825 DEVELOPMENT OF THE ANTICHOLINERGIC MEDICATION INDEX (ACMI)

Abstract Introduction Medications with Anticholinergic (AC) properties, are prescribed to treat a range of conditions. Older people increasingly likely be multiple AC medications, but also more experience unwanted adverse effects, such as falls and delirium. The risks outcomes increase the number potency medications prescribed. aim this study was use prognostic modelling approach develop an Medication Index (ACMI) that identifies patients at high risk medication side effects. Methods model developed using data on aged 65–95 years, registered general practice contributing ‘Connected Bradford’ in 2019. A Time-dependent Cox fitted, hospital admission for delirium or composite outcome age, sex important clinical factors (e.g. dementia, arthritis, urinary incontinence) predictors. Concordance Negalkerke’s R2 derived from five-fold cross-validation were used assess performance. Results There 151,604 included study, whom 47,035 (31.0%) ≥1 during Codeine, Prednisolone, Furosemide Amitriptyline most commonly 7.4%, 4.0%, 3.8% 3.1% these least once 2019, respectively. During 6,078 (4.0%) experienced fall, rate being increased those 2019 (4.8% vs 3.7%; p < 0.001). yielded discrimination statistic 0.86 0.1. Conclusion ACMI shows good discrimination. External validation will soon performed SAIL databank further tool primary care.