Depression and suicidal ideation/attempt/death are major causes of morbidity and mortality from psychiatric illnesses. In 2009, the World Health Organization listed depression as the leading cause of years lost due to disability worldwide. Suicide is the 9th most common cause of death in Canada with 1.6% of Canadians ultimately dying from suicide (Statistics Canada, 2012) and the 2nd most common cause of death in young people after accidental deaths. This information highlights the importance of finding treatments to prevent suicidal deaths.

Ketamine has been shown to provide rapid treatment response for major depressive episodes both in major depressive disorder (MDD) and bipolar disorder (BD), via a single intravenous infusion which persists for at least 72 hours.

The purpose of this study is to conduct a pilot trial of IV ketamine + treatment as usual (TAU) vs. midazolam (an active placebo) + TAU to estimate sample size for a full-scale RCT examining these treatments for decreasing suicidal ideation among depressed inpatients with major depressive disorder and bipolar depression.

A total of 52 patients will be recruited for this trial. All subjects will be inpatients at Sunnybrook Health Sciences Centre with a diagnosis of either major depressive disorder or bipolar disorder type I or II currently depressed. Suicidal ideation must be present at baseline assessment in order to be included in the study. Thirteen subjects will be randomized to each treatment arm in each treatment stream - that is, 13 will be recruited to ketamine + TAU in the major depressive disorder stream, and 13 will be recruited to the midazolam + TAU in the major depressive stream. Likewise, 26 subjects with bipolar depression will be randomized to these two treatments.

Depression and suicidal ideation/attempt/death are major causes of morbidity and mortality from psychiatric illness. The World Health Organization (2009) lists depression as the leading cause of years lost due to disability worldwide. Suicide is the 9th most common cause of death in Canada with 1.6% of Canadians ultimately dying from suicide (Statistics Canada, 2012) and the 2nd most common cause of death in young people after accidental deaths. The investigators' data show that at least 50% of people dying from suicide in Toronto suffer from depression with a small proportion ~12% suffering from BD. These data underscore the urgency of developing new treatments for both MDD and BD but also the suicidality that is often associated with them.

Ketamine treatment represents a potentially viable, safe and effective treatment for MDD/bipolar depression + SI. The Investigators therefore propose to conduct a pilot trial in preparation for a full- scale randomized controlled trial (RCT) which would aim to determine the efficacy of IV ketamine + a standard medication treatment (Treatment As Usual; TAU) vs. midazolam, an "active" placebo + TAU in treating SI among inpatients with MDD and in inpatients with bipolar depression. If the full-scale RCT demonstrates ketamine's efficacy, it would have important implications for both future research as well as inpatient treatment.

The primary objective is to conduct a pilot trial of IV ketamine + TAU vs. midazolam + TAU to estimate sample size for a full-scale RCT examining these treatments for decreasing SI among depressed inpatients with MDD and bipolar depression.

The primary hypothesis is that the effect size for reducing SI in the ketamine group vs. the midazolam group will be in the moderate range or above (d > 0.5 at 14 and 42 days) in terms of reduction in scores on the Scale of Suicidal Ideation (SSI) and the Columbia-Suicide Severity Rating Scale (CSSRS) for both subjects with MDD and subjects with bipolar depression.

Further, there are secondary objectives and secondary hypotheses. The secondary objectives are:

1. To estimate effect size for producing clinical response (≥50% reduction in Montgomery-Asberg Depression Rating Scale (MADRS) scores) and remission (MADRS<12) in the ketamine group vs. the midazolam group at 14 and 42 days in both MDD and bipolar depression.
2. To determine whether the ketamine group produces a more rapid reduction in SSI, CSSRS and MADRS scores compared to the midazolam group in both MDD and bipolar depression.
3. [bipolar depression stream only] To determine whether ketamine or midazolam induces manic symptoms in any subjects and, if so, whether ketamine produces manic symptoms in more subjects than midazolam.

Secondary hypotheses are:

1. Effect sizes for achieving clinical response and remission in the ketamine group vs. the midazolam group will be in the moderate range or above (d > 0.5 at 14 and 42 days).
2. There will be a significantly more rapid reduction in SSI, CSSRS and MADRS scores in the ketamine group than in the midazolam group.
3. [bipolar depression stream only] Manic symptoms will occur in <10% of all subjects and there will be no differences between the two groups.

Finally, the exploratory objectives are:

1. To determine whether mean time to discharge differs between the ketamine and midazolam groups in both MDD and bipolar depression.
2. To determine whether subject satisfaction is differs between the ketamine group and in the midazolam group.

And exploratory hypotheses:

1. Mean time to discharge from hospital will be faster in the ketamine than in the midazolam group in both MDD and bipolar depression.
2. Subject satisfaction will be higher in the ketamine than in the midazolam group in both MDD and bipolar depression.

