Article. An Intensive Treatment Program of Interpersonal Psychotherapy Plus Pharmacotherapy for Depressed Inpatients: Acute and Long-Term Results

Transcription

1 Article An Intensive Treatment Program of Psychotherapy Plus Pharmacotherapy for Depressed Inpatients: Acute and Long-Term Results Elisabeth Schramm, Ph.D. Dietrich van Calker, M.D., Ph.D. Petra Dykierek, Ph.D. Klaus Lieb, M.D. Sabine Kech, D.Clin.Psych. Ingo Zobel, D.Clin.Psych. Rainer Leonhart, D.Clin.Psych. Mathias Berger, M.D. Objective: The purpose of this article was to determine the relative efficacy of a psychotherapy program when combined with pharmacotherapy versus medication and clinical management in more severely depressed patients. Method: A randomized controlled trial was conducted in 124 hospitalized patients with DSM-IV major depressive disorder that compared 5 weeks of interpersonal psychotherapy modified for depressed inpatients (15 individual and eight group sessions) plus pharmacotherapy with a regimen that involved medication plus intensive clinical management. The study included a prospective, naturalistic follow-up 3 and 12 months after acute treatment in 97 of 105 treatment completers. The 17-item version of the Hamilton Depression Rating Scale (HAM- D) was the primary outcome measure. Results: For the intent-to-treat cohort (N=124), analysis of covariance (ANCOVA) showed that patients treated with interpersonal psychotherapy had a significantly greater reduction of depressive symptoms at week 5. Response rates differed significantly between the two treatment conditions, favoring the group that received adjuvant interpersonal psychotherapy (70%) versus clinical management (51%). Remission rates also tended to be higher for patients in the interpersonal psychotherapy group (49% versus 34%). Patients who initially responded to interpersonal psychotherapy exhibited greater treatment gains at the 3-month follow-up evaluation, since only 3% of these subjects relapsed, compared with 25% of the clinical management subjects. Nine months later, this difference lost statistical significance. Conclusions: An inpatient treatment program with both brief and intensive psychotherapy plus pharmacotherapy is superior to standard treatment. The results, which add to a growing body of evidence, suggest that this combination treatment may offer an advantage over treatment with medication and clinical management for more severely depressed patients. (Am J Psychiatry 2007; 164: ) There are numerous pharmacological and psychological alternatives for the treatment of unipolar major depressive disorder. However, a substantial number of depressed patients (50% 70%) either do not fully respond to acute treatment or relapse within a year (1 4). In recognition of the often poor response of depressed patients to monotherapy, the use of combined pharmacological and psychotherapeutic treatment, particularly in more severely depressed patients, is common. Nevertheless, relatively few studies have investigated the benefits of adding psychotherapy to medication in depression, and study results are conflicting (5). While in unselected outpatient cohorts only small effects of combination treatment relative to pharmacotherapy alone have been reported, combined approaches appear to be more effective for severely or chronically depressed patients (5 9). In contrast to the findings among studies of unselected depressed outpatients, two studies of depressed inpatients (10, 11) showed improved efficacy for adding cognitive and behavioral treatments to pharmacotherapy both at posttreatment (10) and long-term follow-up (12). Two German trials on inpatients, however, reported no advantage for combined treatment with medication and cognitive behavior therapy (CBT) versus medication plus supportive therapy (13) or medication and CBT versus both monotherapies (14), respectively. Despite the fact that inpatient treatment is no longer widely used in the United States and in some other countries, it is the standard of care in Germany and many other countries for more severely depressed patients and is therefore still relevant worldwide. psychotherapy has been shown to be effective in the outpatient treatment of mild to more severe depression (15, 16). To our knowledge, there are no controlled trials regarding the efficacy of This article is featured in this month s AJP Audio and is discussed in an editorial by Dr. Weissman on p ajp.psychiatryonline.org Am J Psychiatry 164:5, May 2007

2 SCHRAMM, VAN CALKER, DYKIEREK ET AL. FIGURE 1. Flow Diagram Prescreened for eligibility (N=300 approximately) Screened for eligibility (N=147) Excluded (N=17): Did not meet inclusion criteria (N=9) Refused to participate (N=8) Included in random assignment (N=130) Did not receive treatment (N=2): Comorbid diagnosis as primary (N=1) Developed delusional symptoms (N=1) Did not complete trial (N=10): Lost to posttreatment because of suicidality, delusional symptoms, prominent CT results, and refusal of evaluation (N=3) Discontinued (N=7): Suicidality (N=2) Developed delusional symptoms (N=2) Medical side effects (N=3) Allocated to interpersonal psychotherapy modified for depressed inpatients (N=65) Received treatment (N=63) Allocated to clinical management (N=65) Received treatment (N=61) Did not receive treatment (N=4): Refused after random assignment (N=1) Prematurely discharged (N=1) Acute suicidality (N=2) Did not complete trial (N=9): Lost to posttreatment because of delusional symptoms, noncompliance with questionnaires, preference for psychotherapy, and refusal of evaluation (N=3) Discontinued (N=6): Marital crisis (N=3) Need for additional medication (N=2) Developed delusional