Brief internet-delivered skills training based on DBT for adults with borderline personality disorder – a feasibility study

Abstract

Objective

Borderline personality disorder (BPD) is characterized by instability in emotions, relationships, and behaviors, such as self-injury and suicidal behavior. Dialectical Behavioral Therapy (DBT) is an established intervention for BPD, but there are long waiting times for treatment. This study aimed to explore if a brief internet-delivered DBT skills training program with minimal therapist support is acceptable, that it can be administered, useful, and does not do harm for patients with BPD.

Methods

Acceptability was measured through data on recruitment and attrition, utilization of the intervention, reported impulses to drop out, and through ratings on self-injury and suicidality. Participants were interviewed about their experiences of the intervention; analyzed with content analysis.

Results

Twenty patients on the waiting list for treatment at a DBT-clinic were invited and nine female patients (age 19–37 years) volunteered. The participants completed a large part of the intervention, which did not appear harmful since ratings of suicidal and self-harming behavior were similar before and after the intervention. In the interviews, participants stated that they had gained new knowledge and skills to manage situations, e.g. to stop and think before acting. Some even reported decreased levels of self-injury. The time spent on patient contact was short, and some patients reported difficulties to practice on their own and requested more support.

Conclusions

The intervention seems to be acceptable. Future studies should investigate in what ways some BPD patients are more susceptible to internet-delivered skills training than others, and if this intervention could be delivered within a stepped-care model.

Keywords:

• Borderline personality disorder
• emotionally unstable personality disorder
• dialectical behavior therapy
• internet
• skills training

Introduction

Borderline personality disorder (BPD), also known as emotionally unstable personality disorder, is characterized by a longstanding pattern of instability in emotions, relationships, self-image, and behaviors such as non-suicidal self-injury (NSSI) and suicidal behavior [1]. In addition, individuals with BPD often suffer from comorbidities such as anxiety, mood and/or substance use disorders [2], as well as severe impairments in functioning [3] and reduced quality of life [4]. BPD has an estimated point prevalence of 1% in community settings, 12% in psychiatric outpatient care, and 22% in psychiatric inpatient care [5], and has been linked to increased risk of premature death both from suicide and from non-suicidal causes [6]. Retrospective studies suggest that up to 10% of BPD patients die by suicide [7], making it a major public health concern. Furthermore, BPD is associated with high economic costs, related to both high treatment costs and indirect costs from work disability [8]. Accordingly, it is imperative to implement treatments at an early stage, that are acceptable and effective for this group of patients.

According to the National Institute for Health Care Excellence (NICE), pharmacological treatment should not be used specifically for BPD symptoms such as emotional instability, NSSI, or risk-taking behaviors, but could be considered for managing crises and comorbid symptoms [9]. Instead, the guidelines recommend various psychological treatments for targeting BPD symptoms. The most established psychological treatment for BPD is Dialectical Behavioral Therapy (DBT), which in its full standard form (standard DBT) includes a mix of weekly 1-h sessions of individual therapy, 2 h of skills training in a group, phone coaching between sessions, and team consultation for therapists [10–12]. DBT, which has most commonly been evaluated as a 1-year treatment, has demonstrated positive results in previous studies, evidenced by a reduction in NSSI, suicidal thoughts, plans, and attempts, and acute psychiatric service use/health care utilization [13–18], as well as an increased usage of skills and a reduction of BPD symptoms [17,19]. DBT has also demonstrated positive effects on comorbidities such as depression [16] and substance dependence [20].

Although DBT is the most established psychological treatment for BPD, the demand for evidence-based treatments for BPD is greater than mental health care providers can meet [21], resulting in long waiting times for receiving treatment [22]. The unmet demand for psychological treatments has partly been addressed through internet-delivered interventions, with satisfying results for a variety of psychiatric conditions such as major depression and anxiety disorders [23], internet addiction [24], and problematic drinking [25]. In a recent scoping review, 41 studies on telepsychology in DBT were found [26], and promising results were reported for internet-delivered skills training on skills practice, emotion regulation and mindfulness in a study including adults with mood and/or anxiety disorders [27], and on reduced suicidal ideation, alcohol use and emotion dysregulation in a study including suicidal adults engaged in heavy episodic drinking [28]. In addition to phone coaching, that is already used in standard DBT, the review found that internet-delivered DBT skills training and mobile phone applications for coaching and/or enhancing generalization of skills were the most common digital DBT-interventions [26]. Although none of the studies on internet-delivered DBT skills training in this review included BPD patients [26], an even more recent scoping review [29] found one internet-delivered DBT intervention where BPD patients were emailed PowerPoint slides that covered general information of a topic, related skills, and homework sheets, on a weekly basis. The study reported promising results for improved self-rated emotion regulation at post-intervention assessment [30].

