3753 That’s not my dad: improving delirium care at Leeds teaching hospitals

Abstract Introduction Delirium affects 1 in 4 hospitalised older adults, doubling length of stay and increasing mortality sixfold. Yet, over half of cases go undetected. In 2019, delirium screening compliance at Leeds Teaching Hospitals NHS Trust was under 5%. No standardised care plan for delirium prevention or management existed. We aimed to embed a trust-wide, digital, person-centred strategy to detect and manage delirium in patients aged 65 across all departments. Methods Using the IHI model for improvement, we launched a strategy in March 2023: Spot it, Sort it, Stop it. A 4AT assessment was embedded into our electronic record (PPM+), with a score 0 triggering a mandatory nursing care plan based on the ‘PINCH ME’ framework (Pain, Infection, Nutrition, Constipation, Hydration, Medications/Mobility, Environment). A live run chart tracked 4AT compliance and care plan use by department. Campaigns included ward-based education, short films, podcasts, posters, and governance engagement. A driver diagram describes our multiple targets for improvement and multiple PDSA cycles, Surveys evaluated staff confidence and knowledge. Results 4AT completion rose from 3.6% to 70.4%, with over 1800 patients assessed monthly. Care plan compliance reached 60–90% across departments. Staff confidence in recognising delirium improved from 47% to 88%, and confidence in managing delirium rose from 43% to 86%. Falls reduced from 3.8 to 3.5 per 1000 bed days, and length of stay decreased by 0.5 days. Staff increasingly recognised delirium as a medical emergency, using 4AT scores as a communication tool. Conclusion Our trust-wide strategy significantly improved delirium detection, staff confidence, and care quality for older adults. Sustained cultural and clinical change was supported by education, data visibility, and MDT engagement. Future plans include real-time ward dashboards, revised care plans, and continued rollout through our frailty education programme.