Contingencies Create Capabilities: Adjunctive Treatments in Dialectical Behavior Therapy That Reinforce Behavior Change

Available online at www.sciencedirect.com

ScienceDirect Cognitive and Behavioral Practice 23 (2016) 110-120 www.elsevier.com/locate/cabp

Contingencies Create Capabilities: Adjunctive Treatments in Dialectical Behavior Therapy That Reinforce Behavior Change Adam Carmel, Harvard Medical School, Massachusetts Mental Health Center Katherine Anne Comtois, University of Washington at Harborview Medical Center Melanie S. Harned, University of Washington Rhea Holler, Private Practice, Fullerton, CA Lynn McFarr, Harbor-UCLA Medical Center Dialectical behavior therapy (DBT) has been shown to be effective in the treatment of borderline personality disorder (BPD), a disorder associated with poor functional outcomes and high utilization of behavioral health services. Contingency management strategies are one of the four primary change procedures in DBT. This paper provides an overview of the use of contingency management strategies in DBT with a particular focus on how adjunctive treatments can be utilized as a reinforcer for positive behavior change. We focus specifically on two adjunctive DBT treatments with evidence of efficacy, including the DBT Prolonged Exposure protocol (DBT PE), which targets PTSD, and DBT–Accepting the Challenges of Exiting the System (DBT-ACES), which targets getting off of psychiatric disability by obtaining and maintaining employment. This paper describes how contingency management strategies are used to help clients make the changes necessary to become eligible to receive these adjunctive treatments, as well as the process of clarifying and managing contingencies to maintain and increase adaptive behaviors as these treatments are implemented. Considerations for how DBT therapists and larger health systems can apply contingency management strategies to enhance behavioral capabilities in the treatment of individuals with BPD are discussed.

I

meeting criteria for borderline personality disorder (BPD) use an inordinate amount of psychiatric services, including inpatient admissions, crisis and emergency services, residential and day treatment, and outpatient therapy (Bender et al., 2001; Goodman et al., 2010; Linehan & Heard, 1999). Clients with BPD have been found to use several of these psychosocial treatments at a significantly higher rate compared to a group of individuals with depression (Bender et al., 2001; Linehan, Kanter, & Comtois, 1999). High utilization of acute psychiatric services (defined as three or more inpatient psychiatric hospitalizations or any hospitalization of 30 days or longer in a 2-year period) financially strains many systems of care. Up to 42% of high-service utilizers meet criteria for BPD and few crisis and inpatient services are designed to meet their needs (Comtois & Carmel, 2014). That acute psychiatric services have not been effective in the treatment of clients with BPD is demonstrated by a longitudinal study on the remission NDIVIDUALS

Keywords: dialectical behavior therapy; borderline personality disorder; contingency management

1077-7229/15/© 2015 Association for Behavioral and Cognitive Therapies. Published by Elsevier Ltd. All rights reserved.

rate of BPD symptoms for a period of 16 years after participants’ discharge from an index psychiatric hospitalization. Only 40% to 60% of participants with BPD achieved recovery (defined as diagnostic remission from BPD, working or going to school full time, and having at least one emotionally sustaining relationship) and 44% of those who demonstrated recovery at one point were unable to sustain their recovery gains over time (Zanarini, Frankenburg, Reich, & Fitzmaurice, 2012). Community outpatient treatment of clients with BPD often involves a similar pattern of overtreatment. Clients with BPD features were found to attend more individual and group treatment sessions, require a greater number of providers, and have a significantly higher mean number of treatment sessions compared to individuals without BPD features (Bagge, Stepp, & Trull, 2005). Other studies examining the utilization of outpatient services have determined that clients with BPD have higher rates of mental health and psychopharmacological treatment than control groups with depression or other personality disorders (Bender et al., 2001). In a longitudinal study of 290 individuals with BPD, nearly 75% of those receiving outpatient services at baseline continued to remain in treatment after a 6-year follow-up assessment. (Zanarini, Frankenburg, Hennen, & Silk, 2004).

Contingencies Create Capabilities One potential explanation for this pattern of high utilization of community treatment is that additional treatment can function to reinforce maladaptive behavior. Within many systems of care, receiving higher levels of care are contingent upon more severe maladaptive behaviors. For example, clients may only receive individual therapy sessions by displaying out-of-control behaviors or expressing suicidal ideation, while simultaneously having individual therapy sessions withdrawn when crisis behaviors remit or when they show an increased amount of self-control (Koerner, 2012). This becomes problematic when additional contact with the therapist or extra therapy sessions are desired, as is often the case for clients with BPD (Linehan, 1993). The contingencies in this scenario are likely to strengthen behavioral dyscontrol and crisis behavior, which arguably contributes to both the financial burden associated with overtreatment and the limited improvement of clients with BPD (Koerner, 2012; Linehan et al., 1999; Soeteman, Busschback, Verheul, Hoomans, & Kim, 2011). In contrast to this common pattern in the treatment of clients with BPD, dialectical behavior therapy (DBT; Linehan, 1993) offers an approach that makes the provision of additional treatment contingent upon the display of adaptive rather than maladaptive behaviors. DBT was developed as a treatment for acutely and chronically suicidal individuals and has been shown to reduce suicide attempts, psychiatric hospitalizations, and emotional distress, and it is also the only treatment for BPD that has been widely disseminated (Linehan et al., 1991; Linehan, 1993; Rizvi, Steffel, & Carson-Wong, 2013; Stoffers et al., 2012). DBT has established cost-effectiveness largely due to decreases in acute psychiatric services including hospitalization (Linehan et al., 1999; Wagner et al., 2014; Wunsch, Kliem & Kroger, 2014). DBT was developed by Linehan (1993) and typically includes 1 year of treatment that prioritizes targeting of life-threatening behaviors followed by behaviors that interfere with treatment and those that prevent

111

clients from having a reasonable quality of life. The overall goal is to help suicidal clients not only survive, but to build a life worth living (Linehan, 1993; Salsman & Linehan, 2006). DBT includes a multitude of treatment strategies that are balanced in terms of acceptance (e.g., validation) and change (e.g., problem solving). Contingency management (CM) is one of the four primary change procedures in DBT. CM strategies can be applied in-session (i.e., during therapist-client interactions) as well as more broadly (e.g., using CM plans to shape desired behaviors at home, in social interactions, at the emergency department, etc.). The goal of this paper is to describe the process of clarifying and managing contingencies within DBT, with a particular focus on the use of adjunctive treatments as a way to reinforce behavior change. Overall, we aim to illustrate how strategic and planned reinforcers of adaptive behavior change in therapy eventually give way to natural reinforcers in the client’s life to create capabilities for clients with BPD.

