Borderline personality disorder is a complex mental disorder associated with poor quality of life and high socioeconomic burden (
1). It is also associated with higher burden of illness and premature mortality, with standardized mortality rates of 8.3, compared with the general population (
2). Individuals with borderline personality disorder may have chronic suicidal behaviors (69%−80%), and up to 10% of these individuals die by suicide (
3).
Growing numbers of evidence-based psychological treatments have emerged specifically for patients with borderline personality disorder; dialectical behavior therapy and psychodynamic-based therapies have been shown to be effective in reducing psychopathology, self-harm, and suicide (
4). A meta-analysis indicated that four specialized psychotherapies for people with borderline personality disorder—namely, dialectical behavior therapy, mentalization-based treatment, transference-focused therapy, and schema therapy—are effective in reducing symptom severity and self-injurious behavior among people with borderline personality disorder (
5). A Cochrane review of psychological therapies for people with borderline personality disorder found that psychological interventions were more effective than treatment as usual in reducing symptom severity and self-harm among individuals with borderline personality disorder (
6). The Cochrane review noted that dialectical behavior therapy and mentalization-based treatment were the most studied specific psychotherapies and that mentalization-based treatment was more effective at decreasing self-harm, suicidality, and depressive symptoms.
Mentalization-based treatment is a psychodynamically oriented psychotherapy designed for patients with borderline personality disorder and rooted in attachment theory (
7). Evidence suggests that adverse early life events lead to disorganized attachment among patients with borderline personality disorder, giving rise to its symptoms (
8–
10). These symptoms (for example, emotional dysregulation, impulsivity, and difficulties in interpersonal relationships) are thought to be associated with reduced mentalizing capacity. Mentalizing is “the process by which we make sense of each other and ourselves, implicitly and explicitly, in terms of subjective states and mental processes” (
7). Mentalization-based treatment reinforces the mentalizing capacity of patients with borderline personality disorder to improve their capability to manage daily interpersonal interactions.
A randomized controlled study comparing mentalization-based treatment with standard care for borderline personality disorder found that mentalization-based treatment significantly reduced depressive symptoms, self-harm behaviors, and hospitalizations (
11). At 18-month and 5-year follow-up, these patients were maintained with reduced medication use and service utilization (
12,
13). A randomized controlled trial compared outpatient mentalization-based treatment to structured clinical management over 18 months for patients with borderline personality disorder; patients in both groups demonstrated substantial improvements in self-harm and suicidal behaviors, hospitalizations, and symptoms, with improvements more pronounced in the mentalization-based treatment group (
14). Mentalization-based treatment for borderline personality disorder also yielded similar positive outcomes in other settings (
15–
17).
Cost-effectiveness studies of mentalization-based treatment reported significant cost reductions, compared with usual care, with one-fifth of the costs of the usual-care group (
18). In a preliminary economic evaluation of psychological therapies for borderline personality disorder, mentalization-based treatment was cost-effective, at £5,000 per parasuicide event avoided and £20,000 per quality-adjusted life year (
19).
Mentalization-based treatment is commonly delivered by specialist personality disorder or psychotherapy services with dedicated treatment teams. In Ireland, where there are no such specialist services, mentalization-based treatment is available via two community mental health teams (CMHTs). A feasibility study of mentalization-based treatment delivered in a community adult psychiatric setting showed improvement in symptomatology and functioning (
20). One previous naturalistic study of mentalization-based treatment for borderline personality disorder in a standard psychiatric setting found significant reductions in symptoms (
21).
Few studies have assessed the role of mentalization-based treatment outside the specialist psychotherapy setting. In this study, we aimed to evaluate the outcomes of patients before and after a program of mentalization-based treatment nested within a general adult mental health service, measuring emergency department (ED) attendance as a proxy for self-harm and service utilization as a marker of overall functioning.
Methods
Design
This was a retrospective longitudinal naturalistic study that examined the characteristics of patients with borderline personality disorder 2 years before and after a course of mentalization-based treatment. We compared this cohort with a treatment-as-usual comparison group identified retrospectively and matched for diagnosis, age, and gender. Two years following receipt of mentalization-based treatment or treatment as usual, charts and patient electronic databases were reviewed for all persons who had received the intervention. Ethical approval was granted by Saolta Clinical Research Ethics Committee at University Hospital Galway, Ireland.
