their clients’ spiritual struggles throughout the treatment process when clinically indicated.

Operationalizing technical flexibility as session-to-session adjustments in therapist techniques, this study explored the context-dependent aspects of its effects on treatment outcomes. Specifically, we focused on how session-to-session adjustments in therapist techniques have different influences on subsequent client outcomes depending on the therapist-rated quality of the previous session. We used data collected from 3,970 sessions conducted by 17 doctoral student therapists in a university clinic, where 132 adult community clients received individual psychodynamic psychotherapy. The extents of session-to-session technique adjustment were measured by Euclidean distance scores representing the difference between techniques (i.e., explore, insight, and action skills) used in two adjacent sessions. Dynamic structural equation modeling was used to examine temporal relationships among client-rated therapist technical adjustments, therapist-rated previous session quality, and client functioning. When therapists evaluated their previous sessions as lower than usual in quality, moderate adjustments were associated with worse-than-usual client functioning in the current week, compared to sticking to previous techniques or making significant adjustments. When the therapist-rated previous session quality as higher than usual, more technical adjustments from the previous sessions were associated with better current week client functioning. The implications and limitations of the present study were discussed.

Meta-analyses indicate that patient outcome expectation (OE) and therapeutic alliance quality each correlate positively with improvement across diverse psychotherapies. Moreover, research demonstrates that these constructs relate to one another, including higher OE correlating with better subsequent alliance. However, notable gaps in this area remain. First, most studies that have examined the OE–alliance association have done so from the patient perspective only, despite both constructs having dyadic properties. Second, few studies have examined the connection between these two constructs longitudinally to capture their dynamic interrelations. Third, few studies have parsed OE–alliance links into their within- and between-dyad components, masking interpretability. Finally, most studies derive from controlled trials for specific diagnoses, limiting generalizability. Addressing these gaps, we tested the dyadic, dynamic, and statistically parsed associations among patient and therapist OE and alliance in naturalistic psychotherapy. Patients and therapists rated OE and alliance quality across up to 16 weeks of treatment. For dyads with the requisite data (N = 111 patients nested within 37 therapists), we used a multilevel actor–partner interdependence model that simultaneously tested all within- and between-dyad actor and partner effects of OE on alliance quality. At the within-dyad level, there were no significant actor or partner effects. At the between-dyad level, there were no significant partner effects, but there was a positive OE–alliance actor effect for both patients and therapists. Results suggest that community-based therapists may consider attending most to each participant’s average level of OE across therapy (vs. fluctuations) when forecasting its influence on their own overall relationship experience.

We designed and implemented a multimodal training program for clinicians in how to deliver spiritual psychotherapy for inpatient, residential, and inpatient treatment (SPIRIT), a group-based, spiritually integrated psychotherapy for acute psychiatric settings. The overall goals for this project were to facilitate competency in spiritually integrated psychotherapy among clinicians and meeting established needs for spiritual psychotherapy among patients presenting for acute psychiatric treatment. Our training program in SPIRIT includes the following elements: (a) a training manual, (b) 80-min training video, (c) brief competency exam, and (d) practice component involving the provision of SPIRIT to at least 12 groups of patients. This article provides an overview of the training program and describes its initial dissemination with a multidisciplinary cohort of 17 clinicians within the Massachusetts General Brigham health system, who collectively provided treatment to over 700 diagnostically, demographically, and religiously diverse patients. We also provide preliminary feedback from select clinicians about their experience in the training program, areas for future development, and implications for training clinicians in evidence-based spiritual psychotherapy.