Overview of Contingency Management Strategies in DBT and Principles of Operant Conditioning CM derives from operant conditioning theory and generally refers to the use of consequences to either increase (i.e., reinforce) or suppress (i.e., punish) a specific behavior (Skinner, 1988). There are four general types of contingencies (see Table 1) and examples of how each may be used within DBT are provided below. Positive Reinforcement Positive reinforcement is the primary CM strategy used by DBT therapists to facilitate behavior change. Positive reinforcement involves providing consequences the client finds rewarding in response to desired behaviors in order to increase them. It is essential that positive reinforcers are consistently delivered in response to adaptive behavior, and are not delivered indiscriminately or immediately following maladaptive behaviors (Linehan, 1997). It is

Table 1

Overview of Contingency Management Strategies and Principles of Operant Conditioning Increase Behavior

Decrease Behavior

Add Consequence

Positive Reinforcement Adding a consequence following an individual’s behavior that increases the likelihood of that behavior occurring in the future

Remove Consequence

Negative Reinforcement Removing a consequence following an individual’s behavior that increases the likelihood of that behavior occurring in the future

Positive Punishment Adding a consequence following an individual’s behavior that decreases the likelihood of that behavior occurring in the future Negative Punishment Removing a consequence following an individual’s behavior that decreases the likelihood of that behavior occurring in the future

112

Carmel et al.

equally important that the therapist assess what types of responses are reinforcing for an individual client, as it is often the case that the same response (e.g., therapist praise) is sought after by some clients and is aversive to others. In-session strategies often desired by clients with BPD include validation, in which the therapist communicates that the client’s responses make sense and are understandable, as well as reciprocal communication strategies, in which the therapist expresses warm engagement, is responsive to the client’s stated agenda, and acts in a genuine (rather than therapist-like) manner (Linehan, 1993, 1997). For the client who is motivated to have additional contact with their therapist, it may be useful to create a CM plan whereby additional sessions and/or phone contacts with the therapist are provided in response to the client exhibiting specific adaptive behavior (while blocking their maladaptive behavior). Negative Reinforcement Using this strategy, the DBT therapist removes aversive consequences in response to adaptive behaviors with the goal of increasing their frequency. For example, although chain analysis (i.e., a detailed assessment of the chain of events that led up to and followed a specific dysfunctional behavior such as arriving late to a session) is primarily used as an assessment tool in DBT, it can be aversive for clients to talk about a maladaptive behavior they regret. In such cases, ceasing to conduct a chain analysis when the client stops the maladaptive behavior (e.g., arrives on time) removes an aversive consequence and is likely to increase the functional behavior (e.g., being punctual). Another example in session is the therapist withdrawing attention or engaged body language when a client is changing the subject, complaining, or otherwise avoiding working on a problem in therapy and then fully reengaging as soon as the client constructively engages in solving the problem. Positive Punishment In DBT, the therapist may also apply contingencies in response to undesired behaviors with the goal of reducing their frequency. Note that in behavioral terms, punishment does not (and should not) imply being hurtful or damaging, but instead refers simply to a consequence the individual will work to avoid. An example of applying an aversive consequence is the DBT therapist’s response to a client yelling and cursing at them. As this behavior is ineffective, an appropriate response in DBT is for the therapist, without responding to the content of the client’s complaint, to insist that the client sit down and lower his/her voice. To respond to the content would potentially reinforce the maladaptive yelling and cursing. Being warm and responsive to the client’s request once it is voiced more quietly represents a return to reinforcement, as the aversive (i.e., repeatedly

telling the client to sit and lower their voice) is replaced by reinforcers (i.e., warmth and responsiveness). Another example of positive punishment in DBT would be displaying disappointment or expressing frustration when a client reports having engaged in self-harm behavior. The key in this case is to keep disappointment focused on the problem behavior—self-harm—and not on the individual: In DBT, the client is consistently viewed nonjudgmentally and positively. This will suppress the problem behavior but not suppress attendance or engagement in treatment. Negative Punishment The DBT therapist may also remove reinforcers in response to maladaptive behaviors with the goal of decreasing undesired behaviors. An example of this is the 24-hour rule in DBT, which removes out-of-session access to the primary therapist for 24 hours following suicidal or self-injurious behavior (Linehan, 1993). It should be noted that the duration of punishment is not contingent on the individual’s behavior (e.g., the 24-hour rule is for 24 hours no matter how the client behaves during that period). This differs from negative reinforcement in which the duration of the aversive contingency is under the control of the individual. In negative reinforcement, when the individual engages in an adaptive behavior (e.g., uses skills instead of self-harm, engages constructively in session), the aversive contingency is stopped and thus the duration of the punishment is under the control of the individual. DBT always prioritizes negative reinforcement over punishment. However, in order to enact negative reinforcement, an aversive may need to be present for some duration (e.g., 24 hours) before it can be removed (e.g., the therapist is available for coaching if needed when the 24 hours are up).