symptoms (N=1) Lost to both 3- and 12-month follow-up (N=3): Refused (N=2) Not available (N=1) Completed trial (N=53) Lost to 3-month follow-up only (N=2): Not available (N=2) Lost to 12-month follow-up only (N=2): Not available (N=1) Moved (N=1) Completed trial (N=52) Lost to both 3- and 12-month follow-up (N=5): Moved (N=2) Committed suicide (N=1) Refused (N=1) Not available (N=1) Included in analysis (N=50) Included in analysis (N=47) interpersonal psychotherapy for depressed, hospitalized patients, although several factors make the approach suitable for inpatient treatment (e.g., the medical model, the simple concept, relatively easy to learn for residents). We examined the hypotheses that depressed inpatients treated for 5 weeks with psychotherapy plus pharmacotherapy would have 1) a higher reduction in depressive symptoms and 2) higher response and remission rates compared with patients treated with pharmacotherapy plus clinical management. For the follow-up period, we hypothesized a better symptomatic and psychosocial longterm outcome and lower relapse rates for patients initially treated with psychotherapy plus pharmacotherapy. Method Patients All patients between 18 and 65 years of age referred to our department for acute psychiatric hospitalization by primary care physicians or psychiatrists between Nov and Aug with a diagnosis of major depressive disorder were prescreened during the first 3 days of admission for enrollment in this study. Indications for inpatient treatment included suicidal risk, inability to function, severe conflicts at home, and nonresponse to outpatient treatment. Of the approximately 300 prescreened patients, about 50% were ineligible, mainly because of diagnostic (e.g., comorbid borderline personality disorder or psychotic symptoms) or medication reasons (e.g., nonresponse to a previous appropriate trial of the study medication or clinical indication of the need for additional medications). Inclusion criteria consisted of a primary diagnosis of major depression (single-episode, recurrent, or bipolar II) according to the Structured Clinical Interview for DSM-IV (17) (SCID) (18) and a score of 16 on the 17-item version of the Hamilton Depression Rating Scale (HAM- D) (19). Exclusion criteria were as follows: 1) concurrent diagnosis of bipolar I disorder, primary substance abuse or dependency, other primary axis I disorders, mental disorder because of organic factors, and borderline or antisocial personality disorder; 2) psychotic symptoms; 3) severe cognitive impairment; 4) contraindications to the study medication; and 5) being actively suicidal. The study was approved by the University of Freiburg Ethical Committee. After complete description of the study to the subjects, written informed consent was obtained. In the intent-to-treat analyses, 124 patients who received allocated intervention were included, and 105 patients completed the program (Figure 1). There were no statistical differences between the dropouts and the completers on any of the demo- Am J Psychiatry 164:5, May 2007 ajp.psychiatryonline.org 769

3 PSYCHOTHERAPY PLUS PHARMACOTHERAPY FOR DEPRESSED INPATIENTS TABLE 1. Baseline Demographics and Cohort Characteristics Psychotherapy Clinical Management Analysis Treatment Group and Characteristic a Mean/N SD/% Mean/N SD/% χ 2 /t df p Intent-to-treat group (N=124) Age (years) Female Family status Single Married Separated, divorced, widowed Education (years) b Employed Number of episodes One Two or more Bipolar II Age at onset (years) Chronic (>2 years) Melancholia History of suicidal attempt Comorbidity Axis I Axis II c Disorders Traits Previous treatment d Completers (N=105) Age (years) Female Family status Single Married Separated, divorced, widowed Education (years) b Employed Number of episodes One Two or more Bipolar II Age at onset (years) Chronic (>2 years) Melancholia History of suicidal attempt Comorbidity Axis I Axis II c Disorders Traits Previous treatment d a In the intent-to-treat cohort, the number of interpersonal psychotherapy patients was 63, and the number of clinical management patients was 61. In the completer cohort, the number of interpersonal psychotherapy patients was 53, and the number of clinical management patients was 52. b Data missing for four intent-to-treat patients and three completers. c Data missing for one intent-to-treat patient. d Pharmacotherapy and/or psychotherapy. graphic or clinical variables. Follow-up analyses were restricted to patients for whom termination data were available (N=97). Design Overview Randomization. Patients were randomly assigned by a computer program (dynamic allocation using minimization method) using the following stratification variables: age, gender, unipolar disorder versus bipolar II disorder, comorbidity on axis I, duration of index episode, and number of episodes. These variables were used because they are likely to affect treatment outcome (20). The participants relevant data were entered into the computer database by the enrolling interviewer, and the allocation sequence was unpredictable for any of the investigators. Treatment Conditions. Modifications of interpersonal psychotherapy (21) for an inpatient setting (22) included 1) higher frequency of sessions; 2) eight additional interpersonal psychotherapy group sessions; and 3) inclusion of significant others in some of the sessions, mainly for the purpose of education about the disorder. Fifteen individual sessions (all videotaped) of approximately 50 minutes were administered three times weekly over 5 weeks. For the intent-to-treat cohort (N=124), the average number of interpersonal psychotherapy sessions was 12.8 (SD=2.75; range: 4 15). The treatment completers (N=105) had a mean of 13.5 (SD=1.79; range: 8 15) sessions. Four patients were discharged before the end of the program because of recovery and therefore had less than 15 sessions. 770 ajp.psychiatryonline.org Am J Psychiatry 164:5, May 2007

