An open pilot trial of a behavioral intervention to reduce violence by young adults with early psychosis receiving treatment in an early intervention services setting: a protocol

Abstract

Background:

Despite the public health impact of violence among young adults with psychosis, behavioral interventions to reduce the risk of engaging in violence remain rare. For young adults with early psychosis, cognitive behavioral therapy (CBT)-based psychotherapy has efficacy in reducing impairment and improving functioning. However, no CBT-based intervention to reduce violence has been formally adapted for young adults with early psychosis.

Aims:

This protocol outlines the first clinical trial of a behavioral intervention to reduce violence for young adults with early psychosis. This study is set in an early intervention services (EIS) setting and seeks to adapt and pilot PICASSO, a CBT-based intervention, through an iterative process utilizing mixed-methods assessments.

Method:

All research will occur at OnTrackNY, the largest EIS program in the United States. This study will consist of an open pilot trial, with four EIS clinicians delivering the intervention to one to two EIS participants per round. In this mixed-methods study, both quantitative measures (acceptability, feasibility, and hypothesized mediators of target outcome collected on a weekly basis) and qualitative interviews (with EIS clinicians at weeks 4, 8, and 12) will be conducted. Transcripts will be analyzed using thematic content analysis. Two to three rounds of iterative modifications are anticipated (n = 10-16 EIS participants total).

Results:

Recruitment began in February 2023 and is expected to continue over a 9-to-12-month period.

Conclusions:

Because violent behavior causes interpersonal disruptions such as incarceration and increased caregiver burden, an innovative intervention to reduce violence risk could have broader health impact for this vulnerable population. Adapting the PICASSO intervention to the EIS setting will optimize its acceptability and feasibility by the intended target population.

Trial registration:

This study is registered on 2/23/2023 in Clinicaltrials.gov (NCT05756855).

Introduction

The absolute risk of violence among persons with serious mental illnesses (SMI) is small, but young people experiencing early psychosis are a subgroup at elevated violence risk. Current studies estimate that up to one-third of individuals engage in violence early in their illness.(Large & Nielssen, 2011) Prior violent behavior is widely known to be the strongest predictor of future violence,(Mossman, 1994) but for young adults with early psychosis, violent behavior may first emerge during this high-risk time. Existing research has identified a small number of risk factors for violence in this population, including hostility, criminal justice involvement, lower levels of educational attainment, and substance use, especially cannabis use.(Large & Nielssen, 2011; Rolin et al., 2018) Empirical studies have also identified victimization, impulsivity, and personality disorders as risk factors that may increase risk of violent behavior. (Buchanan et al., 2023; Hoptman et al., 2015; Volavka et al., 2014)

Early psychosis presents a unique opportunity for intervention. Early intervention services (EIS) for young people with early psychosis have been associated with improved outcomes, including: improved quality of life; higher participation in work/school; fewer hospitalizations; and reduced symptoms.(Dixon et al., 2018) OnTrackNY is the largest EIS program in the U.S. and has provided evidence-based coordinated specialty care to over 2,800 patients at 25 clinics in New York State. EIS treatment often includes cognitive behavioral therapy (CBT) for psychosis. CBT is a manualized behavioral intervention teaches people to identify how their emotions and thoughts relate to their behaviors (Munoz & Miranda, 2000). This type of therapy is time-limited and goal-oriented, with a focus on teaching people skills they can use and master to reduce the need for a long-term therapist. CBT interventions have been developed for a wide range of populations, including for people with depression, anxiety, post-traumatic stress, and psychosis.(Johnson & Ceroni, 2020; Landa, 2017; Munoz & Miranda, 2000; Wells, 1997)

Studies have suggested that CBT has a positive effect on people with psychotic disorders by reducing specific symptoms (e.g., through behavioral activation to reduce negative symptoms), increasing coping skills (e.g., distress tolerance), and normalizing experiences that may otherwise be distressing (e.g., hallucinations).(Haddock & Slade, 1995) For young adults with early psychosis, CBT-based psychotherapy has efficacy in reducing impairment and improving functioning and often acts synergistically with medications. (Tarrier et al., 2014; Dixon et al., 2010)

