Outcomes of the Therapeutic Relationship in Treating Individuals Who Have Committed Crimes: A Systematic Review

Abstract

Several studies suggest that the quality of the therapeutic relationship (TR) between the patient and their therapist plays a significant role in predicting the outcomes of various forms of psychotherapy. However, it remains unclear whether the TR also influences the outcomes of treatment programs for justice-involved populations. Therefore, this systematic review aims to identify the impact of the TR on treatment outcomes with justice-involved individuals. A comprehensive search was conducted in B-on, Scopus, Web of Science, PubMed, PsycInfo, and Scielo for peer-reviewed quantitative studies providing data on individuals who committed crimes, available in English, Portuguese, or Spanish. A total of 18 studies that met the inclusion criteria were included in the final synthesis. Most of the included studies comprised predominantly male samples convicted of sexual offenses or intimate partner violence–related crimes. Regarding therapists, both genders were represented, along with a diversity of academic backgrounds. In terms of effectiveness, the findings suggest that the TR may exert a stronger influence on proximal, intrapersonal outcomes—such as motivation to change and well-being—than on more distal, behavioral outcomes, including aggression or recidivism. However, the existing studies are scarce and present several methodological limitations. Although the findings suggest the relevance of considering therapists and their training as potential responsivity factors in treatment, further research is needed to draw firm conclusions.

Keywords

therapeutic relationship individuals who have committed crimes psychological intervention systematic review

Introduction

Over the past few decades, numerous psychotherapies have emerged and evolved, including many evidence-based approaches supported by structured treatment manuals demonstrating effectiveness across a range of mental health conditions (e.g., Alkan et al., 2025; Liu et al., 2023). In this regard, the predominant approaches used with individuals who have committed crimes have also evolved over the years, showing positive effects (e.g., Beaudry et al., 2021; Cunha, Pereira, et al., 2024; Koehler & Lösel, 2024; Sousa et al., 2022; Sousa, Gonçalves, et al., 2025; Sousa, Gouveia, Cunha, et al., 2024). However, challenges such as treatment dropout and lack of patient response persist, and these difficulties are often even more pronounced within populations who committed crimes (e.g., Cunha, Pedrosa, et al., 2024; Pinto e Silva et al., 2024). The literature has shown that between 9% and 67% of individuals who have perpetrated intimate partner violence (IPV) drop out of treatment (Cunha, Pedrosa, et al., 2024), and 27.6% of individuals who committed sex crimes (Olver et al., 2011). To enhance adherence and outcomes, a deeper understanding of psychotherapy’s mechanisms is essential. One mechanism influencing treatment outcomes is the therapeutic relationship (TR).

Therapeutic Relationship

Although researchers differ in their conceptualizations of the TR, most theoretical definitions share three common themes: the collaborative nature of the relationship, the emotional bond between patient and therapist, and the mutual agreement on treatment goals and tasks (Bordin, 1979; Martin et al., 2000; Wampold, 2015). Other terms are often used interchangeably with this concept, such as therapeutic alliance, working alliance, or helping alliance (Catty et al., 2007). However, the term TR will be used predominantly in this work, as it is broader and more inclusive (Catty et al., 2007; Gelson & Kline, 2019; Norcross & Lambert, 2011).

The quality of the interaction between client and therapist constitutes a key determinant of the therapeutic process (e.g., Crits-Christoph & Gibbons, 2021). Historically, this idea emerged in Freud’s (1912/1958) works, where he emphasized the importance of the personal and conscious bond between patient and analyst, while also highlighting the role of the unconscious and transference—the projection of unresolved past relationships onto the therapist. Years later, Rogers et al. (1967) contributed to the development of work about TR by applying empirical methods to study the facilitative conditions of the TR—such as empathy, positive regard, and authenticity—paving the way for the scientific investigation of relational factors. In his seminal work, Bordin (1979) described the TR as being defined by the feelings and attitudes that both the therapist and the client have toward one another and the way these are expressed.

Currently, evidence from several empirical studies, systematic reviews, and meta-analyses conducted with adult populations indicates a consistent and significant link between the TR and treatment outcomes, showing that a stronger TR is associated with more positive therapeutic results (e.g., Cameron et al., 2018; Flückiger et al., 2018; Martin et al., 2000). Specifically, Flückiger et al. (2018) conducted a comprehensive meta-analysis comprising 295 independent studies examining the link between TR and treatment outcomes in face-to-face therapies. Their findings revealed a moderate positive association between TR and symptom improvement. Several potential moderators—such as diagnosis, therapy type, methodological factors, and study context—were tested, but only the timing of alliance assessment significantly influenced results, with stronger correlations observed when TR was measured later in treatment (Flückiger et al., 2018). In a follow-up study, Flückiger et al. (2020) confirmed that this TR–outcome association persisted, even after accounting for patient characteristics and therapist adherence or competence.

The literature has also emphasized that the TR is a reciprocal one—alliance influencing subsequent symptoms and symptoms shaping the quality of the alliance (Flückiger et al., 2020)—as well as dynamic and subject to change over time (e.g., Prusiński, 2024). Specifically, some studies have shown that the TR tends to strengthen as therapy progresses (e.g., Piper et al., 2005; Wu et al., 2024), whereas others have found that it follows a U-shaped pattern: a strong alliance in the early phase, a weaker one during the middle phase, and a renewed strengthening toward the end of therapy (e.g., Luo et al., 2022).

Considering the relevance of the TR in treatment outcomes, as well as its variation throughout the therapeutic process, comprehensive models have been developed to better understand this phenomenon. Ribeiro et al. (2013) proposed the Therapeutic Collaboration Model, which conceptualizes therapeutic collaboration as a joint process in which therapist and client work together within the so-called Therapeutic Zone of Proximal Development (TZPD). This concept, adapted from Vigotsky’s (1978) original framework and introduced to psychotherapy by Leiman and Stiles (2001), represents the gap between what clients can currently manage independently and what they can achieve with therapeutic guidance (Ribeiro et al., 2013, 2021). Because clients’ abilities evolve over the course of therapy, both members of the dyad continuously negotiate and adjust their work within this zone. The therapist assesses whether interventions fall inside or outside the client’s TZPD by observing how the client responds—either validating (accepting) or invalidating (rejecting) the therapist’s input. As clients make progress, their TZPD gradually expands upward, reflecting developmental growth. To promote this progress, therapists alternate between supportive interventions, which consolidate the client’s existing understanding, and challenging interventions, which introduce new perspectives to foster change. The degree of collaboration is inferred from clients’ reactions: validation signals that the interaction is within the TZPD, invalidation indicates work beyond it, and mixed reactions suggest the dyad is operating near its boundary (Ribeiro et al., 2013, 2021).

TR Within the Forensic Context

While TR has been extensively studied in clinical settings, research within forensic contexts, particularly concerning individuals who have committed crimes, remains scarce (e.g., Blasko & Jeglic, 2016; Blasko et al., 2015; Kozar & Day, 2012; Polaschek & Ross, 2010; Santirso et al., 2020).

Developing a TR with individuals who have committed crimes can be challenging (Blasko et al., 2018). First, some crimes evoke negative attitudes even among therapists, as is the case with sexual offenses (e.g., Hardeberg Bach & Demuth, 2018), which influences the development of a TR. More specifically, a study by Sousa, Félix, et al. (2025) found that some participants, prior to taking part in an intervention program, feared they would not be understood because of previous negative experiences with therapists. Second, the fact that intervention with these individuals is often court-mandated may make the development of an RT more difficult (e.g., Hachtel et al., 2019). In this way, research has shown that the development of a strong TR is grounded in a “firm but fair” approach, combining care and fairness with an authoritative style. This relationship is further strengthened when professionals promote interactions based on respect, dignity, fairness, the opportunity for individuals to express themselves, validation of their perspective, clear information, and a genuine concern for their well-being (Hachtel et al., 2019; Wittouck & Vander Beken, 2019; Stinson & Clark, 2017). In fact, research on therapist characteristics in forensic treatment has demonstrated that therapists who display empathy, warmth, and directiveness, as well as reward their clients’ good behavior, have been found to impact changes in the client-therapist relationship with those individuals (Blasko et al., 2018). Nevertheless, the available evidence suggests that even populations considered more challenging can develop positive and robust TRs. This is exemplified by a study by DeSorcy et al. (2016), which examined TR and its associations with psychopathy, treatment completion, and recidivism in a correctional context. Participants with both high and low levels of psychopathy reported strong working alliances with their therapists. Notably, even those with elevated psychopathy scores demonstrated the capacity to establish effective TRs within the context of treatment for sexual offenses (DeSorcy et al., 2016).

Nevertheless, existing studies suggest that the TR is important in interventions with individuals who have committed crimes, significantly influencing treatment outcomes (Blasko et al., 2018). In addition, available evidence indicates that a strong therapeutic alliance plays a key role in reducing treatment dropout rates (Brocato, 2013).

Current Study

Considering the well-established influence of the TR on treatment outcomes in clinical settings (e.g., Cameron et al., 2018; Flückiger et al., 2018), and given that, to date, only one systematic review exists in forensic contexts—focusing exclusively on juvenile populations and treatment-related outcomes (Papalia et al., 2022)—there is currently no systematic synthesis examining this association among justice-involved adults across diverse settings.

Moreover, existing research has not fully examined the impact of the TR on outcomes of interest, such as clinical measures and recidivism. The systematic review of justice-involved youth highlighted the role of treatment readiness and expectations in shaping the TR (Papalia et al., 2022), but did not attempt to examine how these factors relate to key outcomes, including clinical improvement or recidivism.