Setting
The mentalization-based treatment program was delivered by a CMHT in the west of Ireland; the team provides inpatient and outpatient care to a population of 26,000 people. The catchment area is rural, with moderate urban influence (
22). The CMHT provides assessment and treatment to patients with mental disorders ranging from moderate to severe, including personality disorders (
23). The team lead (E.M.) is a consultant psychiatrist with a background in psychoanalytic psychotherapy who is also a mentalization-based treatment trainer and supervisor and who developed the service to meet the needs of patients with borderline personality disorder. Several members of the CMHT are trained in mentalization-based treatment (via Anna Freud Centre, London) and deliver the intervention. These CMHT members receive regular supervision from the team lead and a certified mentalization-based treatment supervisor externally.
Study Participants
Intervention group.
Patients treated by the CMHT who had a diagnosis of borderline personality disorder were referred to the program. The program encompasses the following components: assessment and formulation of a diagnosis, 12 weeks of mentalization-based treatment introductory sessions, an 18-month program of weekly mentalization-based treatment group therapy and individual therapy sessions, and review of mental state and medications in an outpatient clinic every 3 months. There were three intake periods for the mentalization-based treatment group in October 2014, March 2015, and September 2015.
Following completion of the program, patients were either referred to regular outpatient psychiatric care or discharged to primary care. Some patients (N=11) did not complete the entire program because of resolution of symptoms following the mentalization-based treatment introductory intervention (data for these patients were included in the analyses). The introductory intervention is a 12-week manualized program, a low-intensity version of mentalization-based treatment based on guided self-help principles created as a stepped-care approach to improve access to psychotherapy (
24).
Comparison group.
A comparison group of patients who presented to mental health services with a primary diagnosis of borderline personality disorder in the 3 months of intake into the mentalization-based treatment group (October 2014, March 2015, and September 2015) were selected. These patients were not eligible for the program because of their geographical location, and they received treatment as usual (outpatient psychiatric care). Only patients residing within the catchment area of the CMHT are eligible for the mentalization-based treatment program. The groups were matched for age, gender, diagnosis, and time of presentation (these were the only variables available prior to chart review).
In Ireland (similar to the United Kingdom), usual care is delivered by a CMHT, which is publicly funded and free at the point of access. These teams cover a population of 25,000 and have 400–500 active patients across inpatient, day hospital, and outpatient clinic settings. They treat patients with moderate-to-severe mental disorders—for example, mood disorders, psychoses, anxiety disorders, and personality disorders. The team usually comprises a consultant psychiatrist, two psychiatry residents, a psychologist, a social worker, an occupational therapist, and three community mental health nurses. Only a small proportion of patients can be seen by the psychologist, and usually cognitive-behavioral therapy–based interventions are delivered, either by the nurses or via the day hospital program.
Inclusion and Exclusion Criteria
Patients with a primary clinical diagnosis of borderline personality disorder according to ICD-10 diagnostic criteria were included in the study. Patients with a primary axis I mental disorder, primary diagnosis of substance (alcohol or drug) dependence, a diagnosis of intellectual disability, or current engagement in other psychological interventions were excluded.
Data Collection
For both groups, we collected the following socioeconomic information: age, gender, and employment status. Service use was measured by number of ED visits, medical admissions, and psychiatric hospitalizations in the 2 years before and after treatment. All data were collected from patient chart reviews and hospital electronic databases.
Outcome Measures
The primary outcome measure in this study was the number of episodes of severe self-harm. Because data are unreliable by self-report, we used ED attendance as a secondary proxy measure. In line with national guidance, people who engage in severe self-harm primarily present to the ED for treatment. ED attendance was cross-checked with ED electronic notes, patient charts (when available), and patient registers to confirm that the ED visit was a result of self-harm. Similarly, in a study about the efficacy of outpatient mentalization-based treatment, the authors used hospital admissions as proxy measures of suicidal behaviors (
14). Secondary outcomes in our study were the number of ED visits and duration of inpatient admissions in terms of bed-days for both medical and psychiatric admissions.
Health Service Utilization Cost
On the basis of our findings, we performed a crude calculation of health care cost savings from the treatment of borderline personality disorder with mentalization-based treatment, derived from reductions in ED presentations and psychiatric and medical hospitalizations. Calculations were based on Irish public health services’ hospital charges at unsubsidized rates (
25,
26).
Statistical Analysis
Data were analyzed with SPSS Statistics, version 25. We described and compared the baseline characteristics of both groups as means (and standard deviations) for continuous variables and as numbers (and percentages) for categorical variables.
Chi-square tests and analysis of variance were performed, comparing ED presentations, medical admissions, psychiatric admissions, and final clinical outcome between the two groups. Paired t tests were used to compare improvements over time in the two groups. Logistic regression was used to further explore the relationship between the two groups, allowing control for potential confounding factors.
