PRASED - Preventing Relapse After Successful Electroconvulsive Therapy for Depression

Although electroconvulsive therapy (ECT) is a very effective acute treatment for (unipolar as well as bipolar(1,2)) major depression, one of the major challenges in the treatment of mood disorders is to prevent relapse after successful treatment. Without continuation treatment, up to 50% of patients will relapse within 3 months after the last treatment and 6 months relapse rates are as high as 80% (3). Several strategies have been used to decrease these relapse rates. The most commonly used strategy, pharmacological treatment with antidepressants, decrease 6-month relapse rates by 30% (4,5). Nevertheless, as nearly half of patients on antidepressants will relapse within one year after successful ECT, relapse prevention remains a major challenge (6,7). Treatment of patients: Antidepressants: Antidepressant treatment is continued throughout phase 1 of the study. ECT: Acute phase - Treatment-as-usual All patients sign a general ECT informed consent, as routinely used in daily clinical practice, before treatment with ECT is initiated, stating that they agree with the suggested treatment and where they can indicate that routinely collected data during treatment can be used for research purposes to improve ECT practice. Patients are treated twice a week, with standard right unilateral electrode placement (RUL) (d'Elia placement) using either a Somatics Thymatron System IV (Somatics, Lake Bluff, Ill.) or a MECTA spECTrum device (MECTA Corporation, Portland, Ore.) with a pulse width of 0.5 ms (or higher if needed). A dose titration procedure (see appendix \* dose titration) to determine seizure threshold (ST) is conducted at the initial treatment and subsequent treatments are administered at six times the ST. Stimulus dose is raised by first maximizing train duration, followed by increase of frequency, and increase of pulse width, using a constant current (0,8 A). Motor seizure duration is evaluated at sight or with the cuff-method, and two channels of EEG (frontal-mastoid) are recorded. If the IDS score demonstrates a decrease \< 25% from baseline by treatment 6, the stimulus dose will be increased by 50% at the 7th session. If the IDS-C score demonstrated a decrease \<25% from baseline by treatment 9, the stimulus dose will be increased again by 50% or a switch to bitemporal ECT can be chosen, with empirical dose titration and subsequent treatments administered at two times ST. ECT was continued until patients achieved remission or had a plateau. Treatment with ECT is ended when patients show no improvement at all after 10 sessions of BT ECT. Anesthetic medications consist of propofol (1mg/kg) or etomidate (0.2mg/kg), and succinylcholine (1.0 mg/kg), given intravenously. ECT was continued until patients achieved remission or had a plateau in improvement over at least two consecutive evaluations. As part or regular care, cognitive functioning is also evaluated at baseline and after the acute treatment course. After successful ECT (start PRASED): Patients who have achieved remission, confirmed after one week, during acute treatment with ECT (treatment as usual) are eligible to participate in PRASED. After signing informed consent patients continue with personalised M-ECT according to the STABLE algorithm and are randomized to either receive lithium or not. The stimulus dose of the acute ECT course is continued in the continuation phase. After 4 weekly treatment sessions, the frequency of ECT-sessions is determined on the basis of symptom severity as assessed by the patient's IDS scores-C (Table 1) (weeks 5-26). In phase 2 of the study (weeks 27-52), treatment is continued by the treating psychiatrist who decides whether or not ECT, lithium and antidepressants are to be continued or stopped. Patients visit the clinic twice a month to evaluate mood and supplemental IDS-C telephone screenings will be done in the alternating weeks. When telephone IDS-C scores increased significantly based on the STABLE algorithm criteria for moderate to high relapse potential, the patient will be scheduled for an interim confirmatory clinical evaluation within 48 hours. Treatment decision will be based on the clinical confirmation IDS-C. As in the PRIDE study (13), the data centers will enter their results in a web-based program that collects the results for both clinical and telephone visits. The study coordinators at the different study sites will be provided with a description of the appropriate action (plan ECT sessions, schedule interim clinic visit if telephone IDS-C is significantly increased, plan next telephone or clinic visit). The study will be discontinued if the patient no longer wants to participate or in case of adverse events that raise safety concerns. Lithium: Patients in the Lithium-arm will also be treated with antidepressants and M-ECT. Additionally, openlabel lithium will generally be started at 400 mg/day, or lower in older patients. Lithium will be used in moderate doses with a target blood level 0.5-0.7 mEq/L. The day before ECT, the evening dose of lithium will be withheld. On ECT-days, lithium levels will be measured. Lithium