A. Kasi Vishwanadham

The earlier reading: whenever pain appeared it seemed to track the delusion — he attributes it to the electric fields, it eases when he shouts at the hallucinated attacker, and it vanishes on well days — which looked like part of the psychotic layer rather than tissue pain, making orthopaedic work-up low-yield.

The 10 Jul reframe qualifies this. The pain that actually prevents him getting up is movement-evoked, located below the buttocks, and blocks him even when there is no psychosis present. That is behaving like a physical/neurological pain gate, not purely a psychotic symptom — so a structural / neurological cause is back on the table for that pain, and an in-person exam is warranted. The two may be different things (a delusion-linked pain during florid phases vs. a movement-evoked barrier pain), or the "tracks the delusion" claim was over-stated. This is genuinely unresolved and is exactly what a hands-on exam during a bad episode should clarify.

1. In-person exam during a bad episode — the priority for the immobility: does passively moving his leg reproduce the pain? Hip vs nerve-root vs spine? Settles more than any remote reasoning.   2. Levodopa-timing vs stuck-phase diary — free; tests the "off-state" idea and whether good days track a dose.   3. DAT-SPECT — drug-induced vs Lewy-body/degenerative Parkinsonism; decides whether the levodopa lever or the antipsychotic lever is the main one.   4. One "bad-phase" blood/urine panel — sweeps Tier 2 in a single draw.   5. Autoimmune / syphilis / HIV serology — sweeps Tier 3 if 1–4 are unrevealing.

On the lorazepam challenge: it has dropped down the list. It was premised on the immobility being catatonic; the 10 Jul reframe (pain is the gate, volition intact) makes that unlikely, and lorazepam doesn't treat pain — plus his falls/sedation profile (Beers-listed; neuropathy and balance impairment) argues against it. It becomes relevant again only if an in-person exam actually turns up catatonic signs.

Bottom line: the immobility and the psychosis are now two questions. The exam + levodopa-timing diary address the movement-evoked pain; the pimavanserin trial and antipsychotic strategy address the psychosis. Both are low-cost and can run in parallel over the coming weeks.

The two differentials fit together. The 11 Jul list identifies the proximate pain generator (most likely hip / pelvic / gluteal or dynamic nerve-root, with a sacral insufficiency fracture as the must-exclude). The 10 Jul list still contributes the one thing the table doesn't fully resolve — the abrupt, complete day-to-day switching, best explained by a fluctuating state gating the pain on and off (levodopa off-state / dystonic cycling), which is the cheapest, zero-risk thing to test.

Priority order:  1. In-person exam during a bad episode (hip vs root vs spine).  2. Pelvis/hip X-ray — then pelvic MRI if suspicion persists — to exclude a sacral insufficiency fracture the lumbar MRI can't see.  3. Focused labs: CK, vitamin D, Ca/PO₄, ALP, PTH; MMA + homocysteine (the >2000 B12 was post-injection and uninterpretable); thyroid, renal.  4. Levodopa-timing diary — free; tests the off-state contribution.  5. Drop the lorazepam challenge as a near-term step.

The single most useful reframing across both documents: stop treating this as one "stuck" syndrome. Separate the movement-evoked pain (peripheral / mechanical — needs an exam and a pelvis X-ray) from the psychotic overlay (a separate track). That separation, plus the sacral-fracture and hip/gluteal additions, is where the 11 Jul differential improves on the earlier tiered list.

Reproducibility alone doesn't discriminate — it's consistent with two different explanations that predict the identical result:

FND-supportive reading: an alternate, less-automatic motor strategy (or external facilitation) unlocking a blocked pathway is a recognized positive sign of functional neurological disorder (cf. functional gait improving with backward walking or dancing).

Mechanical reading, equally strong: the prone push-up avoids the specific hip-flexion/loading angle a normal sit-up or stand-up transfer requires, and the assisted stand changes the load path through the hips/spine. A load-specific pain generator — hip, nerve root, or a sacral/pelvic source — predicts exactly the same reliable relief from exactly this maneuver.

What would actually discriminate (not yet tested): pain reproduced by passive movement of the hip/leg and focal tenderness on palpation over the sacrum/sit-bones/hip both point mechanical; a positive straight-leg-raise or FABER/FADIR points mechanical; relief generalising to any distraction or novel task, not just this one maneuver, points FND.

This is the first confirmed observation of the abrupt bad→good switch being triggered, rather than arriving spontaneously as on prior good days — new and useful regardless of cause. It does not rule out the physical differential (dynamic nerve-root pain, hip/pelvic/gluteal source, sacral insufficiency fracture, levodopa off-state/dystonic pain) — a mechanical, angle-specific pain generator predicts the same relief pattern seen today.

Not yet captured, worth getting next time: the time of his last levodopa dose relative to when the pain broke (tests the off-state theory at zero cost), whether relief came from the push-up itself vs. the assisted stand vs. simply elapsed time, and a short video of the sequence for a specialist to review later.

Action: log this as one voice on the differential, not a redirect of the workup. The in-person exam still needed is the same one called for since 10–11 Jul — passive hip/leg movement, straight-leg raise, focal palpation, and a pelvis/hip X-ray to exclude a sacral insufficiency fracture — ideally examining specifically for these signs during a bad phase.

Earlier drafts made this the headline test and the first escalation, on the assumption the immobility was catatonic. That assumption is now doubted: the barrier is pain with volition intact, and lorazepam does not treat pain. It also carries real risk here (Beers-listed; boxed warning if combined with a gabapentinoid — relevant again if gabapentin is restarted; falls risk from his neuropathy and balance impairment). It returns to the table only if an in-person exam actually turns up catatonic signs.