In addition to addressing the unmet demand for BPD treatments, internet-delivered interventions targeting BPD patients could help address the lack of access to evidence-based treatments for those living far from treatment centers or in confinement [31]. In internet-delivered treatments offered to BPD patients, consideration must be taken of the high frequency of NSSI and suicidal behaviors in this patient group, making it important to carefully evaluate that such treatments do not result in an increase of NSSI, suicidal behavior, or other detrimental outcomes. Recent systematic reviews have reported promising results for mobile- and internet-delivered psychological interventions targeting self-harm [32] and low risks of harm in internet-delivered treatments targeting suicidality [33], indicating that internet-delivered treatments may be acceptable for BPD patients. There is also some evidence that younger generations prefer internet-delivered health care over traditional face-to-face interventions [34,35], and a wide range of services now integrate telecommunication and digital technologies in interventions based on DBT [26].

To summarize, although standard DBT is the most common intervention for BPD, the demand for DBT exceeds capacity resulting in long waiting-lists, making internet-delivered treatments an important treatment option for BPD patients. There is however a shortage of studies that have explored patients’ experience of stand-alone internet-delivered DBT skills training specifically targeting adult BPD patients within a psychiatric setting. Therefore, we aimed to explore patients’ experiences of a brief and adapted DBT skills training program delivered through the internet. We wanted to investigate if the program is acceptable, that it can be administered and useful and safe for patients with BPD in a psychiatric setting.

Methods

The intervention

The skills training program used in this study was an abbreviated version of the most central psychoeducation, exercises, and Swedish worksheets [36] in the DBT skills training manual [12], developed in cooperation with experiences therapists from the DBT clinic. The skills training consists of four modules teaching skills for managing common BPD symptoms: (i) mindfulness, (ii) distress tolerance, (iii) emotion regulation, and (iv) interpersonal effectiveness [12], Table 1. Each of the four modules were administered over the course of 2 weeks including home assignments which patients were encouraged to work with daily. In the original skills training manual [11], completing all the modules takes around 6 months and it is recommended for those who are new to DBT to repeat the modules by staying in the skills training group for at least 1 year. In our abbreviated version each of the 8 weeks began with an audio exercise in mindfulness, following the standard agenda for the DBT skills group. In the final week (week nine), patients were encouraged to summarize the treatment and create a maintenance plan. All patients were expected to work independently at home with the materials which was reached through Sweden’s national hub for health and healthcare information, advice and e-services, requiring the patient to log in using electronic identification.

Table 1. Included modules of the internet-delivered skills training program.

Module	DBT skills	Description
Mindfulness	‘WHAT’ skills (Observe, Describe, Participate), Wise-mind, ‘HOW’ skills (Non-judgmentally, One mindfully, Effectively)	Skills in perceiving and describing the present
Distress tolerance	STOPP (skill in acting mindful), TIPP (skills in changing body chemistry), Distraction, Turning the mind, Half-smiling, Radical acceptance, Willing hands	Skills in tolerating distress
Emotional regulation	Identifying and labeling primary emotions, Checking facts, Acting opposite to inefficient emotions	Skills in changing inefficient emotions
Interpersonal effectiveness	Clarifying goals and priorities, DEAR MAN (being assertive and saying no), Validation	Skills in maintaining and improving the relationship to others and oneself

Participants

The study group was recruited from the waiting list for standard DBT at the DBT clinic of Uppsala University Hospital in Sweden, excluding those expected to initiate their standard DBT during the study. Prior to asking patients about participation, the local health care providers of each respective patient were asked to evaluate whether an intervention of this kind was appropriate, given the patient’s symptoms and functioning, including sufficient proficiency in the Swedish language. For inclusion, all participants should have a clinical diagnosis of BPD/emotionally unstable personality disorder according to ICD-10. The diagnoses were collected from the medical records and validated using the Structured Clinical Interview for DSM-IV Axis II Disorders (SCID-II). The licensed psychologists who conducted SCID-II were trained by experts. Co-ratings showed a prevalence- and bias-adjusted kappa between 0.6 and 1.0.

