هزینه های اجتماعی بیماری در بیماران مبتلا به اختلال شخصیت مرزی یک سال قبل، در طول و بعد از رفتاردرمانی دیالکتیکی در مراقبت های معمول سرپایی

Societal cost-of-illness in a German sample of patients with borderline personality disorder (BPD) was calculated for 12 months prior to an outpatient Dialectical Behavior Therapy (DBT) program, during a year of DBT in routine outpatient care and during a follow-up year. We retrospectively assessed resource consumption and productivity loss by means of a structured interview. Direct costs were calculated as opportunity costs and indirect costs were calculated according to the Human Capital Approach. All costs were expressed in Euros for the year 2010. Total mean annual BPD-related societal cost-of-illness was €28026 (SD = €33081) during pre-treatment, €18758 (SD = €19450) during the DBT treatment year for the 47 DBT treatment completers, and €14750 (SD = €18592) during the follow-up year for the 33 patients who participated in the final assessment. Cost savings were mainly due to marked reductions in inpatient treatment costs, while indirect costs barely decreased. In conclusion, our findings provide evidence that the treatment of BPD patients with an outpatient DBT program is associated with substantial overall cost savings. Already during the DBT treatment year, these savings clearly exceed the additional treatment costs of DBT and are further extended during the follow-up year. Correspondingly, outpatient DBT has the potential to be a cost-effective treatment for BPD patients. Efforts promoting its implementation in routine care should be undertaken.

Borderline personality disorder (BPD) is a severe mental disorder that is characterized by high instability in affect regulation, impulse control, interpersonal relationships and self-image (Lieb, Zanarini, Schmahl, Linehan, & Bohus, 2004) and occurs in up to 2.7% of the general adult population (Trull, Jahng, Tomko, Wood, & Sher, 2010). There are many indications that BPD is associated with high cost-of-illness. From a societal perspective, cost-of-illness include all costs regardless of the payer and consist of direct costs related to medical and non-medical resource consumption and indirect costs due to loss of productivity. BPD patients use mental health services to a greater extent than patients with major depression (Bender et al., 2001 and Bender et al., 2006) and patients with other personality disorders (Bender et al., 2001 and Hörz et al., 2010). Thereby, on average, BPD patients are first hospitalized for psychiatric reasons at the age of 21 and first seek many other psychiatric and psychotherapeutic treatments between the age of 18 and 25 (Zanarini, Frankenburg, Khera, & Bleichmar, 2001), a critical period for professional development. Accordingly, there seems to be a strong relationship between BPD and poor occupational functioning. Skodol et al. (2002) found that in comparison to patients with major depression, a significantly higher proportion of BPD patients were disabled and significantly fewer were employed. Jackson and Burgess (2004) showed that among all personality disorders, BPD had the strongest association with lost days of role functioning. Furthermore, compared to remitted BPD patients, non-remitted BPD patients were more likely to have quitted or lost their jobs due to their impaired state of health (Frankenburg & Zanarini, 2004). Despite these indications, BPD-related costs have rarely been investigated comprehensively. According to the only state-of-the-art cost-of-illness study assessing direct and indirect costs including 88 BPD patients seeking an outpatient psychotherapeutic treatment, BPD-related societal cost-of-illness in the Netherlands were substantial, amounting to €21120 (€16852)1 per BPD patient per year (Van Asselt, Dirksen, Arntz, & Severens, 2007). It is important to note that slightly more than 50% of total costs were indirect costs, mostly due to work disability, underlining the importance of completing comprehensive cost investigations. Additionally, several studies investigated BPD-related direct costs during the year preceding inpatient or outpatient psychotherapy. On average, these were €15088 in Germany (DM 24000; Jerschke, Meixner, Richter, & Bohus, 1998), €38771 (treatment group: US $44487, treatment-as-usual (TAU) group: US $52562; Bateman & Fonagy, 2003) and €16779 in Great Britain (treatment group: £7860, TAU group: £5240, each reported for a six-month period; Palmer et al., 2006) and €19978 in Australia (AUS $25526; Hall, Caleo, Stevenson, & Meares, 2001). A direct comparison between these results is impeded by a wide range of cost components included in the cost calculations in these investigations. At the same time that BPD is associated with high costs and BPD patients are heavy users of mental health treatments that are seldom specifically designed for their disorder (Jobst, Hörz, Birkhofer, Martius, & Rentrop, 2010), there are several disorder-specific and clinically effective psychological treatments for BPD patients. Of these, Dialectical Behavior Therapy (DBT; Linehan, 1993a and Linehan, 1993b) has been studied most intensely in numerous randomized controlled trials (RCTs) and has the greatest empirical evidence concerning clinical efficacy (Stoffers et al., 2012). In particular, DBT was observed to be effective in reducing self-injurious and suicidal behavior, as well as in reducing psychiatric hospital days and emergency room visits (e.g. Koons et al., 2001, Linehan et al., 1991 and McMain et al., 2009). Moreover, evidence indicates that these improvements can be maintained over a two-year follow-up period (McMain, Guimond, Streiner, Cardish, & Links, 2012). Based on these results related to the clinical efficacy of DBT, one can reasonably assume that DBT also has great economic potential. Accordingly, Heard (2000) found that the total mean annual direct costs of BPD patients were €9210 (US $9889) lower when patients participated in outpatient DBT (€9178 respectively US $9856) compared to TAU (€18388 respectively US $19745) in the US. While treatment costs for outpatient psychotherapy were higher in the DBT group, these were offset by lower psychiatric inpatient and emergency room costs. Similarly, in an Australian trial by Pasieczny and Connor (2011), total direct costs per patient during a six month time interval were €8301 (AUS $12196) in the DBT group and €12335 (AUS $18123) in the TAU condition, resulting in cost savings of €4034 (AUS $ 5927) that again were primarily due to significantly lower inpatient costs. In contrast to these findings, in a sample of self-harming patients with any personality disorder diagnosis, total mean annual direct costs in the DBT group (€6310 respectively £5685) exceeded those in the TAU group (€4167 respectively £3754) by €2143 (£1931; Priebe, Bhatti, & Barn, 2012). Further, Brazier et al. (2006) undertook economic evaluations of four RCTs that had investigated the efficacy of DBT. Total direct costs from the governmental perspective were estimated on the basis of data from the RCTs and a cost regression model. Whereas the estimated total mean annual direct costs of patients treated with DBT and TAU were on a comparable scale in two studies (DBT: €19941 respectively £15691, TAU: €21475 respectively £16898, Linehan et al., 1991; DBT: €22151 respectively £17430, TAU: €21231 respectively £16706, Van den Bosch, Verheul, Schippers, & Van den Brink, 2002), the direct costs in the DBT group were considerably higher than in the TAU group in one study (DBT: €29787 respectively £23439, TAU: €18828 respectively £14815, Koons et al., 2001) and lower in another study (DBT: €20007 respectively £15743, TAU: €24026 respectively £20985, Turner, 2000). In the present study, we assessed from the societal perspective the BPD-related cost-of-illness in a German sample of BPD patients in the 12 months prior to inclusion in a one-year outpatient DBT program. We further investigated BPD-related societal cost-of-illness during that specific year of outpatient DBT administered in the regular health-care system in the city of Berlin and during the following year. Following the results of Heard (2000) and Pasieczny and Connor (2011), we hypothesized that mostly due to the reduction of BPD-related inpatient treatment, the overall societal cost-of-illness during the DBT treatment year are lower than those in the year preceding DBT treatment. Also, in accordance with the findings of McMain et al. (2012), we expected that the presumed cost reduction is maintained during the follow-up year.