Selecting and Applying Contingencies in DBT Reinforcement works better to initiate and maintain new behavior, while punishment is likely to temporarily suppress behavior (Domjan, 2003). Punishment is not as effective as reinforcement in changing behavior as it tends to suppress the problem behavior only when the person(s) doing the punishing are present (Pryor, 2006). Also, punishment tends to create negative associations between the individual being punished and the person(s) providing the punishment that results in the individual making efforts to avoid them. In the case of a therapist (over)using punishment, this can lead to clients missing sessions, coming late to therapy, or avoiding contacting the therapist for coaching when needed, or requesting another therapist. Thus, negative and positive punishment are used infrequently in DBT, and punishment is limited to situations in which reinforcement cannot be used—that is, situations where the only behaviors present are maladaptive and thus there is no

Contingencies Create Capabilities opportunity for reinforcement to be used and have an effect. In those situations, in DBT, positive or negative punishment may be used to suppress problem behaviors temporarily until adaptive behaviors appear and can be reinforced. For example, a client cries so constantly in session that no other adaptive behaviors—talking clearly, listening—occur to be reinforced. An aversive contingency such as steadily coaching the client on how to stop crying (rather than trying to discern what the client is saying while she is sobbing, which the client would prefer) could be used to stop the crying long enough for the client to speak clearly or listen. Once crying is replaced by speaking clearly or listening, the coaching the client finds aversive also stops (negative reinforcement) and the therapist is attentive to what the client is saying and says validating things when the client is speaking clearly (positive reinforcement). In DBT, punishers are also chosen to be as constructive as possible (e.g., coaching on how to stop crying) or easily removed (e.g., disappointment, lack of eye contact) and are stopped and reinforcement restarted as quickly as possible. Behavior is learned much more strongly from reinforcement, where desired consequences are under the individual’s control and can occur in response to changes in the individual’s behavior, than from punishment, where aversive conditions cannot be changed by their behavior (e.g., 24-hour rule, a teenager being grounded for a week). DBT therapists must mindfully maintain consistent responses when applying CM strategies and consider any deviations as likely intermittently reinforcing maladaptive behavior. Therapists use positive reinforcement throughout DBT and must remain alert to adaptive behaviors as they appear in therapy so they can be reinforced. Therapists applying negative reinforcement must be sure to withdraw an aversive as soon as adaptive behavior occurs. If a punisher is in place for a set time, it cannot be removed earlier—always a challenge when working on residential or inpatient settings where others take responsibility for the treatment of the client. As noted in the title and by all behavioral therapies, contingencies create capabilities. That is, all of us are more capable of doing more adaptive things when contingencies reinforce them (e.g., how much easier is it to clean the house when guests are coming or to finish a report on a deadline). This is the key to orienting clients to the use of contingencies in DBT—contingencies are part of DBT because they work to change behaviors to achieve goals. With formal CM plans, such as the 24-hour rule, it is important that the plans are identified and applied collaboratively between therapist and client and the plans are clear and consistent (see Table 2). The client is oriented to these contingencies as part of learning about DBT and asked to agree to them as part

113

Table 2

Considerations for Selecting and Applying Formal Contingency Plans: The Three C’s Clarity

Clarity in CM plans requires that the client is adequately oriented. For example, during pre-treatment sessions clients are given the rationale for a plan and are asked to agree to it as part of treatment. Didactic strategies are used to teach behavioral principles and the effects of contingencies on behavior. The therapist provides transparency about the plans and why they are being considered. The DBT consultation team helps the therapist clarify contingencies operating in therapy and in the client’s life, and to identify the behaviors, reinforcers and aversives while selecting formal contingency plans and implementing them over time. Consistency Consistency is necessary to ensure that a behavioral plan will be effective over time. This often requires consultation from the DBT consultation team to address therapist and therapy-interfering behaviors, particularly a therapist’s urges to avoid or deviate from a plan. The therapist uses a mindful approach to highlight patterns of behavior related to a plan to prevent inconsistencies from occurring and to problem-solve when inconsistencies have occurred. Collaboration A collaborative agreement is necessary to set a plan and to troubleshoot what could get in the way of implementing a plan. Consulting the client’s wise mind on the reasonableness of a plan and its intended effect on obtaining their goals is key. Therapists will attend to the relationship and use commitment strategies to collaborate on a plan. For example, a therapist may ask, “How are we going to tackle this DBT-ACES Work as Therapy requirement? We are two months away from the deadline and it’s important that we come up with a plan now so we don’t have to worry about suspension later.” Phone coaching can be used to help create a collaborative alliance around implementing a plan.

of treatment. For more idiosyncratic situations that are unexpected until they occur (e.g., a client crying in session so she can’t clearly talk or listen), a collaborative agreement could be reached on what the therapist and client should do in the future when disruptive crying arises again. A thorough orientation to the rationale for CM, including clarification on how it will help the client reach their goals and is not intended to be patronizing or punitive, are critical to help clients agree to a specific CM plan. For example, while orienting the client to the rationale for conducting a chain analysis in response to lateness, the therapist might stress that in addition to

114

Carmel et al.

identifying and addressing the factors that interfere with having an entire session, always calling attention to lateness motivates a person to get there on time to avoid this discussion as well as assuring that more time talking about what the client thinks is important will be paired with being on time. Being on time is a key skill in life for both work and relationships because it demonstrates respect, which is reinforcing to others. If nothing were ever done differently in session when the client was late, other pressures in the client’s life would naturally take priority. This would not help the client. Orientation such as this that fits the client’s reinforcers (e.g., getting therapy time, pleasing the therapist) will likely result in collaboration on these contingencies. One problem that can arise for the DBT clinician using CM strategies is encountering the belief among other clinical or administrative staff that doing so is cruel or underhanded. Orienting and gaining commitment on using CM strategies from other staff treating the client can be just as important as obtaining commitment from the client. It is important to emphasize the intended effect of increasing motivation for adaptive behaviors and decreasing motivation for maladaptive behaviors. Contingencies that motivate clients to engage in maladaptive behavior or behavior that is inconsistent with their ultimate goals are not DBT. Because contingencies create capabilities and capabilities create a life worth living, we maintain that effective CM strategies are an important part of a compassionate therapeutic approach. The DBT consultation team is often the best place to consider the specifics for a given therapist, treatment program, or client to assure that contingencies are used to create capabilities for positive outcomes and growth.