The present systematic review, therefore, focuses on adult males and females who perpetrated crimes across diverse settings (prisons, psychiatric hospitals, and community programs) and includes both clinically relevant outcomes and recidivism. In doing so, it provides a comprehensive synthesis of how the TR relates to key outcomes in adult forensic populations, addressing important methodological and conceptual gaps in the literature and offering insights relevant for clinical practice and intervention planning. Specifically, this systematic review seeks to investigate how the TR relates to recidivism reduction, improvements across various therapeutic domains, and treatment dropout. Furthermore, the review adopts a broad scope by including studies conducted in both custodial (prison) and non-custodial (community and/or forensic hospitals) settings, where men and women who committed crimes are engaged in therapeutic interventions.

Methods

Protocol and Registration

This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Page et al., 2021), and was pre-registered on OSF REGISTRIES (reference: 10.17605/OSF.IO/4Z7TN).

Eligibility Criteria

Studies were considered eligible for inclusion if they met the following criteria: (a) involved adult participants (male, female, or mixed samples); (b) were empirical and used a quantitative method; (c) provided data on individuals who have perpetrated crimes and were receiving services to support their mental health and/or prevent recidivism; (d) used a validated measure of therapeutic alliance; (e) examined the therapeutic alliance between the client and a mental health professional in the context of a TR, even if the intervention was not exclusively individual therapy (e.g., psychosocial interventions, group therapies); (f) were published in peer-reviewed journals; (g) were written in English, Portuguese, or Spanish; and (h) were accessible online. This review adopted a broad scope by including both forensic interventions aimed at reducing offending-related risk and psychological interventions targeting mental health difficulties in people who have offended, given the recognized role of psychopathology in offending trajectories and the increasing emphasis on outcomes beyond recidivism alone (see Lutz et al., 2022; Willis & Ward, 2024). Studies were excluded if they met any of the following criteria: (a) were review studies, including systematic reviews, meta-analyses, scoping reviews, psychometric studies, theses, or research protocols; and (b) were case reports.

Search Strategy

The initial search was conducted in February 2025, and an additional search was performed in October 2025, without any date restrictions, by two independent researchers. The following search equation was applied to identify relevant studies: (“therap* relationship” OR “working relationship” OR “helping relationship” OR “therap* alliance” OR “therap* collaboration” OR “helping alliance” OR “working alliance”) AND (offend* OR aggres* OR batterer OR perpetrat* OR “criminal*” OR “felon*” OR “Inmate*” OR “Incarcerat*” OR “Prison*” OR “Sentenced” OR “Detainee*” OR “forens*”) AND (treatment OR intervention OR program* OR rehab* OR therapy). The search was conducted across five electronic databases: B-on, Scopus, Web of Science, PubMed, PsycInfo, and Scielo, focusing on titles, abstracts, and keywords. In addition, the reference lists of relevant and key studies were reviewed to identify any further studies that might meet the inclusion criteria.

Literature Selection Process and Data Extraction

Studies identified through the equation-based search were imported into Rayyan software (Ouzzani et al., 2016), where duplicate entries were removed. Two independent reviewers then screened titles and abstracts against the eligibility criteria, selecting articles for full-text analysis. Any discrepancies were discussed until a consensus was reached, with a third reviewer available if needed.

For the selected articles, a structured codebook was developed to extract relevant data. The development of the coding categories followed a hybrid approach, informed by the objectives of the review and by existing literature. Specifically, categories were defined based on prior research highlighting the relevance of these aspects in the TR, as well as on existing systematic reviews in the field of clinical psychology. Coding was limited to information explicitly reported in the articles; no inferences were made when data were not clearly stated. When relevant information was missing or ambiguously described, it was coded as “not reported.” Specifically, when studies reported multiple intervention components or therapeutic approaches, all modality identified by the authors was coded. Similarly, when more than one therapeutic alliance measure was used, all reported instruments were recorded.

The codebook including reference details (e.g., authors and year of publication); study characteristics (e.g., country of the study, sample size; inclusion and exclusion criteria, study setting), participant demographics (e.g., age, gender; ethnicity; type of crime perpetrated); therapist characteristics (e.g., number of therapists, age, gender, professional qualifications, years of service), intervention details (e.g., type of therapy [e.g., cognitive-behavioral therapy, schema therapy], number of sessions, length of sessions; intervention setting [e.g., prison, community]), therapeutic alliance measure (the instrument used in each study to assess TR); methodological details (e.g., study design, follow-up period); and outcome measures relevant to the review’s objectives. To ensure accuracy, extracted data were cross-checked between the reviewers. All extracted data were cross-checked between the reviewers independently, and any disagreements were resolved through discussion until consensus was reached or, when necessary, by consulting a third reviewer.

Quality Assessment

The methodological quality of the included studies was evaluated using the Mixed Methods Appraisal Tool (MMAT; Hong et al., 2018). This tool assesses different study designs (e.g., qualitative research, randomized controlled trials, non-randomized studies, quantitative descriptive studies, and mixed-methods studies). Each study was screened using two preliminary questions (“Are there clear research questions?” and “Do the collected data allow to address the research questions?”), followed by five criteria rated as “yes,” “no,” or “don’t know,” according to methodological rigor. The five quality criteria, adapted to the study design, focused on the population (e.g., representativeness), participant recruitment (e.g., appropriate randomization), measurement tools (e.g., suitability for outcomes and interventions), completeness of data, and data analysis (e.g., control of confounders). Consistent with previous steps, two independent reviewers conducted the quality assessment. Discrepancies were addressed through discussion, with a third reviewer consulted when necessary. No study was excluded from the present review based on its quality level, in accordance with the recommendations of the scale itself (Hong et al., 2018).

Data Analysis

This study was conducted as a systematic review, with meta-analytic quantitative pooling performed when sufficient and comparable data were available across the included studies. The data were analyzed using a narrative synthesis approach, where both descriptive text and tables were employed to provide a concise summary of the collected data. This method enables reviewers to assess the findings, considering variations in study designs and potential biases across the studies included in the systematic review.

Results

Screening and Selection of Studies

A total of 3,922 articles were retrieved from database searches and supplementary sources. Following the removal of duplicates, 1,499 titles and abstracts were screened for eligibility. From these, 77 studies were retained for detailed evaluation. Fifty-nine studies were excluded as they did not satisfy the inclusion criteria. Specifically, most studies were excluded because they did not relate TR to an indicator of intervention effectiveness (n = 37). In addition, 12 studies were excluded for not being quantitative empirical articles, 4 for not being written in English, Portuguese, or Spanish, and 2 for not being accessible online. Ultimately, 18 studies were included in the final review. Figure 1 illustrates the PRISMA diagram, detailing study inclusion across all screening phases and reasons for exclusion.

Figure 1.

PRISMA flow diagram of the study selection process.

Quality Assessment

Of the included articles, the majority employed quantitative non-randomized designs (n = 15), while the remainder consisted of randomized controlled trials (n = 2), and one was a mixed-methods study (n = 1; see Table 1). Two studies met four out of five criteria of excellent (DeSorcy et al., 2016; Singh et al., 2024), eight met three out of five criteria (Bélanger & Higgs, 2024; Blasko & Jeglic, 2016; Brocato, 2013; DeSorcy et al., 2020; Miles-McLean et al., 2019; Taft et al., 2003; van Tilburg et al., 2025; Walling et al., 2012), five studies met two out of five criteria (Brocato & Wagner, 2008; Brown & O’Leary, 2000; Coleman et al., 2024; Olver et al., 2024; Scanlon et al., 2022), and two studies met one out of five criteria (Polaschek & Ross, 2010; Watson et al., 2018). One study did not meet any of the excellence criteria (Clercx et al., 2021). The majority of the included studies present samples that are not representative of the population under study or omit information regarding the procedure, namely, whether the intervention was administered (or the exposure occurred) as intended. In some cases, the studies include instruments without providing information on their psychometric properties.

Table 1.

Characteristics of the Included Studies.