The current literature for the various mental health professions is replete with studies demonstrating the critical importance of addressing spiritual and religious concerns in mental health treatment, yet many clinicians remain untrained in spiritual and religious competence. In recent years, researchers have developed training programs to support clinicians’ understanding of how spirituality influences mental health and provided resources for how clinicians can address it with clients. The present study investigates the effectiveness of two online training programs, Spiritual Competency Training in Mental Health (SCT-MH) and a contemplative pedagogy-based program called Multicultural Orientation and Contemplative Awareness (MOCA), in enhancing spiritual and religious competence among counselors. We conducted a three-parallel-arm randomized controlled trial with a sample of practicing counselors to compare the SCT-MH (n = 10), MOCA (n = 15), and a wait-list control group (n = 29). We recruited our sample from state licensure boards for mental health practitioners, alumni listservs for counseling programs, and social media marketing. Additionally, the study examined the explanatory role of cultural humility on spiritual competence over time. Linear mixed modeling revealed a significant interaction between group and time, demonstrating that the SCT-MH group had a large effect and MOCA had a moderate effect in improving clinicians’ total spiritual competence score compared with the control group. The findings suggest that both SCT-MH and MOCA can effectively enhance spiritual competence in counselors and cultural humility plays a significant role in this development. Implications and recommendations for implementation are discussed.

Childhood sexual abuse (CSA) affects approximately 30% of American children assigned female at birth. CSA often has a negative impact on survivors, leading to poorer mental health and an increased risk of subsequent sexual victimization. The current proof-of-concept study sought to examine the acceptability, feasibility, and fidelity of a peer-facilitated, online, one-time group intervention model to simultaneously promote recovery and prevent further victimization. We developed a brief intervention model drawing from components of previously tested protocols (i.e., expressive writing, interpersonal skills). Undergraduate students were trained as facilitators. We conducted online group sessions via Zoom with —three to seven participants per group (n = 40 in k = 10 groups). Participants were women (95.7%) or assigned female at birth nonbinary (4.3%) aged 17–30 who had experienced CSA. They were surveyed pre- and postintervention regarding acceptability. Qualitatively, feedback on the group intervention was very positive. Quantitatively, participants reported feeling more comfortable asking someone to stop sexual advances (M = 4.32/5.0), reported a desire to continue using skills learned (M = 4.56/5.0), and felt comfortable working with the peer providers (M = 4.63/5.0). From the participant-feasibility perspective, 100% of participants reported some level of postintervention skill practice during the one-week follow-up. The average, dichotomous, fidelity rating was 84.9%. Our findings indicate that an online group intervention facilitated by peers to improve health after CSA is acceptable, feasible, and meets minimal fidelity standards. Further development and refinement of this intervention model are warranted.

Despite the availability of evidence-based psychosocial and pharmacologic interventions for mental health concerns, access to care remains suboptimal. We present a time-efficient mental health treatment program designed to improve overall community access to evidence-based care. Quality of care within the time-efficient model was prioritized by focusing on a multifaceted program of evidence-based interventions and facilitating connections to long-term care when needed. We present the results of a proof-of-concept open trial that indicates that access to care can be improved while maintaining quality mental health services. The Time-Efficient, Evidence-Based, Accessible, Multidisciplinary approach includes time-limited care provided by a team of psychotherapists, psychiatrists, nurse practitioners, and case managers working in pods, supported by measurement-based care, to develop individualized treatment plans. We used data from the electronic health record to evaluate (a) access to care, (b) engagement, and (c) treatment outcomes for an initial sample of 1,726 patients. Patients waited on average 17 days to see a therapist and 20 days to see a psychopharmacologist. Patients received on average 10 sessions of psychotherapy and four medication management sessions. Fifty-seven percent remained in treatment for at least 3 months. Patients demonstrated large improvements in depression (d = 1.18) and anxiety (d = 1.20) with small to moderate effects for improvements in functioning (d = 0.40). By capitalizing on a package of evidence-based interventions delivered with equitable time limits, the Time-Efficient, Evidence-Based, Accessible, Multidisciplinary clinic demonstrates initial access while yielding good engagement in services and moderate to large treatment effects.