Results Study participants Socio-demographic and clinical characteristics at the time of study inclusion for the 47 patients who completed DBT treatment are shown in Table 2. Additionally, THI-data (average amount of BPD-related hospital and outpatient psychotherapeutic treatment utilization in the ten years prior to study inclusion) are presented (in Table 2). Table 2. Socio-demographic and clinical characteristics at the time of study inclusion and THI-data (average amount of BPD-related treatment utilization in the ten years prior to study inclusion). Variable DBT-completers (N = 47) Age, years: M (SD) 30.1 (8.1) Gender, n (%) Female 43 (91.5) Male 4 (8.5) Marital status, n (%) Single/divorced 18 (38.3) In relationship/married 29 (61.7) Education, n (%) High school (grade 12/13) 14 (29.8) Grade 10 28 (59.6) Grade 9 4 (8.5) No graduation 1 (2.1) General psychiatric symptoms (BSI/GSI), M (SD) (scale range 0–4) 1.9 (0.6) Depression (BDI), M (SD) (scale range 0–63) 31.1 (8.6) Borderline symptom severity (BSL), M (SD) (scale range 0–4) 2.1 (0.5) Number of SCID-II/BPD-criteria, M (SD) 6.5 (1.2) Number of BPD-related hospital admissions during the last ten years (THI), M (SD) 3.4 (3.9) Number of BPD-related inpatient days and days in day program during the last ten years (THI), M (SD) 152.7 (176.1) Number of BPD-related outpatient psychotherapies during the last ten years (THI), M (SD) 1.0 (0.9) Number of BPD-related outpatient psychotherapy sessions during the last ten years (THI), M (SD) 60.6 (89.4) Note. THI = Treatment History Inventory, BSI = Brief Symptom Inventory, GSI = Global Severity Index, BDI = Beck Depression Inventory, BSL = Borderline Symptom List, THI = Treatment History Inventory. Varying n due to missings (BSI: n = 40; BDI: n = 43; BSL: n = 44). Table options Resource consumption and productivity loss In Table 3, the mean annual amount of BPD-related resource consumption and mean annual amount of productivity loss during the three assessment periods is reported. In addition, the percentage of patients who used each service and the percentage of patients who had paid employment or were on work disability are presented for each time interval. Table 3. Amount of BPD-related resource consumption and productivity loss during pre-treatment, DBT treatment and follow-up. Cost Category Pre-treatment year M (SD) (N = 47) (%) DBT-treatment-year (N = 47) (%) Follow-up-year (N = 33) (%) M (SD) M (SD) M (SD) Direct medical and non–medical costs Psychiatric/general hospital, totala 51.3 (74.2) 59.6 6.8 (19.9) 19.1 6.0 (17.3) 15.2 Inpatienta 44.7 (70.5) 51.1 6.8 (19.9) 19.1 6.0 (17.3) 15.2 Day programa 6.6 (17.1) 17.0 0 (0) 0 0 (0) 0 Assisted livinga 32.3 (102.9) 10.6 38.8 (113.8) 10.6 38.3 (109.5) 12.1 Psychotropic drugsb 1.9 (1.9) 68.1 1.5 (1.4) 70.2 1.2 (1.3) 54.5 Emergency roomc 0.8 (2.1) 29.8 0.5 (1.3) 19.1 0.5 (1.5) 18.2 Outpatient psychotherapy, other than DBTc 9.1 (18.2) 38.3 0 (0) 0 0.9 (4.3) 6.1 Individual DBTc 0 (0) 0 33.7 (9.2) 100.0 18.5 (12.8) 97.0 Skills training groupc 0 (0) 0 16.9 (11.1) 89.4 1.2 (2.0) 33.0 Psychiatristc 5.5 (6.8) 74.5 2.7 (3.2) 57.4 3.7 (4.8) 54.5 General practitionerc 3.7 (5.3) 57.4 2.1 (4.3) 38.3 1.5 (5.0) 15.2 Medical specialistc 1.3 (4.0) 19.1 1.1 (3.9) 12.8 0.5 (1.5) 12.1 Community support/Counsellingc 2.9 (6.4) 51.1 1.7 (2.5) 25.5 0.8 (2.5) 15.2 Occupational therapyc 1.0 (3.3) 10.6 2.6 (10.0) 8.5 1.5 (8.4) 3.0 Physical therapyc 1.2 (7.1) 4.3 2.6 (11.1) 6.4 2.5 (9.0) 12.1 Informal cared 77.8 (216.3) 17.0 30.9 (131.7) 12.8 17.3 (19.7) 6.1 Deviant behavioure 0.3 (1.1) 12.8 0.4 (1.1) 19.1 0.2 (0.9) 9.1 Indirect Costs Absence from worka 13.0 (42.2) 23.4 7.4 (22.6) 23.4 2.2 (7.0) 18.2 Work disabilitya 69.9 (145.2) 19.1 69.9 (145.2) 19.1 77.4 (151.5) 21.2 Note. (%) relates to the percentage of participants to whom the item was applicable. a Number of days. b Number of drugs. c Number of visits/sessions. d Number of hours. e Number of incidents. Table options Treatment utilization of most services declined over time (Table 3). The overall reduction in BPD-related hospital days is most visible. In comparison to pre-treatment, hospital days were markedly reduced during DBT treatment. Further, these improvements were maintained during follow-up. Additionally, mean annual number of BPD-related hospital admissions declined from 1.1 (SD = 1.4) during pre-treatment to 0.3 (SD = 0.9) during DBT treatment and remained at similar