Study	Design	Sample characteristics (n, M age, gender, type of crime)	Country	Treatment characteristics (treatment program/models; setting)	Professional characteristics (number of therapists, age, gender, profession/qualification(s), length of experience)	TR measures
Bélanger and Higgs (2024)	Quantitative non-randomized	223, 47 years, male, child sexual abuse	Canada	RP + CBT + other components, correctional institutions	n/a	California Psychotherapy Alliance Scale (Client self-reported)
Blasko and Jeglic (2016)	Quantitative non-randomized	202, 40.87 years, male, sex offenses	United States	CBT, correctional institution	10, 39.94 years, both psychologists with doctorate degrees (n = 2)/psychiatrist (n = 1)/masters-level clinicians (n = 7), minimum of 3 years	Working Alliance Inventory—Client Form (Client self-reported)
Brocato and Wagner (2008)	Quantitative non-randomized	141, n/a, male, drug offenses	United States	n/a, community	6, n/a, n/a, master’s degree–level clinicians, minimum of 2 years	Working Alliance Inventory (Therapist self-reported)
Brocato (2013)	Quantitative non-randomized	43, 28.6 years, male, drug offenses	United States	n/a, community	n/a	Working Alliance Inventory (Client self-reported)
Brown and O’Leary (2000)	RCT	70,^a 37.56 years, male, intimate partner violence	n/a	CBT + psychoeducation, community	n/a, n/a, both, n/a, n/a	Working Alliance Inventory Observer (Observer-rated)
Clercx et al. (2021)	RCT	103, 37.7 years, male, different types of crimes	Netherlands	ST, forensic hospitals	n/a	Working Alliance Inventory (Client and therapist self-reported)
Coleman et al. (2024)	Quantitative non-randomized	94, 35.4 years, both, n/a	United States	n/a, correctional institutions	20, 25.7 years, both, psychologists (41%)/ clinical mental health counselors (55%)/international disaster counselors (4%), M years of experience = 1.83	Working Alliance Inventory—Short Revised (Client self-reported)
DeSorcy et al. (2016)	Quantitative non-randomized	423, 36.8 years, male, sexual offenses	Canada	Clearwater Sex Offender Program (RNR + CBT), forensic hospitals	n/a, n/a, n/a, multidisciplinary team (nurses, psychologists, psychiatrists, social workers, and parole officers)	Working Alliance Inventory (Client self-reported)
DeSorcy et al. (2020)	Quantitative non-randomized	111, 37 years, male, sexual offenses	Canada	Clearwater Sex Offender Program (RNR + CBT), forensic hospitals	n/a, n/a, n/a, multidisciplinary team (nurses, psychologists, psychiatrists, social workers, and parole officers)	Working Alliance Inventory (Client self-reported)
Miles-McLean et al. (2019)	Quantitative non-randomized	293, 36.5 years, male, intimate partner violence	United States	CBT + MT, community	2, n/a, n/a, clinical psychology (1)/agency staff person (1)	Working Alliance Inventory—Short (Client and therapist self-reported)
Olver et al. (2024)	Quantitative non-randomized	317, 37.4 years, male, sexual and non-sexual offenses	Canada	CBT; forensic hospitals	n/a, n/a, n/a, multidisciplinary team (psychiatric nurses, psychologists, psychiatrists, occupational therapists, social workers, parole and correctional ofﬁcers)	Working Alliance Inventory (Client self-reported)
Polaschek and Ross (2010)	Quantitative non-randomized	70, 31 years, male, violent crime	New Zealand	CBT, correctional institutions	11, 35 years, both, clinically or forensically psychologists (5)/diverse backgrounds in rehabilitation work (6), M years of experience = 5.2	Working Alliance Inventory—Short Form (Client, therapist, and observer reported)
Scanlon et al. (2022)	Quantitative non-randomized	265, 35.51 years, both, n/a	United States	Psychosocial, social learning, CBT, community	n/a	Working Alliance Inventory Client Version (Client self-reported)
Singh et al. (2024)	Mixed-Method study	EERSS intervention: n = 17, 40 years, both, n/a; WiP intervention: n = 17, 43 years, both, n/a	United Kingdom	EERSS, correctional institutions	n/a	Working Alliance Inventory—Short Form (Client self-reported)
Taft et al. (2003)	Quantitative non-randomized	107, 36.22 years, male, intimate partner violence	United States	CBT, community	13, n/a, both, at least one of the therapists held a master’s degree in psychology or equivalent training within a doctoral program (2 or more years), and had conducted at least three previous treatment groups with this population under supervision by a doctoral-level psychologist	Working Alliance Inventory (Client and therapist self-reported)
van Tilburg et al. (2025)	Quantitative non-randomized	199, 39.1 years, male, intimate partner violence	Netherlands	“Not Losing it Anymore” (CBT), forensic hospitals	30, n/a, both, multidisciplinary team (psychologists, social workers, and psychiatric nurses); <5 years of experience = 3; 6–10 years of experience = 8; 11–15 years of experience = 4; >15 years of experience = 15	The patient-rated alliance scale: Session Rating Scale; The therapist rated alliance scale: Therapist Satisfaction Scale (client and therapist self-reported)
Walling et al. (2012)	Quantitative non-randomized	107, 36.3 years, male, intimate partner violence	United States	CBT + MI, community	15, n/a, both, graduate students enrolled in a Clinical Psychology doctoral program (14)/licensed psychologist (1), n/a	Working Alliance Inventory + therapist short form (Client and therapist self-reported)
Watson et al. (2018)	Quantitative non-randomized	71, 45.06 years, male, sexual offenses	Australian	n/a, correctional institutions and community	n/a	Working Alliance Inventory Short-Form (Client and therapist self-reported)

Note. RCT = randomized controlled trial; TR = therapeutic relationship; RNR = risk-need-responsivity; EERSS = Enhanced Engagement Relational Support Service; RP = Relapse Prevention; CBT = Cognitive Behavioral Therapy; n/a = not applicable; ST = Schema Therapy; WiP = Women in Prison; MI = Motivational Interview.

The study included couples; however, since the intervention targeted the man, only information regarding the male participant was reported.

Study Characteristics

The publication years of the included studies ranged from 2003 (Taft et al., 2003) to 2025 (van Tilburg et al., 2025). The year 2024 had the highest number of publications (n = 3), followed by 2016 (n = 2; Blasko & Jeglic, 2016; DeSorcy et al., 2016; see Table 1). The majority of the studies were conducted in the United States (n = 8), followed by Canada (n = 4; see Table 1). The remaining studies were conducted in the United Kingdom, the Netherlands, Australia, and New Zealand. Only one study did not explicitly report this information (Brown & O’Leary, 2000).

The sample sizes included in this review range from 34 (Singh et al., 2024) to 317 participants (Olver et al., 2024), with mean ages varying between 28.6 years (Brocato, 2013) and 47 years (Bélanger & Higgs, 2024). Regarding the samples, most studies (n = 15) included only men, while the remaining ones consisted of participants of both sexes. Concerning the crimes committed, there is a greater representation of studies with participants convicted of sexual offenses (n = 5) and IPV (n = 5; see Table 1).

Intervention and Therapist Characteristics

Nine studies applied a cognitive-behavioral program, although some interventions included additional theoretical components such as relapse prevention, motivational interviewing, and the risk-need-responsivity (RNR) model (see Table 1). One study applied schema therapy (Clercx et al., 2021). Some studies were even more specific, mentioning the intervention program used, such as the Clearwater Sex Offender Program (DeSorcy et al., 2016, 2020), the Sexual Offense Treatment Program (SOTP)/Clearwater Program (Olver et al., 2024), the “Te Whare Manaakitanga” (Polaschek & Ross, 2010), the Changing Lives and Changing Outcomes Program (Scanlon et al., 2022), the Enhanced Engagement Relational Support Service (EERSS; Singh et al., 2024), and the “Not Losing It Anymore” (van Tilburg et al., 2025). Four articles did not specify the name of the intervention or the therapeutic approach used (Brocato, 2013; Brocato & Wagner, 2008; Coleman et al., 2024; Watson et al., 2018). Seven programs were implemented in the community, five in prison settings, five in forensic hospitals, and one was implemented both in prison settings and in the community.

The number of professionals implementing the intervention programs ranged from 2 (Miles-McLean et al., 2019) to 30 (van Tilburg et al., 2025), with 6 studies not reporting any information regarding the professional team. All seven studies that reported the gender of these professionals included both genders. The average age of these professionals ranged from 25.7 years (Coleman et al., 2024) to 39.94 years (Blasko & Jeglic, 2016). The basic training of these professionals varied across the included articles, involving intervention with a multidisciplinary team with psychologists, psychiatrists, nurses, staff members of institutions such as parole officers, and social workers (e.g., Blasko & Jeglic, 2016; Coleman et al., 2024; DeSorcy et al., 2016, 2020; Miles-McLean et al., 2019; Olver et al., 2024; Polaschek & Ross, 2010; van Tilburg et al., 2025), and psychologist and/or psychologist students (e.g., Brocato & Wagner, 2008; Taft et al., 2003; Walling et al., 2012). Regarding years of experience, only two studies reported the average number of years, which ranged from 1.83 years (Coleman et al., 2024) to 5.2 years (Polaschek & Ross, 2010). On the other hand, the study by Brocato and Wagner (2008) mentioned that the professionals had at least 2 years of experience, and Blasko and Jeglic (2016) reported a minimum of 3 years of experience. The van Tilburg et al. (2025) study included therapists with varying levels of experience: 3 with fewer than 5 years, 8 with 6 to 10 years, 4 with 11 to 15 years, and 15 with more than 15 years of professional experience.

Instruments Used to Assess TR

Most studies (n = 16) used the Working Alliance Inventory (WAI; Horvath, 1981, 1982) or the short version of WAI (Hatcher & Gillaspy, 2006). Within these studies, eight used only the version of the scale focused on the client’s perspective; five used both the client and therapist versions; one used only the version reflecting the therapist’s view of the TR (Brocato & Wagner, 2008); one used only the observer-rated version (Brown & O’Leary, 2000); and one used all three versions of the scale (Polaschek & Ross, 2010; see Table 1).

One study used the California Psychotherapy Alliance Scale (Bélanger & Higgs, 2024), which allows the assessment of the client’s perspective. Another study used two scales to obtain information from both clients and therapists: the patient-rated alliance scale—the Session Rating Scale, and the therapist-rated alliance scale—the Therapist Satisfaction Scale (van Tilburg et al., 2025).

Intervention Outcomes

Recidivism and risk of recidivism were the most common outcomes (n = 6), followed by retention/ dropout/treatment non-completion (n = 5), aggression/institutional incidents/abusive behavior (n = 4), motivation to change (n = 3), and distress and well-being (n = 3). Four studies focused on miscellaneous problems such as marital satisfaction, alcohol abuse, criminal thinking, and treatment gain (see Table 2).

Table 2.