Deficits in emotional processing are considered to underlie a range of psychopathologies, and thus constitute a crucial target to improve mental health. Given the broad consensus on the importance of emotional processing during psychotherapy to reduce symptomology, it is surprising that there is no self-report scale designed to measure this construct. The purpose of this study was to develop the Emotional Processing Self-Report (EP-SR), a practical easy-to-use tool for assessing client emotional processing during therapeutic sessions. In Study 1, 53 items were developed according to theory and evaluated by experts. Exploratory factor analyses of the newly created EP-SR items were administered in a cross-sectional design to 124 individuals in treatment. In Study 2, multilevel exploratory factor analyses were performed on a sample of 114 patients who completed the EP-SR after each therapy session (1956 measurements). The results indicated that the EP-SR could be used to assess three independent lower order emotional processes: emotional expression, regulation, and understanding emotional meaning. In Study 3, multilevel confirmatory factor analysis was conducted on additional sample of 77 patients who completed the EP-SR after each therapy session (1,423 measurements). Results confirmed the factor structure of the EP-SR. The internal consistency, convergent, incremental, and predictive validity of the EP-SR were confirmed by showing its association with clients’ process and outcome measures. The results of these three studies support the reliability and validity of the EP-SR and its clinical importance. Given its user-friendliness and strong psychometric properties, the EP-SR emerges as a good instrument to assess individuals’ emotional processing.

To test the hypotheses that supervision improves therapeutic skills and therapeutic alliance (Hypothesis 1) and that this effect is stronger for immediate rather than for delayed supervisory feedback (Hypothesis 2). Eighty psychology students (M = 24.29 years; 72.5% identified as female; 75.0% were studying for a bachelor’s degree) conducted three consecutive sessions with a standardized patient in a simulated therapy. These students were randomly assigned to live supervision (LS) or delayed video-based supervision (DS). While the LS group received support from a supervisor during the second session, the DS group received such support after the second session. Several outcomes (students’ skills, self-efficacy, fear of negative evaluation, therapeutic and supervisory alliances, and supervision quality) were assessed from different perspectives (i.e., independent raters, students, patients). From all perspectives, supervision significantly increased students’ skills and self-efficacy, and the therapeutic alliance (d = 0.29–0.49, Hypothesis 1). Moreover, LS was not significantly superior to DS in any of the outcomes (Hypothesis 2). In addition, students subjectively perceived DS as more effective (d = 0.83, 95% confidence interval [0.37, 1.29]), but they identified the supervisory relationship as equally pleasant and the feedback as equally helpful in both conditions. Supervision positively influenced students’ skills and self-efficacy, and the therapeutic alliance, and both LS and DS were viewed as high quality. Thus, live and delayed supervision formats have comparable effects. However, future research should further investigate the specific mechanisms through which they exert their positive influences.

Deliberate practice has been introduced to psychotherapy training to improve therapist performance. The emerging empirical evidence supporting the positive impacts of deliberate practice is complicated by confusion regarding how deliberate practice should be operationalized and applied in the psychotherapy context and wider debate about its role in developing expertise. This systematic literature review aimed to appraise and synthesize the preliminary evidence available regarding the role of deliberate practice in improving psychotherapy outcomes. We searched MEDLINE, EMBASE, APA PsycInfo, CINAHL, Web of Science, reference lists of articles, and other sources (last checked June 2023). Quantitative English language studies that examined deliberate practice interventions in psychotherapy contexts were included. Two authors independently extracted study data and assessed quality using the relevant Joanna Briggs Institute critical appraisal checklist. Randomized controlled trials were further appraised using the Cochrane risk of bias tool. Discrepancies were resolved through discussion with a third author (Stephen Arthey). Twenty studies were included. There was wide variability in how deliberate practice had been operationalized, and only three studies met the contemporary definition that experts in the field have clarified. Significant methodological limitations impacted the reliability of studies, and the evidence was too preliminary to support or refute its use. Further research is required to establish whether deliberate practice is an effective means of improving psychotherapy outcomes. This should include increased fidelity to contemporary operationalizations, more robust methodology, and more reliable measures of long-term client outcomes.

Over the past 2 decades, a growing body of research has revealed that therapists differ in how effective they are at facilitating change depending on their clients’ social identities (see Imel et al., 2011, for example). Most studies examining the phenomenon of variability within therapists’ caseloads are based on the study of client race/ethnicity. However, two were recently published that found that therapists varied in the association between their clients’ sexual orientation and their therapy outcomes (Cabrera et al., 2023; Drinane et al., 2022). The present study sought to replicate and expand upon these findings in a national data set composed of 78,681 clients from university counseling centers who are part of the Center for Collegiate Mental Health. We employed multilevel modeling techniques to detect within-therapist disparities based on client sexual orientation on the following three outcomes: depression scores, an index of distress, and treatment dropout. The association between sexual orientation and all three outcome variables varied significantly across therapists. The presence of sexual orientation disparities within therapist caseloads in such a large and widespread sample is concerning and highlights a need for the analysis of other demographic variables and training that might mitigate this problem.

