Veterans Service Utilization and Associated Costs Following Participation in Dialectical Behavior Therapy: A Preliminary Investigation

Transcription

1 MILITARY MEDICINE, 179, 11:1368, 2014 Veterans Service Utilization and Associated Costs Following Participation in Dialectical Behavior Therapy: A Preliminary Investigation Laura L. Meyers, PhD*; Sara J. Landes, PhD ; Paul Thuras, PhD* ABSTRACT Context: Dialectical Behavior Therapy (DBT) is an evidence-based therapy developed for the treatment of suicidal behaviors and disorders characterized by emotional and behavioral dyscontrol that is effective in veteran populations. The impact of DBT on veterans Veterans Affairs (VA) service utilization and cost is unknown. Evidence Acquisition: This study evaluated the impact of DBT in a VA outpatient mental health setting on VA service utilization and cost of services. Veterans treated for symptoms of Borderline Personality Disorder, who had completed at least 6 months of the DBT program were sampled (N = 41). Use of physical and mental health services during the years prior and following DBT was assessed using medical record information. Results: There was a significant decrease in mental health service utilization. Psychiatric hospitalization dropped in half, and for those with a hospitalization, length of stay decreased significantly. Direct costs associated with all health care were significantly reduced. Conclusion: Changes in service utilization resulted in a significant reduction in direct costs of providing care to veterans with symptoms of Borderline Personality Disorder. Additional research is needed to compare the reduction in overall costs to the cost of implementing DBT and to compare these changes to a control group. INTRODUCTION In the last decade, more than 2.2 million troops have been deployed to Iraq and Afghanistan and many have experienced multiple deployments. As these conflicts come to an end and troop size diminishes, the number of veterans will increase and health care settings, including the Department of Veterans Affairs (VA), will see an increase in the number of veterans seeking health and mental health services. 1 A number of veterans struggle with emotional and behavioral dysregulation and/or chronic issues, such as ongoing suicidal behavior. Suicide is a serious problem for veterans. According to the VA/Department of Defense (DoD) Clinical Practice Guidelines for the Assessment and Management of Suicidality, 2 veterans account for approximately 20% of deaths from suicide in the United States and veterans using VA services account for about 5 suicides per day. For those veterans using VA services with a recorded mental health diagnosis within the past year, the rate of suicide is 70 per 100,000 (as compared to rates of approximately 38 per 100,000 among men and 15 per 100,000 among women in civilian populations). Evidence-based psychotherapies exist for treating suicide, such as cognitive therapy 3 and Collaborative Assessment and *Minneapolis VA Health Care System, University of Minnesota Medical School, One Veterans Drive (116A), Minneapolis, MN National Center for PTSD, VA Palo Alto Health Care System, 795 Willow Road, Menlo Park, CA This article was presented at the annual VA Psychology Leadership Conference, San Antonio, TX, April 11 14, The views expressed reflect those of the authors and not necessarily those of the U.S. Department of Veterans Affairs, the U.S. Government, or the University of Minnesota Medical School. doi: /MILMED-D Management of Suicidality. 4,5 However, in addition to suicidal behavior, some veterans also struggle with concomitant emotional and behavioral dysregulation and/or chronic issues that place them at risk for chronic coping deficits or increased risk of suicidal behaviors. For veterans with additional difficulties beyond acute suicidality, longer term or more comprehensive treatments are necessary. One such treatment is Dialectical Behavior Therapy (DBT). 6 The VA/DoD Clinical Practice Guidelines recommend DBT for treating an underlying disorder (i.e., Borderline Personality Disorder or other personality disorders characterized by emotional dysregulation and a history of suicide attempts and/or self-harm) in patients who are suicidal. With its emphasis on skills training and mindfulness-based emotion regulation, DBT is the most thoroughly studied treatment of the existing psychotherapies for suicidal behavior. 2(p 97) DBT has also been recommended for veterans with complex trauma exposure. 