Results
Demographic Characteristics
Between 2014 and 2015, there were three intake periods of patients for the mentalization-based treatment program. A total of 22 females and 12 males were treated in the mentalization-based treatment program, and the mean±SD age was 33.3±11.2 years (
Table 1). At least half the study participants in both groups were unemployed. No significant differences in sociodemographic variables were noted between the groups. Over 50% of the patients in the treatment group had another psychiatric diagnosis (
Table 2).
Mentalization-Based Treatment Versus Treatment as Usual
Episodes of self-harm and ED presentations.
In the 2 years before treatment, 237 episodes of self-harm were documented in the mentalization-based treatment group. In the 2 years postintervention, the group had a 66% reduction, to 81 self-harm episodes (t=3.21, df=35.40, p=0.03). Self-reported self-harm data were unavailable for the treatment-as-usual group. In the mentalization-based treatment group, there were 27 presentations to the ED 2 years prior to treatment; presentations posttreatment were reduced by 56%, to 12 presentations (not a significant difference). The difference in ED presentations between the two groups was statistically significant—that is, the decrease in the mentalization-based treatment group was significantly greater (p<0.001, partial η
2=0.295) (
Table 3). This difference remained significant after the analysis controlled for age and gender in linear regression analysis (p=0.004) (
Table 4).
Psychiatric admissions.
In the mentalization-based treatment group, psychiatric bed-days utilized were reduced by 69%, from 310 to 95 bed-days, 2 years following treatment (t=2.48, df=41.39, p=0.018). However, the difference between the two groups was not statistically significant (
Table 3).
Medical admissions.
In the mentalization-based treatment group, bed-day occupancy in the 2 years following treatment was reduced by 42%, from 12 to 7 bed-days (not a significant difference). The difference between the two groups was statistically significant; the mentalization-based treatment group experienced a significantly greater decrease (p=0.016, partial η
2=0.092) (
Table 3). This difference remained significant after the analysis controlled for age and gender in linear regression analysis (p=0.040) (
Table 4).
Loss to follow-up and mortality.
At the end of the study, four participants in the mentalization-based treatment group were lost to follow-up, compared with 20 participants in the treatment-as-usual group. The difference between the two groups was statistically significant (χ2=7.59, df=1, p=0.006). In the treatment-as-usual group, three patients died by study endpoint, whereas all participants in the treatment group survived (not a significant difference).
Health Service Utilization Cost
In the mentalization-based treatment group, cumulative reductions in ED presentations and in psychiatric and medical hospitalizations resulted in gross total savings of €202,023 to the health service. Deducting the cost of external supervision for mentalization-based treatment delivery, this treatment approach delivered a mean saving of €5,862 per patient.
Table 5 presents a breakdown of the cost calculation.
Discussion
This study examined the clinical effectiveness of delivering a mentalization-based treatment program for borderline personality disorder in a general adult community mental health setting, compared with usual care.
Dialectical behavior therapy and mentalization-based treatment are the two most researched specialized psychotherapies for borderline personality disorder, with comparable efficacies (
6). The former focuses on skills acquisition and behavioral shaping to reduce emotional dysregulation and manage interpersonal interactions. It is resource intensive and requires a specific skill set within a team to develop the program (
19,
27). Mentalization-based treatment is less resource intensive to deliver and less demanding on patients (
14). Therapists are expected to adopt a mentalizing therapeutic stance (
7), otherwise no specialized psychotherapy skills are required to deliver the treatment, making it more suitable for general psychiatric settings with scarce resources, such as our CMHT.
ED Presentations and Hospitalizations
Both treatment groups experienced statistically significant reductions in ED presentations, measured as a proxy for self-harm behavior. Our finding that use of mentalization-based treatment in the treatment of borderline personality disorder reduced self-harm and suicidal behaviors is similar to findings presented in international literature (
11). Reductions in medical admissions were also observed for both treatment groups, with the mentalization-based treatment group experiencing significantly greater reductions in bed-days. Medical admissions in the population are frequently related to serious self-harm; in the United States, a nationwide ED-based study reported that 1 in 250 ED visits over 15 years were attributable to self-injury, and 31% required critical care admission (
28).
Reductions in psychiatric admissions were also observed for both treatment groups. However, the reduction was significant only for the mentalization-based treatment group. A similar reduction in psychiatric hospitalizations was noted in published studies (
11,
12), and this effect appears to be maintained for a prolonged period following completion of treatment (
13,
14). Psychiatric hospitalizations of patients with borderline personality disorder have been shown to be associated with behavioral regression, recurrent admissions, and accentuation of core symptoms—namely, chronic self-harm and suicidality (
29).