Evidence of the Effectiveness of CM in DBT Several findings provide evidence of the effectiveness of CM strategies in DBT as a method of helping BPD clients increase behavioral control, particularly in terms of reducing suicidal behaviors. One variable examined in the research is the rate of between-session phone calls, which are used in DBT to help generalize skills, prevent maladaptive behaviors, and to receive encouragement for positive behaviors (Linehan, 1993). This component of DBT is applied using the 24-hour rule to ensure that maladaptive behaviors do not increase as a function of reinforcing phone calls with the therapist (Manning, 2011). Within the first randomized controlled trial of DBT, no significant correlation was found between the number of phone calls and suicidal and self-injurious behavior in the DBT condition, whereas clients who engaged in more suicidal and self-injurious behaviors had more phone calls with their therapist in the treatmentas-usual condition (Linehan et al., 1991; Linehan & Heard, 1993). The lack of a relationship between phone contact and suicidal and self-injurious behavior in DBT is intended to

eliminate the possibility that such behaviors may be reinforced by extra contact with the therapist (Linehan & Heard, 1993). Similarly, Bedics et al. (2012) found that higher client ratings of “therapist affiliation”—theoretically reinforcing therapist behaviors on the Structural Analysis of Social Behavior (Benjamin et al., 2006)—were associated with less self-injurious behavior in DBT and more selfinjurious behavior in a community treatment by experts condition. This association could indicate that DBT therapists pay particular attention to providing reinforcers (such being affiliative) contingent on behaviors that indicate clinical progress (reduced self-harm behavior), thereby resulting in adaptive behavior change.

The Use of Adjunctive Treatments to Reinforce Behavior Change In addition to the routine use of contingencies during DBT described above, emerging research suggests that the provision of adjunctive treatments that address specific problems of high priority to clients with BPD can be powerful reinforcers of behavior change. To date, two adjunctive treatments have been developed for clients who have achieved behavioral control (the primary target of Stage 1 DBT) and want to work on additional goals to improve the quality of their lives. These adjunctive treatments are the DBT Prolonged Exposure (DBT PE) protocol, which is designed to target PTSD, and DBT–Accepting the Challenges of Exiting the System (DBT-ACES), which focuses on helping clients get off of psychiatric disability by seeking and maintaining employment. These adjunctive treatments explicitly utilize CM principles by requiring clients to demonstrate control over life-threatening and serious therapy-interfering behaviors before they are eligible to receive the adjunctive treatment and making continued participation in these treatments contingent on maintaining these gains. Emerging evidence has demonstrated that many clients have responded to these contingencies by decreasing long-standing patterns of maladaptive behavior in order to access and complete the adjunctive treatment. The process of managing contingencies within these adjunctive treatments and the effects of contingencies on behavior change will be explained further.

The DBT PE Protocol Treatment Structure The DBT PE protocol (Harned, 2013) is based on Prolonged Exposure (PE) therapy for PTSD (Foa, Hembree, & Rothbaum, 2007) and has been adapted to treat co-occurring PTSD in BPD clients, particularly those with recent suicidal and self-injurious behavior. The primary treatment components include imaginal exposure to trauma memories followed by processing of the exposure experience as well as in vivo exposure to feared

Contingencies Create Capabilities but nondangerous situations. The DBT PE protocol is delivered by the individual DBT therapist and occurs concurrently with DBT if/when clients meet specified readiness criteria. The combined DBT and DBT PE protocol treatment is delivered during 1 year of treatment that begins with DBT focused primarily on reducing behavioral dyscontrol and increasing behavioral skills, and then adds in the DBT PE protocol to target PTSD after sufficient improvement has occurred. Readiness to begin the DBT PE protocol has been defined as: (a) no longer being at imminent risk of suicide (e.g., suicidal ideation with intent and a plan), (b) no recent (past 2 months) suicidal or nonsuicidal self-injurious behavior, (c) able to control urges to engage in suicidal and self-injurious behavior when in the presence of cues for those behaviors, (d) no serious therapy-interfering behavior (e.g., recurring treatment nonattendance or consistent refusal to comply with treatment tasks), (e) PTSD is identified by the client as his/her highest priority target and s/he wants to treat it now, and (f) able and willing to experience intense emotions without escaping. Upon completion of the DBT PE protocol, and assuming time remains in the treatment year, a third phase of treatment may also occur during which the client continues in DBT to address any remaining treatment goals (e.g., improving relationships, seeking employment). Using CM Prior to DBT PE Several CM strategies are used during DBT to increase clients’ motivation and ability to make the changes needed to be eligible to receive the DBT PE protocol, and ideally to make these changes relatively quickly. Clients are oriented to the readiness criteria for starting DBT PE early in treatment (typically during the pretreatment phase of DBT) to establish clear expectations about which behaviors must change in order to receive the treatment. The goal of this discussion is to clarify the contingencies that are in place (i.e., the if-then relationships between a client’s behavior and their ability to receive DBT PE) and to establish a collaborative approach to managing these contingencies. For clients for whom treating PTSD is a treatment goal, this orientation often functions to increase their motivation to quickly gain control over life-threatening and serious therapy-interfering behaviors so that they can receive the DBT PE protocol. As DBT progresses, therapists consistently clarify the contingencies of clients’ behaviors in terms of their effects on their ability to begin DBT PE and, by extension, to reduce suffering caused by PTSD. For example, when higher priority behaviors occur (e.g., self-injury), clients are reminded that continued engagement in these behaviors will delay their ability to receive PTSD treatment. Conversely, progressive steps towards desired behaviors are