levels will also be obtained at weeks 2, 3 and 4 weeks 8 and every 4 weeks after that. TSH and creatinine will be checked at 4, 12 and 24 weeks. Medication changes will be made on the basis of blood levels and clinical side effects. The schedule of patient contacts for clinical and telephone ratings is identical to the schedule in the treatment arm without lithium. METHODS Many tests are part of standard care and will be done once signed up for ECT. Retrospectively, eligibility based on diagnosis and severity of depression before start of ECT will be assessed. The extra assessments in the light of this study are situated after patients remitted after treatment with ECT. The participants are of course allowed to object any or part of the assessments. The time schedule of the project can be found in section 0. Of each patient scheduled for treatment with ECT, several standard variables are registered: * Age * Gender * Educational level * Episode duration * Presence of psychotic features * CORE-defined melancholic features * Baseline IDS-values * Number of previous episodes * ECT-appropriateness Scale (severity, heritability, episodic nature) (24) * Maudsley Staging Method Treatment parameters Treatment parameters of each treatment with ECT are registered. Data on electrode placement, stimulus dose, use of anesthetic, muscle relaxant and insult characteristics are stored in the ECT-files. Treatment effect: The severity of depression (primary outcome) will be tested with the IDS-C (Inventory of Depressive Symptomatology - Clinician) at twice-monthly clinic visits. Self-rating with the IDS-SR (Inventory of Depressive Symptomatology - Self-rating) will also be done once a month at one of the clinic visits. The weeks in between, a telephonic IDS-C will be done. Psychotic symptoms will be assessed with the Psychotic Depression Assessment Scale (PDAS)(25,26) once every month at a clinic visit. We can only aim at an independent assessment of mood (without knowledge of the treatment arm) but realize that the blinding will not be complete when for example Lithium side effect are mentioned by a patient during evaluation of mood. Other strategies are addition of lithium to antidepressant maintenance treatment or continuation of ECT after the acute course (M-ECT), alone or in combination with antidepressants and/or lithium. These strategies decrease relapse rates to different extents. In a large RCT, lithium in combination with Nortriptyline reduced 6-month-relapse rates to 32% (8). A recently published audit also found superior results of the combination of an antidepressant with lithium at six months (relapse rate of 16%) (9). Although it is advised to consider lithium combined with an antidepressant as a possible continuation treatment after ECT (10), its effectiveness remains understudied. M-ECT is a course that begins after the end of the acute ECT course and is intended to prevent relapse of the treated episode occurrence of a new depressive episode (11). Continuation of ECT can be done either at a fixed schedule (slowly tapering of ECT according to a fixed schedule) or personalized (flexible ECT depending on change in the level of depression symptoms). Most studies have used a fixed M-ECT scheme, while clinical practice suggest that a flexible, symptom driven treatment scheme might be more effective in preventing relapse (12). This was confirmed in the recently published PRIDE-study (13). Only 13% of the patients receiving personalized ECT, an antidepressant and lithium had relapsed at six months. This study, however, was limited to elderly patients, who are known to respond better (14) and have a better long-term outcome than the non-elderly (15). Our project is designed to validate results of the PRIDE-study, not limiting inclusions to older patients, as about half of the patients treated with ECT are non-elderly and relapse prevention in this patient group is as least as challenging. Besides that, we would like to assess the additional effect of lithium addition in preventing relapse with a flexible, symptom-driven and personalized ECT algorithm. After successful brief pulse ECT, patients will be randomized to receive either personalized M-ECT in combination with their ongoing (antidepressant) medication and lithium, or personalized M-ECT in combination with their ongoing (antidepressant) medication without lithium. There will be a follow-up period of one year. This project enables us to validate the effectiveness of personalized M-ECT, and to assess the effectiveness of lithium as part of a relapse prevention strategy post-ECT. Improving relapse prevention is 'the most pressing issue in the field' (7), with an added value for the group of patients with depression that were treated with ECT. As this is only scarcely been investigated, it leaves the clinician without clear guidance on the best treatment strategy after successful ECT. Although our main goal is a positive medical impact, reducing relapse rates eventually also has a socio-economic influence by reducing health care costs. As mentioned before, one of the major challenges in the treatment of mood disorders with ECT is to prevent relapse after a