Do Changes in Emotional States Predict Therapy Alliance in Dutch Male Offenders in Mandated Clinical Care?

• Author: Vivienne de Vogel,Marije Keulen-de Vos
• Date: 01 April 2022
• DOI: 10.1177/0306624X211013733
• Published date: 01 April 2022
• Subject Matter: Articles

https://doi.org/10.1177/0306624X211013733

International Journal of

Offender Therapy and

Comparative Criminology

2022, Vol. 66(5) 625 –644

© The Author(s) 2021

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0306624X211013733

journals.sagepub.com/home/ijo

Article

Do Changes in Emotional

States Predict Therapy

Alliance in Dutch Male

Offenders in Mandated

Clinical Care?

Marije Keulen-de Vos1 and Vivienne de Vogel2

Abstract

Therapy alliance has been studied largely in voluntary psychotherapy but less is

known about its predictive factors for positive alliance and treatment outcome in

forensic populations. The aim of this study was to examine the relationship between

offenders’ emotional states and therapy alliance. Moreover, we were interested in

the predictive impact of emotional states early in treatment on alliance at 18 months

into treatment. Self-ratings of emotional states and alliance by 103 male offenders,

and therapist-ratings for therapy alliance were examined using hierarchical multiple

regression analyses. Participants were primarily convicted for violent of sexual

offenses, and were diagnosed with antisocial, borderline or narcissistic personality

disorders. Healthy emotional states were predictive of mid-treatment agreement on

therapy goals and therapist ratings on tasks within the therapy. Unhealthy emotional

states were predictive of patient-rated agreement on tasks. Emotional states were

not predictive for the reported therapist/patient bond or global alliance ratings.

This study emphasizes the importance of healthy emotional states in treatment of

offenders with personality disorders.

Keywords

emotional states, therapeutic alliance, offending, psychotherapy, predict

1Forensic Psychiatric Center de Rooyse Wissel, Venray, Limburg, The Netherlands

2De Forensische Zorgspecialisten, Utrecht, The Netherlands

Corresponding Author:

Marije Keulen-de Vos, Forensic Psychiatric Center de Rooyse Wissel, P.O. Box 433, Venray, Limburg

5800 AK, The Netherlands.

Email: MKeulen-deVos@derooysewissel.nl

1013733IJOXXX10.1177/0306624X211013733International Journal of Offender Therapy and Comparative CriminologyKeulen-de Vos and de Vogel

research-article2021

626 International Journal of Offender Therapy and Comparative Criminology 66(5)

Evidence suggests that non-specific treatment factors, such as the therapeutic relation-

ship between patient and therapist, are robustly related to outcome across therapy

approaches, populations and psychological problems (Fahlgren et al., 2020; Flückiger

et al., 2012, 2018). For example, research shows that the therapy relationship accounts

for up to 30% of the variance in treatment outcomes (Westen et al., 2004). For Bordin

(1994), therapy alliance is the result of the combination between the agreement on the

goals in the therapy, consensus on the tasks that make up the therapy, and the bond

between patient and therapist. The agreement on these components can fluctuate over

the course of treatment; therefore therapy alliance is a dynamic concept. Research on

the role of therapeutic alliance in the field of high-risk violent offender treatment has

been limited although it is an important responsivity factor (Skeem et al., 2007a).

Although therapy alliance is an important part of all therapy, it is the offending treat-

ment population for which responsivity issues and the therapy alliance are vital as

offenders are often very distrustful of professionals. Skeem et al. (2007a) argue that

therapists who provide mandated care have both a caring and controlling role. Because

of their controlling role, patients may feel uncomfortable to disclose information.

Also, offenders may have limited insight into their problems (Trotter, 2015). It is nec-

essary for therapists to overcome this in order to facilitate change (Marshall & Serran,

2004). Overcoming an offender’s lack of insight is important because it refers to the

nature of behavioral problems, the extent to which these problems are self-caused and

ways in which behavior can be changed (Draper et al., 2013; Levenson & Macgowan,

2004; Woods et al., 2003).

Several studies have examined therapeutic alliance in several adult offender popu-

lations, such as sex offenders, partner-violent men, substance abusing offenders,

patients in mandated community treatment, patients on parole, and in adolescent

offenders (Brown & O’Leary, 2000; Florsheim et al., 2000; Taft et al., 2003; Serran &

Marshall, 2010; Skeem et al., 2007a; Sturm et al., 2021; Walters, 2015). These studies

show that the strength of the alliance is positively associated with treatment outcome,

recidivism or probation violations. Little is known, however, about change in alliance

over the course of forensic treatment and factors that predict this change. We know

that, in general, therapy alliance is influenced by a number of factors, such as patients’

attachment style, patients’ interpersonal skills and schemas, therapist’ professional

skills, and patient characteristics and diagnoses (Barrowclough et al., 2010; Kramer

et al., 2009; Ross et al., 2008). For example, Wallner-Samstag et al. (1992) found that

patients who are characterized as submissive, isolated and friendly are more likely to

develop strong alliances, whereas hostility, aggression and dominance is associated

with poor alliance (Puschner et al., 2005). With offenders, personality disorders (PDs)

are common (Langton et al., 2011), especially antisocial, borderline, and narcissistic

PD. According to Ross et al. (2008), forensic patients have high base rates of hostility,

PDs and interpersonal problems. This may have a negative impact on therapy alliance

because they can be anxious or too vigilant to easily form a bond with their therapist

(Fahlgren et al., 2020). Psychopathy is also highly prevalent in offender populations

and a strong predictor of negative treatment outcome (e.g., Leistico et al., 2008). Some

studies have specifically examined whether psychopathy has an impact on developing