Subgroups in group therapy have long been considered an important concern. In this study, we examined how identity-based faultlines (i.e., the likelihood of splitting into subgroups based on member demographic information; Meyer & Glenz, 2013) would moderate the link between therapeutic factors and the reduction of psychological distress. Using data from 26 Chinese counseling groups comprising 141 group members (52.34% women, Mage = 22.41), we investigated the relations between identity-based faultlines (group level), early therapeutic factors (member level, measured at Session 2), and group members’ psychological distress (member level). Results from the two-level linear regression model indicated that being in a group with higher identity-based faultlines did not directly predict members’ symptom reduction. Furthermore, high faultline seemed to facilitate rather than impede group process, such that a member’s perceived helpfulness of the group process (therapeutic factors) in early group more strongly predicted symptom reduction in high-faultline groups and that members with higher levels of pregroup psychological distress reported more symptom reduction in high-faultline groups. Our findings provided new evidence that identity-based faultlines could also have positive effects on group therapy. We discussed the implications of identity-based faultlines in the Chinese context.

This article introduces a special section that addresses the need for culturally and structurally responsive training in psychotherapy. Nine articles explore how to prepare psychotherapists to address cultural and structural factors, in addition to individual and interpersonal factors, in their practice. Key themes include fostering a more critical form of cultural humility, building critical consciousness, and addressing oppressive systems within psychotherapy, supervision, the training environment, and accreditation. The articles also highlight the importance of advocacy for social or structural change, challenging psychotherapists to expand their scope to include systemic influences on client well-being. Together, these contributions provide actionable frameworks to prepare psychotherapists for antioppressive, equity-driven care.

The accessibility of training and fidelity assessment is critical to implementing and sustaining empirically supported psychotherapies like cognitive behavioral therapy for psychosis (CBTp). We describe the development of an online CBTp training tool that incorporates behavioral rehearsal tasks to enable deliberate practice of cognitive and behavioral techniques for psychosis. The development process consisted of designing content, inclusive of didactics, client profiles, and learner prompts; constructing standardized performance tasks and metrics; collecting responses to learner prompts; establishing intraclass correlation (ICC) of responses among trained raters; and training a transformer-based machine learning (ML) model to meet or surpass human ICC. Authenticity ratings of each simulated client surpassed benchmarks. CBTp trainers (n = 12), clinicians (n = 78), and nonclinicians (n = 119) generated 3,958 unique verbal responses to 28 unique prompts (7 skills × 4 simulated clients), of which the coding team rated 1,961. Human ICC across all skills was high (mean ICC = 0.77). On average, there was a high correlation between ML and human ratings of fidelity (rs = .74). Similarly, the average percentage of human agreement was high at 96% (range = 87%–102%), where values greater than 100 indicate that the ML model agreed with a human rater more than two human raters agreed with each other. Results suggest that it is possible to reliably measure discrete CBTp skills in response to simulated client vignettes while capturing expected variation in skill utilization across participants. These findings pave the way for a standardized, asynchronous training that incorporates automated feedback on learners’ rehearsal of CBTp skills.

The two articles by Kopelovich, Brian, et al. (2025) and Kopelovich, Slevin, et al. (2025) mark a new era in psychotherapy research and practice. The articles detail the development and validation of one of the first conversational artificial intelligence- (AI-) enhanced psychotherapy training tools, with profound implications for the future of clinical training. Following the new trail blazed by Kopelovich, Brian, et al. (2025) and Kopelovich, Slevin, et al. (2025), this commentary traces some of the most promising future directions for clinical training and research. In clinical training, trainees will be able to practice therapeutic skills and techniques with virtual clients before working with real ones. After mastering common therapeutic skills and treatment-specific techniques, they will begin treating real clients and receive detailed, immediate, and constructive AI-based feedback on their work to augment supervision sessions. Posttraining, clinicians can maintain and enhance their clinical expertise, acquire new skills, and incorporate the latest evidence-based knowledge into their practice through AI-based solutions. In research, it will be possible to explore the most effective techniques to be used by trainees and therapists at certain moments in a therapeutic session with individual patients, enabling the development of more precise and personalized therapeutic interventions. It will also be possible to explore the most effective trainee-specific supervision approaches to enhance a transformative experience and serve as a catalyst for the trainee’s professional identity development within the supervisor–supervisee relationship, augmented by a systematic mapping of the trainee’s strengths and areas for improvement.