levels (M = 0.4; SD = 1.3) during follow-up. Of the 28 patients who were treated in a hospital during pre-treatment, 10 patients participated in an inpatient DBT treatment that primarily focused on reducing dysfunctional behavior ( Bohus et al., 2004) and 38.2% of all inpatient days during that year involved treatment within such an inpatient DBT program. During follow-up, only one patient took part in an inpatient DBT-program to treat comorbid posttraumatic stress disorder (DBT-PTSD; Bohus et al., 2013). Conversely, mean annual number of individual psychotherapy sessions was considerably higher during DBT treatment compared to pre-treatment. After the end of the DBT treatment year, 32 of 33 patients who later showed up for the follow-up assessment continued individual DBT treatment, but at a lower average frequency. During follow-up, two patients terminated individual DBT and started an outpatient psychodynamic treatment. At the time of the follow-up assessment, 45.5% of the patients were still in individual DBT treatment, whereas 54.5% had terminated DBT. With respect to productivity loss, Table 3 shows that the mean annual number of BPD-related days absent from work constantly decreased from pre-treatment to DBT treatment, and from DBT treatment to follow-up for the entire sample. When calculated using only the subgroup of the eleven patients who had been employed at any time during pre-treatment, DBT treatment or follow-up, the mean annual number of days absent from work declined from 55.5 days (SD = 74.8) during pre-treatment to 31.8 days (SD = 38.8) during DBT treatment and declined again to 11.9 days (SD = 13.3) during follow-up. In contrast, all patients who were work disabled during pre-treatment remained work disabled during DBT treatment and follow-up. Here, the slight mean annual increase in disability days during follow-up was not due to more participants shifting to work disability, but to a different composition of the follow-up sample compared to the pre-treatment and DBT treatment sample. In Table 4, the number and percentage of patients on disability, unemployed, employed, student, trainee, and on parental leave at the time of inclusion in the study, at the end of DBT treatment and at the end of follow-up are presented. Table 4. Employment status at the time of study inclusion, end of the DBT-treatment year, and end of the follow-up year. Employment status, n (%) Time of inclusion in the trial (n = 47) End of DBT-treatment year (n = 47) End of follow-up year (n = 33) Student at university 2 (4.3) 2 (4.3) 1 (3.0) Student at school 2 (4.3) 1 (2.1) 0 (0.0) Trainee 4 (8.5) 4 (8.5) 4 (12.1) Employed 9 (19.1) 9 (19.1) 6 (18.2) Unemployed (on welfare), total 21 (44.7) 21 (44.7) 14 (42.4) With engagementa,b 4 (19.0) 9 (42.9) 6 (46.2) Without engagementb 17 (81.0) 12 (57.1) 7 (53.8) Work disabled 9 (19.1) 9 (19.1) 7 (21.2) Parental leave 0 1 (2.1) 1 (3.0) a Additional income/minor employment, vocational preparation, volunteer work. b Percentages relate to the proportion of the total unemployed. Table options Direct and indirect costs Table 5 presents the total mean annual BPD-related societal cost-of-illness, total mean annual BPD-related direct and indirect costs during pre-treatment, DBT treatment and follow-up. Additionally, mean annual costs for medical and non-medical services, as well as mean annual costs associated with absence from paid work and work disability are detailed. Table 5. Mean annual BPD-related total societal cost-of-illness, total direct and indirect costs Cost category Costs during pre-treatment year in € (n = 47) Costs during DBT-treatment year in € (n = 47) Costs during follow-up year in € (n = 33) M (SD) Median M (SD) Median M (SD) Median Direct medical and non-medical costs Psychiatric/general hospital, total 14167 (20899) 4134 1953 (5720) 0 1719 (4975) 0 Inpatient days 12860 (20292) 288 1953 (5720) 0 1719 (4975) 0 Day program 1307 (3375) 0 0 (0) 0 0 (0) 0 Assisted living 1758 (5743) 0 1713 (5175) 0 1190 (3702) 0 Psychotropic drugs 657 (1319) 96 485 (920) 86 330 (639) 57 Emergency room 72 (182) 0 40 (109) 0 47 (134) 0 Outpatient Psychotherapy, total 895 (1783) 0 5532 (1145) 5765 2729 (1486) 2918 Outpatient DBT 0 (0) 0 5532 (1145) 5765 2632 (1573) 2918 Outpatient Psychotherapy, other than DBT 895 (1783) 0 0 (0) 0 97 (433) 0 Psychiatrist 269 (334) 147 133 (157) 82 180 (236) 74 General practitioner 85 (120) 23 47 (98) 