successful acute course. Although relapse prevention strategies have been investigated in several studies, current guidelines offer only a limited amount of information to clinicians regarding the best form of continuation treatment. Withholding continuation therapy results in relapse rates of up to 80% within 6 months after the last treatment (3-5). Jelovac et al. conclude that the risk of relapse in the year following ECT is substantial, with the first 6 month-period encompassing the greatest risk for relapse (3). The now most commonly prescribed form of continuation treatment are antidepressants, lowering 6- month relapse rates by 30% (4,5). However, 50% of patients will relapse within one year, the majority in the first six months (37%). The largest evidence base for efficacy in post-ECT relapse prevention exists for the tricyclic antidepressants. Other strategies used are addition of lithium to antidepressant maintenance treatment or continuation of ECT after the acute course (M-ECT), alone or in combination with antidepressants and/or lithium. These strategies decrease relapse rates to different extents. Regarding ECT, guidelines about continuation of therapy remain scarce whereas regarding antidepressants, all guidelines recommend a continuation regimen for one year or longer after remission has been achieved. This implicates that ECT often is stopped when remission is achieved, instead of continuing at a lower frequency. Given the limited effectiveness of current continuation treatment protocols, researchers have been looking for alternatives, with following strategies being most promising: Lithium The first promising alternative is combination of lithium with a given continuation treatment (ECT or antidepressants) following an acute ECT course. In a large RCT, lithium in combination with nortriptyline reduced 6-month-relapse rates to 32% (8). A recently published audit also found superior results of the combination of an antidepressant with lithium at six months (relapse rate of 16%) (9). Other studies show promising results but are not as hopeful as the before mentioned studies. A study conducted in the USA compared two maintenance treatment schemes: monotherapy with tricyclic antidepressants (TCA) on the one hand, and combination therapy of TCA-lithium on the other hand (16). Patients were randomly assigned to receive continuation treatment for 24 weeks with placebo (n = 29), nortriptyline (n = 27) or the combination of nortriptyline and lithium (n = 28). The study indicates that without active treatment, virtually all remitted patients relapse within 6 months of stopping ECT (84%). Monotherapy with nortriptyline has limited efficacy (60% relapse rate). The combination of nortriptyline and lithium is more effective (39% relapse), but the relapse rate is still high, particularly during the first month of continuation therapy. A rather large study (N=122) by Prudic e.a. (17) compared maintentance pharmacotherapy of two combination schemes: nortriptyline-lithium and venlafaxine-lithium. No difference in relapse rate (both had relapse rates of 50%) or adverse effects was detected between the treatment schemes. All literature on lithium as a prophylactic against depressive relapse after ECT has recently been reviewed (10). Rasmussen e.a. concluded that there is strong evidence that lithium can help prevent relapses in the first 6 months after successful ECT. However, there are several unanswered questions about its use post-ECT such as the optimal duration of use, and concomitant antidepressant choice. M-ECT is a course that begins after the end of the acute ECT course and is intended to prevent relapse of the treated episode occurrence of a new depressive episode (11). Continuation of ECT can be done either at a fixed schedule (gradual tapering of ECT according to a fixed schedule) or personalized (flexible ECT depending on change in depression symptoms) and starts to find its way into daily clinical practice. Within this context, a Swedish study compared pharmacological continuation treatment with a combination of pharmacotherapy and M-ECT (N=56) (18). In addition, all patients were offered augmentation with lithium. The post-ECT relapse rates were substantial in both treatment groups with a statistically significant advantage, i.e. lower relapse rates, for combination of pharmacotherapy and continuation ECT. Moreover, patients on lithium showed significantly lower relapse rates than the non-lithium group. M-ECT (10 treatments) was also directly compared to efficacy of lithium plus nortriptyline (6 months) (8) by the CORE Consortium for Research in Electroconvulsive Therapy. They concluded that both treatment strategies were equally effective and superior to a historical placebo control group, but both had limited efficacy with more than half of patients either experiencing disease relapse or dropping out of the study (M-ECT group 37% relapse, 17% dropout, lithium + nortriptyline 32% relapse, 22% dropout). The authors concluded that more effective methods to prevent relapse are needed. Most studies have used a fixed M-ECT scheme, while clinical practice suggest that a flexible, symptom driven treatment