Conclusions of the Included Studies.

Study	Type of outcome	Outcomes
Bélanger and Higgs (2024)	Motivation to change	Significant weak positive relationship between TR and contemplation; Significant moderate positive relationship between TR and action; Significant moderate negative relationship between TR and precontemplation.
Blasko and Jeglic (2016)	Risk of recidivism	The Bond Subscale of TR was significantly correlated with risk, r = −.27, p < .001. Risk scores were not significantly related to total TR. The relationship between risk scores and TR depends on gender.
Brocato and Wagner (2008)	Motivation to change	The bond Subscale of TR were significantly associated with changes in motivation (β = .289, p < .05).
Brocato (2013)	Retention and motivation to change	TR demonstrated no predictive power for retention. Significant and positive relationship between TR and motivation scores ate 90 days (r = .51, p = .002).
Brown and O’Leary (2000)	Dropout, aggression and marital satisfaction	TR was not related to dropping out of treatment. The individuals’ level of TR significantly predicted reductions in mild and severe psychological aggression and in mild and severe physical aggression after treatment. TR did not predict marital satisfaction.
Clercx et al. (2021)	Institutional incidents (verbal aggression, threats, physical aggression, and violation of hospital rules)	No effects TR on incidents during the first 18 months. Models including covariates, incidents in the first 18 months, and TR and motivation measured at the start significantly predicted incidents during 18–36 months.
Coleman et al. (2024)	Distress	TR was significantly associated with psychological distress (b = −0.43, p < .01; ES = −0.36), and relationship distress (b = −0.41, p < .01; ES = 0.38) at the end of the treatment.
DeSorcy et al. (2016)	Completion, length of program stays and recidivism	TR total score and TR subscales scores were negative and significantly correlated with treatment non-completion. TR total scores and Bond and Goal subscales were significantly related to length of stay in program. TR scores did not predict any recidivism.
DeSorcy et al. (2020)	Treatment noncompletion and recidivism	TR did not predict noncompletion (b = 0.07; p = .155; Wald = 2.02) and sexual recidivism (b = 0.01; p = .804; Wald = 0.06).
Miles-McLean et al. (2019)	IPA and GV Recidivism; Alcohol use	Late client perception of TR and therapist perception of TR were not significantly associated with IPA and GV recidivism. Alcohol use was not associated with TR.
Olver et al. (2024)	Recidivism	TR scores were significantly associated with increased sexual recidivism but not with violent recidivism.
Polaschek and Ross (2010)	Risk factors	There is a small-to-medium positive relationship between change in TR and behavioral change on dynamic risk factors for violence (r = .32, p < .01).
Scanlon et al. (2022)	Mental Illness and Criminal Thinking	TR significantly moderates reductions in anxiety, manic, anger, and depression symptoms and general, reactive, and current criminal thinking. TR was not a significant moderator of change in suicidal ideation, psychosis, or substance use.
Singh et al. (2024)	Well-being	Significant positive association between TR and well-being (b = 0.39, 95% CI [0.07, 0.72], p = .020).
Taft et al. (2003)	Abuse Behavior	Therapists’ ratings of the TR predicted lower levels of physical and psychological abuse at the 6-month follow-up and emerged as the strongest predictors of treatment outcomes.
van Tilburg et al. (2025)	Treatment noncompletion	Therapists’ assessments of the TR were strong predictors of both session attendance and treatment completion, while patients’ own TR ratings showed no predictive value for either outcome.
Walling et al. (2012)	Physical Aggression	The TR’s average improvement does not determine how much violence decreases. For White participants, the TR trajectory did not influence the reduction in violence. For minority participants, an increase in TR was associated with a robust reduction in violence (similar to White participants). However, if the alliance did not increase, minority participants showed worse outcomes, with even a slight increase in violence at follow-up.
Watson et al. (2018)	Treatment gain	Late offender-rated TR were positively correlated with treatment gain (b = 0.30). Late offender-rated therapeutic alliance did not predict Treatment Gain (β = .17, t(43) = 1.12, p = .27).

Note. IPA = intimate partner abuse; GV = general recidivism; TR = therapeutic relationship.

Recidivism and Risk of Recidivism

The findings across studies were inconsistent regarding the relationship between TR and risk: Blasko and Jeglic (2016) reported that TR is not correlated with risk of recidivism, whereas the study by Polaschek and Ross (2010) showed the opposite. Specifically, their study found a positive small-to-medium correlation (r = .32).

Regarding recidivism, the studies showed that TR is not correlated with reoffending for any type of crime (DeSorcy et al., 2016, 2020; Miles-McLean et al., 2019), regardless of whether TR was assessed by the client or the therapist (Miles-McLean et al., 2019). However, this absence of correlation does not apply to TR subscales. For example, the Bond subscale was significantly negatively correlated with risk, regardless of the small effect sizes, r = −.27, p < .001 (Blasko & Jeglic, 2016). Curiously, Olver et al. (2024) reported that TR was actually associated with an increased risk of sexual recidivism (Harrell’s C = 0.62).

All programs were based on a cognitive-behavioral approach; however, no differences were found in TR depending on whether the program adhered to RNR principles (e.g., DeSorcy et al., 2016, 2020) or incorporated motivational approaches (Miles-McLean et al., 2019). Regarding the intervention setting, no specific pattern was observed in the relationship between TR and recidivism or risk of recidivism.

Retention/Dropout/Treatment Non-Completion

The results showed that the TR is not a predictor of treatment retention—that is, staying in or dropping out of therapy (Brocato, 2013; Brown & O’Leary, 2000; DeSorcy et al., 2020). However, the study by DeSorcy et al. (2016) revealed a different pattern: participants with a poorer TR were more likely to drop out of treatment, whereas those with a stronger TR—especially regarding bond and goal alignment—tended to remain in the program for a longer period. In addition, the relationship between TR and treatment completion seems to depend on who evaluates the relationship. Specifically, the study by van Tilburg et al. (2025) found that therapists’ ratings of the TR reliably predicted session attendance and treatment completion, whereas patients’ ratings did not predict either outcome.

No differences were found in the relationship between TR and retention/dropout/treatment compliance across different intervention settings, with studies showing mixed findings.

Aggression/Institutional Incidents/Abuse Behavior

The included studies have produced contradictory findings regarding the impact of TR on aggression variables. Brown and O’Leary (2000) and Taft et al. (2003) found that the TR significantly predicts reductions in psychological aggression and physical aggression after treatment. However, the study by Walling et al. (2012) showed that the extent to which TR influences the decrease in violence depends on racial background. Specifically, for minority participants, an increase in TR was associated with a substantial reduction in violence. In addition, the study by Clercx et al. (2021) examined the effect of TR on the number of institutional incidents involving verbal and physical aggression. The authors found that the impact of TR varied depending on the assessment period and the influence of concurrent variables. Specifically, while baseline TR did not predict incidents during the first 18 months, it significantly predicted incidents in the subsequent 18 months. TR assessed at 18 months was not a significant predictor.

When focusing specifically on outcomes related to aggression, institutional incidents, or abusive behavior, Clercx et al. was the only study to examine a non-cognitive-behavioral intervention, conducted in a forensic hospital setting. This study found no associations between TR, assessed at baseline or post-treatment, and aggressive incidents during the first 18 months. However, the TR assessed at baseline was associated with aggressive incidents occurring after the first 18 months. In contrast, studies of cognitive-behavioral programs conducted in community settings reported significant associations (Brown & O’Leary, 2000; Taft et al., 2003; Walling et al., 2012).

Motivation to Change

The studies indicated that the quality of the TR is associated with motivation to change (Bélanger & Higgs, 2024; Brocato, 2013; Brocato & Wagner, 2008), with studies suggesting that these associations vary depending on the stage of change (Bélanger & Higgs, 2024) or the specific dimension of TR examined (Brocato & Wagner, 2008). Specifically, Bélanger and Higgs (2024) found a positive but weak association between TR and the contemplation stage (i.e., when individuals begin to recognize the need for change), a moderate positive association between TR and the action stage (i.e., when individuals are actively engaged in making behavioral changes), and a moderate negative association between TR and precontemplation (i.e., when individuals do not yet acknowledge a need to change their behavior). Brocato and Wagner (2008), who assessed TR based on therapists’ self-reports, found that only the bond dimension was associated with motivation to change.

TR results did not vary according to the intervention setting (i.e., community or correctional institution).

Distress and Well-being

The included studies showed that TR was moderately associated with lower psychological and relationship distress at the end of treatment (Coleman et al., 2024) and with higher well-being (Singh et al., 2024). Specifically, the study by Scanlon et al. (2022) further showed that TR is a moderator of change for certain dimensions of mental health, such as anxiety, mania, anger, and depression symptoms. The results did not vary by intervention approach or setting.

Other Outcomes

The results indicated that TR did not predict outcomes related to marital satisfaction (Brown & O’Leary, 2000), alcohol use (Miles-McLean et al., 2019), or improvement in treatment (Watson et al., 2018), but it was significantly associated with reductions in general, reactive, and current criminal thinking (Scanlon et al., 2022).

The only program in which TR reached significance was based on a cognitive-behavioral model that also incorporated a psychosocial rehabilitation framework and a social learning paradigm (Scanlon et al., 2022), specifically in relation to anxiety, manic symptoms, anger, depression symptoms, and general, reactive, and current criminal thinking. No differences were observed based on the intervention setting.

