This investigator-initiated trial is registered as: "A comparison of the Cognitive Behavioural Analysis System of Psychotherapy against supportive psychotherapy for early onset chronic depression" at ClinicalTrials.gov (NCT00970437). Figure illustrates the study design of this multisite, observer blind, prospective, parallel-group, randomized, controlled trial with an active control and two treatment phases (acute and continuation). A stratified block randomization with randomly varying block size is performed, stratified by trial center.
Trial design. CBASP: Cognitive Behavioral Analysis System of Psychotherapy. SP: Supportive Psychotherapy. Wk: week.
The intervention includes 20 weeks of acute treatment with 24 individual sessions followed by 28 weeks of continuation treatment with another 8 sessions in each arm of the trial.
Oversight of research with human participants
The study is being conducted in compliance with the protocol, Good Clinical Practice and the applicable regulatory requirements. This research was approved by the Institutional Review Board/Institutional Ethical Committee (IRB/IEC) of the University of Freiburg and the ongoing trial is under continued review by the IRB/IEC. The local IRB/IECs of each participating site were asked for confirmation prior to initiation. Written informed consent is provided by all participants prior to clinical interview, randomization and intervention.
Adverse Events and Serious Adverse Event (AE/SAE) are reported to the project's Data Safety and Monitoring Board (DSMB) and SAEs to each IRB/IEC. Criteria for discontinuation include for individuals: a) active suicidality; b) the physical health of the patient is at risk according to clinical judgment; c) occurrence of an AE/SAE with therapeutic implications incompatible with the study; d) newly occurring exclusion criteria, or e) the informed consent is withdrawn. Parts of the trial or the entire trial will be discontinued if: a) an investigator has serious ethical concerns because of the performance at one of the sites. b) severe safety concerns become apparent to the DSMB.
In addition, the DSMB will conduct regular phone conferences and visits at trial sites to ensure compliance with ethical principles and the study protocol, as well as to check data quality and accuracy.
Data collection sites
The patients are recruited and treated at eight clinical sites in Germany:
1) Department of Clinical and Developmental Psychology, University of Tuebingen, (Site Principal Investigator: Martin Hautzinger, PhD);
2) Department of Psychiatry, University of Heidelberg (Site Principal Investigator: Matthias Backenstraß, PhD);
3) Central Institute of Mental Health in Mannheim (Site Principal Investigator: Josef Bailer, PhD);
4) Psychological Outpatient Clinic, University of Marburg, (Site Principal Investigator: Katrin Wambach, PhD);
5) Department of Psychiatry and Psychotherapy, University of Luebeck, Germany (Site Principal Investigator: Philipp Klein, MD);
6) Department of Psychiatry; University Medical Center Bonn (Site Principal Investigator: Dieter Schoepf, MD);
7) Department of Psychosomatic Medicine and Psychotherapy, University Medical Center Hamburg-Eppendorf and Clinic Center Eilbek (Site Principal Investigator: Bernd Löwe, MD)
8) Department of Psychiatry and Psychotherapy, University Medical Center Freiburg (Site Principal Investigator: Elisabeth Schramm, PhD).
Recruitment methods include project promotion through reports in the media and by letters/announcement to private practitioners. Most patients are expected to be referred from practitioners. Patients have to be unmedicated for at least 2 weeks prior to study entry. In addition, the treating or a collaborating physician has to attest that there is no contraindication for psychotherapeutic treatment. When a review of medical history necessitates doing so, a physical examination and appropriate laboratory tests are obtained to ensure that patients are diagnostically eligible. Patient recruitment started in March 2010.
All Site Principal Investigators will continue to meet regularly by telephone (at least monthly or if necessary more frequently) and at least twice annually in person until study completion. Each research procedure is described in detail in a procedure manual to facilitate the implementation of the procedures in a consistent manner across sites.
Inclusion and exclusion criteria
Key inclusion criteria are:
- a DSM-IV diagnosis of: 1) chronic Major Depressive Disorder/MDD (at least 2 years duration), or 2) current MDD superimposed on a pre-existing dysthymic disorder (so called "double-depression"), or 3) recurrent MDD with incomplete remission between episodes. Further inclusion criteria are:
- early onset (before the age of 21) according to DSM-IV,
- age between 18 and 65, and
- a score of at least 20 on the 24-item Hamilton Rating Scale of Depression (HRSD).