This article reports program evaluation findings of the effectiveness of a pilot training program integrating key principles from the psychological framework for radical healing (French et al., 2020) and models aimed at decreasing internalized stigma (Israel et al., 2021). Our main goal of the training program was to increase psychotherapists’ knowledge and skills related to competent and affirming practice with two-spirit, transgender, and nonbinary clients who are Black, Brown, and people of color adults. Pre- and posttraining qualitative and quantitative data were collected from 82 psychotherapist attendees to assess the acceptability of the training for learning goals and to provide preliminary evidence of training effectiveness for the training’s learning objectives. Descriptive content analysis of qualitative data was used to assess the degree to which the training addressed psychotherapists’ goals and increased knowledge and skills. Regression analyses of quantitative data found significant increases in psychotherapists’ self-reported knowledge of key concepts and frameworks and confidence to apply new learning to psychotherapy practice. Program evaluation results largely support the acceptability and effectiveness of this training for increasing self-rated competency for psychotherapy practice with two-spirit transgender, and nonbinary Black, Brown, and people of color adult clients. Implications are discussed.

Interpersonal problems have been identified as a potential risk factor for a weaker therapeutic alliance during psychotherapy, yet their temporal relationship and underlying dynamics remain unclear. To address this, the present study explores these associations during the first 20 sessions of cognitive behavioral therapy. The data from 2,123 patients undergoing treatment in an outpatient clinic were analyzed. Interpersonal problems were assessed every fifth session with the 12-item version of the Inventory of Interpersonal Problems–12, while therapeutic alliance was measured after every session with the Session Rating Scale. Temporal associations were modeled using both a random-intercept cross-lagged panel model and an autoregressive latent trajectory model with structured residuals, as they allow the differentiation of within- and between-patient components. The random-intercept cross-lagged panel model produced more reliable and interpretable estimates. At the within-patient level, contemporaneous associations were significant, indicating that higher-than-usual interpersonal problems within a session were associated with lower-than-usual experienced therapeutic alliance at the same time point (β = −.067 to −.074, p = .005). Over time, higher-than-usual interpersonal problems negatively influenced therapeutic alliance at the next assessment (β = −.052 to −.063, p = .032), while higher-than-usual therapeutic alliance predicted reductions in interpersonal problems five sessions later (β = −.051 to −.083, p = .002). These findings suggest a reciprocal dynamic between interpersonal problems and therapeutic alliance, where improvements in one construct are associated with beneficial changes in the other over time, highlighting the importance of addressing interpersonal difficulties to strengthen the therapeutic alliance.

In this preregistered Prospero (CRD42022334071) review, the authors updated a meta-analysis review that examined the relation between therapist facilitation of patient emotional experience/expression and outcome in psychodynamic psychotherapy. Computer, manual, and backwards searches were conducted for relevant publications, and 14 independent samples of short-term dynamic psychotherapy were included in the meta-analysis. Data analysis included calculation of an overall effect size of the relationship between therapist affect focus and outcome, statistical significance, and test for homogeneity. In addition, moderator analyses were conducted to examine the potential impact of the methodological quality of individual studies. The overall weighted average r was .265, which was statistically significant, p < .001 (k = 14; 95% confidence interval [.130, .392]), indicating that therapist affect focus was associated with greater degree of patient change over the course of psychodynamic therapy. The results were not demonstrably heterogeneous, Q(13) = 14.787, p = .321, I² = 12.085, and publication bias analyses did not indicate cause for concerns regarding the results (all ps > .05). Trim-and-fill results indicated an adjusted weighted average r of .250 (decrease of 5.660%). None of the moderator analyses examining study methodology quality were statistically significant (all ps > .10). These data indicate that therapist facilitation of patient affective experience/expression is associated with patient improvement over the course of psychodynamic psychotherapy. The size of this relation was not significantly related to methodological quality.