Superiority of Mentalization-Based Treatment Over Usual Care
A direct comparison of mentalization-based treatment and treatment-as-usual groups showed that mentalization-based treatment was superior in effecting significant reductions in ED presentations and medical admissions. Patients allocated to treatment as usual did not have access to mentalization-based treatment. Following assessment in the ED, when admission was not indicated, these patients were referred to their local CMHT for follow-up. Different CMHTs have different clinical approaches to the management of borderline personality disorder. Some teams may have more structured clinical resources in place to deliver consistent and well-organized treatment to patients with borderline personality disorder—thereby delivering “good enough” treatment in managing borderline personality disorder, which may explain the nonsignificant difference in psychiatric admissions between mentalization-based treatment and usual-care groups (
14,
30–
34).
Even though the reductions were not significant, the treatment-as-usual group experienced reductions in ED presentations and in medical and psychiatric hospitalizations. In the absence of evidence-based psychotherapy, usual care delivered by clinicians may have value in stabilizing patients with borderline personality disorder. There is a role for psychoeducation (
35,
36) and supportive therapy (
17,
37) in limiting the symptoms of borderline personality disorder, and both interventions are incorporated in routine psychiatric care in Ireland. It is possible that interpersonal interaction with a clinician, regardless of the intervention delivered, helped to contain the disorder in these cases. Despite the positive effects of usual care, where resources allow, evidence-based psychotherapy should be offered to patients with borderline personality disorder (
6).
Loss to Follow-Up and Mortality
A significantly higher proportion of patients were lost to follow-up in the treatment-as-usual group, compared with the mentalization-based treatment group. Our findings differed from those of a recent Cochrane review, which found that specialized psychotherapies for people with borderline personality disorder did not reduce attrition at treatment conclusion, compared with treatment as usual (
6). The lower dropout rate in the mentalization-based treatment group suggests better satisfaction with treatment. A previous study yielded similar results (
15). The structured program of weekly individual and group therapies may facilitate therapeutic rapport, resulting in a lower dropout rate in this intervention group (
20).
Although the difference in mortality rates between the two groups was not statistically significant, it is clinically significant that there were three deaths among 51 participants (6%) in the treatment-as-usual group, compared with no deaths in the mentalization-based treatment group. The mean age of those who died was 39.7 years (range 28–49). Perhaps the lack of structured engagement with treatment in the treatment-as-usual group made it difficult for these patients to be contained in times of crisis, resulting in self-harm in response to stressors.
Health Service Utilization Cost
The average annual direct health care cost to treat one person with borderline personality disorder in Ireland was €10,844, with an estimated annual total cost of €311.50 million in direct health care costs (i.e., acute care, community care, and medication) (
23). Our calculations focused only on hospital-based treatments, which resulted in a total savings of nearly €200,000. Our calculations did not factor in medication or the higher cost of admission to an intensive care unit for patients who engaged in severe self-harm, such as serious self-poisoning.
Limitations
This study included a small sample, compared with existing studies of the effectiveness of mentalization-based treatment. This may have reduced the statistical power of our measured outcomes between the groups.
The treatment and comparison groups were matched for age, gender, diagnosis, and time of presentation, because these were the only variables available prior to chart review. However, chart review of the selected comparison group revealed significant differences in service use prior to treatment. There were difficulties accessing clinical charts of patients in the treatment-as-usual group. As a result, data for individual self-harm episodes were not collected in the treatment-as-usual group.
Patients were given clinical diagnoses according to ICD-10 criteria, and a structured clinical instrument was not used. As a naturalistic study reflecting real-life, time-limited clinical settings, where psychiatrists make a diagnosis of a mental disorder on the basis of clinical assessment using ICD-10 criteria, our findings are likely generalizable to the community mental health setting.
In the treatment-as-usual group, it was not possible to standardize the treatment being delivered. This may have affected the clinical outcomes measured in this group.
Our basic economic analysis simply considered service factors—i.e., bed-days avoided—which may have underestimated the overall cost savings generated by the mentalization-based treatment program. A full cost-effectiveness study would include quality-adjusted life years to better assess the value of the intervention.
Further studies of mentalization-based treatment in naturalistic settings are needed, including those utilizing semistructured instruments and patient-reported measures to demonstrate reproducibility and to examine subcohorts of patients who best respond to treatment.
Conclusions
Our findings suggest that mentalization-based treatment embedded within a general adult CMHT is associated with reductions in self-harming behaviors, ED presentations, and hospitalizations. There is value in building on this study with more prospective, systematic research and patient-reported outcomes to assess the practical significance of this intervention in general psychiatric settings.