Lithium's Anti-Suicide Evidence Against a Patient's Own Overdose History
A single patient, three years past a serious suicide attempt and now facing recurrent depressive breakthroughs. Lithium is the one mood stabilizer with evidence for reducing completed suicide — and the one whose own overdose toxicity maps most directly onto her documented history.
J.R. is a 41-year-old woman, a single mother of two teenagers who separated from her husband of fifteen years last spring. She was diagnosed with bipolar I disorder eight years ago and has been maintained on quetiapine monotherapy since a serious suicide attempt three years ago, when she took a deliberate overdose of acetaminophen during a severe depressive episode and required several days of inpatient medical treatment before psychiatric stabilization. Quetiapine has kept her largely well since, but she has had two milder depressive breakthroughs in the past year, both with passive suicidal ideation that resolved without a plan or attempt, most recently six weeks ago in the weeks surrounding her separation.
Her psychiatrist is now weighing whether to add or switch to lithium, and the reason is exactly the fact that makes the decision hard rather than easy: lithium is the only mood stabilizer with trial evidence for reducing completed suicide, and she is a patient with a documented prior attempt and recent recurrence of ideation — precisely the population that evidence base describes. But her prior attempt was a deliberate overdose, and lithium has a narrow therapeutic index with real lethality in overdose, including cardiac and neurologic toxicity that acetaminophen at the quantities she took would not have caused. Withholding the drug with the best suicide-prevention evidence from a patient specifically because of her suicide history is its own kind of contradiction, and her psychiatrist has been reluctant to prescribe it for exactly that reason for three years running.
Weighing lithium against her own history
The reluctance to prescribe lithium because of her attempt history has, in effect, kept her three years on a regimen without the one agent shown to reduce suicide mortality specifically in patients like her. That reasoning protects against a hypothetical lithium overdose while leaving the actual, already-demonstrated risk — recurrent depressive breakthroughs with ideation — undertreated.
Overdose risk from a prescribed supply is manageable through dispensing quantity and visit frequency. It is not a reason to withhold the drug category altogether, any more than a hepatotoxic-in-overdose drug would be withheld from every patient with a prior overdose on a different substance.
I weight her actual history more heavily than the class-level argument. She has already shown she will act on suicidal impulse under enough pressure, and I have watched her go through a divorce this year. A drug that is dangerous specifically in the way lithium is dangerous, in a patient who has already demonstrated capacity for a deliberate overdose, is a real and specific risk to her, not an abstract one.
I take the undertreatment argument seriously, which is why I am not proposing we do nothing — augmenting quetiapine with lamotrigine addresses her recurrent depressive symptoms without introducing an agent whose overdose profile maps directly onto her prior method of choice.
Both of you are right about different halves of the same problem, and the actual fix is operational, not pharmacologic. A 15-day supply with no automatic refill, pharmacy-level quantity limits, and her sister holding the medication and administering it daily would cut the amount available for any single overdose attempt well below a dangerous threshold while still giving her the drug with the strongest suicide-prevention evidence.
That only works if she agrees to it and if her sister is actually willing and available — this is a plan that depends on her support system holding up, not just on the prescription itself.
Lithium 300mg twice daily started with a 15-day supply and no automatic refill, quantity limits flagged at the pharmacy, and her sister agreeing to hold and administer the medication daily during the initial months. Quetiapine continued unchanged. A lithium level and renal/thyroid panel were scheduled at day 5 and again at week 4.
The disagreement about whether the overdose risk justified withholding lithium at all was not fully resolved — it was managed around, not settled. The psychiatrist's concern was documented explicitly as a standing condition of the plan: if the sister's involvement becomes unreliable, or if a further depressive breakthrough occurs, lithium is to be reconsidered rather than assumed safe by default.