Procedures

The intervention was initiated during a meeting at the DBT clinic with the therapist in charge. Self-rated pre- and post-measures were collected, see below. At the initial meeting, the skills training was presented, as was the project’s framework. All participants were informed, both verbally and in writing, that their local health clinic kept their duty of care during the online intervention and that the function of the DBT therapist was to support the patient’s use of the online intervention. Patients were introduced to the national hub, where the intervention was administered, and given a folder containing all the home assignments and weekly diary cards. Patients were informed that they were expected to report their weekly diary cards and home assignments prior to a weekly fixed time when the therapist also gave feedback and access to the following week’s material, regardless of whether the patient had completed their homework or not. This was followed by an introduction to mindfulness, including an exercise. Lastly, a personalized crisis plan was written on a worksheet, in accordance with DBT praxis, and all patients were encouraged to have their crisis plan easily accessible. All crisis plans included contacting the emergency clinic for adult psychiatry as a last step. In cases where a patient reported suicidal plans and/or drastic changes in suicidality at the weekly check-up, the therapist directed the patient to their crisis plan and the emergency clinic for adult psychiatry, and informed the patient’s local health clinic. After the intervention, all participants, regardless of their amount of engagement in the program, took part in a semi-structured interview, either at the DBT clinic or via Skype or phone. The study was approved by the Swedish Ethical Review Authority (ref. no 2019-02265). All participants provided written informed consent to participate in the study.

Interviews

The semi-structured interviews were conducted by two licensed psychologists and a graduate student in medicine, none of whom had had prior contact with the participants. All interviewers had prior training in interviewing techniques. The interviews lasted between 16 and 46 min. An interview guide was made by the authors and included open-ended questions with follow-up questions, targeting participants’ experiences of the internet-delivered skills training program, including the initial meeting at the DBT clinic, working with the platform, content, length, and any outcomes of the intervention.

Measures

For baseline characteristics, BPD symptoms were assessed using the Borderline Symptom List-23 (BSL-23), a shortened version of the self-rating instrument originally named the Borderline Symptom List [37]. The BSL-23 includes 23 items targeting typically reported experiences and feelings among BPD patients during the preceding week. Items range from 0 (‘not at all’) to 4 (‘very strong’), resulting in a total mean score between 0 and 4, where a higher score indicates more BPD symptoms. A mean score of 2.05 has been reported among BPD patients [38]. In this study, Cronbach’s alpha for BSL-23 was 0.94.

Time spent by the therapist on patient contact was monitored at the weekly check-up for each participant. Information on opened online material, submitted home assignments, and impulses to drop out of treatment were retrieved from the national hub. Impulses to drop out from treatment (monitored using diary cards) were rated between 0 (indicated no impulse at all) to 5 (the strongest possible impulse) and a value of 1–5 were regarded as presence of an impulse.

Frequencies of NSSI, suicidal communication, and suicide attempts during the preceding week were assessed before and after the intervention using three complementary items to the BSL-23, measuring suicidal and self-harm behaviors. The items consisted of the following three statements: ‘During the last week, I hurt myself through cutting, burning, strangling, etc.’, ‘During the last week, I told others that I would commit suicide’, and ‘During the last week, I attempted suicide.’ The participants answered on a five-point scale: Not at all (0), one-time last week (1), two-times last week (2), daily (3), several times a day (4). Thus, higher scores indicated higher frequencies of the respective behaviors.

Data analyses

Acceptability was measured through data on recruitment and attrition, utilization of the intervention, reported impulses to drop out, and through self-ratings on NSSI and suicidality items, as well as through qualitative data on participants’ experiences of the intervention gathered from interviews. In order to assess that the intervention does not do harm, differences in pre-post data on NSSI, suicidal communication, and suicide attempts were tested with Wilcoxon’s matched-pair test using IBM SPSS Statistics (version 23.0). Missing data were imputed with mean values for each respective subscale/scale. Two participants had missing data on BLS-23, each for one item. A p-value < 0.05 was regarded as significant.