Key exclusion criteria include:
- acute risk for suicide (as opposed to suicidal thoughts) assessed according to clinical practice guidelines; suicidal patients are eligible, as long as outpatient treatment is deemed safe by the clinician,
- a history of psychotic symptoms, bipolar disorder, or organic brain disorders,
- a primary diagnosis of another axis I disorder,
- antisocial, schizotypical, or borderline personality disorder (SCID-II),
- severe cognitive impairment,
- a serious medical condition,
- absence of a response to a previous adequate trial of CBASP and/or SP, and
- other ongoing psychotherapy or medication.
CBASP as the experimental intervention follows a manual [19
] and is a highly structured approach based on an interpersonal contemporary learning acquisition model [21
]. It is the only psychotherapy model developed exclusively for the treatment of chronic forms of early-onset depression. The founder of the approach, James McCullough, suggests that chronically depressed patients suffer from a reduction of perceived functionality, i.e. the ability to detect the effects of their own behavior on other persons. This results in a pervasive degree of social isolation, which worsens the depressive mood. In addition, in chronic depressives early interpersonal traumatization resulted in an inhibition of maturation in childhood. CBASP focuses on the patient's central mechanisms of derailed affective and motivational regulation by using the therapeutic relationship in a personal, disciplined way to shape dysfunctional interpersonal behaviour. Further specific techniques (e.g. Interpersonal Discrimination Exercise, Situation Analysis) are applied to aid acquisition of perceived functionality in the patient. In summary, CBASP integrates behavioral, cognitive, and interpersonal strategies to help the patient recognizing the consequences of their behavior and interacting more effectively with others. Other goals of CBASP include the transferal of social problem solving strategies to the daily living, the interpersonal healing of earlier trauma and to generate authentic empathy.
The alternative treatment, supportive psychotherapy, is an active but less specific, manualized [22
] control intervention previously used in several comparative trials [23
]. SP - defined as non-interpersonal and non-cognitive-behavioral therapy - resembles supportive clinical management or client-centered counseling, and includes psychoeducational elements. SP is also defined as a psychotherapy wherein the therapist strives to create a supportive relationship by emphasizing non-specific therapeutic interactions and techniques that convey to the patient the therapist's interest, concern, and understanding. It utilizes the so-called common factors that have been assumed to account for much of the effect of all tested psychotherapies. These common or non-specific factors include the facilitation of affect, helping the patient to feel understood, to provide a framework for understanding, empathy, a treatment ritual, success experiences, hope and therapeutic optimism. It is assumed that many clinicians in private practice proceed in this unstructured manner. The explanatory mechanism for treatment effect offered to the patient focuses on the antidepressive effects of a supportive and understanding relationship, on the benefits of exploring and expressing emotions, the patient-directed structure of session and focus on personally relevant themes.
The number and duration of sessions as well as the experience of the therapists is equivalent in both study conditions. According to a meta-analysis of Baskin et al. [24
] structurally equivalent "placebos" produced negligible effects compared to active treatments [25
]. But there are also studies with depressed subjects which show significant effects for supportive interventions [26
]. Both CBASP and SP are conducted with two weekly individual sessions of 50 minutes each for the first 4 weeks and 1 weekly session for the remaining 16 weeks in the acute phase (=24 sessions, see Figure ), followed by 8 continuation sessions over the next 28 weeks (2 sessions in the first 4 weeks, and 1 session every 4 weeks thereafter). A 12-months naturalistic follow-up is planned for a second study phase since sustainment of response is particularly relevant given the chronic nature of the disorder.
Medication/treatments during trial: In cases of severe sleep problems, zolpidem is allowed for a maximum of 3 weeks. Central acting drugs are not allowed during the study. Single dosages of non-steroid analgesics and other non-centrally acting drugs for medical conditions are permitted.
Description of risks
Psychotherapeutic treatment with CBASP as well as with SP involves the chance of improvement of the depressive symptomatology. Side effects of evidence-based psychotherapies are fortunately rather rare [27
]. Possible undesired „side-effects" may include transient worsening of symptoms and transient risk of suicidality at the beginning of therapy (due to breaking out of avoidance behavior), but was rarely observed [28
Therapist training and monitoring of adherence
CBASP and SP are implemented by two separate groups of psychotherapists, both trained (in a 2-day training workshop and at least 1 practice day) in one of the methods and meeting the criteria for mastery of CBASP or SP procedures as assessed by evaluation of their performance during two videotaped pilot cases. All psychotherapists have completed a 3-year psychotherapy training program or are in an advanced stage of training. All sessions will be videotaped and site-supervisors continue to review the videotapes regularly on a random basis to assess psychotherapists' adherence to the treatment procedures using specific rating scales [19
]. In addition, a separate team of independent raters trained to reliability will randomly evaluate several of the tapes from early, middle, and late therapy phase of each treatment for adherence and therapist competence. Site-supervisors will be directly supervised by the trial-supervisors (E. Schramm & M. Hautzinger) in terms of bi-weekly conference calls and meetings (twice a year or more if needed).