CBT-based interventions to reduce violence have promise. In populations with and without SMI, CBT has demonstrated efficacy in reducing violent behavior and is considered the most promising non-pharmacological approach to reducing violence.(Rampling et al., 2016) Various cognitive-behavioral interventions have found efficacy in improving coping and problem solving, as well as reducing disruptive behaviors and self- reported violent attitudes. (Rees-Jones et al., 2012; Yip et al. 2013; Jotangia et al. 2015). A review of the literature identified no studies of CBT to reduce violence for young adults with early psychosis though did identify two promising trials. Haddock et al. (2009) compared a CBT intervention called the Psychological Intervention for Complex PTSD and Schizophrenia-Spectrum disorder (PICASSO), targeting violence, anger, and psychosis (n=38), to a control condition, i.e., social activity therapy (n=39). Over the 12-month follow-up, fewer people in the CBT group had been aggressive, including less physical aggression. (Haddock et al., 2009) This study population was limited to participants with a history of violent behavior, persistent hallucinations, and an antipsychotic medication prescription. In 2012, Cullen et al., randomized 84 “mentally disorder[ed] offenders” (inpatients at forensic hospitals) to cognitive skills training to reduce violence and antisocial behavior vs. treatment-as-usual. Verbal aggression and antisocial behavior were reduced in the intervention group, though retention was problematic given the setting. (Haddock et al., 2009)

This study will address this treatment gap by building on existing research, utilizing findings from formative research to adapt and pilot a behavioral intervention to reduce violence risk for young adults receiving treatment in an EIS setting. The focus of this study is reducing psychosis-related violence, or violence among EIS participants with no prior history of antisocial behavior. This distinction is made as the prior research has identified two subgroups of people with schizophrenia who commit violent acts: 1) people whose violent behavior predates the development of psychosis, and 2) people who engage in violence only after illness onset. Overall, research on violence has suggested that different treatments are needed to manage violence risk for these two groups. (Moran & Hodgins, 2004; Volavka & Citrome, 2011; Mueser et al., 2006; Adams & Yanos, 2020).

Aims

In this study, an existing CBT intervention called PICASSO (Haddock et al., 2009) will be adapted and tested in an EIS setting. PICASSO is a CBT-model that targets violence, anger, and psychosis, developed by Haddock et al. (2009) and tested in a randomized controlled trial (RCT). Over the one-year follow-up period, fewer people in the PICASSO group (n=38) had been physically aggressive compared to a control group receiving a social activity therapy (n=39).(Haddock et al., 2009) However, since PICASSO was tested in a sample of primarily forensic inpatients, this intervention will first be adapted to the EIS setting and then tested in an open pilot trial, utilizing a mixed methods assessment of feasibility and acceptability.

Methods

Study design

This is an open pilot trial examining the acceptability and feasibility of a manualized behavioral intervention (an adapted version of PICASSO). The primary endpoint is at the end of the 12-week treatment; follow-up will occur on a weekly basis (see Table 1). The study has been approved by the New York State Psychiatric Institute IRB on 7/28/2023. The study sponsor is the National Institutes of Health (K23MH126312 and L30MH120711). The study is registered with Clinicaltrials.gov (NCT05756855).

Table 1:

Proposed Schedule of Assessments

Variable type and/or instrument		Method of collection	Week Number
			1	2	3	4	5	6	7	8	9	10	11	12
EIS participants
Measure of acceptability
	Cognitive interview	Self-report	X	X	X	X	X	X	X	X	X	X	X	X
Measure of feasibility
	Attendance, completion of homework	Chart review	X	X	X	X	X	X	X	X	X	X	X	X
Measures of target engagement		Self-report				X				X				X
EIS Clinicians
Measure of acceptability
	AIM	Completed by EIS clinician	X	X	X	X	X	X	X	X	X	X	X	X
	Qualitative interview	Completed by EIS clinician				X				X				X
Measure of feasibility
	Recruitment, retention in treatment	Chart review	X	X	X	X	X	X	X	X	X	X	X	X
	FIM	Completed by EIS clinicians	X	X	X	X	X	X	X	X	X	X	X	X
	Qualitative interview	Completed by EIS clinicians				X				X				X

Setting

This pilot study will occur at OnTrackNY, the largest EIS program in the United States. All study participants will be recruited from the OnTrackNY clinics where they work (EIS clinicians) or receive treatment (EIS participants). The OnTrackNY Stakeholder Workgroup, which has strong established relationships with OnTrackNY clinics throughout the state and has successfully supported the recruitment of clients and clinicians for research projects in the past, will assist by identifying appropriate clinics (based on interest, clinical availability, staffing concerns, and other factors affecting availability).

Inclusion and Exclusion Criteria

All EIS participants in this study will meet the following criteria for treatment at OnTrackNY: (1) 16–30 years old; (2) diagnosis of a non-affective psychosis, with a history of psychotic symptoms lasting at least one week; (3) enrolled within two years of onset of first psychotic symptoms; and (4) resident of New York State. Additionally, all EIS participants will meet the study-specific inclusion criteria of being 1) an adult age 18 and older, 2) able to speak and read English, 3) willing to participate in audio-recorded research interviews, 4) without a history of pre-existing antisocial behaviors, and 5) able to provide informed consent.