For the qualitative analysis, an inductive content analysis [39] was carried out for the purpose of identifying categories and subcategories in the manifest content derived from the interviews. First, two researchers independently read the anonymized transcripts of each interview several times to fully comprehend the essence. Subsequently, all units of meaning were extracted from the transcripts, and the essence of each meaning unit was condensed. This was followed by an abstraction conducted collectively by all six members of the research group, where contents with similar meaning were grouped together into main categories. Disagreement in the research group was resolved through discussion. Main categories were subsequently divided into subcategories based on similarity or dissimilarity. To validate the data, the authors used a back-and-forth approach throughout the process.

Results

Recruitment and attrition

Twenty-one patients on the waiting list were not expected to initiate their standard DBT during the study period and therefore potentially available for inclusion. Twenty (of 21) health care providers were reached, and all twenty found the intervention appropriate for their patient. Therefore, 20 patients were offered the internet-delivered skills training, and nine (45%) chose to take part in the intervention and evaluation. The eleven patients who did not participate in the intervention and evaluation did not differ significantly from the study population in terms of gender and age. All nine participants had a BPD diagnosis and a majority (n = 7) also had comorbid psychiatric diagnoses according to ICD-10, see Table 2. Mean value on the BSL-23 was 1.99 (SD = 0.89).

Table 2. Characteristics of the study population (n = 9).

Gender	Female n = 9
Age Mean (SD)	27.89 (6.31)
Marital status	Living together with partner n = 5
	Single n = 3
	Living apart from partner n = 1
Education	Elementary school n = 4
	High school n = 2
	University n = 3
Occupation	Unemployed n = 1
	Sickness benefits n = 2
	Part-time employment (25–75%) n = 4
	Student n = 2
Number of met BPD-criteria Mean (SD)	5.78 (0.83)
Criterion 1: Uncontrolled efforts to avoid	n = 6
real/Imagined abandonment
Criterion 2: Unstable and intense relationships	n = 3
Criterion 3: Identity disturbance	n = 4
Criterion 4: Impulsive/Self-destructive	n = 5
behaviors
Criterion 5: Recurrent suicidal behavior	n = 8
Criterion 6: Affective instability/Intense	n = 9
emotional swings
Criterion 7: Chronic feelings of emptiness	n = 5
Criterion 8: Inappropriate/Intense anger or	n = 5
difficulty controlling anger
Criterion 9: Stress-related paranoid ideation or	n = 5
severe dissociative symptoms
Comorbidities
Any	n = 7
Depressive disorders	n = 6
Anxiety disorders	n = 4
Asperger Syndrome	n = 1
Attention Deficit Disorder	n = 1
Bipolar Disorder	n = 1
Dissociative Disorder	n = 2
Schizophrenia, schizotypal and delusional
disorders	n = 1
Substance Abuse	n = 1

Note. BPD: Borderline personality disorder.

Utilization and potential negative outcomes

The average clinical time per patient and week was 16.2 min (range 7.8–23.9 min/week), including responding, motivating for improving compliance, directing a patient to other health care clinics, and, if needed, contacting the local health clinic. On average, participants opened 75.4% (range 35.0–100%) of the online material and submitted their home assignments at the weekly check-up for 6.2 (SD 2.9, range 2–9) out of 9 weeks. On average, participants reported presence of impulses to drop out from the intervention on 1.8 days/week (SD = 2.6). As shown in Supplementary Table 1, self-ratings showed no increased (or decreased) levels of NSSI, suicidal communication, and suicide attempts at a group level. At an individual level, the majority had no reported NSSI or suicidal communication the last week pre- and post-intervention, two participants reported fewer NSSI during the week after the intervention while one reported more NSSI, as compared to the week before the intervention; two participants reported less suicidal communication after the intervention while one reported more. No participant reported any suicide attempts.

Qualitative results

Five categories and 17 subcategories on how the intervention was experienced were identified in the qualitative analysis. The categories and subcategories, with illustrative quotes, are summarized in Table 3.

Table 3. Main categories and subcategories with illustrative quotes.