7 DIALECTICAL BEHAVIOR THERAPY DBT is an evidence-based therapy that was developed for the treatment of suicidal behaviors and disorders characterized by emotional and behavioral dyscontrol (e.g., Borderline Personality Disorder) and may therefore be particularly well suited to treating the kinds of complex emotional and behavioral dysregulation seen in veteran populations. DBT addresses five core areas of dysregulation: emotional, interpersonal, self, behavioral, and cognitive. In the standard model of DBT, patients are seen for individual therapy 1 hour per week, skills group 2 hours per week, and have access to 24/7 telephone coaching for assistance with skills generalization. All DBT therapists also participate in weekly consultation teams, in which the focus is therapy for the therapist and to 1368 MILITARY MEDICINE, Vol. 179, November 2014

2 promote adherence to skills use and practice. Patients are also encouraged to engage in other ancillary treatments (e.g., medication management, substance abuse support). 6 DBT has been recognized by the Substance Abuse and Mental Health Services Administration 8 as an evidencebased treatment for reducing suicide attempts, nonsuicidal self-injury, substance abuse, symptoms of eating disorders, anger, and depression. Currently, there have been at least 11 randomized controlled trials demonstrating the efficacy of DBT, including trials conducted independently of the treatment developer. Within the system, there is also evidence that DBT is effective in treating suicidal ideation, hopelessness, depression, dissociation, and expression of anger among female veterans diagnosed with Borderline Personality Disorder. 9 Although many treatment studies of DBT report outcomes that include service utilization (i.e., frequency of non-dbt medical and mental health care, such as emergency room visits, inpatient treatment days), they do not report the cost (i.e., dollar amount spent by the medical system) associated with those outcomes. Therefore, there is limited data published on the cost-effectiveness of DBT, and a comprehensive literature review did not identify any studies among the veteran population. In a review of the effects and costs of treatments for personality disorders in general, Bartak et al 10 concluded that the evidence for treatments is favorable and that cost-effectiveness data are still scarce. Two recently published studies of DBT in the National Health Service (NHS) in the United Kingdom have examined cost related to DBT. Priebe et al 11 found that although the mean total cost of DBT was higher than treatment as usual (TAU) ( 5,685 versus 3,754), it was not significantly different. As this was a pragmatic trial, TAU included any kind of treatment other than DBT offered by the NHS to provide realworld alternative treatments. 11(p 358) However, they only looked at the cost of treatment during the treatment year and did not compare service utilization the year before DBT to the year following DBT, where there may have been a total cost savings (e.g., individuals using fewer services as a result of receiving DBT). Amner 12 examined the effect of providing DBT in reducing the cost of adults with symptoms of Borderline Personality Disorder in a specific region of the NHS. She compared pre- and post-treatment service costs associated with 1 year of DBT and found a 36,000 savings by the end of the post-treatment year. Cost savings resulted mainly from fewer inpatient hospital admissions and community nursing contacts. Although DBT is not one of the nationally rolled out evidence-based practices in VA, 13 it has been implemented locally at sites across VA, with the full model available at 13 of the 143 VA medical centers. DBT has been offered at the Minneapolis VA since 1996, 14 with the addition of 24/7 telephone coaching added in Before this article, there does not appear to have been any published data on the costeffectiveness of DBT with veterans. These data would be enormously helpful to clinics and programs considering implementing DBT, as its implementation is both time and resource intensive. 