Discussion

Building on the previously noted relevance of the TR for clinical outcomes (e.g., Cameron et al., 2018; Flückiger et al., 2018), this review further investigates how this relational factor operates within forensic treatment. In particular, the present study examines the associations between the TR and recidivism, therapeutic progress, and treatment attrition in intervention programs for individuals who have committed crimes.

Despite the considerable amount of research on intervention with individuals who have committed crimes (e.g., Beaudry et al., 2021; Gannon et al., 2019; Pinto e Silva et al., 2025; Sousa et al., 2022; Sousa, Féliz, et al, 2025; Sousa, Gonçalves, & de Castro-Rodrigues, 2024), in this systematic review, only 18 studies focused on the TR. This reflects the small investment and focus on TR among psychological interventions with individuals who have committed crimes, in contrast to non-forensic settings, where research on TR is of particular importance (e.g., Cameron et al., 2018; Flückiger et al., 2018; Luong et al., 2020). In fact, in clinical settings, more than 300 studies have investigated the alliance-outcome relationship, encompassing over 30.000 patients (Flückiger et al., 2018; Wampold, 2021). The extensive body of research on TR in clinical settings is associated with its recognition as a common factor across all treatments that plays a central role in determining positive therapeutic outcomes and is, at times, considered more influential than the treatment itself (Messer & Wampold, 2002; Norcross & Lambert, 2018; Wampold, 2021). Furthermore, our data further revealed that the included studies were conducted between 2003 and 2025, also reflecting a delay in the study of TR in forensic settings, given that interest in the clinical context dates back to 1978 (see Flückiger et al., 2018). In this regard, the studies conducted also exhibit low methodological quality, with only two studies employing a randomized controlled trial design. The low methodological quality observed is consistent with that of studies included in the systematic review on TR outcomes among justice-involved youth (see Papalia et al., 2022). The limited attention afforded to the TR in forensic contexts may be largely attributable to the dual role of forensic practitioners, who must integrate therapeutic work with risk assessment and legal reporting (Merkt et al., 2021; Ward, 2013), thereby creating tensions with foundational relational principles such as trust and confidentiality. Historically, forensic interventions have prioritized risk management, control, and program fidelity over relational processes, reflecting actuarial and correctional paradigms that conceptualized individuals who committed crimes as resistant, manipulative, or insufficiently motivated for change (Bonta & Andrews, 2023). Furthermore, the coercive nature of treatment, pronounced power asymmetries, and ethical constraints have complicated both the development and empirical examination of therapeutic alliance in forensic settings (Skeem et al., 2007).

In terms of sample characteristics, the included studies predominantly focused on men convicted of sexual offenses or IPV. Although the near-exclusive focus on the male population aligns with trends in scientific research (e.g., Cunha, Pereira, et al., 2024; Sousa et al., 2022; Sousa, Gouveia, Freitas, et al., 2024), it is surprising that there is a particular emphasis on individuals convicted of sexual or IPV-related crimes. This may be due to the continued high prevalence of these crimes and their psychological, physical, and social repercussions for victims, their families, and society both in the short and long term, as well as to their substantial social impact and the associated medical and legal costs (Hailes et al., 2019; Sardinha et al., 2022; Stubbs & Szoeke, 2022). Consequently, there remains an ongoing scientific and political debate regarding the effectiveness of interventions for these populations (e.g., Annison & Guiney, 2022; Gouveia et al., 2025; Sousa et al., 2026), highlighting the need to improve intervention strategies and identify the key components that can maximize outcomes (Lussier et al., 2023).

In relation to the characteristics of therapists, the number of professionals varied across studies, with different professional backgrounds such as psychologists, psychiatrists, nurses, institutional staff members (e.g., parole officers), and social workers (e.g., Blasko & Jeglic, 2016; Coleman et al., 2024; DeSorcy et al., 2016, 2020; Miles-McLean et al., 2019; Olver et al., 2024; Polaschek & Ross, 2010; van Tilburg et al., 2025). These findings are concerning, as a meta-analysis revealed that intervention programs targeting individuals convicted of sexual offenses as well as IPV-related crimes show lower recidivism rates when the programs are delivered by psychologists (Gannon et al., 2019). Although the present review does not allow for an assessment of whether therapists’ professional background has a consistent impact across the included studies, findings from Gannon et al. suggest that this factor may be relevant and, therefore, warrants caution when selecting professionals to deliver intervention programs. The overuse of paraprofessionals to deliver intervention programs, that is, professionals who are not licensed psychologists for clinical practice, has been associated with the growth of the prison population and the need to provide interventions to a larger number of individuals (Gannon & Ward, 2014).

The impact of the TR on treatment outcomes in justice-involved adults showed marked heterogeneity. Mixed findings were observed concerning the influence of the TR on recidivism and the risk of perpetrating new crimes (e.g., Blasko & Jeglic, 2016; DeSorcy et al., 2016, 2020; Miles-McLean et al., 2019; Olver et al., 2024; Polaschek & Ross, 2010), as well as its influence on the reduction of aggressiveness (e.g., Brown & O’Leary, 2000; Taft et al., 2003). Nevertheless, the included studies showed that TR is linked to motivation to change (Bélanger & Higgs, 2024; Brocato, 2013; Brocato & Wagner, 2008), as well as with reductions in psychological and relational distress (Coleman et al., 2024) and improvements in well-being (Singh et al., 2024). In addition, only one study found it to be significant for participants to remain in treatment (DeSorcy et al., 2016). Taken together, these findings suggest that the TR may exert a stronger influence on proximal, intrapersonal outcomes, such as motivation to change and well-being, than on more distal behavioral outcomes, such as aggression or recidivism. The TR did not show a robust link to treatment retention, which is somewhat surprising given evidence from recent qualitative studies suggesting that therapist characteristics are essential for treatment retention in programs for men who have perpetrated child sexual abuse (see Sousa, Félix, et al., 2025). Furthermore, the absence of an effect at the behavioral level may occur because recidivism, risk, and levels of aggression are complex, multifactorial outcomes influenced by a wide range of individual and therapeutic factors beyond the TR (Sturm et al., 2022). However, the available evidence base is too small and limited to allow firm conclusions to be drawn.

In addition, no differences were observed in the influence of the TR across the different outcomes as a function of the theoretical approach or the intervention setting. This may be because of the lack of variability, as most studies employed a CBT-based approach, which is in accordance with the literature (e.g., Gannon et al., 2019).

Limitations and Strengths

There are some limitations that should be considered. First, the number of included studies was limited, and the resulting evidence was frequently inconsistent. Further research is therefore needed, which would be facilitated by therapists and program evaluators routinely incorporating measures of the TR into interventions with justice-involved adults. Second, the overall quality of the studies is low, with limitations related to sample representativeness, the instruments used, and the procedures employed. In this regard, more randomized controlled trials are needed, with greater rigor in the selection of instruments with good psychometric properties and clearer reporting of procedures. In addition, the diversity of participants in the included studies was limited. Most studies included only men, and where both sexes were included, detailed demographic information such as race/ethnicity or socioeconomic status was generally not reported. This restricts the ability to assess how findings may vary across different populations and limits the generalizability of the results. Future research should aim to include more diverse samples and systematically report demographic characteristics to better understand the applicability of findings across different groups and cultural contexts.

The review also highlighted key knowledge gaps and directions for future research, some of which have already been addressed. From a practical perspective, the findings appear to suggest, albeit based on a limited body of evidence, that the TR functions as a responsivity factor that may influence participants’ engagement, motivation to change, and well-being, as it allows for the creation of a learning environment—considered a crucial element of responsivity—in which change is promoted and initiated and client engagement is enhanced (Bourgon & Bonta, 2014). This variable is particularly relevant, given that motivation to change in this population is generally low because of the mandatory nature of participation, which could influence recidivism rates (Pinto e Silva et al., 2022). In this regard, the findings also point to the importance of ensuring that professionals involved in these interventions receive appropriate clinical training, including training in relational and person-centered skills. Therapists’ person-centered characteristics (i.e., empathy, respect, warmth, and accepting and caring behavior) are associated with the formation of positive TRs, whereas a confrontational approach (i.e., hostility, negative attitudes, therapist rigidity, and arrogance) hinders the development of the therapeutic alliance (Norcross & Lambert, 2014; see Table 3).

Table 3.

Key Findings of the Systematic Review.

Existing studies are limited in number and present several methodological shortcomings.

Most studies included male samples convicted of sexual crimes or intimate partner violence.

Most studies involved therapists of different genders and varied academic backgrounds.

Six out of 18 studies did not report any information about the professional team.

TR appears to have an impact on proximal intrapersonal outcomes (e.g., motivation, well-being) but not on distal behavioral outcomes (e.g., aggression and recidivism).

No differences in TR effects across outcomes were found based on theoretical approach or intervention setting.

Note. TR = therapeutic relationship.

From a research perspective, further studies with greater methodological rigor, particularly randomized controlled trials, are needed to examine the influence of the TR in the treatment of diverse populations of individuals who have committed crimes. Future research should also move beyond examining whether TR is associated with treatment outcomes and instead clarify under which conditions, for whom, and through which mechanisms these associations emerge. Greater methodological consistency in the assessment of TR is needed, including the use of validated measures, repeated assessments across different stages of treatment, and the consideration of both participant and therapist perspectives. It is also important that future studies distinguish between different types of outcomes, such as treatment engagement, dropout, aggressive incidents, psychological well-being, and recidivism, to better understand the specific contribution of TR. Moreover, further research should include more diverse forensic samples, including individuals with different offense profiles and participants from prison, forensic psychiatric, and community-based settings. Finally, qualitative and mixed-methods studies may be particularly useful in clarifying the interpersonal and contextual processes through which TR develops in mandated treatment, including the role of perceived coercion, trust, responsivity, and the balance between care and control.