Category	Subcategory	Illustrative quotes
Working with internet-delivered skills training	The online format	‘I thought it was very good that it was online based, because I did not have to keep any appointments… because I think that is very hard… I thought that it was good that you had access to the material whenever you wanted.’
	DBT skills	‘I liked the audio exercises because it felt more like someone was actively participating, because someone was talking and telling you what to do.’
	Diary card	‘The diary cards were a little bit like this, I kind of thought that they were unnecessary. Or I cannot say that but for me it was kind of… I had gotten a long way in not self-harming and in the majority of them there was not any change …’
	The length of the skills training program	‘I actually think that it needs to be a little bit longer, because if I were to think of why it did not fit me, it was because it takes a while before you get into how you should answer the questions and how you should think.’
	Technical issues	‘…it was a little bit frustrating to do the assignments, because you sat and waited for something to happen, sometimes it crashed…’
Allocated time		‘… I have a pretty stressful everyday life. Things are happening all the time and then you have to squeeze in the skills training into this.’
The personal contact	Initial meeting with the therapist	‘…it was good as well to get a face on her… so then it felt easier when you wrote, knowing who it is, that was nice’.
	Helpful but scarce	‘It felt like it might as well have been Artificial Intelligence kind of… it did not feel so personal…’
Personal struggles	Difficulties in initiating and maintaining skills practice on their own	‘There was no checking and then I did not really feel like I wanted to do it.’ ‘…when it is online you can choose the time you work with it, and that was not good for me. Because it always came last on the to-do list.’
	Stress and/or pressure in doing home assignments	‘I think it was more my responsibility, that I just did not give myself enough time, and therefore it got stressful.’
	Distressing to practice skills	‘There was one skill that triggered me… I got impulses to self-harm and anxiety… but then afterwards I have looked back and thought but this is kind of good’.
	Distress from the feeling of using skills incorrectly	‘…if this does not work on you then there is something wrong with you, kind of.’
Outcomes of the skills training program	Perceived well-being	‘As you have thought that okay, the waiting time is long, but we can do something meanwhile to make it easier in some way and I thought that was very good…’
	Gained and/or maintained knowledge and skills	‘that you get yourself some time and… that you do not only act on your feelings either, and try to kind of think that okay, now I will stop here, instead of…’
	Decreased levels of self-harm	‘…so I have been able to tell my mother or my father that now… ‘I saw a pair of scissors, remove them or stay at home I do not want to be alone’, that I otherwise would have told them afterwards…’
	Personal growth	‘I felt as if I kind of grew a lot in a short time… as a person… I had a pretty immature behavior before…’
	Feelings of hope	‘I think that the whole idea with this treatment is really good, and I think that it can give a lot of people a feeling of hope, because when you hear … it is 15 months of waiting time, then you just feel a little bit like oh well.’
	Opinions on their upcoming standard DBT treatment	‘…I look forward to it after doing this short one, now I know what it is about, and what I like.’

Working with internet-delivered skills training

Participants perceived the online format in different ways. Those who were critical of the format stated that they lacked someone to discuss the material with and reported too little personal contact overall. Some perceived that they did not have the self-discipline that the online intervention required. Those who were positive experienced that the online format gave them freedom to work with the material whenever they wanted and hence could avoid taking time off work. Some appreciated not being obliged to participate in groups, and it was valued that the work from previous weeks was still available, making it possible to go back and repeat old material.

All participants reported that they acquired knowledge about DBT skills, while they had opposing views on which specific skills training they preferred. Some appreciated the audio exercises in mindfulness, partly because they felt that the therapist participated actively. On the other hand, others perceived mindfulness as ‘wimpy’, and some struggled to understand the purpose of it. Distress tolerance skills, which are used to cope with and tolerate distress without acting on destructive impulses, were perceived as helpful by some, while others found them difficult and wished they had been presented later in the program. It was mentioned that the distress tolerance skills and crisis plan could be perceived as meaningless in times of depressed mood. The emotion regulation module was described as helpful in capturing specific emotions. The last module, interpersonal effectiveness, was experienced as difficult by some, and it was also mentioned that if one could not relate to the examples in the material, one would have a hard time finding other situations to apply the skill to. Some participants reported that the overall material was too basic. One suggestion was to give access to more in-depth material.

The diary card that was used for screening of drop out inclinations, self-harm, and suicidality, was experienced as either helpful or pointless. It motivated some participants to activate themselves and was reported to decrease self-harm impulses and suicidal thoughts. Self-harm was less prevalent among those who reported the diary card as pointless, and therefore there was no change in ratings over time. One suggestion for improvement was to include ‘weekly goals’ in the diary card, mimicking the standard DBT praxis.

