A Long-Acting Injectable Behind Bars: Whose Decision Is It
The team agrees the injectable is the right clinical call. What they don't agree on is whether recommending it and conditioning his release on it are the same decision wearing different words.
D.W., a 34-year-old man, is eleven days from his third release in eighteen months, and everyone on his treatment team, himself included, already knows roughly how this one goes if nothing changes. Diagnosed with schizophrenia at 24, he was stable and working a warehouse job for nearly two years before his first arrest — a disorderly-conduct charge during a florid relapse that followed his decision, some months earlier, to stop the oral risperidone a community clinic had started him on. The two cycles since have run almost identically: released, stable for roughly a month, medication stopped, symptoms return, a new charge within ten weeks. He is not oppositional about the diagnosis or the medication in the abstract — inside, on a supervised schedule, his adherence has been essentially perfect every time. The pattern breaks specifically at the point of release, not before it.
His own account of why is consistent across both post-release relapses: once he's out, without someone checking, the daily pill becomes the first thing that slips when housing, probation check-ins, and finding work all compete for the same mornings — not a delusional refusal of treatment, an ordinary adherence failure of exactly the kind daily oral dosing is most vulnerable to in an unstructured transition. His symptoms this admission — auditory hallucinations, disorganized speech, a two-week gap in memory for the days immediately preceding his arrest — have resolved fully on the same oral risperidone dose that has worked for him twice before. The clinical question in front of the team isn't which drug controls his illness; that's settled. It's whether an injectable formulation of the same drug, given before release rather than after, is the way to break a cycle that oral dosing alone has now failed to break twice in a row.
Discharge planning, eleven days out
I want to offer him the injectable, strongly, as the clinical recommendation — not as a condition of anything. The PRIDE trial randomized 444 patients with schizophrenia and a history of incarceration — his population, not a general outpatient sample — to paliperidone palmitate or one of seven daily oral antipsychotics. Its primary endpoint was time to first treatment failure, a composite counting arrest or incarceration, psychiatric hospitalization, and discontinuation for inadequate efficacy or tolerability; the injectable delayed that failure significantly, hazard ratio 1.43. Arrest and hospitalization were the two commonest components on both arms. I'd flag that it was open-label, so I'm not calling it airtight — but a monthly injection removes the daily decision point where his pattern keeps breaking down, and that's a genuinely stronger recommendation than "keep taking the pill" has turned out to be, twice.
I don't dispute the trial data, and I'm not questioning your clinical judgment about which drug helps him most. What I'm watching is the room this conversation is happening in — the same institution making the treatment recommendation is the one whose report goes to the parole board eleven days from now. "Strongly recommended" from the person who also writes your release evaluation doesn't land the same way it would from an outpatient clinic with no stake in his liberty date.
I know the offer as written doesn't condition anything on his accepting it — but a man in his position, weighing a recommendation against his release timeline, may not experience that distinction the way we intend it, and I think that's worth naming directly to him rather than assuming good phrasing settles it.
I think the disagreement is actually smaller than it sounds, because you're answering two different questions with the same word. The adherence and reoffending data — PRIDE, and the Chang et al. 2016 JAMA cohort of over 22,000 released prisoners, which found lower violent reoffending during periods dispensed antipsychotics — genuinely supports a strong clinical recommendation. It says nothing at all about the separate legal question of whether accepting it could ever be made a formal release condition. Those are different claims resting on the same numbers.
My own answer: recommend it as strongly as the psychiatrist proposes, document the reasoning plainly for the parole board, and never let the report imply his release is contingent on his answer — recommend, don't condition, and say so explicitly in what goes in front of the board.
Agreed: the LAI is offered as a documented clinical recommendation, both initiation doses given before release with the oral risperidone stopped at the first injection, and the parole report states plainly that release is not conditioned on his acceptance.
Not agreed: whether the report should also name the relapse pattern itself as a factor in the board's own separate housing-stability determination, since unstable post-release housing is independently in his file. The psychologist argued that doing so risks exactly the coercive blending she flagged, now working through a different door; the psychiatrist thought withholding a genuinely relevant clinical fact from a board making its own release decision was its own kind of distortion. He accepted the injection before the disagreement was resolved — a decision the team recorded as his own, made with the record showing he was told plainly it changed nothing about the board's timeline.