Conclusion

The primary aim of the present systematic review was to provide an overview of the outcomes associated with the TR in the treatment of adults who have committed crimes. Overall, the findings indicate that the TR plays a significant role in influencing proximal therapeutic factors, such as motivation to change and well-being. Nevertheless, further research with this specific population remains essential. In particular, additional studies are needed to deepen our understanding of the impact of the TR on the effectiveness of individuals’ rehabilitation (see Table 4).

Table 4.

Implications for Research, Practice, and Policy.

Therapists should be qualified psychology professionals with appropriate training and a person-centered orientation.

More methodologically rigorous studies, particularly randomized controlled trials, are needed.

Future research should include both male and female offender populations.

Qualitative studies could help clarify interactional processes and factors shaping the therapeutic relationship.

Footnotes

Acknowledgements

The authors acknowledge the support of the doctoral student Teresa Pinto e Silva for the draft prepared for the development of the “Introduction” section.

ORCID iDs

Marta Sousa

Sónia Caridade

Andreia de Castro-Rodrigues

Olga Cunha

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded by Fundação para a Ciência e Tecnologia (FCT), under HEI-Lab R&D Unit (UID/05380/2025, UIDP/05380/2020, and UID/PRR/05380/2025; , https://doi.org/10.54499/UIDP/05380/2020, and https://doi.org/10.54499/UID/PRR/05380/2025 https://doi.org/10.54499/UIDB/05380/2020). This work was also conducted at CIPsi, School of Psychology, University of Minho, supported by Fundação para a Ciência e a Tecnologia (FCT; UID/01662/2020) through the Portuguese. This work is funded with national funds from FCT, I.P., in the context of the R&D Unit: UID/04810/2025—William James Center for Research.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Author Biographies

Marta Sousa, Psychologist, PhD, is an assistant professor at the Lusófona University of Porto. Member integrated of the HEI-Lab: Digital Human-Environment Interaction Labs, Porto, Portugal. Her major research interests include intervention with men who perpetrated sexual violence. Forensic Psychologist at the Justice Psychology Unit of the Psychology Service of the University of Minho.

Sílvia Félix, PhD, is a clinical psychologist and researcher at the P5 Digital Medicine Center. She holds a PhD in Psychology (specialization in Psychotherapy and Psychopathology) by University of Minho. She is also a visiting lecturer at the Faculty of Nutrition and Food Sciences, University of Porto. Her research interests include emotion regulation, psychological intervention programs, eating behavior, and obesity.

Catarina Padinha, MSc, holds a Master’s degree in Forensic Psychology from ISPA. Her research interests focus on deviant behavior.

Sónia Caridade, Psychologist, PhD in Psychology of Justice. Assistant professor at School of Psychology, University of Minho. Researcher at the Psychology Research Center at the University of Minho. Her research interests include intimate partner violence, ciber, and in-person (adult and youth), and has authored several national and international publications on this topic.

Andreia de Castro-Rodrigues, PhD, is an assistant professor at ISPA—University Institute and a Researcher at William James Center for Research (Lisbon, Portugal). She developed her post-doc research and her PhD in Psychology of Justice, respectively, in the area of users’ perceptions and effectiveness of penal sanctions and sentencing.

Olga Cunha, Psychologist, PhD in Psychology of Justice. Assistant professor at School of Psychology, University of Minho. Researcher at the Psychology Research Center at the University of Minho. Her research interests include intervention with perpetrators of violence in intimate relationships. Forensic Psychologist at the Justice Psychology Unit of the Psychology Service of the University of Minho.

References

Alkan

Kumar

Ravichandran

Kaushal

S. R.

de Pablo

G. S.

Alerci

Michaud-Feinberg

Gutiérrez-Rojas

Zorzi

Klauser

Golay

Kramer

Alameda

(2025). Effectiveness of mindfulness-based interventions in reducing depressive symptoms across mental disorders: A meta-analysis of randomized controlled trials. Psychiatry Research, 348, Article 116473.

Annison

Guiney

(2022). Populism, conservatism and the politics of parole in England and Wales. The Political Quarterly, 93(3), 416–423. https://doi.org/10.1111/1467-923X.13170

Beaudry

Perry

A. E.

Fazel

(2021). Effectiveness of psychological interventions in prison to reduce recidivism: A systematic review and meta-analysis of randomised controlled trials. The Lancet Psychiatry, 8(9), 759–773. https://doi.org/10.1016/S2215-0366(21)00170-X

Bélanger

Higgs

(2024). Factors related to motivation to change and therapeutic alliance among men early in treatment following perpetration of child sexual abuse. Journal of Sexual Aggression, 30(3), 427–440. https://doi.org/10.1080/13552600.2023.2249542

Blasko

B. L.

Friedmann

P. D.

Rhodes

A. G.

Taxman

F. S.

(2015). The parolee–parole officer relationship as a mediator of criminal justice outcomes. Criminal Justice and Behavior, 42(7), 722-740.

Blasko

B. L.

Jeglic

E. L.

(2016). Sexual offenders’ perceptions of the client–therapist relationship: The role of risk. Sexual Abuse, 28(4), 271–290. https://doi.org/10.1177/1079063214529802

Blasko

Serran

Abracen

(2018). The role of the therapeutic alliance in offender therapy. In New frontiers in offender treatment: The translation of evidence-based practices to correctional settings (pp. 87-108). Springer International Publishing.

Bonta

Andrews

D. A.

(2023). The psychology of criminal conduct (7th ed.). Routledge.

Bordin

E. S.

(1979). The generalizability of the psychoanalytic concept of the working alliance. Psychotherapy: Theory, Research & Practice, 16(3), 252–260. https://doi.org/10.1037/h0085885

10.

Bourgon

Bonta

(2014). Reconsidering the responsbility principle: A way to move forward. Federal Probation, 78, 3.

11.

Brocato

(2013). The impact of acculturation, motivation, and the therapeutic alliance on treatment retention and outcomes for Hispanic drug-involved probationers. Journal of Ethnicity in Criminal Justice, 11(3), 150–180. https://doi.org/10.1080/15377938.2012.756845

12.

Brocato

Wagner

E. F.

(2008). Predictors of retention in an alternative-to-prison substance abuse treatment program. Criminal Justice and Behavior, 35(1), 99–119. https://doi.org/10.1177/00938548073094

13.

Brown

P. D.

O’Leary

K. D.

(2000). Therapeutic alliance: Predicting continuance and success in group treatment for spouse abuse. Journal of Consulting and Clinical Psychology, 68(2), 340–345. https://doi.org/10.1037/0022-006X.68.2.340

14.

Cameron

S. K.

Rodgers

Dagnan

(2018). The relationship between the therapeutic alliance and clinical outcomes in cognitive behaviour therapy for adults with depression: A meta-analytic review. Clinical Psychology & Psychotherapy, 25(3), 446–456. https://doi.org/10.1002/cpp.2180

15.

Catty

Winfield

Clement

(2007) The therapeutic relationship in secondary mental health care: A conceptual review of measures. Acta Psychiatrica Scandinavica, 116, 238–252. https://doi.org/10.1111/j.1600-0447.2007.01070.x

16.

Clercx

Voge

Lance

Vos

M. K.

(2021). The influence of therapy alliance and treatment motivation in patients with Cluster B personality disorders on incidents in forensic hospitals. Journal of Forensic Practice, 23(3), 272–284. https://doi.org/10.1108/JFP-05-2021-0022

17.

Coleman

J. J.

Drinane

J. M.

Owen

Sinha

Porter

E. F.

Agorsor

DeBlaere

Davis

D. E.

(2024). Psychotherapy with clients who are incarcerated: Therapists’ multicultural orientation, alliance, and outcomes. Professional Psychology: Research and Practice, 55(1), 39–47. https://doi.org/10.1037/pro0000522

18.

Crits-Christoph

Gibbons

M. B. C.

(2021). Psychotherapy process outcome research: Advances in understanding causal connections. In Barkham

Lutz

Castonguay

L. G.

(Eds.), Bergin and Garfield’s handbook of psychotherapy and behavior change (50th ed., pp. 263–296). Wiley.

19.

Cunha

Pedrosa

Silva Pereira

Caridade

de Castro Rodrigues

Braga

(2024). Intervention program dropout among perpetrators of intimate partner violence: A meta-analysis of correlated variables. Trauma, Violence, & Abuse, 25(4), 2735–2751. https://doi.org/10.1177/15248380231224036

20.

Cunha

Pereira

Sousa

Rodrigues

A. C.

(2024). Cognitive behavioural “third wave” therapies in the treatment of justice-involved individuals: A systematic review. Aggression and Violent Behavior, 76, Article 101923. https://doi.org/10.1016/j.avb.2024.101923

21.

DeSorcy

D. R.

Olver

M. E.

Wormith

J. S.

(2016). Working alliance and its relationship with treatment outcome in a sample of aboriginal and non-aboriginal sexual offenders. Sexual Abuse: A Journal of Research and Treatment, 28(4), 291–313. https://doi.org/10.1177/1079063214556360

22.

DeSorcy

D. R.

Olver

M. E.

Wormith

J. S.

(2020). Working alliance and psychopathy: Linkages to treatment outcome in a sample of treated sexual offenders. Journal of Interpersonal Violence, 35(7–8), 1739–1760. https://doi.org/10.1177/0886260517698822

23.