Subjects will be inpatients at Sunnybrook Health Sciences Centre with a diagnosis of either MDD or BD type I or II currently depressed. To be included in the study, SI must be present at the time of baseline assessment. Subjects will be recruited on the first regular week day (non-weekend/holiday) after their admission. This is the day on which a comprehensive inpatient treatment plan is typically developed. Subjects will be randomized in a 1:1 double-blind fashion to two groups in both a MDD and a BD stream. That is, 13 subjects will be randomized to ketamine + TAU and 13 subjects will be randomized to midazolam + TAU in the stream for subjects with MDD. Likewise 26 subjects with bipolar depression will be randomized to these two treatments.

Subjects in the ketamine IV groups will receive infusions 3 times weekly for two weeks (0.5 mg/kg infused over 40 minutes on approximately days 1, 3, 5, 8, 10 and 12 of admission). The exact schedule of dates is referred to as approximate since it may need to be adjusted slightly depending on the timing of admission/weekends etc. Subjects in the midazolam group will have the same dosing schedule but will instead receive midazolam 0.045 mg/kg IV infused over 40 minutes. A sub-anesthetic dose of midazolam was chosen as an active placebo because it has CNS effects including sedation and amnestic effects, thus making it more difficult for subjects to guess which group they are in. Vital signs including pulse, respiratory rate and arterial oxygen saturation will be monitored throughout the ketamine/midazolam infusion and for one hour post-infusion as has been the standard in the published literature. If the subject experiences side effects, the protocol will allow for the infusion to be slowed to up to 90 minutes.

Subjects will also receive TAU during the course of the study in addition to IV ketamine or midazolam treatment. In the MDD group, TAU may include a newly initiated or longstanding antidepressant. In the BD group, TAU may include a mood stabilizer such as lithium or valproate that is a first or second line agent as per Canadian Network for Mood and Anxiety Treatments (CANMAT) guidelines (Yatham et al., 2013). This is done in an attempt to mitigate the risks of relapse after cessation of ketamine therapy and also makes the use of midazolam treatment ethically justifiable. The duration of the study is two weeks. All subjects will receive all 6 treatments, regardless of whether their symptoms have remitted, given recent evidence that 6 IV ketamine treatments led to a more robust and lasting response compared to only 1-3 treatments (Aan Het Rot et al., 2012). If subjects are discharged before 2 weeks, they will be asked to return to hospital as outpatients for any remaining ketamine treatments as well as for outcome measures at 2-weeks and 42 days. TAU will be maintained after discharge.

Depression, suicidal ideation measures (MADRS, SSI, CSSRS) and, in the BD stream, mania measures (Young Mania Rating Scale; YMRS) will be administered on admission, on treatment days in the morning prior to ketamine/midazolam administration and on days 14 and 42. At both days 14 and 42, subjects will also be asked to rate their satisfaction with the study.

Inclusion Criteria:

1. Provision of written informed consent
2. [MDD stream only] Diagnosis of major depressive disorder, currently depressed as determined by DSM-IV diagnostic criteria (confirmed using the MINI)
3. [BD stream only] Diagnosis of bipolar disorder, type I or type II, currently depressed as determined by DSM-IV diagnostic criteria (confirmed using the MINI)
4. Both females and males, aged 18 to 65 years
5. Inpatient status
6. Female patients of childbearing potential must have a negative urine human chorionic gonadotropin (hCG) test at enrolment and must be taking or willing to take some acceptable form of birth control during the course of the study if they are or plan to be sexually active
7. The ability to understand and comply with the requirements of the study and capable of providing informed consent
8. Suffering from suicidal ideation/attempts as evidenced by a score of >0 on either of the SSI or CSSRS or both.

Exclusion Criteria:

1. Current or past psychotic symptoms
2. Substance or alcohol dependence at enrollment (except dependence in full remission, and except for caffeine or nicotine dependence), as defined by DSM-IV criteria
3. Opiates, amphetamine, barbiturate, cocaine, cannabis, or hallucinogen abuse by DSM-IV criteria within 4 weeks prior to enrollment
4. Any pervasive developmental disorder (according to DSM-IV criteria)
5. Diagnosis of dementia (according to DSM-IV criteria)
6. Known intolerance or hypersensitivity to ketamine or midazolam as judged by the investigator
7. Significant medical condition that would contraindicate the use of ketamine, midazolam or that is untreated and would need urgent attention (as determined by treating physician)
8. Medical conditions that would significantly affect absorption, distribution, metabolism, or excretion of ketamine or midazolam
9. Unstable or inadequately treated medical illness (e.g. congestive heart failure, angina pectoris, hypertension) as judged by the investigator
10. Any clinically significant deviation from the reference range in clinical laboratory test results as judged by the investigator
11. Pregnancy (or female of child-bearing age not using adequate contraception) or lactation
12. A positive β-hCG test at enrollment
13. Involvement in the planning and conduct of the study
14. Previous enrollment or randomisation of treatment in the present study
15. Participation in another drug trial within 4 weeks prior enrollment into this study or longer in accordance with local requirements