0 33 (113) 0 Medical specialist 51 (146) 0 40 (132) 0 18 (53) 0 Community support/Counselling 182 (394) 62 107 (322) 0 51 (162) 0 Occupational therapy 26 (85) 0 67 (255) 0 37 (213) 0 Physical therapy 15 (87) 0 27 (118) 0 25 (91) 0 Informal care 815 (2265) 0 323 (1378) 0 181 (949) 0 Deviant behaviour 46 (191) 0 57 (147) 0 9 (38) 0 Total direct Costs 19038 (25207) 5793 10524 (9321) 7100 6549 (8251) 4111 Indirect Costs Abscence from work 1570 (5383) 0 816 (2424) 0 248 (763) 0 Work disability 7418 (15485) 0 7418 (15485) 0 7953 (15566) 0 Total indirect Costs 8988 (15651) 0 8234 (15273) 0 8201 (14453) 0 Total societal Costs 28026a (33081) 12502 18758 (19450) 7228b 14750 (18592) 4594 Note. All costs are adjusted for the 2010 price level. a Total societal cost-of-illness during pre-treatment year for the n = 33 patients who showed up for follow-up-interview was M = €27696 (SD = €32692; Mdn = €12502). b Total societal cost-of-illness during DBT treatment year for the n = 33 patients who showed up for follow-up interview were M = €18803 (SD = €16632; Mdn = €7707). Table options Apparently, total mean annual societal cost-of-illness for the DBT treatment year was considerably less than that for the pre-treatment year, and further declined during follow-up. Overall, compared to pre-treatment, costs were reduced by €9268 during DBT treatment and by €13276 during follow-up. Cost savings were especially due to marked reduction in the costs of BPD-related hospital days. Already during DBT treatment, those savings exceeded the costs of the total DBT program. Of the total mean annual DBT treatment costs during DBT treatment shown in Table 5, €3367 (SD = €873) were due to individual therapy, including transport costs, €1016 (SD = €612) were due to skills training group, €1074 (SD = €365) were due to therapists' consultation team and €75 (SD = €0) were due to telephone coaching. During follow-up, the mean annual cost for individual DBT was €1807 (SD = €1242), €70 (SD = €126) for skills training group, €702 (SD = €370) for consultation team and €52 (SD = €31) for telephone coaching. In contrast, overall indirect costs hardly decreased over time. Costs due to absence from work declined from pre-treatment to DBT treatment and again from DBT treatment to follow-up. It is important to note, however, that less than one fifth of the sample was employed. When calculated within the subgroup of employed patients, mean annual cost due to work absenteeism was €6707 (SD = €9766) during pre-treatment, declined to €3488 (SD = €4096) during DBT treatment, and fell further to €1393 (SD = €1422) during follow-up. At the same time, costs due to work disability slightly increased during follow-up compared to DBT treatment and pre-treatment in the entire sample, but remained unchanged within the sample of work disabled patients. Finally, when changes in costs were tested for statistical significance, the decrease in total mean annual societal cost-of-illness reached statistical significance over the three years of our study (χ2(2) = 13.15, p = .001). Whereas total societal cost-of-illness did not significantly change from DBT treatment to pre-treatment (z = −1.41, p = .159), societal cost-of-illness was significantly reduced during follow-up relative to pre-treatment (z = −2.53, p = .011). Regarding the single cost components included in our analysis, costs of BPD-related hospital days decreased significantly over the three years of our study (χ2 (2) = 25.52, p = .000), as well as between pre-treatment and DBT treatment (z = −4.32, p = .000) and between DBT treatment and follow-up (z = −3.77, p = .000). Additionally, cost reductions related to visits to psychiatrists were statistically significant over the study period (χ2(2) = 7.76, p = .021), but when pairwise comparisons were examined, only the reduction from pre-treatment to DBT treatment was significant (z = −3.13, p = .002). Further, significant overall reductions were observed for costs associated with visits to general practitioners (χ2 (2) = 17.75, p = .000). Here, there was a significant reduction from DBT treatment to pre-treatment (z = −3.13, p = .002) and follow-up was significantly lower than pre-treatment (z = −2.53, p = .011). Ultimately, costs related to informal care were significantly reduced over the three years of the study (χ2(2) = 7.05, p = .029), but none of the pairwise comparisons reached statistical significance.