15 Current Study The purpose of the present retrospective study was to evaluate veterans service utilization and associated costs following participation in DBT provided by VA. It was expected that veterans who completed at least 6 months of a 6- to 18-month course of DBT treatment would utilize less mental and physical health care services in the year following treatment as compared to the year before treatment. METHODS After receiving institutional review board approval, all veterans who had successfully completed DBT at a Midwestern VA medical center with a well-established DBT program were identified through existing administrative databases. Data were gathered from computerized medical records for a period 1 year before the start of treatment and 1 year after the conclusion of treatment (from August 2006 to January 2011). Encounters during treatment were not included. Data included encounter location and treatment procedures. Encounters were coded by stop code, which indicate whether the encounter took place in primary care, emergency department, or mental health (includes all mental health stop codes). Mental health encounters were further subdivided into those occurring in the Partial Day Hospitalization Program (PPH) and those occurring in mental health outside of PPH. Cost data were pulled by the Decision Support Services (a division of VA s business operation support) for these patients for the same period of time as the encounter and procedure data. Patient Characteristics The patients sampled were 41 outpatients at a Midwestern VA medical center being treated for significant levels of symptoms of Borderline Personality Disorder. Upon referral to DBT, each patient was assessed by his/her assigned DBT therapist through a clinical interview that included the 5 core areas of emotional, interpersonal, self, behavioral, and cognitive dysregulation. 6 Clinical judgment was used to determine if patients demonstrated a significant level of dysregulation in 3 of the 5 areas to warrant the level of intensity of care of DBT. The average age was 47.1 years of age (standard deviation [SD] = 10.7). The sample was 54% (22) male and 89% (32/36) of those disclosing ethnic identity were Caucasian. All patients (100%) had a chart diagnosis of depression or dysthymia; 81% had a comorbid diagnosis of post-traumatic stress disorder (PTSD), 70% had a comorbid substance use diagnosis, and 95% had been diagnosed with an Axis II personality disorder. Table I lists the demographic and clinical characteristics of the patients sampled. Formal inclusion and exclusion criteria were not utilized; however, veterans with active psychosis, unmedicated bipolar disorder, and severe uncontrolled substance dependence are typically not enrolled MILITARY MEDICINE, Vol. 179, November

3 TABLE I. Demographic and Clinical Characteristics of Patients Sampled Characteristic N Percentage Age (M + SD) Gender Male Female Ethnic Identity Caucasian African American 3 7 Asian 1 2 Unknown/Refused 5 12 Axis I Diagnoses Present Depression or Dysthymia PTSD Substance Abuse At least 1 Axis II Diagnosis Present in DBT. Reliable data regarding veterans who initiated but did not complete DBT were not available. Treatment Veterans were referred for DBT by their primary mental health care provider. An assessment interview was conducted to determine appropriateness of fit for DBT. Specifically, a diagnosis of Borderline Personality Disorder, or significant symptoms that warranted this level of care, was the ground for inclusion. Following Linehan s 6 recommendations, veterans with significant levels of dysregulation in core areas of emotional, interpersonal, self, behavioral, and/or cognitive dysregulation were included, regardless of formal diagnosis. Treatment was conducted by psychologists, social workers, and advance practice nurses who were intensively trained in DBT (i.e., 2-week intensively trained) 15 or who were trained locally by providers who had completed intensive training. This local training consisted of a day-long overview of DBT philosophy and skills, a 6-month 90-minute weekly didactic and group supervision that reviewed the Linehan 6 text, weekly 90-minute consultation team, and coleading a DBT skills group and carrying one individual patient under supervision. Treatment was also provided by psychology or social work interns who had completed a comprehensive training workshop and participated in a 6-month didactic series. Veterans receiving DBT completed at least 6-months of the full model of DBT (i.e., weekly 1-hour individual session, weekly 2-hour DBT skills group, access to 24/7 DBT phone coaching, all therapists participated in weekly 90-minute consultation group). Analyses Pre- and post-treatment comparisons were made on cost data for all primary stop codes in mental health, the emergency department, and primary care. For continuously distributed data (e.g., PTSD and depressive symptoms, cost data) matched-pairs t tests were conducted to look at change before and after treatment. Within group, pre-to-post effect sizes were calculated using Cohen s d. 16 Mental and physical health care utilization data as well as some cost data were highly positively skewed (skew > 1.5) and Wilcoxon signed ranks tests were conducted to examine pre post