None of the participants said that the length of the skills training program should have been shorter, but many indicated that they would have appreciated if the program had been longer. One concern regarding the length was that the format could be too demanding during periods when a patient suffered from more symptoms.

All participants reported technical issues during the intervention, such as inaccurate notifications, and that the platform was slow, causing frustration when working with home assignments. A suggestion for improvement was to be able to choose the frequency of notifications.

Allocated time

The allocated time for the skills training varied. Some reported that they only worked through approximately half of the material, though others reported that they spent a significant amount of time on the material each week. Some claimed that they worked with the skills training ‘kind of all the time’, that they logged in to the platform several times each week, and that they often went back to old material from previous weeks. Others reported that they mainly worked with the material on the last day before the check-up, and skipped some parts. Most participants reported reading and working with the material 1–2 h per week.

The personal contact

Participants expressed satisfaction with the initial meeting with the therapist. They appreciated getting a personal contact with her and argued that it made it easier for them to later correspond with her via the platform. Participants reported that the initial meeting clarified the framework of the intervention. However, some reported that there was too much information given during this meeting, and one suggestion for improvement was to provide a timeline, to give participants a better sense of the framework.

Overall, the online communication with the therapist was experienced as helpful but scarce. Most participants wanted more frequent and personal contact with the therapist either through live chat, phone and/or meetings in person. A couple of participants stated that the feedback was too impersonal and generic, like ‘artificial intelligence.’ Difficulties in proceeding with the training when feeling low were reported, which also decreased the motivation to write to the therapist for the weekly check-up. One suggestion for improvement was to increase the proportion of live communication, to make it easier to ask questions. An alternative that was mentioned was the possibility to schedule a live chat if a patient expressed a need for this.

Personal struggles

Many participants experienced difficulties in initiating and maintaining skills practice on their own. Although participants reported that they had enough time to work with the skills, some still did not manage to allocate their time and practice on their own. Several participants described increased difficulties during periods with more symptoms, and expressed the need for more support during such periods. Some reported the need for more support from their household and/or a need for fewer things being scheduled in their life during the skills training.

Several participants also said that they experienced stress and/or pressure in doing home assignments, even though they could not remember anything that was said or done that made them feel that way. Some described that the home assignments generated a feeling of schoolwork.

It was mentioned that it could be distressing to practice skills, especially in the beginning, when participants were not accustomed to them. Emotion regulation and interpersonal skills were reported to cause a lot of distress, but it was also stated that they were necessary to learn.

It was mentioned that, when the skills were not working or not deemed efficient, this generated distress from the feeling of using skills incorrectly. When expressing this to the therapist, it was felt that the feedback was to try more or do it in a different way, which made them feel like there was something wrong with them.

Outcomes of the skills training program

Most participants reported no evident change in perceived well-being during the skills training program, but all participants reported some positive experiences from the intervention. While some reported periods of decreased well-being, they did not link it directly to the program. At the same time, some participants reported that they were feeling a lot better during the skills training program, and that they were sad when it came to an end. Some also stated that they would like to redo the program.

All participants reported that they appreciated at least a few of the skills and many that they gained and/or maintained knowledge and skills. More specifically, participants reported that they learned to manage situations in new ways, for example to stop, take a step back, and think before acting, and that the skills training served as a good distraction and/or occupation. A few patients had completed DBT as adolescents and said that this skills training helped them a lot in that it was a reminder to use the skills.

Some participants reported decreased levels of self-harm during the skills training program, including being free from self-harm for a period of up to 2 months at the time of the interview. None of the participants reported an increase of NSSI or suicidality due to the skills training program.

Others reported personal growth claiming that the skills training, ‘in a good way’, made them think in a completely different way from what they were used to, that they grew a lot as human beings, and that others had said that they perceived them as calmer after the intervention.

It was also reported that the skills program generated feelings of hope, as they were prone to feel hopelessness because the waiting time for standard DBT was so long, and the reason for wanting to participate in the skills training program was to be able to do something in the meantime.

Patients had varying opinions on their upcoming standard DBT treatment. Some said that they looked forward to their standard DBT more now that they had been through the skills training program and knew what it was all about. On the other hand, some participants reported the reverse, that the skills training program was not helpful, because they had undergone similar treatments in the past and that they therefore did not know what effect doing it again might have, or that they would have preferred to do the internet delivered intervention post standard DBT.

