A High-Dose Steroid Pulse and an Assault Charge: Whose Act Was It
The dose-response data make the psychosis itself easy to believe. What the same data can't settle is whether the specific person he hurt was chosen by the delusion, or by something closer to him.
T.A., a 39-year-old man, has no psychiatric history of any kind and no history of violence toward his wife of twelve years or anyone else — the closest thing in his record is a single traffic citation from 2019. He was admitted just over two weeks ago for a lupus nephritis flare severe enough that his rheumatologist started him on three days of IV methylprednisolone pulse therapy, 1,000mg daily, transitioning to oral prednisone 60mg on day four — an aggressive but standard, guideline-consistent regimen for a flare threatening his kidney function. By day three of the pulse he was sleeping two or three hours a night and had begun telling his wife, calmly at first, that hospital staff were altering his chart to hide a diagnosis they didn't want him to know about. By day five he was floridly delusional, accused his wife specifically of being part of the concealment, and struck her twice when she tried to call his physician — the only violent act in either of their histories.
Steroids were stopped that day. He was started on a brief course of olanzapine, and by day twelve — seven days off the corticosteroid entirely — his psychosis had resolved fully, and by his own account afterward he has no clear memory of believing any of it, only a confused, distressed recollection of feeling that something was deeply wrong that he could not get anyone to take seriously. His wife, though shaken, does not believe the man who hit her is the man she has been married to for twelve years, and has said so directly to the evaluating team. The assault charge is proceeding regardless, and defense counsel has asked for a forensic evaluation of whether his mental state at the time supports an involuntary-intoxication defense — a doctrine ordinarily built around drugs or alcohol someone didn't knowingly or willingly take, being asked here to reach a medication he took exactly as a physician directed.
The forensic evaluation, drafting the report
The dose-response literature here is unusually strong for a psychiatric causation question. The Boston Collaborative Drug Surveillance Program's classic data found psychiatric disturbance in 18.4 percent of patients above 80mg/day prednisone-equivalent, 4.6 percent between 41 and 80mg, and 1.3 percent at or below 40mg — a real dose-response curve, not an anecdotal association, and he sat well above that top threshold during the pulse. Onset within days of a high-dose course, full resolution within days of stopping it, zero psychiatric or violence history before or since. Piel's 2015 appellate case review of the involuntary-intoxication-by-prescribed-medication defense doesn't grade fact patterns, but it sets out what the defense actually turns on: that the defendant did not know of the drug's potential to produce this effect when he took it. He had no prior exposure and no warning that a lupus regimen could do this, which is the element the case law asks about.
I want to be careful about one thing before this goes further — I prescribed that pulse because his renal function was declining fast enough that under-treating the flare carried its own serious risk. This wasn't an excessive or careless dose; it's the standard guideline-consistent regimen for lupus nephritis at his severity. I don't think that undercuts the psychiatric argument, but I want the record to reflect that "involuntary" here means he didn't choose the psychiatric effect, not that the medical decision behind it was questionable.
I'd also push back gently on treating this as settled just because the dose-response numbers are real. Your top figure rests on seven events in thirty-eight patients — a genuine gradient, and also a small enough denominator that I'd be careful calling it precise. Those percentages describe a population risk, not a guarantee that his particular reaction on his particular day happened the way we're now reconstructing it after the fact.
I agree with both of you on the diagnosis and the dose-response argument — I want to name a gap neither of you has addressed yet. The chart shows diffuse symptoms in the days before: insomnia, non-specific paranoia about staff. What actually happened on day five was a targeted act against one named person. Global causation — the steroid caused a psychotic state — is well supported. Specific causation — that THIS act, aimed at his wife specifically rather than at staff or at no one, was itself a direct product of the psychosis rather than something closer to his own reasoning, however impaired by it — is the question the court actually needs answered, and it's thinner in this record than the diagnosis itself.
His wife's own account — that the delusion named her specifically as complicit in the "concealment" — actually does connect the two, and I think that detail, not the dose-response curve alone, is what should carry the causation argument in the final report.
Agreed: the report will document the dose-response and temporal evidence for corticosteroid-induced psychosis in full, name the specific delusional content connecting his wife to the perceived conspiracy, and state plainly that the underlying prescribing decision was medically appropriate — leaving the ultimate legal characterization of "involuntary" to the court.
Not agreed: how strongly the report should characterize the causation link between the psychosis and the specific act, given the second psychiatrist's flagged gap. The team recorded both positions rather than resolving them into one voice — the dose-response and diagnostic argument stated with high confidence, the specific-causation argument stated as supported but not conclusively established, leaving the fact-finder to weigh the distinction the two evaluators themselves could not fully close.