4 SCHRAMM, VAN CALKER, DYKIEREK ET AL. TABLE 2. Post-Treatment Scores for Psychotherapy Modified for Depressed Inpatients and Clinical Management Groups Baseline Week 5 Absolute Difference a Analysis Effect Therapy Group and Measure Mean SD Mean SD Mean SD F df p Size b Intent-to-treat group (N=124) HAM-D therapy , Clinical management Beck Depression Inventory therapy Clinical management Global Assessment of Functioning therapy , Clinical management Completers (N=105) HAM-D therapy , Clinical management Beck Depression Inventory therapy , Clinical management Global Assessment of Functioning therapy , Clinical management a Absolute difference between baseline and week 5 (pre-and posttreatment). b Effect size calculated using Cohen s d. All interpersonal psychotherapy therapists (five psychiatric residents and five clinical psychologists) were in an advanced or completed stage of a 3-year psychotherapy training program, had attended training in interpersonal psychotherapy, and had received weekly group supervision. The patients in the clinical management condition also received three weekly sessions according to a brief guideline (adapted from Guideline for Medication Clinic, by C.F. Reynolds, III and J.M. Perel). Clinical management was defined as a psychoeducative, supportive, and empathic intervention of 20 to 25 minutes and was delivered by psychiatric residents. All residents (N=26) had a didactic training in clinical management for at least one-half day. Audiotaped clinical management sessions were continually checked for adherence using the Therapist Rating Scale (23). Both treatment groups received standardized pharmacotherapy (described in a data supplement that accompanies the online version of this article) in addition to interpersonal psychotherapy or clinical management by the treating physician who was not blind to the patients treatment status. The first-line treatment was sertraline (intent-totreat cohort: N=81; completers: N=68). In case of a known nonresponse to sertraline in a patient s treatment history, amitriptyline or amitriptyline-n-oxide was applied as the second line treatment (intent-to-treat cohort: N=35; completers: N=30). Eight patients (six clinical management and two interpersonal psychotherapy patients) who started treatment with sertraline received additional amitriptyline or amitriptyline-n-oxide after the second week because of an unsatisfactory response to sertraline. In the intent-to-treat cohort, the mean final dosage for sertraline was 90.2 (SD=43.9) mg/day (range: mg/day), and among completers it was (SD=41.9) mg/day (range: mg/ day). For amitriptyline or amitriptyline-n-oxide, the mean final dosage in the intent-to-treat cohort was (SD=66.9) mg/day (range: mg/day), and among completers it was (SD=69.7) mg/day (range: mg/day). The number of patients using sertraline or amitriptyline in both treatment conditions was similar (in both the intent-to-treat and completer cohorts). Furthermore, there was no significant difference in the mean final dosage between interpersonal psychotherapy and clinical management patients for both types of medication in both cohorts. All patients participated in usual hospital activities such as daily ward rounds, occupational therapy groups, and physiotherapy. Assessments The 17-item version of the HAM-D (19) served as the primary outcome measure. Secondary measures included the Beck Depression Inventory (24), the Global Assessment of Functioning Scale (25), and the Social Adjustment Scale Self-Report (26). In the follow-up period, weekly psychiatric status ratings, based on DSM-IV symptoms of depression, were recorded retrospectively by the patient. Criteria for the psychiatric status ratings (according to the Longitudinal Interview Follow-up Evaluation [27]) format ranged from 1=no symptoms to 6=meets full criteria of major depression. Treatment response was defined a priori as a reduction in symptom severity of 50% or higher on the HAM-D and remission as a score of 7 or less on the HAM-D. Following initial posttreatment response, patients were considered to have relapsed if they had a HAM-D score of 15 in combination with a psychiatric status ratings score of 5 for a minimum of 2 weeks. Sustained response was defined as responding to acute inpatient treatment and maintaining the response as well as staying free from relapse and not being rehospitalized during the course of follow-up. Sustained remission was defined as a score of 7 or less on the HAM- D plus a psychiatric status ratings score of 1 or 2 for at least 2 weeks following initial remission with no relapse and no rehospitalization during the follow-up period. The SCID (18) was conducted by trained and experienced clinical psychologists. The efficacy of treatment was evaluated 5 weeks after the beginning of treatment and 3 and 12 months posthospitalization in terms of a naturalistic follow-up. The evaluations were performed by blind and independent raters (clinical psychologists) who were not involved in the inpatient care and not located in the hospital. The patients were instructed not to mention anything that might reveal their treatment group to the evaluator. Interrater reliability for the HAM-D was obtained from all evaluators rating a random cohort of 13 audio- or videotaped HAM-D interviews (intraclass correlation coefficient: 0.94). Am J Psychiatry 164:5, May 2007 ajp.psychiatryonline.org 771