Since early psychosis is considered a high-risk period for violence, all EIS participants being treated by participating clinicians who do not have histories of pre-existing antisocial behaviors will be eligible to participate in the study. Pre-existing antisocial behaviors will be identified with six questions from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study to identify major conduct problems before the age of 15 (i.e., skipping school often, running away from home more than once, deliberately destroying property, often starting fights, having ever been arrested or in juvenile court, or having ever been suspended from school).(Swanson et al., 2008) Individuals scoring a 0 or 1 (i.e., below the median of the CATIE study sample) will be included in the study. This approach was successfully used in CATIE to stratify individuals with psychosis-related violence (that is, violence which appears influenced by the symptoms of schizophrenia) versus individuals with violence related to pre-existing antisocial behavior.(Swanson et al., 2008) Current estimates suggest that fewer than 15% of OnTrackNY participants have pre-existing antisocial behavior (Rolin 2021); therefore, the majority of EIS participants will not be excluded. Exclusion based on this characteristic will be tracked and recorded.

EIS clinicians will be masters- or doctoral-level clinicians with extensive experience working with patients with psychotic disorders. Formal refresher training in CBT will be offered prior to intervention delivery.

Discontinuation and Withdrawal

EIS clinicians must remain employed with OnTrackNY (or an affiliated clinic) for the duration of the study. Similarly, EIS participants must remain in treatment with OnTrackNY (or an affiliated clinic) for the duration of the study. If an EIS clinician or participant is discontinued early, the week 12 follow-up assessments will be completed time of discontinuation.

Patients who engage in violence or are hospitalized during the study will be allowed to continue in the study if they return to treatment with OnTrackNY, in order to allow researchers to examine patterns of violence, including tracking the frequency and severity of multiple violent incidents over time.

Intervention

In the initial study by Haddock et al. (Haddock et al., 2009), the PICASSO intervention (a CBT-model targeting violence, anger, and psychosis) consisted of 25 sessions and was offered in an individual format to a population of primarily forensic inpatients. This intervention is individualized to address unique participant needs, which may include (but are not limited to) coping strategies, reality monitoring skills, and belief modification related to behaviors identified during treatment. Violence will be assessed through the MacArthur Community Violence Interview, which measures includes physical acts of violence against another person such as pushing, kicking, or choking someone. (Cartwright et al., 2020)

In an earlier phase of research, formative research was conducted to inform the adaptation of the PICASSO intervention.(Rolin et al., 2023) Figure 1 displays a proposed model of violence based on both this formative research and risk factors identified in existing literature. See Figure 2 for examples of anticipated adaptations based on preliminary results from qualitative interviews with EIS staff and EIS participants. Further modifications are anticipated based on the results of the formative research phase to produce an intervention that has relevance, acceptability, and potential effectiveness for young adults with early psychosis receiving treatment in an EIS setting.

Figure 1.

Proposed Model of Violence

Figure 2:

Potential adaptions of the PICASSO intervention

All proposed modifications will be systematically evaluated by an Advisory Council that has been specifically convened to guide this study. This group meets monthly and includes representatives from OnTrackNY, as well as members with lived experience relating to topics addressed in the intervention.

Outcome measure

During the study intervention, the research team will obtain feedback about the feasibility and acceptability of the PICASSO intervention from both EIS clinicians and EIS participants. The primary outcomes of this open pilot trial will be feasibility and acceptability of the intervention, with collection of mixed methods data to inform further adaptation (see Table 1 for proposed schedule of assessments). Quantitative measures of hypothesized mediators of target outcomes will also be collected on a weekly basis.

Feasibility will be assessed by tracking data pertaining to recruitment, retention, and assessment procedures following CONSORT guidelines, including the duration and number of contacts per participant. EIS clinicians will provide feedback on feasibility and acceptability in weekly assessments, using the Feasibility of Intervention Measure(Weiner et al., 2017) (FIM) and the Acceptability of Intervention Measure(Weiner et al., 2017) (AIM), as well as semi-structured interviews at weeks 4, 8, and 12. EIS participants will provide feedback on feasibility and acceptability weekly in brief post-session cognitive interviews, with questions regarding the content’s appropriateness, strengths and weaknesses of the therapy, value of the current adaptations, and recommendations for improvement.

Safety and AEs

A Data Safety and Monitoring Board has been convened for this study. Information concerning AEs will be collected starting after informed consent is obtained. All EIS participants will continue to receive treatment at their OnTrackNY clinic as usual, along with the intervention. They will continue to have access to usual patient resources including clinical coverage for any emergencies.

Procedures

Trained research personnel will screen interested EIS clinicians for eligibility by phone or in-person. If a clinician is deemed eligible, personnel will describe the study in detail, answer all questions, and schedule a session to complete the consent process (either in-person or remotely, based on participant preference).