differences. Analyses were conducted with SPSS version 19. RESULTS Pre- and post-treatment measures of health care encounters are displayed in Table II. We found a decrease in the use of mental health services from an average of 92.8 (SD = 64.5) visits in the year before treatment to an average of 48.2 (SD = 52.9) visits in the year after treatment (z = 3.62, p < 0.001). Because of the large number of encounters that accompany use of partial day hospital programs (i.e., multiple group sessions each day, Monday-Friday, for 3 weeks) we removed those encounters from the analysis as a small number of individuals (i.e., 4) could distort the number of encounters observed. We re-examined pre post differences in mental health utilization and significant pre post differences remained with an average of 57.0 encounters before treatment and 40.4 after treatment (z = 2.82, p = 0.005). We also examined TABLE II. Number of Encounters 1 Year Pre- and Post-Therapy (N = 41) Percentiles Mean SD Minimum Maximum 25th 50th (Median) 75th z Significance Pre-all Mental Health Post-all Mental Health <0.001 Pre-PPH Post-PPH Pre-all Mental Health W/O PPH Post-all Mental Health W/O PPH Preprimary Care Postprimary Care Pre-emergency Department Postemergency Department PPH, Partial Day Hospitalization Program MILITARY MEDICINE, Vol. 179, November 2014

4 psychiatric hospitalizations and length of stay 1 year before and 1 year after DBT. Admissions dropped in half from 34.1% in the year before treatment to 17.1% in the year after treatment (z = 1.92, p = 0.052). Among those with a hospitalization either before or after treatment (n = 17), length of stay also dropped from an average of 5.2 days to 1.6 days (t [16] = 2.91, p =.01). For the 4 patients with hospital admissions both before and after DBT, length of stay dropped from 5.25 to 3.25 days (t [3] = 2.00, p = 0.14). We also examined the use of primary care and the emergency department pre- and post-therapy. The decline in primary care use was modest (4.0 to 3.1 visits) and not significant, z = 1.58, p = Emergency department usage trended downward as well from 1.95 to 1.34 visits, z = 1.69, p = Direct costs associated with all health care were significantly reduced in the 1-year period following DBT when compared to the same period before treatment, t (40) = 3.13, p < This reduction from an average of $21, (SD = 11,010.56) to an average of $15, (SD = 10,325.64) per patient yields a savings of 28.2% and an effect size of d = Regarding mental health treatment, we observed a significant decline in both outpatient direct costs from an average of $ ($4363.8) per patient to $ ($ ), a reduction of 41.3% (z = 3.62, p < 0.001, d = 0.60) and inpatient directs costs from an average of $ ($ ) per patient to $ ($ ) a reduction of 71.2% (z = 2.49, p = 0.013, d = 0.39). We did not observe a significant change in primary care direct costs (z = 0.41, p = 0.68) but did observe a trend toward lower emergency department direct costs (z = 1.83, p = 0.067) with costs declining from an average of $ ($651.12) per patient to $ ($384.39). The average cost to provide DBT in this VA outpatient setting was $ per patient per week. This was calculated using the cost of 2 hours of group and 1 hour of individual therapy per week. The average length of time in DBT treatment for this sample was 46 weeks (SD = 20.9). Therefore, the average cost of providing a course of DBT was $ per patient. Of note, this did not include other treatment encounters during the course of DBT (e.g., medication management visits). DISCUSSION In the current study, veterans who received DBT services in a VA setting had significantly decreased the use of outpatient mental health services in the year following DBT by 48%, as compared to the year before DBT. This information is extremely important for clinicians and administrators who may be considering implementing a DBT program in their setting, as one of the primary barriers to doing so is a perceived need to increase services to implement DBT. DBT is a comprehensive treatment, including weekly skills group and individual therapy. However, in this study in the year before DBT, patients were attending an average of 1.78 sessions per week of services, just under the 2 sessions per week required by DBT. Following treatment in DBT, patients were attending an average of 0.93 mental health sessions per week. Therefore, although DBT does increase services during the 1-year treatment period, this study provides preliminary data that the long-term service utilization decreases in highservice-utilizing patients. Perhaps more importantly, this study demonstrated