5 PSYCHOTHERAPY PLUS PHARMACOTHERAPY FOR DEPRESSED INPATIENTS FIGURE 2. Response, Remission, and Relapse Rates in the Completer Cohort Using HAM-D 80 Response a Remission a Relapse Percent Posttreatment Month 3 Month Psychotherapy Clinical Management Psychotherapy Clinical Management Psychotherapy Clinical Management a For follow-up data sustained response/sustained remission. TABLE 3. Follow-Up Scores for Psychotherapy Modified for Depressed Inpatients and Clinical Management Groups Measure and Psychotherapy Clinical Management Absolute Difference a Analysis Effect Follow-Up Mean SD Mean SD Mean SE t df p Size b HAM-D 3 months months Beck Depression Inventory 3 months months Global Assessment of Functioning 3 months months a Absolute difference between groups. b Effect size calculated using Cohen s d. Statistical Methods Cohort size was based on a power analysis for repeated-measures analysis of variance (ANOVA) with an alpha set at 0.05 and an assumed moderate effect size (HAM-D: f 2 =0.15). A cohort of 53 participants per treatment would have yielded an estimated power of 95%. The cohort size for a two-group comparison (one sided) should be 51 patients per treatment to detect an effect size of d=0.50 at a power of 80% and alpha <0.05. Demographic and clinical characteristics of treatment groups were compared using chi square or Fisher s exact tests and unpaired t tests. Posttreatment analysis: Analyses for both the intent-to-treat and the completer cohorts are reported. Overall efficacy of treatment was assessed by conducting analysis of covariance (AN- COVA) controlling for pretreatment scores on all outcome measures by treatment condition at termination. The effects of time were analyzed in a repeated-measures ANOVA. When attrition occurred between baseline and posttreatment, last observation carried forward was used. Follow-up analysis: Completer analyses of the HAM-D and the Beck Depression Inventory at 3 and 12 months were conducted to examine differences between the treatment groups. For the follow-up analysis, we adopted a multiple imputation approach (28 30). Response, remission, and relapse data for the percentage of patients who reached cutoff scores were analyzed using Fisher s exact test. Treatment effect sizes were calculated using Cohen s d statistic (31) and for categorical measures using procedures described by Rosenthal (32) for better comparability with previous studies. All statistical tests were conducted using a twotailed alpha level of Results Baseline Patient Characteristics Most of the patients (95%) were evaluated as markedly and severely ill by the treating physician using Clinical Global Impression ratings (33) as a part of clinical routine. In addition, the majority of the patients were characterized by features of severe depressive disorder (Table 1), such as initial HAM-D scores of 20 or more (mean HAM-D score: 23.6 [SD=4.8]) as employed by Elkin et al. (15), scores of 49 and below on the Global Assessment of Functioning Scale (25) (mean Global Assessment of Functioning Scale score: 46.4 [SD=8.2]), and melancholic subtypes (64.5%) as suggested by Thase (34). Almost one-third of the patients (N=65) suffered from marked suicidal thoughts using a HAM-D-item score of 3 or higher. The majority (83%) had received outpatient treatment before admission, and 44% of the pa- 772 ajp.psychiatryonline.org Am J Psychiatry 164:5, May 2007

6 SCHRAMM, VAN CALKER, DYKIEREK ET AL. tients had at least one previous hospitalization for the treatment of depression. Despite the controversy about patients with bipolar II disorder representing a distinct group, we included them in all analyses, since we found no significant differences between unipolar (N=119) and bipolar II (N=5) currently depressed patients in terms of demographic- and disorder-related variables and in terms of pre- and posttreatment as well as follow-up HAM-D and Beck Depression Inventory scores. The treatment conditions of both the intent-to-treat and the completer cohorts did not differ in background demographic and clinical characteristics (Table 1), except for previous treatments that were more common in the interpersonal psychotherapy group of the completer cohort. Secondary axis I diagnoses, according to the SCID I, among the intent-to-treat patients at baseline were as follows: 23 diagnoses of dysthymia, 14 of substance use (not primary), 10 of anxiety disorder, five of somatoform, five of attention deficit hyperactivity disorder, and two of eating disorders. TABLE 4. Post-Hospital Treatment and Rehospitalization Rates in the Naturalistic Follow-Up Period a Psychotherapy Clinical Management Treatment and Rehospitalization N % N % p Pharmacotherapy 3 months months Psychotherapy 3 months months Rehospitalization 3 months months a Of the 97 patients evaluated at 3 months, 50 had initially been given interpersonal psychotherapy, and 47 had initially received clinical management. Of the 95 patients evaluated at 12 months, 48 had initially been given interpersonal psychotherapy, and 47 had initially received clinical management. Acute Phase Outcomes General acute treatment effects. At baseline, there was a significant group difference in depressive symptoms on the HAM-D (intent-to-treat: t=3.74, df=122, p<0.001; completers: t= 3.47, df=103, p=0.001). Over time both treatment conditions showed significant improvements (from pre- to posttreatment) for clinician-rated (intent-totreat: F=343.27, df=1, 122, p<0.001; effect size : interpersonal psychotherapy, d=3.17, clinical management, d= 2.53) and self-rated depression (intent-to-treat: F=246.30, df=1, 122, p<0.001; effect size : interpersonal psychotherapy, d=1.91, clinical