Flückiger

Del Re

A. C.

Wampold

B. E.

Horvath

A. O.

(2018). The alliance in adult psychotherapy: A meta-analytic synthesis. Psychotherapy: Theory, Research & Practice, 55(4), 316–340. https://doi.org/10.1037/pst0000172

24.

Flückiger

Del Re

A. C.

Wlodasch

Horvath

A. O.

Solomonov

Wampold

B. E.

(2020). Assessing the alliance–outcome association adjusted for patient characteristics and treatment processes: A meta analytic summary of direct comparisons. Journal of Counseling Psychology, 67(6), 706–711. https://doi.org/10.1037/cou0000424

25.

Freud

(1912/1958). The dynamics of transference. In Starchey

(Ed.), The standard edition of the complete psychological works of Sigmund Freud (Vol. 12, pp. 99–108). Hogarth Press.

26.

Gannon

T. A.

Olver

M. E.

Mallion

J. S.

James

(2019). Does specialized psychological treatment for offending reduce recidivism? A meta-analysis examining staff and program variables as predictors of treatment effectiveness. Clinical Psychology Review, 73, Article 101752. https://doi.org/10.1016/j.cpr.2019.101752

27.

Gannon

T. A.

Ward

(2014). Where has all the psychology gone? A critical review of evidence-based psychological practice in correctional settings. Aggression and Violence Behavior, 19, 435–446. https://doi.org/10.1016/j.avb.2014.06.006

28.

Gelso

C. J.

Kline

K. V.

(2019). The sister concepts of the working alliance and the real relationship: On their development, rupture, and repair. Research in Psychotherapy: Psychopathology, Process and Outcome, 22(2), 142–149. https://doi.org/10.4081/ripppo.2019.373

29.

Gouveia

Silva

Seto

Cunha

Caridade

(2025). A systematic review on the effectiveness of interventions for individuals who have committed online sexual crimes involving children. Journal of Forensic and Legal Medicine, 113, Article 102889.

30.

Hachtel

Vogel

Huber

C. G.

(2019). Mandated treatment and its impact on therapeutic process and outcome factors. Frontiers in Psychiatry, 10, Article 219. https://doi.org/10.3389/fpsyt.2019.00219

31.

Hailes

H. P.

Danese

Fazel

(2019). Long-term outcomes of childhood sexual abuse: An umbrella review. The Lancet Psychiatry, 6(10), 830–839. https://doi.org/10.1016/S2215-0366(19)30286-X

32.

Hardeberg Bach

Demuth

. (2018). Therapists’ experiences in their work with sex offenders and people with pedophilia: A literature review. Europe’s Journal of Psychology, 14(2), 498–514. https://doi.org/10.5964/ejop.v14i2.1493

33.

Hatcher

R. L.

Gillaspy

J. A.

(2006). Development and validation of a revised short version of the Working Alliance Inventory. Psychotherapy Research, 16(1), 12–25. https://doi.org/10.1080/10503300500352500

34.

Hong

Q. N.

Pluye

Fàbregues

Bartlett

Boardman

Cargo

Dagenais

Gagnon

M. P.

Griffiths

Nicolau

O’Cathain

Rousseau

M. C.

Vedel

(2018). Mixed methods appraisal tool (MMAT), version 2018. Registration of copyright (#1148552). Canadian Intellectual Property Office.

35.

Horvath

(1981). An exploratory study of the working alliance: Its measurement and relationship to outcome [Unpublished doctoral dissertation, University of British Columbia].

36.

Horvath

(1982). Users’ manual of the Working Alliance Inventory. Unpublished manuscript, No. 82:2, Simon Fraser University.

37.

Koehler

Lösel

(2024). A meta-evaluative synthesis of the effects of custodial and community-based offender rehabilitation. European Journal of Criminology, 22(1), 3–29. https://doi.org/10.1177/14773708241256501

38.

Leiman

Stiles

(2001). Dialogical sequence analysis and the zone of proximal development as conceptual enhancements to the assimilation model: The case of Jan revisited. Psychotherapy Research, 11(3), 311–330. https://doi.org/10.1093/ptr/11.3.311

39.

Liu

C. I.

Hua

M. H.

M. L.

Goh

K. K.

(2023). Effectiveness of cognitive behavioural-based interventions for adults with attention-deficit/hyperactivity disorder extends beyond core symptoms: A meta-analysis of randomized controlled trials. Psychology and Psychotherapy: Theory, Research and Practice, 96(3), 543–559.

40.

Luo

Bugatti

Molina

Tilley

J. L.

Mahaffey

Gonzalez

(2022). Conceptual invariance, trajectories, and outcome associations of working alliance in unguided and guided internet-based psychological interventions: Secondary analysis of a randomized controlled trial. JMIR Mental Health, 9(6), Article e35496. https://doi.org/10.2196/35496

41.

Luong

H. K.

Drummond

S. P.

Norton

P. J.

(2020). Elements of the therapeutic relationship in CBT for anxiety disorders: A systematic review. Journal of Anxiety Disorders, 76, Article 102322. 10.1016/j.janxdis.2020.102322

42.

Lussier

Chouinard Thivierge

Fréchette

Proulx

(2023). Sex offender recidivism: Some lessons learned from over 70 years of research. Criminal Justice Review, 49(4), 413–452. https://doi.org/10.1177/07340168231157385

43.

Lutz

Zani

Fritz

Dudeck

Franke

(2022). A review and comparative analysis of the risk-needs-responsivity, good lives, and recovery models in forensic psychiatric treatment. Frontiers in Psychiatry, 13, Article 988905. https://doi.org/10.3389/fpsyt.2022.988905

44.

Martin

D. J.

Garske

J. P.

Davis

M. K.

(2000). Relation of the therapeutic alliance with outcome and other variables: A meta-analytic review. Journal of Consulting and Clinical Psychology, 68(3), 438–450. https://doi.org/10.1037/0022-006X.68.3.438

45.

Merkt

Haesen

Eytan

Habermeyer

Aebi

M. F.

Elger

Wangmo

(2021). Forensic mental health professionals’ perceptions of their dual loyalty conflict: Findings from a qualitative study. BMC Medical Ethics, 22(1), 123. https://doi.org/10.1186/s12910-021-00688-2

46.

Messer

S. B.

Wampold

B. E.

(2002). Let’s face facts: Common factors are more potent than specific therapy ingredients. Clinical Psychology: Science and Practice, 9(1), 21–25. https://doi.org/10.1093/clipsy.9.1.21

47.

Miles-McLean

H. A.

LaMotte

A. D.

Semiatin

J. N.

Farzan-Kashani

Torres

Poole

G. M.

Murphy

C. M.

(2019). PTSD as a predictor of treatment engagement and recidivism in partner abusive men. Psychology of Violence, 9(1), 39–47. https://doi.org/10.1037/vio0000161

48.

Norcross

J. C.

Lambert

M. J.

(2011). Evidence-based therapy relationships. In Norcross

J. C.

(Ed.), Psychotherapy relationships that work: Evidence-based responsiveness (pp. 3–21). Oxford University Press.

49.

Norcross

J. C.

Lambert

M. J.

(2014). Relationship science and practice in psychotherapy: Closing commentary. Psychotherapy, 51, 398–403. http://dx.doi.org/10.1037/a0037418

50.

Norcross

J. C.

Lambert

M. J.

(2018). Psychotherapy relationships that work III. Psychotherapy, 55(4), 303–315. https://doi.org/10.1037/pst0000193

51.

Olver

M. E.

Stockdale

K. C.

Riemer

E. K.

(2024). The risk, need, and responsivity relevance of working alliance in a Sexual Offense Treatment Program: Its intersection with psychopathy, diversity, and treatment change. Sexual Abuse: A Journal of Research and Treatment, 36(4), 383–417. https://doi.org/10.1177/10790632231172161

52.

Olver

M. E.

Stockdale

K. C.

Wormith

J. S.

(2011). A meta-analysis of predictors of offender treatment attrition and its relationship to recidivism. Journal of Consulting and Clinical Psychology, 79(1), 6–21. https://doi.org/10.1037/a0022200

53.

Ouzzani

Hammady

Fedorowicz

Elmagarmid

(2016). Rayyan—A web and mobile app for systematic reviews. Systematic Reviews, 5, 210. https://doi.org/10.1186/s13643-016-0384-4

54.

Page

M. J.

McKenzie

J. E.

Bossuyt

P. M.

Boutron

Hoffmann

T. C.

Mulrow

C. D.

Shamseer

Tetzlaff

J. M.

Akl

E. A.

Brennan

S. E.

Chou

Glanville

Grimshaw

J. M.

Hróbjartsson

Lalu

M. M.

Loder

E. W.

Mayo-Wilson

McDonald

. . . Moher

(2021). The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. bmj, 372.

55.

Papalia

Dunne

Maharaj

Fortunato

Luebbers

Ogloff

J. R. P.

(2022). Determinants and outcomes of the therapeutic alliance in treating justice-involved youth: A systematic review of quantitative and qualitative research. Clinical Child and Family Psychology Review, 25(4), 658–680. https://doi.org/10.1007/s10567-022-00407-2

56.

Pinto e Silva

Cunha

Caridade

(2022). Motivational interview techniques and the effectiveness of intervention programs with perpetrators of intimate partner violence: A systematic review. Trauma, Violence, & Abuse, 24(4), 2691–2710. https://doi.org/10.1177/15248380221111472

57.

