Mentalization-based treatment (MBT) is an evidence-based, fully manualized, specialist form of psychotherapy for personality disorders (Bateman & Fonagy, 2008; Bateman & Fonagy, 2009; Laurenssen et al., 2018; Juul et al., 2023). MBT is rooted in psychodynamic and attachment theory and aims to alter personality pathology by improving mentalizing and enhancing epistemic trust. Eligible programs must adhere to the minimum standards described in the evidence-based literature. MBT programs may differ in intensity and duration, even at the same site, but must include a minimum of either 40 sessions or a duration of 12 to 24 months. In additon, in order to be eligible, an MBT program must be of higher intensity or longer duration than the generalist treatment (GIT-PD) offered at the same site.

GIT-PD is a non-theoretical and non-methodological approach, based upon principles derived from common factors across evidence-based treatments for PD. GIT-PD has a flexible approach, enabling tailoring treatment to clinical needs of different types of patients. The treatment duration (start-end) is 12- 18 months. The treatment consists of three distinct phases. First, an assessment phase. Second, a modular treatment phase that offers either group or individual interventions. Third, a follow-up phase with limited treatment and focus on relapse prevention. In order to be eligible, GIT-PD should be less intensive than specialist psychotherapies offered in the control arm at the same location (i.e. MBT or ST). GIT-PD will not include evidence-based psychotherapy for personality disorders as described in previous RCTs.

The BSI is a self-report questionnaire investigating symptom severity, consisting of 53 items which can be answered using a scale ranging from 1 (not at all) to 5 (extremely). The total score represents the number of symptoms that participants experienced during the past week. A global severity index and sub scores on all 9 domains are calculated by summing the values on all items and dividing by the number of items/domains. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

Suicidality is assessed using the C-SSRS. This self-report questionnaire covers 6 items and has been recommended by ICHOM as a standard outcome measurement in personality disorder research, due to its adequate psychometric properties (Prevolnik Rupel et al., 2021). The 6 items are answered with either 'yes' or 'no'. The total number of positive answers indicates suicide risk. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

The SCID-5-PQ (First et al., 2016) is a self-report measure assessing different criteria for DSM-5 personality disorders according to Section II. Participants are presented with 106 items describing whom they are as a person providing them with a dichotomous 'yes' or 'no' answer scheme. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

Social and professional functioning will be measured using the WHODAS 2.0 self-report questionnaire, which includes 12 items. This questionnaire has been recommended by ICHOM for its good reliability, adequate validity and high sensitivity to change (Prevolnik Rupel et al., 2021). Participants answer 12 questions across 6 domains on their daily functioning during the last 30 days. Answers across the 6 domains are weighted and summed to create a total score of disability ranging from 0 to 100. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

The EQ-5D-5L is a self-report questionnaire that covers 5 items measuring quality of life across 5 domains. Answer options consist of 5 levels ranging from 'no problems' to 'extreme problems'. The level indicates the health state for that domain and can be combined into a 5-digit number indicating the overall quality of life of the participant on that day. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

The Q-LES-Q-SF is a self-report questionnaire consisting of 16 items.. Participants rate their satisfaction with different aspects of quality of life on a 5-point scale from "not at all or never" (1) to 'frequently or all the time" (5). The possible scores range from 14 to 70 with higher scores indicating greater satisfaction with life (Stevanovic, 2011). Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

The TIC-P for Adults will be used to assess healthcare consumption and productivity loss during the last 3 months. Through 57 items the costs of different types of care are determined, e.g. for medication use, sessions with therapists, doctor's appointments and more indirect costs like loss of productivity and informal care by the participants network. The first section comprises 8 general questions. The second section includes 22 items regarding healthcare consumption. The last part consists of12 items on informal care and productivity loss. Some items feature multiple follow up questions resulting in the total number of 57 items. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

The PCL-5 aims to screen for post-traumatic stress disorder using 20 items that correspond to the DSM-5 criteria (Blevins et al., 2015). Likewise, the four subscales correspond to the DSM-5's clusters B through E for PTSD. The 20 items are answered on a 4-point scale ranging from 0 (not at all) to 4 (extremely) resulting in scores ranging between 0 and 80. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

The 5-item version of the Autism Spectrum Quotient (AQ-5) is a very brief screener, which has been derived from the AQ-10. The AQ-5 measures social communication deficits, corresponding to Criterion A of the ASD-diagnosis. The scale has acceptable internal consistency and convergent validity. Items are rated on a 4 point scale ranging from 'definitely disagree' to 'definitely agree'. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

Patient commitment to treatment is measured with a selection of items from the subscale Motivation to Engage of the TMS-F. The four items can be rated on a 7-point Likert scale . Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

A modified version of the SRIS is used to assess self- reflection and insight. The SRIS contains 20 5-point Likert scale items ranging from 1 (strongly disagree) to 5 (strongly agree). Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

Internal locus of control, defined as the extent to which a person experiences an outcome as the result of their own behavior or personal characteristics rather than external circumstance, is assessed by the Locus of Control scale (IE) . The IE contains 10 5-point Likert scale items. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

The MHC-SF was developed by Keyes (2008) to measure hedonic and eudaimonic well-being, which can be seen as a separate, but related, dimension to mental "ill-being" (Franken et al., 2018; Keyes, 2007). The questionnaire contains 14 items and three dimensions: emotional well-being (3 items), psychological well-being (6 items) and social well-being (5 items). Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

By using the social problems list, derived from the (IAPT) program social problems (e.g., financial problems, housing problem, and unemployment) are assessed (Clark, 2011). Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

The MSPSS (O'Toole et al 2012) is assessed to investigate perceived support from three sources: significant others, family, and friends. The MSPSS contains 12 items which can be rated on a 7-point Likert scale ranging from 1 (very strongly disagree) to 7 (very strongly agree). Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

PD has been associated to stigma of immutability (Aviram 2006) and we will use the five items scale as developed by Wibbelink et al (2022), measuring the extent to which participants believe that PD is resistant to treatment. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

Verbal Intelligence will be measured by the Dutch Reading Test for Adults ('NLV'). Psychometric properties have been shown to be satisfactory among various populations (Hermans et al., 2022). Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

By administering the 24-item QET, insight is gained into a person's developmental capacity to accept information as authentic, trustworthy, generalizable and relevant to the self (i.e. epistemic trust; Knapen et al., 2023). The instrument sheds light on four factors, referring to hypervigilance, curiosity/openness, expectation of help and openness to help. Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.

The PID-5-SF is a shortened 100 item version of the original PID-5 (Maples et al., 2015). It measures 25 facets across 5 domains of pathological personality traits, which corresponds to criterium B of the AMPD (American Psychiatric Association, 2022). Data collection is done digitally via Castor or with pen and paper, based on participants preference. Pen and paper results are copied into Castor by hand by research assistants. Raw data is digitalized and stored on a local drive.