The State of Overmedication in Borderline Personality Disorder: Interpersonal and Structural Factors

Abstract

a). Purpose of review:

This review paper describes the state of prescribing practice in Borderline Personality Disorder (BPD), wherein medications are prescribed far more than either evidence or practice guideline would recommend. First, we describe the frequencies of medication use and polypharmacy in people with BPD.

b). Recent findings:

In subsequent sections, we elaborate two main categories of factors that lead to overmedication of people with BPD: the interpersonally mediated and the structural. We consider interpersonally mediated factors to arise from communications of patients in distress and the well-meaning efforts of their prescribers to provide relief for certain overwhelming affective states. We are particularly focused on patterns of countertransference in prescribing that are directly linked to specific aspects of BPD pathology. We consider structural factors to arise from the complexities of medical and medicolegal systems and the contemporary patterns of financing medical care; we postulate that these complexities often compel prescribers to start medications, with associated disincentives for decreasing or discontinuing those medications over time.

c). Summary:

More research is needed to understand how to best use medications in BPD, for example in targeted combination with psychotherapeutic and psychosocial interventions. However, current practice often departs markedly from the evidence. We recommend the dissemination of accessible, generalist BPD-treatment models in outpatient and inpatient practice; increased early detection of BPD; and increased diagnostic disclosure. We also recommend for individual providers and systems to implement prospective treatment plans that draw from BPD-specific psychosocial models. This approach can employ tiers of interventions to minimize reactive prescribing by anticipating high affect and offering BPD patients steadily empathic evidence-supported care.

Introduction

Current treatment guidelines for Borderline Personality Disorder (BPD) emphasize disorder-specific psychotherapies and limited use of psychotropic medications. However, empirical evidence of prescribing practices in the treatment of patients with BPD across settings reveals high rates of psychopharmacology and polypharmacy.

This paper highlights key interpersonally mediated and structural factors leading to this discrepancy and suggests possible points of intervention to help bring real-world prescribing practices in line with clinical guidelines. In particular, we examine 1) countertransference in prescribing; 2) barriers to accessing BPD-focused psychotherapy; 3) stigma, late diagnosis, and reluctance to share diagnosis; and 4) the impact of inpatient hospitalizations. We argue for 1) the dissemination of generalist BPD treatment models; 2) increased, early diagnosis of BPD; and 3) broader efforts to help providers identify and manage countertransference-driven prescribing decisions. We comment on the particular challenges of the inpatient setting and suggest possible interventions to minimize the resultant overmedication. These interventions include prospective planning for complex medical team and medical-insurance interactions as well as post-discharge deprescribing.

Psychotherapy, not medication, is the preferred treatment for BPD

Multiple psychotherapeutic approaches have been found to be effective at treating BPD, reviewed by Stoffers-Winterling et al [1]. These range from the widely disseminated Dialectical Behavioral Therapy (DBT [2]) and the psychoanalytically informed (e.g., Transference Focused Psychotherapy (TFP [3]), Mentalization-Based Treatment (MBT [4])) to the generalist-friendly Good Psychiatric Management (GPM [5]). However, no medication has been officially indicated for BPD, and in fact, medication treatment of psychiatric conditions is less effective in people with co-morbid BPD [6].

Major treatment guidelines all emphasize psychotherapy as the first-line treatment for BPD and recommend psychotropic medications either as an adjunctive treatment to psychotherapy targeted at specific symptoms, to treat co-morbidities, or to assist in short-term crisis stabilization. For the long-term treatment of BPD, guidelines for use of psychotropic medication diverge. This may be due to the limited available evidence base and to differing tolerance for the risks of psychotropic medication. The 2001 American Psychiatric Association (APA) practice guideline recommends pharmacotherapy to help address specific symptom areas: affective instability, impulsivity, psychotic-like symptoms, and self-destructive behavior [7], as does the subsequent guideline watch [8]. Specific medication classes and impacts on symptoms have been reviewed in [9–11]; in sum, modest beneficial effects have been found for some mood stabilizers and second-generation antipsychotics, but most studies have been small and un-replicated. In contrast to the APA’s recommendation of symptom-targeted adjunctive medications, the 2009 United Kingdom’s National Institute for Health and Care Excellence (NICE) guideline recommends pharmacotherapy only in the case of co-morbidities or during a crisis, stating specifically that “drug treatment should not be used specifically for borderline personality disorder or for the individual symptoms or behavior associated with the disorder (for example, repeated self-harm, marked emotional instability, risk-taking behavior and transient psychotic symptoms)” [12]. Australian guidelines concur that “medicines should not be used as a primary therapy for BPD, because they have only modest and inconsistent effects, and do not change the nature and course of the disorder,” and recommend time-limited use of medications only to manage specific symptoms as an adjunct to psychotherapy and within acute crises [13].