scheme might be more effective in preventing relapse (12). To improve relapse rates, Odeberg e.a. (19) introduced individually tapered M-ECT with concomitant medication after an acute course of ECT. The mean duration of administered continuation ECT at follow-up was 1 year, but for most patients (63%), ECT had been terminated after the acute course. For 49% of patients, adjustments between ECT sessions had been made due to early signs of relapse. Two weeks was the most common interval between sessions for patients with ongoing ECT. The need for hospital care 3 years before and after the initiation of maintenance ECT + medication was evaluated. The number of patients hospitalized, number of admissions, and total days in hospital were all significantly reduced. Hospital days were reduced by 76%. This study presents the evidence that supports previous findings that individually tapered ECT combined with medication can maintain initial response to ECT and serve as a bridge to long-term relapse prevention. As there was no clear adjustment scheme in the study done by Odeberg e.a., it is hard to confirm their results. Lisanby e.a. (12) proposed Symptom-Titrated, Algorithm-Based Longitudinal ECT (STABLE) as a novel patient-focused approach to perzonalize the ECT schedule. In STABLE, the ECT schedule adapts to symptom fluctuations to prevent over-treatment of those who do not need it, and to re-capture response in those who might have otherwise relapsed with a rigid dosing schedule. It has taken several years to test this algorithm in an RCT but results of treatment according to the STABLE algorithm in an elderly population were published a few months ago in the PRIDE study (13). The algorithm proved to be very effective with only 13% of relapse in the personalized ECT plus medication (Venlafaxine + Lithium) group at six months. The group that was treated with (the same) medication, e.a. Venlafaxine + Lithium, had 20% relapse at six months. This study, however, was limited to elderly patients, who are known to respond better (14) and have a better long-term outcome than nonelderly (15). Another difference with ECT practice in Belgium is that the USA study used ultra-brief pulse ECT. This is a treatment method well tolerated but not as effective (20) as brief pulse ECT that is used for our acute and continuation treatment schedules. Besides that, they were treated with venlafaxine and not with a TCA. It can however be concluded that additional ECT after remission (only as needed) was beneficial in sustaining mood improvement for most patients. Literature on M-ECT was also reviewed a few years ago (21). The conclusion was that maintenance ECT was efficacious for the prevention of relapse of major depression and that the efficacy increased when combined with antidepressant medication and at flexible treatment intervals, responsive to early signs of recurrence. To conclude, there is a substantive amount of evidence for a role of antidepressants in relapse prevention after an ECT course. Tricyclic antidepressants and Venlafaxine seem to be promising antidepressants and can contribute to sustaining remission after successful ECT. As this is often not sufficient to retain remission of prevent recurrence of a new depressive episode, a symptom driven form of continuation ECT and addition of lithium to the treatment can be considered as potential strategies to improve treatment after successful ECT. Confirmation of their potential in a depressed sample of all ages has never been done and would be valuable. Although electroconvulsive therapy (ECT) is a very effective acute treatment for depression, one of the major challenges in the treatment of mood disorders is to prevent relapse after successful treatment. Several strategies have been used to reduce relapse rates but most of them seem to be relatively unsuccessful. We selected the two strategies that we consider to be least investigated and most promising in relapse prevention. An RCT was designed to determine their potential to reduce relapse after successful ECT. Study Design: All patients that respond well to treatment with ECT for unipolar depression in one of the four treatment centers, are screened for eligibility in the study. Patients that have reached the remission criterion (strict (IDS-C ≤ 12) or liberal (IDS-C ≤ 17) on two consecutive measurements) after this ECT course are considered eligible to participate in this RCT. Retrospectively, the pre-treatment depression characteristics are evaluated because the presence of a unipolar depressive disorder (MINI International Neuropsychiatric Interview) of at least moderate severity (IDS ≥ 29) at baseline is necessary for inclusion in the study. Furthermore, they have to be treated with either tricyclic antidepressants or venlafaxine. When patients provide their written informed consent for the study, phase 1 starts - open label antidepressants are continued and personalized ECT (see Figure 1) is given for the next six months. Patients will be randomized to addition of lithium or no lithium. Personalized ECT in the maintenance phase of the treatment is given according to the so-called STABLE-algorithm (Symptom-Titrated