Discussion

The present study examined the acceptability of a brief internet-delivered skills training based on DBT, with minimal therapist support, for psychiatric patients with BPD. Main findings showed that nine of the 20 initially invited patients accepted participation in the intervention and evaluation. A majority of the online material was opened and most home assignments were submitted. Importantly, for valuation of safety, the intervention did not result in statistically increased levels of suicidal or self-harm behavior when comparing pre- and post-measures, which is in line with previous studies on internet-delivered treatments targeting suicidality [33]. In the interviews, participants stated that they had gained knowledge and skills to manage situations in new ways. Some even reported decreased levels of NSSI and expressed feelings of hope. Several suggestions for changes were reported; most commonly, participants stated that they wanted more therapist support.

As reported above, more than 50% of the patients chose not to participate in the intervention. As BPD is a serious mental disorder, working independently with only minimal therapist support through the internet could be too challenging for many BPD patients within psychiatry. Other studies have reported that drop out from interventions are common from standard DBT interventions [40], but also from internet-delivered DBT skills training interventions [29], including interventions for suicidal and heavy episodic drinkers [28], as well as using smartphone applications targeting BPD symptoms [41]. In the study by Alavi et al. [30] 49% of the invited BPD patients accepted to participate, and of these 44% completed the email-based DBT skill-building program. Hence, there is a need to refine predictions of which patients might be susceptible to internet-delivered interventions.

Those who participated in our study completed more than two-thirds of the online material and most submitted their home assignments at the weekly check-ups, despite reported impulses to drop out. Patients’ motivation to follow the program fluctuated, underlining the importance of motivational efforts from the therapist to prevent drop out. The average clinical time per patient and week, 16.2 min, was far less than the average patient contact time in standard DBT including individual weekly 1-h sessions, 2 h of skills training in groups, and phone coaching between sessions [10–12]. This difference may explain why patients felt that the therapist support was insufficient and that they would have appreciated if the program was longer. At the same time, qualitative data showed that those who participated in the intervention and evaluation were happy that they took part, that it gave some support, and that they – in line with research findings from standard DBT [17,19] – perceived that they had acquired knowledge of DBT skills, with a reported average of 1–2 h/week spent on reading and working with the material. Hence, internet-delivered interventions of this kind could be feasible, and potentially address the demand in excess of the supply of mental health care providers administering evidence-based psychological treatments for BPD [21], increase access to evidence-based treatment for those living far from the treatment centers or in confinement [31], and meet the demands of younger patients who show a preference for working with digital media [26]. Some participants also reported that the program generated feelings of hope while waiting for standard DBT. In addition, the intervention seemed to be a way to conduct booster sessions and remind patients with prior DBT experience of how to work with the skills.

In the interviews some of the participants reported that they reduced or completely stopped self-harming, as they found new ways to address their self-harm impulses. In the pre- and post-ratings, most participants rated no NSSI or suicidal communication, whereas two reported fewer NSSI and suicidal communication in the week after the intervention. These results are in line with previous data, suggesting that there is a low risk of negative effects in internet-delivered interventions targeting NSSI and suicidality [33], and preliminary evidence of a reduction in NSSI [32]. However, one of the patients rated more NSSI and one rated more suicidal communication after the intervention, and some reported negative experiences when practicing skills, particularly in the beginning when they were not used to them or when the skills were still ineffective. Several patients also reported that they experienced stress and/or pressure in doing home assignments. These results highlight the potential importance of having access to additional support from a therapist. The support could include access to pre-recorded videos addressing common experiences and frequently asked questions, and telephone coaching with a therapist.

For the future, some changes in the interventions could be discussed. As patients generally struggled with compliance throughout the program, and most participants did not participate in the whole program, future studies should focus on how to maintain patients’ motivation and compliance throughout the program. The therapist support was experienced as helpful, but many reported that it was too scarce. Most participants wanted more frequent and more personal contact, particularly during burdensome periods with more symptoms. One suggestion for improvement was to increase the proportion of live communication to make it easier to ask questions and so those in need could get support. In the future, this could potentially be addressed through online group meetings, getting one step closer to the standard DBT praxis. In addition, all participants reported technical issues, which seemed to have caused frustration, and since technical issues have been identified as contributors to drop out [42], possibly a decrease in patients’ motivation to complete the program. Participants expressed the need for a timeline in order to get a better sense of the framework and the option to choose the frequency of notifications. Many also stated that they would have appreciated if the program had been longer. Considering that this intervention was immensely scaled down from its original format, future studies should aim to offer programs longer than 9 weeks.