that utilization of high-cost inpatient services decreased following treatment in DBT. In this sample, the number of patients requiring psychiatric hospitalizations decreased by 50%. Of those who were admitted, their length of stay was 69% shorter than before treatment, resulting in an average savings of $ in inpatient costs per patient in the 12-months following DBT treatment. Use of primary care services and emergency department visits declined, but were not significant. In summary, each patient in the program cost $ less in total services in the year following DBT treatment as compared to the year prior. In this sample of 41 patients, this is a total cost decrease of $244, Of note, the average length of treatment for those who completed DBT was 46 weeks. Given the challenges of engaging and retaining patients with complex problems that may include Borderline Personality Disorder so that they receive an adequate dose of treatment, this length of time in treatment can be viewed as compelling evidence that DBT engages and retains patients in treatment. This is especially important in considering a population that is often considered difficult to treat and are more likely to drop out of most treatments. 17 These findings extend the literature available on the impact of DBT on service utilization and cost. This study is the first available with a veteran population. It also extends research by Priebe et al 11 who compared the cost of DBT to TAU, but did not compare use and cost of services the year before and the year following DBT. The findings are similar to those found by Amner, 12 who found that the cost of services was lower in the year following DBT than the year before. The average cost of providing a course of DBT in this outpatient setting was $ per patient. This amount is nearly equivalent to the average outpatient direct costs from the year before DBT treatment, which was $ per patient. This comparison is hindered, however, because of the fact that no other treatment encounters were included in the estimate of the cost of providing DBT (e.g., medication management visits) and should be further explored in future research. This information is helpful to show that although DBT is a complex and lengthy treatment, it may not cost more than the services provided to clients with chronic difficulties, such as Borderline Personality Disorder. This may be especially true when considering how long a patient has been a high utilizer of services. These findings are relevant to clinicians and administrators who are considering offering DBT in their outpatient setting. The requirements of developing a DBT program can MILITARY MEDICINE, Vol. 179, November

5 initially appear daunting, though the band-aid alternative of crisis-response services leads to significantly greater financial costs in the long run. The results of this study provide initial support to clinicians and administrators advocating for the development of DBT programming at their facility from a cost-savings perspective. Limitations The primary limitation of this study is that there was no comparison group. It is possible that the use of services in this sample would have decreased over time, regardless of participation in DBT. This is unlikely, given the chronic nature of the symptoms treated in DBT, though that cannot be determined without a control group. The use of a control group is suggested for future studies. An additional limitation is the absence of non-va service utilization data. The generalizability of these findings to other VA settings may be limited given the large proportion of the sample that was female. Although it is unknown who DBT is predominantly offered to across VA settings, previous research on DBT in VA was conducted with female veterans with Borderline Personality Disorder. However, this sample may be more similar to a civilian sample, given the proportion of female patients. Future Directions Given this promising initial data, it is encouraged that future studies examine all aspects of the short-term costs versus the long-term gains of implementing the full model of DBT in the VA system and include a control group for comparison. When working with such an intensive, high-risk, population, initial efforts may feel daunting. Long-term benefits of cost, symptom reduction, increase in quality of life, and decreases in provider burnout and increases in job satisfaction are crucial to motivating providers and administrators to develop DBT programming. Beyond clinical efficacy, DBT needs to demonstrate cost efficacy to build support for its implementation. Future research could examine a