Pinto e Silva

Dinis

M. A.

Pedrosa

Sousa

H. F.

Cunha

Caridade

(2024). Short motivational program for perpetrators of intimate partner violence: A feasibility study. Journal of Police and Criminal Psychology, 40, 719–731. https://doi.org/10.1007/s11896-024-09680-z

58.

Pinto e Silva

Gouveia

Santirso

F. A.

Cunha

Caridade

(2025). Effectiveness of motivational interviewing with justice-involved people: A systematic review and meta-analysis. Psychosocial Intervention, 34(2), 89–102. https://doi.org/10.5093/pi2025a8

59.

Piper

W. E.

Ogrodniczuk

J. S.

Lamarche

Hilscher

Joyce

A. S.

(2005). Level of alliance, pattern of alliance, and outcome in short-term group therapy. International Journal of Group Psychotherapy, 55(4), 527–550. https://doi.org/10.1521/ijgp.2005.55.4.527

60.

Polaschek

D. L.

Ross

E. C.

(2010). Do early therapeutic alliance, motivation, and stages of change predict therapy change for high-risk, psychopathic violent prisoners?. Criminal Behaviour and Mental Health: CBMH, 20(2), 100–111. https://doi.org/10.1002/cbm.759

61.

Prusiński

(2024). The linear pattern of therapeutic alliance development: Exploring the relationship between alliance trajectories over time and the outcomes of systemic psychotherapy treatment for adaptation disorders. Counselling and Psychotherapy Research, 24(4), 1322–1336.

62.

Ribeiro

Gonçalves

M. M.

Pinto

(2021). Responsiveness and therapeutic collaboration in narrative therapy. In Watson

J. C.

Wiseman

(Eds.), The responsive psychotherapist: Attuning to clients in the moment (pp. 195–217). American Psychological Association. https://doi.org/10.1037/0000240-010

63.

Ribeiro

A. P.

Gonçalves

M. M.

Horvath

A. O.

Stiles

W. B.

(2013). How collaboration in therapy becomes therapeutic: The therapeutic collaboration coding system. Psychology and Psychotherapy: Theory, Research and Practice, 86(3), 294–314. https://doi.org/10.1111/j.2044-8341.2012.02066.x

64.

Rogers

C. R.

Gendlin

E. T.

Kiesler

D. J.

Truax

C. B.

(1967). The therapeutic relationship and its impact: A study of psychotherapy with schizophrenics. University of Wisconsin Press.

65.

Santirso

F. A.

Lila

Gracia

(2020). Motivational strategies, working alliance, and protherapeutic behaviors in batterer intervention programs: A randomized controlled trial. European Journal of Psychology Applied to Legal Context, 12(2), 77-84.

66.

Sardinha

Maheu-Giroux

Stöckl

Meyer

S. R.

García-Moreno

(2022). Global, regional, and national prevalence estimates of physical or sexual, or both, intimate partner violence against women in 2018. The Lancet, 399(10327), 803–813. https://doi.org/10.1016/S0140-6736(21)02664-7

67.

Scanlon

Hirsch

Morgan

R. D.

(2022). The relation between the working alliance on mental illness and criminal thinking among justice-involved people with co-occurring mental illness and substance use disorders. Journal of Consulting and Clinical Psychology, 90(3), 282–288. https://doi.org/10.1037/ccp0000719

68.

Singh

Short

Minoudis

(2024). Evaluation of the Enhanced Engagement and Relational Support Service (EERSS) within the offender personality disorder (OPD) pathway. The Journal of Forensic Psychiatry & Psychology, 35(4), 629–644. https://doi.org/10.1080/14789949.2024.2356817

69.

Skeem

J. L.

Eno Louden

Polaschek

D. L. L.

Camp

(2007). Assessing relationship quality in mandated community treatment: Blending care with control. Psychological Assessment, 19(4), 397–410. https://doi.org/10.1037/1040-3590.19.4.397

70.

Sousa

Andrade

de Castro-Rodrigues

Gonçalves

R. A.

(2022). The effectiveness of psychological treatment in adult male convicted for sexual offenses against children: A systematic review. Trauma, Violence, & Abuse, 24(3), 1867–1881. https://doi.org/10.1177/15248380221082080

71.

Sousa

Félix

Gonçalves

R. A.

de Castro-Rodrigues

(2025). In their voices: Perceptions of individuals who perpetrated child sexual abuse about the INSIGHT program. Journal of Child Sexual Abuse, 34(8), 983–1005. https://doi.org/10.1080/10538712.2025.2603982

72.

Sousa

Gonçalves

R. A.

de Castro-Rodrigues

(2024). INSIGHT intervention for individuals who sexually offended against children: Preliminary results of a randomized pilot study. Psychology, Crime & Law, 32(3), 580–602. https://doi.org/10.1080/1068316X.2024.2381115

73.

Sousa

Gonçalves

Caridade

de Castro-Rodrigues

Maia

Â.

Cunha

(2025). The effectiveness of trauma-focused interventions with individuals who perpetrated crimes: A systematic review. Trauma, Violence, & Abuse. Advance online publication. https://doi.org/10.1177/15248380251397413

74.

Sousa

Gouveia

Cunha

de Castro Rodrigues

(2024). The effectiveness of schema therapy in individuals who committed crimes: A systematic review. Trauma, Violence, & Abuse, 25(5), 3631–3642. https://doi.org/10.1177/15248380241254082

75.

Sousa

Gouveia

Freitas

Caridade

Cunha

(2024). The effectiveness of psychological intervention for women who committed child sexual abuse: An empty systematic review. Trauma, Violence, & Abuse, 26(1), 156–166. https://doi.org/10.1177/15248380241277274

76.

Sousa

Pinto e Silva

Almeida

T. C.

Santirso

F. A.

Caridade

Soares Cunha

(2026). The effectiveness of intervention with intimate partner violence perpetrators: An umbrella review. Journal of Family Violence. Advance online publication. https://doi.org/10.1007/s10896-026-01115-z

77.

Stinson

J. D.

Clark

M. D.

(2017). Motivational interviewing with offenders: Engagement, rehabilitation, and reentry. Guilford Publications.

78.

Stubbs

Szoeke

(2022). The effect of intimate partner violence on the physical health and health-related behaviors of women: A systematic review of the literature. Trauma, Violence, & Abuse, 23(4), 1157–1172. https://doi.org/10.1177/1524838020985541

79.

Sturm

de Vogel

Menger

Huibers

M. J.

(2022). The price of mistrust: A study into the working alliance as predictor for recidivism. Journal of Police and Criminal Psychology, 37(3), 576–586. https://doi.org/10.1007/s11896-021-09450-1

80.

Taft

C. T.

Murphy

C. M.

King

D. W.

Musser

P. H.

DeDeyn

J. M.

(2003). Process and treatment adherence factors in group cognitive-behavioral therapy for partner violent men. Journal of Consulting and Clinical Psychology, 71(4), 812–820. https://doi.org/10.1037/0022-006X.71.4.812

81.

van Tilburg

C. A.

de Ruiter

Lodder

Bachrach

van Dam

. (2025). Trust the therapists’ trust: Working alliance development in relation to treatment attendance and completion in intimate partner violence perpetrator group therapy. Psychotherapy Research, 24, 1–18. https://doi.org/10.1080/10503307.2025.2548501

82.

Vigotsky

(1978). Mind in society. Harvard University Press.

83.

Walling

S. M.

Suvak

M. K.

Howard

J. M.

Taft

C. T.

Murphy

C. M.

(2012). Race/ethnicity as a predictor of change in working alliance during cognitive behavioral therapy for intimate partner violence perpetrators. Psychotherapy (Chicago, Ill.), 49(2), 180–189. https://doi.org/10.1037/a0025751

84.

Wampold

B. E.

(2015). How important are the common factors in psychotherapy? An update. World Psychiatry, 14(3), 270–277. https://doi.org/10.1002/wps.20238

85.

Wampold

B. E.

(2021). Healing in a social context: The importance of clinician and patient relationship. Frontiers in Pain Research, 2, Article 684768. https://doi.org/10.3389/fpain.2021.684768

86.

Ward

(2013). Addressing the dual relationship problem in forensic and correctional practice. Aggression and Violent Behavior, 18(1), 92–100. https://doi.org/10.1016/j.avb.2012.10.003

87.

Watson

Daffern

Thomas

(2018). The impact of sex offender’s interpersonal style, treatment readiness and the therapeutic alliance on treatment gain. The Journal of Forensic Psychiatry & Psychology, 29(4), 635–655. https://doi.org/10.1080/14789949.2018.1432673

88.

Willis

G. M.

Ward

(2024). Evidence for the good lives model in supporting rehabilitation and desistance from offending. In Craig

L. A.

Dixon

Gannon

T. A.

(Eds.), The Wiley handbook of what works in correctional rehabilitation: An evidence-based approach to theory, assessment and treatment (pp. 299–309). Wiley-Blackwell.

89.

Wittouck

Vander Beken

(2019). Recovery, desistance, and the role of procedural justice in working alliances with mentally ill offenders: A critical review. Addiction Research & Theory, 27(1), 16–28. https://doi.org/10.1080/16066359.2018.1518434

90.

M. S.

Wickham

R. E.

Chen

S. Y.

Varra

Chen

Lungu

(2024). A large-scale evaluation of therapeutic alliance and symptom trajectories of depression and anxiety in blended care therapy. PLoS One, 19(11), Article e0313112. https://doi.org/10.1371/journal.pone.0313112