Long-Acting Injectable Antipsychotics in First-Episode Psychosis: Early Use vs. Reserved for Demonstrated Non-Adherence
A newly diagnosed patient with real insight into his own illness still has a family history of stopping antipsychotics against medical advice. Does his LAI conversation start now, or only after he actually misses doses?
D.K., a 21-year-old man, moved out of his family's house two years ago to live independently for the first time, taking pride, his discharge nurse notes, in managing his own bills and groceries even on an unpredictable warehouse-shift income. He was hospitalized three weeks ago for a first psychotic break marked by auditory hallucinations and paranoid ideation, responded well to oral risperidone, and is being discharged today with good insight into his diagnosis and a stated intention to keep taking his medication.
He lives alone in a studio apartment two bus transfers from the clinic, works irregular shifts at a warehouse that make a fixed appointment schedule genuinely difficult, and mentions, when asked directly about family history, that his older brother was diagnosed with schizophrenia at nineteen and stopped his own medication twice in the years since, both times followed by hospitalization — a history D.K. brings up unprompted, saying he doesn't want the same thing to happen to him.
Traditional practice reserves long-acting injectable antipsychotics for patients who have already demonstrated non-adherence on oral therapy — effectively, LAIs as a second-line response to a documented failure. A growing body of first-episode-specific evidence has pushed against that sequencing: relapse after a first episode carries real, measurable risk to long-term trajectory, oral non-adherence in this population is common and frequently invisible to the treatment team until a relapse has already occurred, and LAIs remove the day-to-day adherence burden before that failure happens rather than after. D.K. himself is not demonstrating non-adherence right now — his brother's history and his own logistical barriers are the argument for considering an LAI anyway, before either becomes the reason he's back in the hospital.
LAI conversation now, or wait
He has good insight and no adherence problem to point to yet. Raising the LAI conversation now risks him hearing "you don't trust me" at the exact moment his engagement with treatment is strongest — I'd rather watch how the first several weeks on oral risperidone actually go before introducing that.
That risk is real — I'm not going to pretend raising an injectable this early costs nothing with a patient this engaged. But I'd put a number next to what waiting costs, too. Subotnik and colleagues' randomized trial, published in JAMA Psychiatry, is the one first-episode-specific study we actually have comparing long-acting risperidone to oral risperidone, and it found a relapse or symptom-exacerbation rate of five percent on the injectable against thirty-three percent on oral, over twelve months. In plain terms: roughly one in three patients on the oral arm had a real setback within the year, versus about one in twenty on the injectable.
You said you'd rather wait and see how the first few weeks actually go — but the reason that trial matters is that non-adherence in this population is usually invisible until it's already produced the relapse we're trying to avoid. "Waiting to see" is functionally the same strategy the trial tested against, and it lost.
I don't think that number settles who decides, though. It's a real, striking difference, but it comes from a single-site trial of eighty-six patients who'd already agreed to be randomized to an injectable — not obviously the same population as a patient with strong insight and no demonstrated adherence problem yet. What it does settle, for me, is that this isn't a case for defaulting to either the traditional wait-for-failure sequencing or the newer evidence on his behalf. He's told us directly he doesn't want his brother's outcome, he has the insight to weigh a real relapse-rate difference against an injection-site or stigma concern himself, and his own commute and shift schedule are exactly the kind of quiet adherence barrier this evidence describes. I'd offer it honestly, numbers included, and let him decide.
Agreed: D.K. was given a full, honest explanation of both LAI and continued-oral pathways, including his brother's history as part of the reasoning, and asked to decide without a deadline forced onto the conversation.
Transition planned with an oral overlap; first injection timed before hospital discharge if logistics allow.
Adherence checked explicitly at each visit, with the LAI conversation revisited immediately at any sign of a missed dose, not after a second one.