115

shaped through reinforcement and linked to the goal of treating their PTSD. For example, therapists may provide praise for achieving abstinence from self-injury at first for several days, then for a week, and then for a month while consistently highlighting the link between these improvements and their increasing readiness for DBT PE. In addition, when relevant, therapists link clients’ broader goals (e.g., improving relationships, reducing anger) to the need to treat their PTSD. In this way, clients are also made aware of the if-then relationships between treating their PTSD and achieving improvements in other problems they view as important, particularly those that are being maintained or exacerbated by PTSD. Importantly, the ability of any reinforcer to motivate behavior change is largely dependent on how much the person wants to receive the promised outcome. In the case of the DBT PE protocol, research has shown that a majority of suicidal and self-injuring clients with BPD and PTSD prefer to receive a combined DBT and PE treatment (73.8%) compared to either DBT alone (26.2%) or PE alone (0%; Harned, Tkachuck, & Youngberg, 2013). When asked to provide reasons for this treatment preference, clients most often reported a desire to receive specific DBT and PE treatment components (75.9%) and to obtain relief from PTSD- and BPD-related distress (65.5%). In addition, the preference for a combined DBT and PE treatment was predicted by a childhood index trauma and more severe PTSD reexperiencing symptoms. Taken together, these findings indicate that the DBT PE protocol is desired by a majority of BPD clients with PTSD, particularly those whose PTSD is more severe and longstanding, and is therefore likely to function as a reinforcer for most clients. However, for the minority of BPD clients with PTSD who are not interested in receiving DBT PE, the above strategies are unlikely to work to help them gain control over their behavior and may even have the opposite effect; namely, clients may work to avoid having to receive DBT PE by continuing to engage in behaviors that will make them ineligible. To mitigate against this possibility, therapists are encouraged to assess clients’ degree of interest in receiving DBT PE early in treatment and, for clients who are not or only minimally interested, to make it clear that they do not have to do the treatment and to identify other reinforcers to motivate behavior change. Using CM During DBT PE Once clients meet the required readiness criteria and the DBT PE protocol is started, additional contingencies are put in place to strengthen clients’ motivation to remain in control of life-threatening and other higherpriority behaviors while they receive PTSD treatment. Specifically, clients are told that the DBT PE protocol will be stopped if they engage in suicidal or self-injurious

116

Carmel et al.

behavior or any other behavior that significantly interferes with the treatment (e.g., severe dissociation during sessions, failure to complete homework tasks). This contingency often functions to increase clients’ motivation and ability to maintain control over higher-priority behaviors during the course of DBT PE. For example, clients who experience increased urges to self-injure or commit suicide during DBT PE often report that their awareness that engaging in these behaviors would cause them to have to stop DBT PE helps them to refrain from the behavior. Therapists also consistently reinforce clients’ efforts to maintain behavioral control using whatever strategies clients find rewarding (e.g., praise, increased warmth, between-session contact). If clients do engage in a higher-priority behavior that causes the DBT PE protocol to be stopped, they return to DBT to target the behavior that occurred. DBT PE is only resumed if/when clients demonstrate the ability to control the problem behavior, change any circumstances that may have caused the behavior to occur, express strong commitment to preventing future occurrences of the behavior, and make repairs (when appropriate). In this way, the ability to resume DBT PE is used as a reinforcer for working to regain control over the behavior that caused the treatment to be stopped. Although a majority of clients experience the ability to receive DBT PE as a positive outcome that they will work to obtain, in a minority of cases clients remain (or become) ambivalent about receiving DBT PE after they have started the treatment. In such cases, continued receipt of DBT PE may be an aversive outcome that clients will work to avoid by, for example, engaging in maladaptive behaviors in an effort to cause the treatment to be stopped (e.g., increased hostility toward the therapist, refusal to complete homework). When possible, therapists should work to continue delivering DBT PE to these clients to prevent potential negative reinforcement of such behaviors while also encouraging clients to skillfully ask for the treatment to be stopped if they do not wish to continue. Finally, therapists use CM strategies informally throughout DBT PE sessions. As in DBT more broadly, reinforcement is used whenever possible to increase adaptive behaviors, particularly those that are likely to make exposure-based treatment successful. A required element of the treatment is that therapists provide positive reinforcement during in-session imaginal exposure by saying things like, “You’re doing great, stick with it!” In addition, clients are reinforced for completing exposure in effective ways by, for example, reducing avoidance behaviors during exposure, allowing emotions to be experienced to their full intensity, and completing the full dose of in vivo homework. Punishment is used infrequently to reduce or suppress problem behaviors, particularly those that are likely to interfere with the treatment’s efficacy, and removed as

negative reinforcement as soon as feasible. This may include strategies such as expressing irritation when clients do not complete homework, withdrawing warmth when clients begin to dissociate during imaginal exposure, or ending sessions early if clients come to session under the influence of drugs or alcohol. CM Strategies and the Effectiveness of DBT PE To date, DBT with the DBT PE protocol has been evaluated in an open trial (n = 13; Harned, Korslund, Foa, & Linehan, 2012) and a pilot randomized controlled trial (RCT) that compared DBT with and without the DBT PE protocol (n = 26; Harned, Korslund, & Linehan, 2014). Across both studies, 90% of clients who completed the full year of DBT demonstrated the behavior change necessary to start the DBT PE protocol, and this occurred after an average of 20 weeks of DBT. Among those who started the DBT PE protocol, only 16.7% relapsed into suicidal or self-injurious behavior during this portion of the treatment and 73% completed the full protocol (M = 13 sessions). Moreover, in the pilot RCT, clients assigned to DBT + DBT PE reported significantly lower urges to commit suicide and self-injure before and after therapy sessions and were 1.3 to 2.4 times less likely to engage in these behaviors than those in DBT alone. These improvements in suicidal and self-injurious urges and behaviors were obtained while simultaneously doubling the rate of PTSD remission among treatment completers (80% vs. 40%) and achieving large reductions in shame, guilt, dissociation, depression, anxiety, and global impairment compared to clients who received DBT alone. Taken together, these findings suggest that, even among high-risk and severe BPD clients, the promise of receiving an effective treatment for a problem viewed by the client as high-priority (PTSD) appears to function as a powerful reinforcer for quickly gaining control over suicidal, self-injurious, and other high-priority behaviors. Indeed, although clinicians often report concerns that traumafocused treatment will exacerbate suicidality and selfinjury, these findings demonstrate that adding the DBT PE protocol to DBT in a manner that makes it contingent on positive behavior change may actually decrease the frequency of these behaviors.

