Evidence-Based Psychiatric Medicine: Lithium for Suicide Prevention

Article excerpt

The Problem

You are working in one of America's contemporary long-term psychiatric care facilities, also known as the state prison. One of your patients suffers from some type of affective instability because of a primary mood disorder, an Axis II disorder, or a combination of both.

It is difficult to diagnose accurately as the inmate/patient provides a somewhat unreliable history, collateral data are not available or of questionable credibility, and records are either unavailable or in shambles.

The patient has a history of what might be suicide attempts or self-injurious behaviors, though again, it is difficult to tell for similar reasons. You consider prescribing lithium to target affective instability and possible suicide prevention.

The Question

How effective is lithium in preventing suicidal behavior?

The Analysis

We first searched the Cochrane Database of Systematic Reviews (www.cochrane.org/reviews) and found no reviews. We then searched Medline combining "lithium" and "suicide" and "prevent."

The Evidence

We were able to locate a meta-analysis published in 2005 and two original articles published since that time.

For the meta-analysis, the authors included randomized, controlled trials comparing lithium with placebo or other medications used in long-term treatment (at least 3 months) for various mood disorders that included major depression, bipolar disorder including rapid cycling, dysthymia, and schizoaffective disorder (Am. J. Psychiatry 2005;162:1805-19).

Outcomes were suicide; deliberate self-harm, including suicide attempts; and deaths from all other causes (the latter was to include deaths from suicide that had been incorrectly classified and to incorporate the toxic effects of treatment).

A total of 32 studies were included in the meta-analysis--19 comparing lithium with placebo and the rest with amitriptyline, imipramine, maprotiline, nortriptyline, fluvoxamine, carbamazepine, divalproex, or lamotrigine. In total, 1,389 patients were randomly assigned to lithium and 2,069 to the other medications.

Seven trials reported on suicides: two for lithium, compared with placebo; two for lithium, compared with amitriptyline, two for lithium, compared with carbamazepine; and one for lithium, compared with lamotrigine. Patients who received lithium were less likely to die by suicide (odds ratio 0.26).

However, the difference between groups was not statistically significant, because few deliberate self-injurious behavior (SIB) events were reported.

When suicide and SIB were considered together, fewer patients receiving lithium experienced this outcome (OR 0.21).

With regard to all-cause mortality, six trials were useful in comparing lithium and placebo, six for lithium and tricyclic antidepressants, two for lithium and carbamazepine, and one for lithium and lamotrigine. …