Once an EIS clinician decides to enroll in the study, they will be asked to identify up to five interested and eligible EIS participants they are currently working with. The screening, eligibility determination, and consent process for patients will occur as described above for EIS clinicians. To increase the likelihood of the study’s completion, EIS participants will complete the study intervention with their own EIS clinicians, with whom they have an existing therapeutic alliance. The appointments will take place with their EIS clinician (either in-person at their clinic location or remotely, based on their clinic’s guidelines), at a time of mutual convenience.

Based on clinic policy, EIS clinicians will not be directly compensated, as their clinics receive funding to facilitate study participation and clinic protocols prohibit direct compensation. EIS participants will be paid $50 for the baseline assessment and week 4, 8 and 12 assessments and $25 each for the other weekly assessments. There will be no compensation for psychotherapy intervention visits.

To increase the likelihood of retention, EIS participants will be completing this therapy with their own EIS clinicians, with whom they have an existing therapeutic alliance. The appointments will be at their EIS clinic, at a time of their convenience. Barriers such as site location inconvenience will be addressed by providing reimbursement for transportation. Scheduling conflicts will be addressed by offering make-up sessions for missed sessions. Make-ups with also be offered for a variety of extenuating circumstances, including (but not limited to) forgotten appointments or illness as per clinic policy.

Statistical analysis

Qualitative data analysis will be guided by standard methods: immersion, crystallization, and thematic content analysis. Together, two trained coders (SAR and MGF) will develop an open coding schema based on a priori and emergent themes, and will present it to the qualitative methods expert (LP) to discuss emergent themes, refine codes, and develop a final analytic framework and codebook. To establish coding reliability, the coders will independently code up to 10% of transcripts, meeting weekly during this process to review definitions and assignments of codes and to resolve differences. The coding scheme will remain flexible and open, with some codes predefined and others developed iteratively to accommodate expansion. All coding will be performed in Dedoose. Quantitative data about the hypothesized mediators will be presented descriptively, with analysis of inter- and intra-individual patterns of behavior over the time of the intervention. Two to 3 rounds of iterative modifications are expected during the open pilot trial, with 4 EIS clinicians delivering the intervention to 1-2 EIS participants per round (total 10-16 EIS participants). This mixed methods framework was selected to holistically draw inferences from both the quantitative and qualitative measures conducted in this study.

Data management

Data collection and collation will following the guidelines of Good Clinical Practice.(McFadden, 2007) Monitoring of the trial will follow the procedures outlined by the DSMB.

Results

Recruitment will begin in December 2023 and is expected to occur over a 9-to-12-month period.

Discussion/Significance

This protocol presents the first clinical trial of a behavioral intervention to reduce violence for young adults with early psychosis. Though most people with SMI, including psychosis, are not violent, this study’s potential impact is great given the cumulative public health impact of violence among all adults with psychosis. One-third to two-thirds of caregivers to individuals with psychosis report one or more incidents of physical violence or aggression within a given year.(Chan, 2008) If a caregiver calls emergency services for assistance, a person with mental illness is at increased risk of being killed by police in their homes, even if not armed with a firearm.(Saleh et al., 2018) Overall, adults with schizophrenia are more likely to be convicted of a violent crime than their age-matched peers, and are over-represented in the criminal justice system.(Kooyman et al., 2007) Violence also contributes to intergenerational trauma, which imposes a lasting impact on families and communities.

Though innovative, this study design has a few limitations. First, the study is based on a state-wide EIS program, and it is possible that findings will not be generalizable to other settings. However, this pilot study seeks to ascertain new information relevant to the intervention rather than transfer conclusions to other settings. In addition, there is stigma about discussing violence and it is possible this stigma will limit the willingness of research participants to engage in a study about violence. In response, the research team hopes to limit the effect of this stigma by normalizing discussion about violence in treatment, as data suggest that up to a third of participants experience violent ideation or violent behavior.

The intent of this trial is to develop a behavioral intervention that could be offered on a voluntary basis to young adults with early psychosis, a population at higher risk of experiencing violent ideation or violent behavior. An intervention incorporating attention to the risk factors for violence identified by existing research could have life-altering impacts for young adults with psychosis in EIS treatment. Currently, few interventions are available for individuals who are thought to be at higher risk of violence due to mental illness, and even fewer of these interventions are offered as voluntary preventive measures. While there is public support for the use of coercive methods to manage violence risk for people with schizophrenia (Pescosolido et al., 2019), controversy persists as to the effectiveness of these strategies.(Morris & Kleinman, 2022) Future research will seek to test this adapted intervention in a RCT.

Data Availability:

Data availability is not applicable to this article as no new data were created or analyzed in this study.