In the absence of consensus, we believe caution is warranted. People with BPD appear to be more susceptible to the adverse effects of medications [14]. Many side effects are demoralizing or functionally impairing in the short-term (e.g., weight gain, sexual side effects, excessive sedation, and the subjective experience of numbness). Some medications prescribed in BPD also have significant and concerning long-term side effects, particularly second-generation antipsychotics [15]; these are increasing in use for BPD [16]. Long-term antipsychotic medication use may contribute to the significant decrease in life expectancy for those with serious mental illness, in particular causing cardiovascular and metabolic disease [17, 18], including in BPD [19]. Many medications commonly prescribed in BPD, including benzodiazepines, stimulants, and pain medications also carry risks of abuse and dependence [20–22]. Polypharmacy carries additional risk of adverse events, as well as drug-drug interactions, and cumulative toxicity [23]. Additionally, evidence for combination treatments is more limited than that of individual medications [24], particularly in BPD [25].

Divergence of Clinical Practice from Recommendations

In practice, however, cross-sectional studies show high rates of psychotropic medication use and polypharmacy in BPD across settings (Table 1), even in patients without comorbidities [26, 27]. Despite lack of evidence in favor of prescribing, people with BPD are prescribed a higher number of concurrent psychotropic medications than those with what used to be called Axis I disorders or other personality disorders [28, 29]; higher prescribing persists over more than a decade of follow-up after index hospitalization in one large prospective study [30]. Medications from a variety of classes are commonly prescribed, including antidepressants, antipsychotics, mood stabilizers, and anxiolytics. This approach may result from efforts to treat individual symptoms that arise in people with BPD, as suggested by the American Psychiatric Association 2001 Practice Guideline [7].

Table.

Rates of Psychotropic Medication and Polypharmacy in BPD

	Date of Collection	N	Setting	Psychotropic Medications	Polypharmacy (≥2)	Polypharmacy (≥3)
Bridler et al., 2015 [32]	2001-2011	2195	In-patient, across Europe	90%	80%	54%
Paton et al., 2015 [26]	2012	1776 786	UK, 88% outpatient	92% (across sample) 87% (sample with no recorded comorbidities)	Not reported 66% more than 1 class	Not reported 38% more than 2 classes
Romanowicz et al., 2019 [33]	2013-2015	569	In-patient, US	55%	35%	19%
Martin-Blanco et al., 2017 [16]	2001-2016	457	Outpatient, Spain	88%	Not reported	54%
Zanarini et al., 2001 [29]	c. 2001	290	In-patient, US	84%	66%	46%
Pascual et al., 2010 [34]	2010	226	Outpatient, Spain	94%	Not reported	56%
Riffer et al., 2019 [27]	2011-2016	110	In-patient, Austria	96%	Not reported	62%
Timaus et al., 2019 [35]	2008-2012 1996-2004	87 139	In-patient, Germany	94% 72%	Not reported Not reported	Not reported Not reported
Haw and Stubbs, 2011 [36]	2010-2011	79	In-patient, UK	80%	48%	26.6%

The resultant accumulation of medications can lead people with BPD to feel like “guinea pigs,” as they are plied with increasing numbers of ineffective medications [31]. People with BPD can also experience blame from staff members when they do not experience improvements while medicated [31]. This experience can be painful and invalidating. Even worse, the lack of effect can lead providers to add additional medications in hopes of better results; paradoxically, the lack of medication response leads to increasing polypharmacy.