management, d=1.46) after 5 weeks. On the HAM-D (but not on the Beck Depression Inventory), we found a statistically highly significant interaction between treatment and time (intent-to-treat: F=17.06, df= 1, 122, p<0.001) in favor of interpersonal psychotherapy. ANCOVA showed a significant advantage for the interpersonal psychotherapy group on the HAM-D in the intentto-treat cohort (interpersonal psychotherapy, 8.9 [SD= 6.4], versus clinical management, 11.8 [SD=7.9]; p=0.009), demonstrating a moderate effect size (d=0.62) of the interpersonal therapy condition relative to the comparison group (Table 2). Response and remission rates. Response rates in the intent-to-treat and the completer cohort based on the HAM-D differed significantly between both groups after 5 weeks, favoring therapy with interpersonal psychotherapy (intent-to-treat: 70% versus 51%; p=0.043; completers: 77% versus 58%; p=0.038). For the intent-to-treat cohort, these response rates equal a number needed to treat of 6 (95% confidence interval [CI]=3 21). The difference in remission rates did not reach statistical significance in the intent-to-treat cohort (interpersonal psychotherapy: 49%; clinical management: 34%; p=0.105), but there was a tendency in the completer cohort for interpersonal psychotherapy patients to have higher rates of remission (interpersonal psychotherapy: 58%; clinical management: 40%; p=0.079) (Figure 2). Global functioning. Both treatment groups showed an improvement in overall functioning on the Global Assessment of Functioning Scale at week 5 (intent-to-treat: F= , df=1, 122, p<0.001; completers: F=232.00, df=1, 101, p<0.001). The ANCOVA results for the main treatment effect significantly favored therapy with interpersonal psychotherapy (intent-to-treat: F=4.25, df=1, 121, p=0.041; completers: F=4.07, df=1, 100, p=0.046) (Table 2). The Social Adjustment Scale Self-Report was only completed at the follow-ups. Follow-Up Phase Data by means of the HAM-D were collected on 97 (92.4%) of the 105 completers (Figure 1), on 93 (89%) patients who participated in both the 3- and 12-month follow-ups and on 95 (90%) patients who participated at either 3 or 12 months. There were no statistically significant differences on any demographic or clinical variables between those patients who participated in the follow-up evaluations and those who did not. One patient who was initially treated with standard therapy committed suicide 10 days after discharge from the hospital. No other suicide attempts were reported during the entire follow-up period. Ninety-eight percent of all evaluated patients (95 of 97) received further treatment after discharge from the clinic. We found no significant differences between both treatment groups regarding the use of posthospital pharmacotherapy or psychotherapy. Only 17% (16 of 95) of all patients had discontinued any form of treatment at 12 months. General treatment effects. A significantly greater reduction in HAM-D scores was found at the 3- and 12- month follow-up assessments for the patients initially treated with interpersonal psychotherapy compared with those treated with clinical management (3 months: p= 0.016; 12 months: p=0.008) (Table 3). At this time, the dif- Am J Psychiatry 164:5, May 2007 ajp.psychiatryonline.org 773

7 PSYCHOTHERAPY PLUS PHARMACOTHERAPY FOR DEPRESSED INPATIENTS Patient Perspectives Fred, a 58-year-old veterinarian experiencing his second depressed episode, was referred by his psychiatrist for inpatient treatment because of suicidal ideas. After suffering a disc prolapse, he retired early from his job and fell into a hole. It was as if my high speed train stopped abruptly in the middle of nowhere. I did not know what to do with myself, and I didn t want to be a burden for others. Symptoms management was part of the initial phase in the interpersonal psychotherapy modified for depressed inpatients program, focusing on suicidal thoughts and rumination. Role transitions were negotiated as the main problem area associated with the patient s current depressed episode. Setting positive and negative aspects of the old role in perspective, the patient realized that he idealized what he had lost. The reality was that Fred had been overwhelmed by his job requirements for a long time. I guess I lost some things (e.g., recognition, physical fitness), and it makes me sad. On the other hand, if I am honest with myself I do feel relieved. In the interpersonal psychotherapy group sessions, strategies were developed to adapt to this life change, such as planning his retirement together with his wife and exploring possible activities related to his former profession (e.g., consulting colleagues). In one of the individual sessions where his wife participated, she recognized that her life had also changed as a consequence of his retirement. We learned to communicate more and in a different way than before, which honored the new phase of our partnership. One year after Fred was discharged in a remitted state, he remained symptom free. The therapy helped me to get back in the driver s seat. I know now what I want to do during this phase of my life, and I began to enjoy the journey. Lisa, a 34-year-old mother of two boys, was referred to the hospital after failing to respond to various antidepressant medications and feeling no longer capable of caring for her children. The first depressive symptoms began after her husband was promoted within his company, which included moving to a small town. First, I felt furious, but I kept it to myself, then I felt helpless and guilty. Besides having to give up relationships that were important to her and the children, she resented her husband for not giving