Severity of depression
The 24-item-version of the HRSD has been implemented in the study by Keller et al. [12
] and will therefore provide internationally comparable study results. In addition, the applied 16-item Quick Inventory of Depressive Symptomatology, clinician-rated (QIDS-C16) and the self-rated 30-item Inventory of Depressive Symptomatology, Self-Report (IDS-SR) [30
] have highly acceptable psychometric properties [31
In a modified version of the IDS-SR (E-IDS), patients will be asked what they expect to answer in the IDS-SR at the end of therapy.
Severity of anxiety
In order to identify anxiety as a possible predictor for treatment outcome, symptoms of anxiety will be measured using the anxiety scale and the phobic anxiety scale from the Brief Symptom Inventory (BSI). The BSI is a validated self-report scale with strong test-retest and internal consistency reliabilities. Factor analytic studies of the internal structure of the scale have demonstrated its construct validity [32
]. Further, the GAD-7 [33
] is used. It is a valid and efficient tool with 7 items for screening for generalized anxiety disorder.
Quality of life
The Medical Outcome Study 36-item Short Form Health Survey (SF-36) [34
] is an internationally approved, generic instrument to assess Health-Related Quality of Life (HRQoL). The 12-item Short Form Survey (SF-12), derived from the SF-36, has been demonstrated to be reliable and valid in clinical and population-based applications in the U.S. and other countries [35
]. A more disease-specific instrument is the Quality of Life in Depression Scale (QLDS). This 34-item measure was developed to measure the impact of depression symptoms and treatment on quality of life [39
]. The QLDS has evidence of reliability, construct and content validity, and sensitivity to change in depressed patients [40
Interpersonal problems will be measured with the German translation of the 64-item self-report Inventory of Interpersonal Problems (IIP-64) [41
]. Psychometric research on the instrument in English-speaking communities as well as in German-speaking populations [41
] demonstrated the validity and the reliability (good internal consistency and test-retest reliability) of the IIP-64.
To measure global psychological, social, and occupational functioning, the widely utilized Global Assessment of Functioning (GAF, Axis V in DSM-IV) scale will be used. Another, more specific measurement is the Social Adaptation Self-Evaluation Scale (SASS) [42
], particularly for self-assessment of social functioning by patients with depression. It contains 21 items covering the different aspects of social interactions, global social attitude, and self-perception. The SASS has been validated and found to be simple to use and sensitive to changes in the different areas of social functioning [43
At baseline, the Childhood Trauma Questionnaire (CTQ) [44
] will be completed. The CTQ is a 28-item retrospective self-report questionnaire which determines 4 severity categories of emotional/physical/sexual abuse, and emotional/physical neglect. "Early trauma" is defined as one of these experiences before the age of 18 to a degree of at least "moderate to severe". In addition, the Early Trauma Inventory (ETI) [45
], a 56-item semi-structured interview is used. The ETI assesses also the domain of general trauma (not assessed by CTQ). The psychometric properties of both instruments have shown to be favorable [46
Evaluation by a relative
At the end of the therapy, a relative of the patient will be asked to evaluate 14 items of depressive symptoms before and after therapy.
Process analyses of therapies
After each session, the patients and therapists will fill out the Helping Alliance questionnaire (HAQ) developed by Luborsky [47
]. The instrument assesses two types of alliance: patient's experience of feeling helped and supported by the therapist and patient's experience of working together with the therapist in a joint effort in order to overcome the difficulties. The HAQ is correlated with other well validated instruments [48
], is also highly correlated with treatment outcome [49
] and shows similar psychometric properties to other alliance instruments [50
Screening, socio-demographic, and medical data
The initial screening visit consists of a medical and psychiatric history. Diagnoses will be derived using the Structured Clinical Interview for DSM-IV (SCID-I and II) [51
] during the screening evaluation as well as after 20 and 48 weeks of treatment. Sociodemographic data include sex, age, nationality, marital status, education, occupation, measure of household income, and employment. Medical data refer to previous or present diseases, outpatient and/or inpatient psychiatric and/or psychotherapeutic treatments; suicidal attempts and risk factors for suicide.