possible dose response relationship between months of DBT and change in service use to determine a possible minimum dose of DBT needed. Additional health economic analyses should examine incremental cost-effectiveness ratios, include a measure of quality of life, and examine quality-adjusted life years. These more refined examinations of health economics would help to make an even stronger argument for the cost-effectiveness of such a comprehensive treatment. In addition, data on missed work days and lost productivity should be included. CONCLUSIONS This study examined the changes in service utilization and associated costs in the year before DBT treatment to the year following treatment in a sample of 41 patients who completed DBT in a VA setting. Findings highlight that after a course of DBT, patients utilized less mental health services, reduced psychiatric hospitalization admissions by half, and decreased length of psychiatric hospital stays. These changes resulted in a significant reduction in direct costs of providing care to veterans with significant symptoms of Borderline Personality Disorder. Additional research is needed to compare the reduction in overall costs to cost of implementing DBT and to compare these changes to a control group. Future research should examine this, as well as other factors that impact the implementation and provision of DBT for veterans. This knowledge is helpful to clinicians and administrators making decisions regarding the provision of evidence-based treatments for mental health in VA settings. REFERENCES 1. Tuerk PW, Wangelin B, Rauch SAM, et al: Health service utilization before and after evidence-based treatment for PTSD. Psychol Serv 2013; 10(4): The Assessment and Management of Risk for Suicide Working Group: VA/DoD Clinical Practice Guideline for Assessment and Management of Patients at Risk for Suicide. Washington, DC, Office of Quality and Performance, Available at MH/srb/;accessed November 5, Brown GK, Ten Have T, Henriques GR, Xie SX, Hollander J E, Beck AT: Cognitive Therapy for the prevention of suicide attempts: a randomized controlled trial. JAMA 2005; 294(5): Jobes DA: Managing Suicidal Risk: A Collaborative Approach. New York, NY, Guilford Press, Jobes DA, Wong SA, Conrad AK, Drozd JF, Neal-Walden T: The collaborative assessment and management of suicidality vs. treatment as usual: a retrospective study with suicidal outpatients. Suicide Life Threat Behav 2005; 35(5): Linehan MM: Cognitive-Behavioral Treatment of Borderline Personality Disorder. New York, NY, Guilford Press, Landes SJ, Garovoy ND, Burkman KB: Treating complex trauma among veterans: three stage-based treatment models. J Clin Psychol 2013; 69(5): Substance Abuse and Mental Health Service Administration: National Registry of Evidence-Based Programs and Practices. Available at accessed December 10, Koons CR, Robins CJ, Tweed JL, et al: Efficacy of dialectical behavior therapy in women veterans with borderline personality disorder. Behav Ther 2001; 32(2): Bartak A, Soeteman DI, Verheul R, Busschbach JVJ: Strengthening the status of psychotherapy for personality disorders: an integrated perspective on effects and cost. Can J Psychiatry 2007; 52(12): Priebe S, Bhatti N, Barnicot K, Bremmer S, Gaglia A, Katsakou C: Effectiveness and cost-effectiveness of dialectical behaviour therapy for self-harming patients with personality disorder: a pragmatic randomised controlled trial. Psychother and Psychosom 2012; 81(6): Amner K: The effect of DBT provision in reducing the cost of adults displaying the symptoms of BPD. Br Journal of Psychother 2012; 28(3): Karlin BE, Cross G: From the laboratory to the therapy room: national dissemination and implementation of evidence-based psychotherapies in the U.S. Department of Veterans Affairs Health Care System. Am Psychol 2014; 69(1): Spoont MR, Sayer NA, Thuras P, Erbes C, Winston E: Adaptation of dialectical behavior therapy by a VA medical center. Psychiatr Serv 2003; 54(5): MILITARY MEDICINE, Vol. 179, November 2014

6 15. Landes SJ, Linehan MM: Dissemination and implementation of dialectical behavior therapy: an intensive training model. In: Dissemination and Implementation of Evidence-based Psychological Interventions, pp Edited by David Barlow H., Kathryn McHugh R. New York, NY, Oxford University Press, Cohen J: Statistical Power Analysis for the Behavioral Sciences, Ed 2. Hillsdale, NJ, Lawrence Erlbaum Associates, Ben-Porath DD: Strategies for securing commitment to treatment from individuals diagnosed with Borderline Personality Disorder. J Contemp Psychother. 2004; 34(3): MILITARY MEDICINE, Vol. 179, November