her a chance to make her own decision about moving. He is very dominant and would never consider anything else more important than his own career. Both therapist and patient agreed on interpersonal role disputes as the main problem area associated with the onset of the depression. To get beyond the impasse, the husband was invited to participate in a few sessions. The imbalance of dominance and submission was identified as the main source of the dispute. In one of the interpersonal psychotherapy group sessions, Lisa practiced to communicate her feelings: I feel ignored by you and that makes me mad. I did not say anything because I was too scared to get you upset. From now on I want you to negotiate with me decisions that affect me or our family. Her husband was surprised by her lucid words and accepted her request. Following her hospital discharge with minimal symptoms, she continued psychotherapeutic treatment to keep [herself] on track so that [she didn t] fall back into the old patterns of avoidance. One year after her discharge, Lisa reported, there are still ups and downs but I feel I acquired more room for my needs in the marriage, and we are dealing more with each other. ference between both groups also became statistically significant on the Beck Depression Inventory (p=0.032). Sensitivity analyses using the completer data set did not reveal any differences compared with the results of statistical tests using multiply imputed data sets. Relapse rates and rehospitalizations. As presented in Figure 2, only one of 39 (2.6%) patients who responded to acute combined treatment with interpersonal psychotherapy plus pharmacotherapy relapsed during the first 3 months of the follow-up period compared with seven of 28 (25%) clinical management patients (number needed to treat=5, odds ratio [OR]=0.08, 95% CI=3 18; p=0.008). For those who initially remitted with interpersonal psychotherapy, the relapse rate was 3.4% (1 of 29 patients) compared with 26% (five of 19 patients) in the clinical management group (number needed to treat=5, OR=0.10, 95% CI=3 50; p=0.029). Within the 12-month follow-up phase, five (13%) responders to interpersonal psychotherapy and eight (29%) responders to standard treatment experienced a relapse (number needed to treat=7, OR=0.37, 95% CI=3 23; p=0.209). For initial remitters, two (7%) of the interpersonal psychotherapy patients and six (32%) of the clinical management patients relapsed 1 year after discharge (number needed to treat=5, OR=0.16, 95% CI=3 69; p=0.049). In the first 3 months, there was a higher tendency for patients initially given standard treatment to be rehospitalized (13%) compared with those who received prior combination treatment with interpersonal psychotherapy plus pharmacotherapy (2%) (one interpersonal psychotherapy patient versus six clinical management patients; p=0.054) (Table 4). Sustained response and remission. Three months after discharge, 73% (35 of 48) of the patients initially responsive to interpersonal psychotherapy showed a sustained response compared with 47% (22 of 47) of the patients who responded to clinical management (p= 0.012) (Figure 2). At 12 months, the percentage of patients with sustained response was still significantly higher (p= 0.002) in the interpersonal psychotherapy condition (33 of 48 [69%] patients) than in the clinical management condition (17 of 47 [36%] patients). Rates of sustained remission did not show statistically significant differences between both treatment groups at 3 (interpersonal psy- 774 ajp.psychiatryonline.org Am J Psychiatry 164:5, May 2007

8 SCHRAMM, VAN CALKER, DYKIEREK ET AL. chotherapy: 19 of 52 [36%] patients; clinical management: 12 of 50 [24%] patients) and 12 months (interpersonal psychotherapy: 18 of 52 [35%] patients; clinical management: 10 of 50 [20%] patients). Global and social functioning. Compared with the clinical management condition, interpersonal psychotherapy given in the acute phase of treatment was found to be significantly more effective in increasing social functioning at both follow-up assessments by means of the Global Assessment of Functioning Scale (3 months: p= 0.034; 12 months: p=0.026) (Table 3). On the Social Adjustment Scale-Self-Report, the t tests show a significant difference in overall social adjustment 1 year after acute treatment between the two prior treatment conditions (interpersonal psychotherapy: 1.87 [SD=0.49]; clinical management: 2.11 [SD=0.59]; t=2.01, df=85, p=0.048). Discussion Given the short time period of acute treatment in our study and a strong comparison condition, the short- and long-term advantage of adding psychotherapeutic intervention to pharmacotherapy is impressive. The high response (70%) and remission rates (49%) and the low relapse rates (13% in a 12-month period) may reflect the high intensity of the multicomponent treatment program with 15 individual and eight additional group sessions. The intensive delivery of therapy in the acute phase possibly conveys a unique treatment advantage for more severely depressed patients. However, the brevity of the intervention may have imposed a ceiling on the maximum efficacy of the treatment. Miller et al. (35) concluded that current treatments are not very efficacious in the aftercare of recently discharged major depressive disorder patients. They suggest that beginning psychotherapy while in the hospital may be more advantageous than delaying initiation until discharge. In fact, the timing of additional psychotherapy initiation seems to play an important role. Frank et al. (36) also suggest not waiting for the period of disorder stabilization to add psychotherapy. Our data clearly support this