Limitations

There are several limitations in this study. As this was a feasibility study, including a small number of participants, quantitative outcome measures should be interpreted as preliminary and explorative. Some patients had parallel contact with their local health clinic throughout the program, mainly for medical follow-up, which further limits the possibility to draw any conclusions regarding effects. As only around half of the patients chose to participate in the intervention and evaluation, it is possible that this study did not capture all potential risks and benefits of an intervention of this kind. It would have been interesting to compare those who participated with those who did not, in greater depth. In addition, all patients included in this study were women aged between 19 and 37 years, which may limit generalizability, while also keeping in mind that women are overrepresented (75%) among those diagnosed with BPD in clinical psychiatry [1]. The main author, who served as the therapist, also analyzed the qualitative data. To counteract this potential interpretation bias, all interview transcripts were anonymized and analyzed through triangulation. Data were analyzed independently by two of the researchers from the start and the manifest content in the last interviews did not result in any new categories, suggesting saturation of data.

Conclusions

Approximately half of the invited patients agreed to participate, and based on their ratings and interviews, the intervention could be regarded as feasible in several aspects. On average, the participants completed a large part of the intervention, despite impulses to drop out. The time spent on patient contact during the intervention was short, as expected. The intervention did not result in increased levels of suicidal or self-harm behavior. At the same time, since only half of the invited patients agreed to participate, an intervention of this kind may be insufficient for BPD patients in psychiatric settings. Some patients requested additional support, which could be addressed by incorporating live chats, phone calls and/or meetings in person to the intervention. Future studies should investigate in what ways some BPD patients are more susceptible to internet-delivered skills training in order to, for example, more quickly reach those who can manage to reduce and/or stop self-harming with the help of this intervention. There is a lack of studies on the added value of telepsychology and standard DBT, including internet-based interventions in comparison to standard DBT [26]. Since brief treatment has been suggested as a first step for the majority of patients with BPD [43], future studies should also explore if internet-delivered skills training before, during, or after DBT could improve outcomes, as a stepped care model, or as an add-on to other treatments for patients presenting with similar symptoms, but not meeting the criteria for BPD.

Acknowledgements

We are grateful to all patients who participated and to the DBT clinic of Uppsala University Hospital for its contributions in the recruitment and treatment procedure. We are also grateful to Tone Winqvist for her valuable contribution to the research.

Disclosure statement

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

Data availability statement

The data that support the findings of this study are available on reasonable request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Additional information

Funding

This research was financially supported by funds from the Uppsala University Hospital. The financiers had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review or approval of the manuscript.

Notes on contributors

Sara Vasiljevic

Sara Vasiljevic is a licensed psychologist at the Child and Adolescent Psychiatry Research Center in Stockholm and is affiliated to the Department of Clinical Neuroscience, Karolinska Institutet. Her research focuses on immunopsychiatry among children and adolescents.

Martina Isaksson

Martina Isaksson, PhD, is a licensed psychologist and a researcher at the Department of Medical Sciences, Uppsala University. Her research focuses on interventions for adolescents and adults with eating disorders.

Martina Wolf-Arehult

Martina Wolf-Arehult, PhD, is a licensed psychologist and a researcher at the Department of Medical Sciences, Uppsala University, and Centre for Psychiatry Research, Department of Clinical Neuroscience, Karolinska Institutet, Stockholm. Her research focuses on emotion regulation, dialectical behavior therapy and borderline personality disorder.

Caisa Öster

Caisa Öster, PhD, is a senior lecturer and researcher at the Department of Medical Sciences, Uppsala University. Her research focuses on impact of psychiatric morbidity in adaptation after injury and relationships between mental health and quality of life.

Mia Ramklint

Mia Ramklint, PhD, is a professor and researcher at the Department of Medical Sciences, Uppsala University. Her research focuses on difficulties in emotional regulation and impulse control among patients within psychiatry units.

Johan Isaksson

Johan Isaksson, PhD, is an associate professor and researcher at the Department of Medical sciences, Uppsala University, and the Department of Women’s and Children’s Health, Karolinska Institutet. His research focuses on neurodevelopmental disorders, emotional regulation and physiological stress patterns among children and adolescents