Countertransference Prescribing in BPD

People with BPD experience strong and labile emotions as a core feature of the disorder [37, 38] and can evoke similarly strong emotions in their providers within psychotherapeutic and pharmacologic treatments [39]. Mental health providers are likely subject to common, expectable reactions when repeatedly asked to help those with highly distressing situations and symptoms. Most providers respond with efforts to help, some going out of their way to try to improve the patient’s situation. That said, prescribers can understandably become frustrated or even angry, when their efforts to help relieve a patient’s symptoms are met with persistent failure or prominent critical or even derogatory responses. In other words, people with BPD can engage with their providers in alternating overvaluation (engaging with providers as the solution to overwhelming problems) and devaluation (engaging with providers as barriers to treatment, unable to fully and quickly resolve the problems); providers may unwittingly join patients in enacting this split. Overuse of medications can arise in both phases of the interaction: during the overvaluation phase, when providers can become overly eager to be helpful in resolving symptoms and enact fantasies of “fixing” the patient, and during the devaluation phase, when the intensity and negative valence of the patient’s emotion can quickly become intolerable to the provider. In high-affect situations, prescribers understandably seek to communicate care to patients in distress and may seek to offer hope through medication even in situations where medication use is not necessary or indicated.

Mintz and Flynn have used the term “countertransference prescribing,” to describe the situation when the use of medication aims to manage the experience of the prescriber as much or more than that of the patient [40]. While this could sound like providers are acting unprofessionally, by contrast this often occurs when providers are working hard to sit with and understand the emotions of highly distressed patients. In the course of this work, the evocation of strong emotions in the prescriber is an expected part of the process and with support and planning, these emotions can be used very productively. As Mintz and Flynn write, countertransference experienced in the prescribing context can be a useful tool when examined, with these dynamics becoming “valuable sources of data about the patient’s experience and relationships” outside of treatment [40]. However, as we have previously noted, prescribers who understand their role in treatment as more biological—for example, those who embrace the term psychopharmacologist—may find this form of psychodynamic thinking less intuitive and may not fully appreciate the contribution of countertransference risk [41, 42]. The same may well be true for those clinicians, psychiatrists and others, who see themselves as “therapists,” and in their frustration turn to prescribers to help with challenging patients, often overpromising the benefits of medications for personality disorder symptoms. Lack of attention to these forms of countertransference can risk reactive prescribing, particularly in the absence of adequate consultation and supervision.

Structural Factors

In addition to the interpersonal factors that attend BPD symptomatology, several structural factors promote the use of medications where they may not be indicated.

1. Access to evidence-based treatment

First, it is often difficult for people with BPD to access evidence-based psychotherapy due to a shortage of trained providers, stigmatization, marginalization within mental health care systems, and financial barriers [43, 44]. BPD-trained providers are scarce, and many mental health professions feel underprepared to diagnose and treat people with the disorder. Meanwhile, it can often be difficult to obtain insurance coverage for the long-term psychotherapy needed in BPD [44].

If prescribing psychiatrists felt that BPD-specific treatments were within reach, as Paris writes in his commentary on overmedication in severe personality disorders, they might be “slower to reach for their prescription pad,” and instead make referrals to appropriate psychotherapeutic treatment [45]. This would require system-level shifts in mental health care to support greater access to, and reimbursement for, psychotherapy. Additionally, at the provider-level, it would require further dissemination of evidence-based BPD treatment methods.

The Good Psychiatric Management (GPM) model is particularly well-poised to address the shortage of trained BPD treaters. A generalist approach that combines case management, psychoeducation, and a focus on return to life outside of treatment, GPM holds promise in lowering the barriers to clinician training and providing non-specialists the tools to treat most BPD patients [5]. GPM training is far less costly and time-intensive than for other evidence-based treatments, allowing for clinicians who do not wish to specialize in BPD to gain competence in the area. Even brief GPM workshops generate lasting improvements in clinician attitudes towards BPD [46]. In addition to GPM, psychoeducation programs following BPD diagnostic disclosure offer promise as an effective, low cost-intervention and a form of pre-treatment [47]. Psychoeducational programs for families also offer important intervention to increase positive support in the patient’s environment [48]. Given that people with BPD are estimated to make up 10-12% and 20-22% of the outpatient and inpatient population respectively [49], it is crucial that these high-quality evidence-based treatment methods are disseminated in mental health contexts beyond the specialized clinic.