view. Compared with previous studies with depressed inpatients (10 14), this study differed in having utilized a larger cohort size, a more comprehensive treatment program, and a different psychotherapy method that may have affected treatment response more favorably. Several factors may account for the discrepancy between clinician and patient ratings. First, the two measures show divergent factor structures (37). Furthermore, Lambert et al. (38) demonstrated larger effect sizes and a higher sensitivity for the HAM-D compared with the Beck Depression Inventory. An additional possibility addresses the familiar phenomenon that more severely depressed patients require a longer course of treatment to recognize their progress because of a negative thinking bias (39). This study has several limitations. First, the entire cohort was heterogeneous in terms of subgroups of depression being chronic, recurrent, or comorbid. In terms of generalizability and external validity, however, this could be considered a strength of the study, since the population of depressed, hospitalized patients is usually comorbid, more severely impaired, and disturbed in a complex way. Second, the terminology of remission was inconsistent with other work (40), but this was because of the short treatment period defined on the basis of a single assessment covering only 1 week. In the follow-up phase, there were only two assessment points using the HAM-D. On the other hand, weekly psychiatric status ratings were recorded by the patient to identify change over time in depressive symptoms. Third, since both treatment groups did not receive comparable amounts of therapeutic attention, it is unclear whether the effect is attributable to interpersonal psychotherapy modified for depressed inpatients per se as opposed to extra time with a therapist. Nevertheless, we find the amount of time spent with a health professional as a main efficacy factor unlikely, since both treatment groups reached ceiling effects in terms of the amount of therapeutic contact because of the inpatient setting. Finally, these findings were obtained for inpatient treatment and may not be directly transferable to other medical systems in other countries. However, the current cohort appears to be comparable with those seen in acute psychiatric outpatient settings in the United States. In summary, our results, although having some limitations, support a growing body of evidence (4 6, 8) suggesting that while for the average patient the additional benefits of combined treatment with interpersonal psychotherapy plus pharmacotherapy may be limited, this combination treatment offers a strong advantage for more severe and complex patients. Our data suggest that treating depression aggressively during the acute phase of treatment is recommended. Future studies should examine whether it is more beneficial to add psychotherapy while in the hospital rather than beginning combination interpersonal psychotherapy plus pharmacotherapy treatment after hospital discharge. It remains uncertain whether an equally frequent level of treatment provision on an outpatient basis could be similarly effective. Presented in part at the First International Conference on Psychotherapy, Pittsburgh, June 2004, and the Third Conference of the International Society for Affective Disorders, Lisbon, March Received Oct. 12, 2006; revision received Dec. 9, 2006; accepted Jan. 8, From the Departments of Psychiatry, Psychotherapy, and Psychology, University of Freiburg, Freiburg, Germany. Address correspondence and reprint requests to Dr. Schramm, University Medical Center Freiburg, Department of Psychiatry and Psychotherapy, Hauptstrasse 5, Freiburg, Germany; Elisabeth.Schramm@uniklinik-freiburg.de ( ). Drs. Schramm, Lieb, Dykierek, Kech, Zobel, Leonhart, and Berger report no competing interests. Dr. van Calker has received honoraria Am J Psychiatry 164:5, May 2007 ajp.psychiatryonline.org 775

9 PSYCHOTHERAPY PLUS PHARMACOTHERAPY FOR DEPRESSED INPATIENTS for lecturing from AstraZeneca, Pfizer, Eli Lilly, Merz, Sanofi, Organon, Neuraxpharm, Wyeth, and Bristol-Myers Squibb. Funded by grants SCHR443/4-1 and 4-2 from the German Research Society, Bonn, Germany. The authors thank Charles F. Reynolds, III, M.D., Scott Stuart, M.D., and Christiane Kühner, Ph.D., for their comments on an earlier version of this article. References 1. Westen D, Morrison K: A multidimensional meta-analysis of treatments for depression, panic, and generalized anxiety disorder: an empirical examination of the status of empirically supported therapies. J Consult Clin Psychol 2001; 69: Rush AJ, Trivedi MH, Wisniewski SR, Stewart JW, Nierenberg AA, Thase ME, Ritz L, Biggs MM, Warden D, Luther JF, Shores-Wilson K, Niederehe G, Fava M: Bupropion-SR, sertraline, or venlafaxine-xr after failure of SSRIs for depression. N Engl J Med 2006; 354: Khan A, Warner HA, Brown WA: Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: an analysis of the Food and Drug Administration database. Arch Gen Psychiatry 2000; 57: Hollon SD, Jarrett RB, Nierenberg AA, Thase ME, Trivedi M, Rush AJ: Psychotherapy and medication in the treatment of adult and geriatric depression: which monotherapy or combined treatment? J Clin Psychiatry 2005; 66: Friedman MA, Detweiler-Bedell JB, Leventhal HE: Combined psychotherapy and pharmacotherapy for the treatment of major depressive disorder. Clin Psychol Sci Prac 2004; 11: Thase ME, Greenhouse JB, Frank E, Reynolds CF III, Pilkonis PA, Hurley K, Grochocinski V, Kupfer DJ: Treatment of major depression with psychotherapy or psychotherapy pharmacotherapy combinations. Arch Gen Psychiatry 1997; 54: Jindal RD, Thase ME: Integrating