The facilitative interpersonal skills (FIS) task is a performance-based task designed to assess clinicians’ capacity for facilitating a collaborative relationship. Performance on FIS is a robust clinician-level predictor of treatment outcomes. However, the FIS task has limited scalability because human rating of FIS requires specialized training and is time-intensive. We aimed to catalyze a “big needle jump” by developing an artificial intelligence- (AI-) based automated FIS measurement that captures all behavioral audiovisual markers available to human FIS raters. A total of 956 response clips were collected from 78 mental health clinicians. Three human raters rated the eight FIS subscales and reached sufficient interrater reliability (intraclass correlation based on three raters [ICC3k] for overall FIS = 0.85). We extracted text-, audio-, and video-based features and applied multimodal modeling (multilayer perceptron with a single hidden layer) to predict overall FIS and eight FIS subscales rated along a 1–5 scale continuum. We conducted 10-fold cross-validation analyses. For overall FIS, we reached moderate size relationships with the human-based ratings (Spearman’s ρ = .50). Performance for subscales was variable (Spearman’s ρ from .30 to .61). Inclusion of audio and video modalities improved the accuracy of the model, especially for the Emotional Expression and Verbal Fluency subscales. All three modalities contributed to the prediction performance, with text-based features contributing relatively most. Our multimodal model performed better than previously published unimodal models on the overall FIS and some FIS subscales. If confirmed in external validation studies, this AI-based FIS measurement may be used for the development of feedback tools for more targeted training, supervision, and deliberate practice.

Group therapy has evolved as a powerful therapeutic approach, facilitating mutual support, interpersonal learning, and personal growth among members. However, the complexity of studying communication dynamics, emotional expressions, and group interactions between multiple members and often coleaders is a frequent barrier to advancing group therapy research and practice. Fortunately, advances in machine learning technologies, for example, natural language processing (NLP), make it possible to study these complex verbal and behavioral interactions within a small group. Additionally, these technologies may serve to provide leaders and members with important and actionable feedback about group therapy sessions, possibly enhancing the utility of feedback-informed care in group therapy. As such, this study sought to provide a proof of concept for applying NLP technologies to automatically assess alliance ratings from participant utterances in two community-based online support groups for weight stigma. We compared traditional machine learning approaches with advanced transformer-based language models, including variants pretrained on mental health and psychotherapy data. Results indicated that several models detected relationships between participant utterances and alliance, with the best performing model achieving an area under the receiver operating characteristic curve of 0.654. Logistic regression analysis identified specific utterances associated with high and low alliance ratings, providing interpretable insights into group dynamics. While acknowledging limitations such as small sample size and the specific context of weight stigma groups, this study provides insights into the potential of NLP in augmenting feedback-informed group therapy. Implications for real-time process monitoring and future directions for enhancing model performance in diverse group therapy settings are discussed.

We developed an asynchronous online cognitive behavioral therapy (CBT) training tool that provides artificial intelligence- (AI-) enabled feedback to learners across eight CBT skills. We sought to evaluate the technical reliability and to ascertain how practitioners would use the tool to inform product iteration and future deployment. We conducted a single-arm 2-week field trial among behavioral health practitioners who treat outpatients with psychosis. Practitioners (N = 21) were invited to use the AI-enabled CBT training tool over a 2-week (15 days, inclusive) period. To enable naturalistic observation, no adjustments were made to their workloads nor were prescriptions on use provided. We conducted daily assessments and collected backend analytics for all users. At end point, we assessed acceptability, appropriateness, feasibility of implementation, perceived usability, satisfaction, and perceived impact of training. We observed four types of technical issues: broken links, intermittent issues receiving AI-enabled feedback, video replay errors, and an HTML error. Participants averaged 6.57 logins over the 2 weeks, with more than half engaging daily. Most participants (44.7%) engaged for <30-min increments. Usability scores exceeded industry standard and satisfaction scores indicated good promotion of the tool. All participants endorsed high feasibility, acceptability, and appropriateness. Twelve participants (57%) used the AI-enabled feedback feature; those who did tended to report improved satisfaction, feasibility, and perceived impact of the training. The training tool was used by practitioners in a routine care setting, met or exceeded conventional implementation benchmarks, and may support skill improvement; however, data suggest that practitioners may need support or accountability to fully leverage the training tool.