DBT-ACES Treatment Structure DBT-ACES was developed at Harborview Medical Center as an advanced level of DBT focused on achieving living-wage employment outside of the public mental health system for psychiatrically disabled clients with BPD. To reduce clients’ dependency on social services and family for basic needs, DBT-ACES focuses on selfsufficiency, teaching skills such as goal-setting, problemsolving, troubleshooting, dialectics, and reinforcement to

Contingencies Create Capabilities address deficits which repeatedly appeared during treatment development whenever clients were encouraged to seek and maintain employment (Comtois et al., 2010). DBT-ACES was developed over a decade of work with psychiatrically disabled clients with severe BPD. It became clear that major changes were more reinforcing than small ones; that is, clients strived harder and made more difficult changes for living-wage employment than they did for a few hours of employment or other smaller outcomes. At the same time, fear, shame, and frustration quickly arose when planning major changes, resulting in objections and cautions from social service and vocational staff, other clinicians, and friends and family. These aversive internal and external responses resulted in clients avoiding new behaviors (e.g., looking for work, investigating vocational training) and retreating to behaviors that (although no longer life threatening) were familiar, calming, and/or approved of by others (e.g., hanging out at home, helping out friends or family). Contingencies for new behaviors had to be created in DBT-ACES to compete with the increased negative emotions and many systemic disincentives that new behaviors trigger. Incentives were also needed to help DBT therapists to persist in the face of clients' increased anticipatory anxiety and other negative emotions—especially for new therapists who had not experienced how quickly these emotions resolve once the person begins the new behavior (such as starting work). Using CM Prior to DBT-ACES The first set of contingencies in DBT-ACES is similar to the DBT PE protocol: that is, contingencies for admission to DBT-ACES. Admission to DBT-ACES is contingent on (a) completion of an application, (b) performance evaluations by the individual and group therapists, (c) creating a resume ready to use, (d) passing a test of DBT skills, (e) engaging in 20 hours per week of normative and productive activity (i.e., scheduled activities outside their home where the client acts like they don’t have emotional problems around other people acting like they don’t have emotional problems), and (f) at least 2 months of no life-threatening or significant therapy-interfering behavior. The therapist begins working on these requirements by 6 months into the year of DBT and application materials are due one skills module before the end of the year for those wanting to continue directly into DBT-ACES. Returning to DBT-ACES at a later time or attending DBT-ACES after completing DBT elsewhere require the same criteria at the time of enrollment. These criteria are always required, and interested clients are encouraged to keep working to meet these criteria until they achieve them. As with the DBT PE protocol, the effectiveness is dependent on how reinforcing

117

entering DBT-ACES is to the individual client. For interested clients, these contingencies have led to dramatic improvements in client behavior. Using CM During DBT-ACES The second set of contingencies are applied during the DBT-ACES year. Two series of graduated contingencies facilitate living-wage employment. The first, Career Development, refers to the activities that will prepare clients for the profession they have chosen. These activities include attending college, internships, employment in their field or in junior or volunteer versions of their field (e.g., working as a nursing assistant or nursing volunteer as experience toward nursing), vocational-technical training (e.g., a certificate program), temporary or contract work, or self-employment or business startup. Clients must be doing this 10 hours a week by 4 months into DBT-ACES and 20 hours/week by 8 months. However, Career Development is not all that is needed to succeed in living-wage employment and be financially independent. If clients aren’t effective and successful employees, then their career skills will go to waste. And even if they can work in a particular career of their choosing, there may come a time when they lose their job, need more money, move to a new city, or otherwise have to start over. When they do this without financial support from the government or others, it means very quickly finding and maintaining a job. In DBT-ACES, these skills are achieved through the “Work as Therapy” contingency. That is, clients are required to find and then maintain a job on the open market for which their employer files official employee tax documents (e.g., a W2 in the U.S.) and work in this job for a minimum of 10 hours/week for at least 6 months during their DBT or DBT-ACES year. Since there is often variability of the timing and extent to which disability payments and insurance coverage plans are discontinued if an individual becomes employed, DBT-ACES therapists and their clients consider the implications of Work as Therapy so they can problem solve and troubleshoot in advance. Making continuing in DBT-ACES contingent on Work as Therapy has greatly improved the number of clients who worked during and then after DBT-ACES. As noted earlier, clients in DBT-ACES often experience increased symptoms such as anxiety and substantial distress in response to an approaching deadline to begin or increase work or career activities; however, the deadlines function to reinforce persistence as the client is suspended from treatment (aka put on a DBT “therapy vacation”) if the deadline is missed until the criteria is met. The suspension functions as negative reinforcement because it is under the control of the client—it can be avoided by effective behavior and stops as soon as the employment criteria are met within 1 week. (To assure it

118

Carmel et al.

is not the economy in control of the client’s employment, clients are required to submit 1, then 5, then 10, then 15 applications per week leading up to the 4-month deadline. This almost always results in employment on time. In cases where it doesn’t, clients who have done this much work can apply for an extension of the deadline.) Suspension from DBT-ACES is selected as a negative reinforcer in large part because it mirrors real-world environmental contingencies. For example, the client who doesn’t complete their tasks at their job will likely get suspended or fired. Therefore, if the client doesn’t do their DBT-ACES related tasks of homework, skills practice, or the steps towards the Work as Therapy requirement, they will have a chain analysis completed by their therapist in session, receive reminders about the consequence of problem behaviors, and eventually get suspended from DBT-ACES. In the case of suspension, behavioral change generally comes from wanting to get back in touch with the reinforcers of the therapeutic relationship with the therapist and the support of DBT-ACES skills group. Eventually, DBT-ACES helps clients replace the reinforcers within the therapeutic relationship with the reinforcers from employment. The DBT-ACES approach to increasing self-sufficiency and reducing financial dependency also includes a third set of contingencies: a strong emphasis on positive reinforcement of effective behaviors in-session. The individual and group skills therapists constantly provide cheerleading around the client’s ability to achieve living-wage employment and self-sufficiency, to meet each of the criteria, and for all evidence of hope, willingness, and interest in DBT-ACES. Therapists also do a lot of planned ignoring of hopelessness, willfulness, and negativity (i.e., ignoring is almost always aversive). This generally results in a client’s movement back to hope, willingness, and interest, which results in the return of the therapist’s attention (negative reinforcement). Some clients do not have a vision of living-wage employment—either from fear, other obligations, or lack of interest. Making DBT-ACES a requirement for positive reinforcement or focusing too much positive reinforcement on DBT-ACES when the client is not genuinely interested generally results in half-hearted DBT-ACES behavior, which does not result in finding and keeping a job or career. Therefore, when clients are more hopeful, willing, or interested in a plan other than DBT-ACES, therapists reinforce this and help clients move in a direction other than DBT-ACES after their DBT year. CM Strategies and the Effectiveness of DBT-ACES DBT-ACES shows promise as an advanced treatment for graduates of DBT to increase employment and live largely free of psychiatric symptoms based on the findings of a feasibility study in the treatment of 30 clients with BPD

