Day-1 post-hemiarthroplasty confusion — screen, bundle, drugs and prognosis

Model answer

(A) Hypoactive delirium until proven otherwise, high-risk by rule and cohort, HELP by the numbers (4 marks). Quiet inattention with disorientation on day 1 is hypoactive delirium — the predominant PACU phenotype (64.2% of 19.2% affected) — not baseline dementia, and gestalt cannot exclude it, so screen now with 4AT (ICU AUC 0.879; sensitivity 74.0%, specificity 95.4%, matching CAM-ICU) rather than guessing.^[18][11][14] She sits in Marcantonio's high stratum: age over 70, poor cognition and function, orthopaedic surgery in an 82-year-old — the rule's 50% group — with the cohort multipliers stacked (dementia odds 3.42, frailty 2.39, delayed surgery 1.74, malnutrition 2.16, high 4AT 3.04).^[7][16][19] The HELP bundle targets six factors — cognitive impairment, sleep deprivation, immobility, visual and hearing impairment, dehydration — cutting delirium from 15.0% to 9.9% (odds 0.60) in the original trial and to odds 0.47 across its meta-analysis, with falls and costs down too.^[1][2]

(B) Orders, drugs, and the anaesthetic honesty (3 marks). Write delirium-friendly orders executable by ward nurses — orient, mobilise, hydrate, feed, protect sleep, remove the catheter, invite family to the bedside — the order set that cut hip-fracture delirium from 51% to 33% (strongest with dementia) with flexible visitation cutting orthopaedic-ICU delirium from 44.4% to 29.5%.^[29][30] Review morphine (postoperative opioids nearly double delirium odds; meperidine triples them — never substitute it) and every anticholinergic (burden erases even haloperidol's hint of help) — multimodal analgesia with fascia-iliaca blockade replaces dose escalation.^[42][43] Her spinal did not cause this: ten randomised trials found neuraxial versus general delirium odds 1.10 — no difference — so technique follows indication and prevention follows the patient.^[52]

(C) No routine antipsychotic, honest prognosis, booked follow-up (3 marks). Neither prevent nor treat with antipsychotics by default: haloperidol equals placebo for prevention at high strength of evidence (and failed in thoracic surgery, 22.1% versus 28.4%), while MIND-USA showed haloperidol and ziprasidone identical to placebo for shortening established delirium (p=.26) — reserve them for dangerous agitation only, lowest dose, shortest course.^[35][36][57] Counsel with numbers: 30-day mortality odds 3.2 with lost independence, 1-year dementia odds 13.9 with mortality tripled — then book cognitive follow-up before discharge, because delirium is a prognosis, not an episode.^[17][61]