Postpartum Psychosis: The Diagnosis on Trial

There's an illness in psychiatry with a published treatment path and a remission rate above ninety percent — and the treatments that matter most usually arrive last, if they arrive at all. I'm Justin Ray. I'm a psychiatric nurse practitioner, I've worked in mental health for over thirty years, and this is one of the hardest topics I've had to cover: postpartum psychosis.

You may already know how the Lindsay Clancy case ended. I recorded this while the jury was still out — and the verdict doesn't change a single word of what follows. Clancy, a labor and delivery nurse, killed her three children in January 2023, then attempted to end her own life. The trial ran twenty-one days. The defense said postpartum psychosis. The prosecution said depression, exhaustion, and a plan. But experts on both sides reached for the same clinical term: altruistic filicide. I never examined her, and I won't diagnose a woman I've never met. I'd be suspicious of any clinician who does. But at the level of the category, I'll be direct. A fixed conviction that three healthy children cannot survive your absence is not an intensity of sadness. We'd name that anywhere else: a failure of reality testing.

Postpartum psychosis follows one to two of every thousand deliveries, typically within the first two weeks — often within days. And early on, it rarely looks like the word suggests. It looks like insomnia that has stopped behaving like exhaustion. A mother who can't sleep even when the baby sleeps — and doesn't seem to need to. Perplexity. Mood swinging hour to hour. An almost delirium-like quality — intact at nine in the morning, unrecognizable by noon. Then come the beliefs, frequently centered on the infant. "She seemed fine an hour ago" is compatible with this illness. It is not evidence against it. The biology stacks the deck: the steepest hormonal cliff in human physiology, landing on enforced sleep deprivation, often in a brain already carrying a bipolar diathesis.

Two distinctions matter at the bedside. Intrusive thoughts in postpartum OCD are common, terrifying — and experienced as horrifying intrusions. The mother hides the knives. Very low risk of action. Psychotic beliefs are the opposite. They are not experienced as intrusions. They are experienced as truths. And the only way to tell is to ask the follow-up questions. Do the thoughts feel like yours? Do they feel true? What do you believe would happen if you never acted on them? The second distinction is disclosure. Mothers conceal these thoughts precisely because they fear their children will be taken. A checkbox denial is worth exactly the paper it's printed on.

The stakes: older cohort data put suicide risk near five percent, and infanticide risk near four percent. New psychotic symptoms in a postpartum woman are a psychiatric emergency. Full stop. Hospitalization, safety planning for mother and infant, and immediate definitive treatment. This is not watch-and-wait.

The strange, hopeful part: we've known the sequence for a decade. Sixty-four consecutive women, first-onset. Benzodiazepines at night to restore sleep, then an antipsychotic, then lithium — ECT the fourth step. Ninety-eight point four percent reached full remission within the first three steps. None required ECT. And in women at high risk, lithium started immediately after delivery can prevent the episode.

Now hold that against American prescribing. Across twenty years, antipsychotic use in bipolar disorder rose from twelve percent of visits to fifty-one. Lithium and anticonvulsants fell by more than half. In all my years prescribing, no representative has ever bought my staff lunch to talk about lithium. No booth. No speaker program. These treatments have no advocate except us.

And ECT: in Swedish registry data, postpartum women responded at eighty-seven percent — above ninety in the psychosis subset. It's fast, and compatible with continued breastfeeding. When hours matter — catatonia, refusal of food and fluids, active suicidal or infanticidal drive — nothing else we have moves at its speed. So why does it arrive last? Not the evidence. Everything around it. Stigma with a seventy-year head start. Anesthesia slots. A suite two hours from the delivery ward. My position is simple. ECT should be named at admission, out loud, to the family. Considering it immediately is not the same as using it immediately.

One more thread. The hours after delivery are biologically modifiable — in recent trials, a single dose of esketamine right after childbirth cut later depression dramatically. And I want to be blunt here, because I operate an esketamine center, and I feel the pull of extending that logic. Those are depression prevention trials — not psychosis. Ketamine and esketamine are dissociatives with psychotomimetic potential. Nobody has established they prevent or treat psychosis in this window. For psychosis, the proven tool is thirteen years old: lithium.

So the intervention starts in the second trimester, with three questions in every intake. Any personal history of bipolar disorder? Any prior postpartum psychosis? Any family history of either? A yes to any of the three buys a delivery plan with psychiatry attached, sleep protection on the unit, and a prophylaxis conversation before discharge.

We hold a treatment sequence that returns nearly every one of these women to themselves. Its anchor drug has spent twenty years in retreat. The courtroom must decide what she knew. The clinic must answer for what we know. Only one of those verdicts is ours to deliver.

If you or someone you care about is struggling during pregnancy or after childbirth, the National Maternal Mental Health Hotline is there around the clock: 1-833-TLC-MAMA. If you're in crisis, call or text 988.