2. Reluctance to Diagnose

Second, due to lack of training and stigma, clinicians can also be reluctant to diagnose BPD, to record it in the medical record, and to disclose the diagnosis to patients. In recent years, increasing interest in BPD-focused education has led to a growing literature on teaching strategies and setting-specific approaches [50, 51]. People with BPD are commonly underdiagnosed [52, 53] and misdiagnosed, particularly with bipolar disorder [54, 55]. These inaccurate diagnoses can lead to years of struggle and missed opportunities for treatment while risking the excessive accumulation of medication [44].

Some reluctance to diagnose BPD may also be a repercussion of DBT dismantling. Following the publication of the Linehan et al. dismantling study [56], many programs in the United States have shifted from a traditional DBT model with individual therapy, skills group, and skills coaching calls to providing skills-only groups. Clinical experience suggests that group leaders may be less likely to share the BPD diagnosis than those clinicians providing both individual DBT and DBT skills sessions.

In addition to the dissemination of less-intensive, generalist treatment models, structured interviews and screening instruments can be helpful in accurately diagnosing BPD early in the treatment course (rather than after repeated challenges in treatment intended solely for comorbidities, e.g. major depression). Several brief screening instruments have been validated for use in BPD, including a one-item screener for affective lability for BPD [38]. We anticipate that use of this brief screen in primary care and mental health treatment settings would be a resource-effective way of identifying patients for further assessment, thereby increasing accurate and timely diagnosis. Additionally, best practice supports diagnostic disclosure [57], which many providers shy away from: GPM helps provide tools for effective, non-stigmatizing disclosure of diagnosis [58]. Given the limited resources of many BPD-focused providers and programs, we imagine that “prescribing” online psychoeducational resources (e.g. [47]) as an adjunctive intervention could ensure that the diagnosis is named and the patients have access to detailed non-stigmatizing and hopeful information as a part of their treatment plan. Diagnostic disclosure can not only help patients understand their symptoms as an understandable, shared experience, but can also minimize the potential for inappropriate interventions, including medications.

3). Inpatient initiation and outpatient maintenance of polypharmacy

Psychotropic medication lists often lengthen during hospitalizations, while post-discharge deprescribing is not widely practiced. For instance, in a large US patient sample, Romanowicz found that hospitalization was significantly associated with increased psychiatric prescribing [33]. At admission, most patients (74.9%) were not taking any psychotropic medications; at discharge, most patients were medicated (44.6% unmedicated). These data replicate previous findings from a smaller sample of inpatients with BPD in Germany [35]. In this sample, medication use also significantly increased from admission to discharge, rising from 67% to 94% over the course of hospitalization. Hospitalization was associated with an increase in multiple prescription classes: SSRIs, SNRIs, antipsychotics, mood stabilizers, hypnotics, and anxiolytics. It was also associated with increased polypharmacy. While at admission only 15.6% of patients were taking more than one psychotropic medication, at discharge the frequency had increased by almost 20% (34.5%).

The urge to rapidly add multiple medications can be particularly acute in the inpatient setting, where interdisciplinary and team dynamics are complex, and decisions are under more time pressure. Staff and patients on inpatient units also see that other patients, e.g. people with acute psychosis and depression, experience marked relief after new prescriptions are initiated; they reasonably request similar treatment. Front-line staff must also care for both individuals and the unit as a whole. In service of maintaining the safety and atmosphere of the unit--their workplace and the patients’ therapeutic milieu--they may ask prescribers for medications to help particular patients act more manageably. These dynamics can result in a situation we might call countertransference prescribing by proxy wherein the wishes of the frontline staff and the patient for resolution of distress flow through the prescriber and onto the prescription pad. Preparatory meetings engaging the perspectives of prescribers and people who do frontline care, especially with those with less voice in treatment decisions (e.g. the overnight team) can help manage these dynamics by preparing together for strong feelings and helping to make them tolerable [59].