Complementing the oft-studied construct of interpersonal isolation, research has increasingly focused on existential isolation (EI), or the subjective feeling of separateness in one’s experience. In the clinical realm, several studies have demonstrated that higher EI is associated with more severe mental health problems at a single cross-section of time. Moreover, one study showed that higher pretreatment EI predicted worse psychotherapy outcomes. However, it remains unknown whether an average level of EI across all of psychotherapy (a type of during-treatment “trait”) and/or fluctuations in EI during psychotherapy (a type of during-treatment “state”) relate to broader treatment outcomes. Addressing EI in this more nuanced and complex manner, the present study parsed the EI–outcome association in its between- (trait) and within-patient (state) components in the context of naturalistic outpatient psychotherapy (Constantino et al., 2021). Participants were 46 therapists treating 144 patients who provided enough EI and outcome data to establish average, longitudinal, and temporal associations across treatment. Using a random intercept cross-lagged panel model, the between-patient results showed that higher average EI was associated with worse average outcomes (standardized association = .60, p < .001). However, although within-patient EI demonstrated significant variability over time, such fluctuations were unexpectedly unrelated to subsequent changes in outcome—when accounting for prior EI and outcome changes (p = .617). The findings suggest that although EI seems to possess both traitlike and statelike qualities, the former component may have the greatest influence on treatment outcomes, whereas the latter may be more of an outcome in itself.

A key mechanism of psychopathology and change proposed by the theory of interpersonal reconstructive therapy (IRT; Benjamin, 2003, 2018) is termed the “gift of love” (GOL). The GOL hypothesis is that wishes to receive love and acceptance from specific internalized attachment figures shape and maintain problem patterns and their associated symptoms for many patients across a wide range of psychopathology. According to IRT theory, optimal intervention is defined by therapist alignment, or “adherence,” to a core algorithm of principles that are tailored individually and bring awareness to (a) attachment-based yearnings for love and acceptance from internalized figures and (b) how those yearnings shape and motivate current problems and symptoms. The method then seeks to enhance choice about those relationships and their attendant feelings, hopes, and fears. The study sample includes 30 patients, referred while receiving inpatient treatment and followed on an outpatient basis, with complex, high-acuity clinical needs (i.e., histories of multiple psychiatric hospitalizations, recurrent suicidality, ineffectiveness of prior treatment, and significant personality pathology). Reliable measures were developed to track therapist adherence to IRT principles, as well as patient stages of change coming to terms with the GOL. IRT adherence was associated with retention, reduced depression and anxiety, and improved self-treatment. Mediation analyses support the proposition that change in IRT is contingent upon patient progress at grieving the losses and associated wishes linked to attachment figures. Implications for theory-guided research and practice at the level of underlying principles are discussed.

A number of studies have shown an association between therapist skills (particularly insight skills) and the working alliance, but few studies have examined the directionality of this relationship. In addition, studies have used either the client or therapist report of the working alliance rather than a dyadic perspective. Thus, we examined whether (a) dyadic insight skills are indirectly related to client outcome through the working alliance and (b) the dyadic working alliance is indirectly related to client outcome through insight skills. The dyadic working alliance was measured as a latent, dyadic average of both the client and therapists’ reports of the working alliance. Therapist use of insight skills was operationalized via the use of skills such as interpretations, immediacies, challenges, and disclosures of insight. We used dynamic structural equation modeling to analyze longitudinal data in long-term, psychodynamic treatment. Results indicated that at the within-client level, the use of insight skills was associated with the working alliance in the next session, and the working alliance was associated with therapist use of insight skills in the next session; however, there were no mediation effects. However, at the between-client level, the pathway from working alliance at T-1 to therapist use of insight skills at T-2 to client outcome at T-3 was significant but only for clients in longer term treatments. Findings reveal the importance of the working alliance as a signal for therapists to use insight skills for client improvement in long-term, psychodynamic therapy.