Algorithm-Based Longitudinal ECT, (Table 1) that was translated to a version that can be used with the Inventory of Depressive Symptoms (IDS), the depression severity scale that we use in our study instead of the HDRS24 on which the STABLE-algorithm was originally based. Scores were changed in to the IDS-equivalents of the HDRS24-scores (22,23). After six months, phase 1 ends and patients enter phase 2 of the study. In this phase, we no longer control treatment. Care will be handed over to the treating psychiatrist. In consultation with the patient, treatment with lithium can be continued and there is a possibility to continue M-ECT when this seems beneficial for the patient. In phase 2, mood and (change in) treatment of the previous three months will be assessed at 9 and 12 months after randomization. The design of the PRIDE study (13) was used to inspire our protocol, although the treatment arms we compare are somewhat different. Adaptations were made according to European ECT practices and feasibility in our centers. Study Population: Patient recruitment strategy Before starting treatment with ECT, patients for whom a course of ECT is indicated will routinely undergo a diagnostic workup. As part of this diagnostic workup, a MINI diagnostic interview is done and when a unipolar major depression diagnosis is confirmed by this interview, patients will be considered eligible for the PRASED-study. After an acute course of ECT with antidepressants (TCA dosed according to target plasma level or venlafaxine (target dose of ≥ 225mg/day)), patients will be screened for eligibility, depending on whether or not they reached remission after treatment with ECT. The research team will be notified when the last treatments of the acute course are planned. In the last week of the acute treatment, mood is evaluated and when remission is confirmed, patients will be informed about the PRASED study. When remission is confirmed one week later, informed consent for the PRASED study will be asked. For the power-analysis, we have looked at the relapse rates in two studies. The first one compared antidepressant maintenance treatment with a TCA with TCA+lithium (16) - with respectively 60% and 39% relapse after 24 weeks. The second one was the PRIDE study (13), that compared antidepressants+lithium with antidepressants+lithium and the personalized M-ECT according to the STABLE-algorithm, the proportion of relapse in this study was remarkably lower (20 vs 13%), possibly due to the fact that the treated patients were elderly and that M-ECT was additionally used in the group with the lowest relapse rate. Based on our study design and targeted study population (not only elderly), we expect relapse rates to be somewhere in between those of the two studies. As we consider a difference in relapse rates of 15% between the groups clinically relevant, our power calculation was done using relapse rates of 15% in the one treatment arm and 30% in the other one. We did a power calculation with an α of 0.05 and a standard power value of 0.8. With a sampling ratio of 1 (the same number of patients in the two groups), we would need a sample size of 95 in each group for our primary outcome measure. That makes a total of 190 patients. Because of the intensity of the study protocol (weekly evaluation of mood), we some dropout and aim at inclusion of 240 patients at randomization so that a dropout of 20% would leave enough patients to do meaningful statistical analyses at the start of phase 2 (after 6 months) of the study. Given the size of the ECT units, we expect to randomize 88 patients in Leuven, 64 in Duffel, 64 in Brugge and 24 in Rotterdam. Treatment of patients: Antidepressants Patients are eligible for participation in the study when they are treated with either an adequately dosed TCA (with therapeutic blood levels) or venlafaxine (target dose of ≥ 225mg/day) at the start of the study. Antidepressant treatment is continued throughout phase 1 of the study study. ECT: Acute phase - Treatment-as-usual All patients sign a general ECT informed consent, as routinely used in daily clinical practice, before treatment with ECT is initiated, stating that they agree with the suggested treatment and where they can indicate that routinely collected data during treatment can be used for research purposes to improve ECT practice. Patients are treated twice a week, with standard right unilateral electrode placement (RUL) (d'Elia placement) using either a Somatics Thymatron System IV (Somatics, Lake Bluff, Ill.) or a MECTA spECTrum device (MECTA Corporation, Portland, Ore.) with a pulse width of 0.5 ms (or higher if needed). A dose titration procedure (see appendix \* dose titration) to determine seizure threshold (ST) is conducted at the initial treatment and subsequent treatments are administered at six times the ST. Stimulus dose is raised by first maximizing train duration, followed by increase of frequency, and increase of pulse width, using a constant current (0,8 A). Motor seizure duration is evaluated at sight or with the cuff-method, and two channels of EEG (frontal-mastoid) are recorded. If the IDS score demonstrates a decrease \< 25% from baseline