(Comtois et al., 2010). Within the 2-year treatment period that included 1 year of DBT followed by 1 year of DBT-ACES, there was a significant improvement in participants’ odds of being employed or in school, working at least 20 hours per week, subjective quality of life, and a decrease in the number of inpatient admissions. Comparing the end of DBT to a year after DBT-ACES during which the majority received no DBT, the latter two outcomes were mostly retained, but the findings were not significant. One year after leaving DBT-ACES, only 36% of DBT-ACES clients were still receiving public mental health services. Emergency room admissions, inpatient psychiatry admissions, and medically treated self-inflicted injuries all decreased during DBT and remained low during and following DBT-ACES (Comtois et al., 2010). There are two additional active DBT-ACES programs outside of the University of Washington at Harborview Medical Center, one at the Harbor-UCLA Medical Center, and the other in Lengerich, Germany.

Conclusion DBT PE and DBT-ACES are adjunctive treatments of DBT that make the availability of treatment contingent on behavior change. This appears to create significant positive changes in the lives of individuals with BPD, the most important of which is the quicker and more complete reduction of life-threatening behaviors in order to start these adjunctive treatments. Both treatments place emphasis on the benefit of exposure, the importance of not reinforcing phobic avoidance, utilizing skills in the client’s environment, and on creating more naturalistic environmental reinforcers. By employing more strategic and planned reinforcers in therapy, this leads the client to increase natural reinforcers in their life to enhance behavioral capabilities (i.e., planning for and accomplishing a DBT-ACES Career Development task will increase capabilities to plan for and accomplish a task at work). As in DBT, DBT-ACES and DBT PE therapists create an environment rich in reinforcement, with positive reinforcers delivered in response to behavior that indicates clinical progress, rather than provide reinforcers indiscriminately or immediately following maladaptive behaviors that are often maintained by their tendency to elicit validation or other reinforcing responses from the environment (Linehan, 1997). This is in contrast to the pattern seen in many health systems of providing extra therapy sessions or contact with providers following maladaptive client behaviors, such as suicide crisis behaviors, which often serves to inadvertently reinforce these behaviors (Koerner, 2012). This illustrates a primary dialectical tension within the DBT treatment model of working with the client to create a warm and supportive relationship that validates the client’s emotional experience while being careful to guard the therapeutic process from mood-dependent reactions that

Contingencies Create Capabilities unintentionally reinforce maladaptive behaviors. The DBT therapist uses CM strategies frequently in session by identifying client behaviors as adaptive versus maladaptive as they occur in the moment, followed by a quick response involving positive reinforcement whenever possible. More formal contingency plans are identified and applied in a collaborative manner with the goal of shaping the client into becoming more active in creating and implementing their own behavioral plans. The DBT-ACES Career Development deadline is an example of a formal contingency plan to set up the client to build a life that is motivated by reinforcement instead of variable punishment and short-term relief with no long-term change. Individuals with BPD have a chronic pattern of engaging in mood-dependent behavior, which is often accompanied with a belief that one is unable to follow through with many steps to achieve greater quality of life. Both clients themselves, and others in their social network, are often surprised that if the client is held to task using contingencies, it can be possible to remain consistent in treatment, gradually tolerate increased stressors, block avoidance responses, and build in natural reinforcers into life outside of therapy. Clients whose behavior is shaped by CM and who find environments that reinforce them for being effective achieve more. Contingencies create capabilities and capabilities create a life worth living. As seen in these adjunctive treatments, clients are able to achieve many life goals including reducing posttraumatic stress responses or increasing employment, self-sufficiency, and financial independence.

References Bagge, C. L., Stepp, S. D., & Trull, T. J. (2005). Borderline personal disorder features and utilization of treatment over two years. Journal of Personality Disorders, 19(4), 420–439. Bedics, J. D., Atkins, D. C., Comtois, K. A., Linehan, M. M. (2012). Treatment differences in therapeutic relationship and introject during a 2-year randomized controlled trial of dialectical behavior therapy versus nonbehavioral psychotherapy experts for borderline personality disorder. Journal of Consulting and Clinical Psychology, 80(1), 66–77. Bender, D. S., Dolan, R. T., Skodol, A. E., Sanislow, C. A., Dyck, I. R., McGlashan, T. H., . . . Gunderson, J. G. (2001). Treatment utilization by clients with personality disorders. American Journal of Psychiatry, 158(2), 295–302. Benjamin, L. S., Rothweiler, J. C., & Critchfield, K. L. (2006). The use of structural analysis of social behavior (SASB) as an assessment tool. Annual Review of Clinical Psychology, 2, 83–109. Comtois, K. A., & Carmel, A. (2014). Borderline personality disorder and high utilization of inpatient psychiatric hospitalization: Concordance between research and clinical diagnosis. Journal of Behavioral Health Services & Research. Advance online publication. http://dx.doi.org/ 10.1007/s11414-014-9416-9 [ePub ahead of print.]. Comtois, K. A., Kerbrat, A. H., Atkins, D. A., Harned, M. S., & Elwood, L. (2010). Recovery from disability for individuals with borderline personality disorder: A feasibility trial of DBT-ACES. Psychiatric Services, 61(11). Domjan, W. (2003). The principles of learning and behavior (p. 302). Stamford, CT: Thompson Learning.