Depressive symptoms 20 weeks after randomization (after acute treatment phase) as measured by the 24-item HRSD.
a) Depressive symptoms after 12 and 48 weeks measured by the HRSD;
b) Remission rates after 12, 20, and 48 weeks utilizing the IDS-SR and defining remission a priori as a score of 13 or less for at least three consecutive weeks.
c) HRSD-remission rates (HRSD ≤ 8) and HRSD-response rates (HRSD score by at least 50 percent from baseline) will be calculated for the main measurement time points.
d) Changes in QIDS-C16 (from baseline to week 12, 20, and 48);
e) Temporal changes in IDS-SR - Total sum score between baseline assessment and follow-up assessments (time course of 27 time points).
Measures taken to minimize/avoid bias
The internet-based randomization will be conducted according to a central computerized randomization schedule, with a 1:1 treatment allocation ratio, stratified by centre, in blocks of variable size, to guarantee concealment. No-one can delete records from the randomization database, so that all randomizations have to be accounted for. Audit log files detailing all activity on the randomization system are available to the trial coordinator.
All clinical ratings will be completed by trained and independent evaluators blinded to treatment assignment. Each of the sites implements procedures to mask a patient treatment assignment from the person who will evaluate the results of the clinical ratings through the following: 1) locating the rater at a separate physical location, and 2) reminding the patients at each visit not to mention anything that might reveal their treatment condition to the independent evaluator. The baseline and the HRSD interview at 20 weeks are videotaped and will be evaluated by another rater.
Control of therapy allegiance
Several recommendations for how to minimize the allegiance effect are considered: involving several investigators who represent a "mix of therapy allegiances", comparing interventions of the same length and duration, using blinded raters for process and outcome analyses, and conducting both interventions in all sites [53
Control for overlapping treatments
The following measures are taken to prevent confounding of treatment conditions through the overlap of treatment methods:
Each therapist will conduct only one of the two treatments. The therapists are obligated to adhere to the therapeutic procedures and interventions described within the manuals. Adherence to the treatment manuals will be continuously supervised by rating videotapes of the sessions on a randomized basis using adherence scales [19
]. The validity of the therapist's statements will be checked through external assessment of the video recordings. In the post-hoc-analysis it will be checked if the expected differences regarding intervention characteristics appear within the therapies.
Control for confounding factors
The influence of the trial site upon the effectiveness of the respective treatment approaches will be investigated as a separate factor. In addition, patients are asked not to engage in off-study psychosocial (e.g., group therapy) or psychiatric interventions (e.g. antidepressive medication) during the treatment period.
The sample size calculation is based on the primary hypothesis testing CPASP against SP with regard to mean HRSD-scores at the end of the acute treatment phase (null hypothesis: identical expected HRSD). We considered a difference of five points on the HRSD between mean post-treatment scores after 24 sessions of the treated groups as clinically relevant. In similar studies, standard deviations of post-intervention HRSD-scores of groups receiving CBASP, SP, or combinations range from 5.4 to 10.4 points [12
]. Assuming a common standard deviation of 10 points for two-group comparisons yields a medium-sized effect of 0.5 (Cohen's d). To detect this effect by a two-tailed t-test with a power (1-β) of 0.95 and type I error probability level of α = 0.05 for significance, 210 patients (105 per group) are needed.
Assuming a drop-out rate of approximately 20% from baseline to week 20, the maximum sample size is fixed as 268 patients to be randomized. This is the number needed for an appropriately powered per-protocol analysis (only completers are analyzed). In the intention-to-treat analysis, the higher number of patients is expected to be compensated by a potential dilution of treatment effects, so that the power will be approximately the same.
The final analyses will be performed in the intention to treat (ITT) population, analyzing patients in treatment groups to which they were randomized, and using the last observation carried forward (LOCF) method in case of missing outcome data at week 20. Because of the chronic nature of the disorder, spontaneous remission is unlikely to happen.
At main analysis, the null hypothesis of equal efficacy will be tested (two-sided test) using analysis of covariance (ANCOVA) controlled for pre-treatment scores and site. Secondary analyses of the primary endpoint will include a per-protocol approach, regression controlling for additional factors, and exploratory analyses of treatment effect modifiers. To examine changes over time, a mixed model approach to repeated measures with 2 treatments by 4 measurement points will be used: baseline, after 12 weeks, after 20 weeks (acute intervention), and after 28 further weeks of continuation treatment. Analyses of continuous secondary variables will be performed using linear mixed models. For remission rates, chi-squared tests and logistic regression will be used. The analysis of time to remission and time to response will be analyzed using standard survival analysis techniques. Level of significance will be set at α = .05