psychotherapy and pharmacotherapy to improve outcomes among patients with mood disorders. Psychiatr Serv 2003; 54: Otto MW, Smits JAJ, Reese HE: Combined psychotherapy and pharmacotherapy for mood and anxiety disorders in adults: review and analysis. Clin Psychol Sci Prac 2005; 12: Keller MB, McCullough JP, Klein DN, Arnow B, Dunner DL, Gelenberg AJ, Markowitz JC, Nemeroff CB, Russell JM, Thase ME, Trivedi MH, Zajecka J: A comparison of nefazodone, the cognitive behavioral-analysis system of psychotherapy, and their combination for the treatment of chronic depression. N Engl J Med 2000; 342: Miller IW, Norman WH, Keitner GI, Bishop S, Dow M: Cognitivebehavioral treatment of depressed inpatients. Behav Ther 1989; 20: Miller IW, Norman WH, Keitner GI: Cognitive-behavioral treatment of depressed inpatients: six- and twelve-month followup. Am J Psychiatry 1989; 146: Bowers WA: Treatment of depressed inpatients: cognitive therapy plus medication, relaxation plus medication, and medication alone. Br J Psychiatry 1990; 156: de Jong-Meyer R, Hautzinger M, Rudolf GAE, Straub W, Frick U: The effectiveness of antidepressants and cognitive-behavioral therapy combined in patients with endogenous depression: results of analyses of variance regarding main and secondary outcome criteria. Zeitschr Klin Psychol 1996; 25: (German) 14. Hautzinger M, de Jong-Meyer R, Treiber R, Rudolf GAE, Thien U: Efficacy of cognitive behavior therapy, pharmacotherapy, and the combination of both in non-melancholic, unipolar depression. Zeitschr Klin Psychol 1996; 25: (German) 15. Elkin I, Shea T, Watkins JT, Imber ST, Sotsky SM, Collins JF, Glass DR, Pilkonis PA, Leber WR, Docherty JP, Fiester SJ, Parloff MB: National Institute of Mental Health Treatment of Depression Collaborative Research Program: general effectiveness of treatments. Arch Gen Psychiatry 1989; 46: Hollon SD, Thase ME, Markowitz JC: Treatment and prevention of depression. Psychol Science 2002; 3: American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC, American Psychiatric Press, First M, Spitzer R, Gibbon M, Williams J: Structured Clinical Interview for DSM-IV Axis I Disorders, Patient Edition (SCID-P), version 2. New York, New York State Psychiatric Institute, Biometrics Research Department, Hamilton MA: A rating scale for depression. J Neurol Neurosurg Psychiatry 1960; 23: Boland RJ, Keller MB: Course and outcome of depression, in Handbook of Depression. Edited by Gotlib IH, Hammen CL. New York, Guilford, Klerman GL, Weissman M, Rounsaville BJ, Chevron ES: Psychotherapy of Depression. New York, Basic Books, Schramm E: psychotherapy for outpatient and inpatient treatment of depression. Psychotherapie im Dialog 2001; 4: (German) 23. Wagner E, Frank E, Steiner SC: Discriminating maintenance treatments for recurrent depression: development and implementation of a rating scale. J Psychother Prac Res 1992; 1: Beck AT, Ward Ch, Mendelson M: An inventory for measuring depression. Arch Gen Psychiatry 1961; 4: Endicott J, Spitzer RL, Fleiss JL, Cohen J: The Global Assessment Scale: a procedure for measuring overall severity of psychiatric disturbance. Arch Gen Psychiatry 1976; 33: Weissman M, Bothwell P: Assessment of social adjustment by patient self-report. Arch Gen Psychiatry 1976; 33: Keller MB, Lavori PW, Friedman B, Nielsen E, Endicott J, Mc- Donald-Scott P, Andreasen NC: The Longitudinal Interval Follow-Up Evaluation: a comprehensive method for assessing outcome in prospective longitudinal studies. Arch Gen Psychiatry 1987; 44: King G, Honaker J, Joseph A, Scheve K: Analyzing incomplete political science data: an alternative algorithm for multiple imputation. Am Political Science Rev 2001; 95: Schafer JL: NORM: Multiple imputation of incomplete multivariate data under a normal model, version 2. Software for Windows 95/98/NT, 1999, Rubin DB: Multiple Imputation for Nonresponse in Surveys. New York, John Wiley and Sons, Olejnik S, Algina J: Measures of effect size for comparative studies: applications, interpretations, and limitations. Contemp Edu Psychol 2000; 25: Rosenthal R: Meta-Analytic Procedures for Social Research. London, Sage Publications, National Institutes of Mental Health: Clinical Global Impression Scale. Psychopharmacol Bull 1985; 21: Thase ME: Treatment of severe depression. J Clin Psychiatry 2000; 61(suppl 1): Miller IW, Keitner GI, Ryan CE, Solomon DA, Cardemil EV, Beevers CG: Treatment matching in the posthospital care of depressed patients. Am J Psychiatry 2005; 162: Frank E, Kupfer DJ, Thase ME, Mallinger AG, Swartz HA, Fagiolini AM, Grochocinski V, Houck P, Scott J, Thompson W, Monk T: Two-year outcomes for interpersonal and social rhythm therapy in individuals with bipolar I disorder. Arch Gen Psychiatry 2005; 62: ajp.psychiatryonline.org Am J Psychiatry 164:5, May 2007

10 SCHRAMM, VAN CALKER, DYKIEREK ET AL. 37. Steer RA, Beck AT, Riskind JH: Relationships between the Beck Depression Inventory and the Hamilton Psychiatric Rating Scale for Depression in depressed outpatients. J Psychopathol Behav Assessment 1987; 9: Lambert MJ, Hatch DR, Kingston MD, Edwards BC: Zung, Beck, and Hamilton Rating Scales as measures of treatment outcome: a meta-analytic comparison. J Consult Clin Psychol 1986; 54: Rush AJ, Hiser W, Giles DE: A comparison of self-reported versus clinician-rated symptoms in depression. J Clin Psychiatry 1987; 48: Frank E, Prien RF, Jarrett RB, Keller MB, Kupfer DJ, Lavori PW, Rush AJ, Weissman MM: Conceptualization and rationale for consensus definitions of terms in major depressive disorder: remission, recovery, relapse, and recurrence. Arch Gen Psychiatry 1991; 48: Am J Psychiatry 164:5, May 2007 ajp.psychiatryonline.org 777