119

Foa, E., Hembree, E., & Rothbaum, B. (2007). Prolonged exposure therapy for PTSD: Emotional processing of traumatic experiences. New York: Oxford University Press. Goodman, M., Patil, U., Steffel, L., Avedon, J., Sasso, S., Triebwasser, J., & Stanley, B. (2010). Treatment utilization by gender in clients with borderline personality disorder. Journal of Psychiatric Practice, 6(12). Harned, M. S. (2013). Treatment of posttraumatic stress disorder with comorbid borderline personality disorder. In D. McKay & E. Storch (Eds.), Handbook of treating variants and complications in anxiety disorders (pp. 203–221). New York: Springer. Harned, M. S., Korslund, K. E., Foa, E. B., Linehan, M. M. (2012). Treating PTSD in suicidal and self-injuring women with borderline personality disorder: Development and preliminary evaluation of a dialectical behavior therapy prolonged exposure protocol. Behaviour Research and Therapy, 50, 381–386. Harned, M. S., Korslund, K. E., Linehan, M. M. (2014). A pilot randomized controlled trials of Dialectical Behavior Therapy with and without the Dialectical Behavior Therapy Prolonged Exposure protocol for suicidal and self-injuring women with borderline personality disorder and PTSD. Behaviour Research and Therapy. Harned, M. S., Tkachuck, M. A., & Youngberg, K. A. (2013). Treatment preference among suicidal and self-injuring women with borderline personality disorder and PTSD. Journal of Clinical Psychology, 69, 749–761. Koerner, K. (2012). Doing dialectical behavior therapy: A practical guide. New York: Guilford Press. Linehan, M. M. (1993). Cognitive-behavioral treatment of borderline personality disorder. New York: Guilford Press. Linehan, M. M. (1997). Validation and psychotherapy. In A. Bohart & L. Greenberg (Eds.), Empathy reconsidered: New directions in psychotherapy (pp. 352–392). Washington, DC: American Psychological Association. Linehan, M. M., Armstrong, H. E., Suarez, A., Allmon, D., & Heard, H. L. (1991). Cognitive-behavioral treatment of chronically parasuicidal borderline clients. Archives of General Psychiatry, 48, 1060–1064. Linehan, M. M., & Heard, H. L. (1993). Impact of treatment accessibility on clinical course of parasuicidal clients: In reply to R.E. Hoffman. [Letter to the editor]. Archives of General Psychiatry, 50, 157–158. Linehan, M. M., & Heard, H. L. (1999). Borderline personality disorder: Costs, course, and treatment outcomes. In N. Miller & K. Magruder (Eds.), The cost effectiveness of psychotherapy: A guide for practitioners (291–305). New York: Oxford University Press. Linehan, M. M., Kanter, J. W., & Comtois, K. A. (1999). Dialectical behavior therapy for borderline personality disorder: Efficacy, specificity, and cost-effectiveness. In D. S. Janowsky (Ed.), Psychotherapy: Indications and outcomes (pp. 93–118). Washington, DC: American Psychiatric Press. Manning, S. Y. (2011). Common errors made by therapists providing telephone consultation in dialectical behavior therapy. Cognitive and Behavioral Practice, 18, 178–185. Pryor, K. (2006). Don't shoot the dog! The new art of teaching and training (3rd ed.). Devon, U.K.: Ringpress Books. Rizvi, S. L., Steffel, L. M., & Carson-Wong, A. (2013). An overview of dialectical behavior therapy for professional psychologists. Professional Psychology: Research and Practice, 44, 73–80 Salsman, N. L., & Linehan, M. M. (2006). Dialectical behavior therapy for borderline personality disorder. Primary Psychiatry, 13(5), 51–58. Skinner, B. F. (1988). The operant side of behavior therapy. Journal of Behavior Therapy and Experimental Psychiatry. 19(3), 171–179. Soeteman, D. I., Busschback, J. J., Verheul, R., Hoomans, T., & Kim, J. J. (2011). Cost-effective psychotherapy for personality disorders in the Netherlands: The Value of further research and active implementation.. Value in Health, 14, 229–239. Stoffers, J. M., Vollm, B. A., Rucker, G., Timmer, A., Huband, N., & Lieb, K. (2012). Psychological therapies for people with borderline personality disorder. Cochrane Database of Systematic Reviews, 8. Wagner, T., Fydrich, T., Stiglmayr, C., Marschallio, P., Salize, H. J., Renneberg, B., Fleba, S., Roepke, S. (2014). Societal cost-of-illness in clients with borderline personality disorder one year before, during and after dialectical behavior therapy in routine outpatient care. Behaviour Research and Therapy, 61, 12–22.

120

Carmel et al.

Wunsch, E. M., Kliem, S., & Kroger, C. (2014). Population-based cost– offset estimation for the treatment of borderline personality disorder: Projected costs in a currently running, ideal health system. Behaviour Research and Therapy, 60, 1–7. Zanarini, M. C., Frankenburg, F. R., Hennen, J., & Silk, K. R. (2004). Mental health service utilization by borderline personality disorder clients and Axis II comparison subjects followed prospectively for 6 years. Journal of Clinical Psychiatry, 65(1), 28–36. Zanarini, M. C., Frankenburg, F. R., Reich, D. B., Fitzmaurice, G. (2012). Attainment and stability of sustained symptomatic remission and recovery among patients with borderline personality disorder and

axis II comparison subjects: A 16-year prospective follow-up study. American Journal of Psychiatry, 169(5), 476–483. Address correspondence to Adam Carmel, Ph.D., Massachusetts Mental Health Center, 75 Fenwood Road, Boston, MA 02115.; e-mail: [email protected]. Received: October 23, 2014 Accepted: April 5, 2015 Available online 14 April 2015