Another factor that may contribute to overmedication in inpatient settings is the incentive posed by hospital utilization management [33, 45, 60]. Under pressure to prove the need for often-scarce psychiatric hospital beds to utilization management (UM) committees and, in turn, insurance companies, providers and insurance reviewers may “view medications as ‘doing something’ that justifies hospitalization,” [33]. While UM requirements can often be obscure to providers and challenging to clarify, it is clear that managed care often results in significantly shorter hospital stays than providers request [60]. In other words, prescribing can practically assist in maintaining care for patients who need it, which generates tension for providers seeking to enact evidence-based care.

In both inpatient and outpatient settings, structural factors support the impulse to prescribe as evidence of assertive response to risk. When patients present with self-harming and suicidal behavior, prescribing providers may imagine a future reader (perhaps a Joint Commission reviewer or a judge in a malpractice proceeding). That prescriber may add medications in order to ensure that the record reflects specific actions taken to decrease risk. However, adding medication in this setting does not necessarily actually decrease risk or improve outcomes, as we have discussed. More research is needed to guide crisis management in BPD: in 2012, the Cochrane review of evidence regarding crisis interventions for people with BPD found “no RCT-based evidence for the management of acute crises in people with BPD” [61]. As it emerges, research evidence for beneficial crisis interventions and best practices for implementation will make a critically important contribution to the clinical literature.

In the meantime, and as better evidence emerges, we hope that providers will be able to collaborate with UM experts to understand how treatment efforts and progress can be evidenced and compensated. Much care for BPD can and should take place outside of the hospital: however, when hospitalization is necessary for safety or in the case of acute decompensation, it would be helpful to have agreed upon, reimbursable standards that align with the population’s needs.

In particular, we would suggest that conversations with UM experts may facilitate appropriate documentation of “doing something” and compensation of treatment beyond psychopharmacology. While acute crises in BPD may at times merit the use of medication, even in the emergency department, psychotherapeutic stance and specific interventions have been used to good effect. Victor Hong, an emergency room psychiatrist, proposes that interventions like Good Psychiatric Management (GPM) have an important place in managing overwhelming affect that can be so distressing to patient and provider alike [51].

Where required, medications initiated during inpatient stays should be brief and time-limited, with plans for medication tapering and discontinuation. Post-discharge, outpatient providers often maintain medications that were started to bridge a crisis over the long-term, perhaps out of fear of “rocking the boat” [62]. Therefore, in addition to increasing support for the use of non-medication-based interventions with BPD patients in the emergency setting, we also see a critical need to improve handoffs between acute care settings (emergency department and inpatient) and outpatient providers to avoid unnecessary long-term maintenance of medications. Within the current prescribing landscape, in which overmedication is prevalent, reducing polypharmacy can be conceptualized as a treatment goal in and of itself [63]. This goal is particularly suited to outpatient settings in the context of a strong therapeutic alliance. One approach we have previously suggested for BPD is deprescribing [42]. First developed in geriatric medicine, deprescribing is an active intervention in which patients and providers collaborate to assess the evolving risk-benefit of each medication a patient takes. In psychiatry and in BPD in particular, we believe this approach can increase patient and provider agency in treatment planning and help bring prescribing practice in line with the best available evidence. Shortening medication lists minimizes the risks of polypharmacy while making space for the crucial narrative work of psychotherapy.

Conclusion

People with BPD are prescribed far more medication than is indicated, recommended, or helpful. Though further research may define new indications for psychopharmacology in BPD—we would be particularly interested in evidence for how to best combine medications with psychotherapies—currently, the best evidence favors supporting non-medication interventions at all levels of care. Both the intense distressing affects that people with BPD experience and the related countertransference responses of their providers are expected; good prospective planning can help to decrease reactive prescribing that contributes to overmedication. A similar approach to teamwork with interdisciplinary providers and insurance companies can help to plan for expected situations with empathic and evidence-supported interventions.