by treatment 6, the stimulus dose will be increased by 50% at the 7th session. If the IDS-C score demonstrated a decrease \<25% from baseline by treatment 9, the stimulus dose will be increased again by 50% or a switch to bitemporal ECT can be chosen, with empirical dose titration and subsequent treatments administered at two times ST (Figure 2). ECT was continued until patients achieved remission or had a plateau. Treatment with ECT is ended when patients show no improvement at all after 10 sessions of BT ECT. Anesthetic medications consist of propofol (1mg/kg) or etomidate (0.2mg/kg), and succinylcholine (1.0 mg/kg), given intravenously. ECT was continued until patients achieved remission or had a plateau in improvement over at least two consecutive evaluations. As part or regular care, cognitive functioning is also evaluated at baseline and after the acute treatment course. After successful ECT (start PRASED) Patients who have achieved remission, confirmed after one week, during acute treatment with ECT (treatment as usual) are eligible to participate in PRASED. After signing informed consent patients continue with personalised M-ECT according to the STABLE algorithm and are randomized to either receive lithium or not. The stimulus dose of the acute ECT course is continued in the continuation phase. After 4 weekly treatment sessions, the frequency of ECT-sessions is determined on the basis of symptom severity as assessed by the patient's IDS scores-C (Table 1) (weeks 5-26). In phase 2 of the study (weeks 27-52), treatment is continued by the treating psychiatrist who decides whether or not ECT, lithium and antidepressants are to be continued or stopped. Patients visit the clinic twice a month to evaluate mood and supplemental IDS-C telephone screenings will be done in the alternating weeks. When telephone IDS-C scores increased significantly based on the STABLE algorithm criteria for moderate to high relapse potential, the patient will be scheduled for an interim confirmatory clinical evaluation within 48 hours. Treatment decision will be based on the clinical confirmation IDS-C. As in the PRIDE study (13), the data centers will enter their results in a web-based program that collects the results for both clinical and telephone visits. The study coordinators at the different study sites will be provided with a description of the appropriate action (plan ECT sessions, schedule interim clinic visit if telephone IDS-C is significantly increased, plan next telephone or clinic visit). The study will be discontinued if the patient no longer wants to participate or in case of adverse events that raise safety concerns. Lithium: Patients in the Lithium-arm will also be treated with antidepressants and M-ECT. Additionally, openlabel lithium will generally be started at 400 mg/day, or lower in older patients. Lithium will be used in moderate doses with a target blood level 0.5-0.7 mEq/L. The day before ECT, the evening dose of lithium will be withheld. On ECT-days, lithium levels will be measured. Lithium levels will also be obtained at weeks 2, 3 and 4 weeks 8 and every 4 weeks after that. TSH and creatinine will be checked at 4, 12 and 24 weeks. Medication changes will be made on the basis of blood levels and clinical side effects. The schedule of patient contacts for clinical and telephone ratings is identical to the schedule in the treatment arm without lithium. METHODS: Many tests are part of standard care and will be done once signed up for ECT. Retrospectively, eligibility based on diagnosis and severity of depression before start of ECT will be assessed. The extra assessments in the light of this study are situated after patients remitted after treatment with ECT. The participants are of course allowed to object any or part of the assessments. The time schedule of the project can be found in section 0. Of each patient scheduled for treatment with ECT, several standard variables are registered: * Age * Gender * Educational level * Episode duration * Presence of psychotic features * CORE-defined melancholic features * Baseline IDS-values * Number of previous episodes * ECT-appropriateness Scale (severity, heritability, episodic nature) (24) * Maudsley Staging Method Treatment parameters Treatment parameters of each treatment with ECT are registered. Data on electrode placement, stimulus dose, use of anesthetic, muscle relaxant and insult characteristics are stored in the ECT-files. Treatment effect The severity of depression (primary outcome) will be tested with the IDS-C (Inventory of Depressive Symptomatology - Clinician) at twice-monthly clinic visits. Self-rating with the IDS-SR (Inventory of Depressive Symptomatology - Self-rating) will also be done once a month at one of the clinic visits. The weeks in between, a telephonic IDS-C will be done. Psychotic symptoms will be assessed with the Psychotic Depression Assessment Scale (PDAS)(25,26) once every month at a clinic visit. We can only aim at an independent assessment of mood (without knowledge of the treatment arm) but realize that the blinding will not be complete when for example Lithium side effect are mentioned by a patient during evaluation of mood.