Anticholinergic prescribing in hospitalised older adults: an uncontrolled before-and-after study of clinician education

Background Anticholinergic medications are associated with medication-related harm in older adults. Hospitalisation provides an opportunity to review prescribing, but evidence about clinician education alone remains uncertain. Objective To compare anticholinergic prescribing among hospitalised older adults during periods before and after clinician education. Methods This exploratory, single-site, uncontrolled retrospective before-and-after study included unique patients aged ≥65 years admitted for ≥24 h with medications documented at admission and discharge. Patient-level data were obtained for two 3-month periods before and after education was delivered to doctors and pharmacists. Outcomes were admission-to-discharge changes in Anticholinergic Cognitive Burden (ACB) score and number of anticholinergic medications. Linear regression estimated the adjusted mean between-period difference in discharge ACB. Results The study included 686 patients (368 pre-intervention; 318 post-intervention). Median ACB score decreased from admission to discharge in both periods (pre: 2 [IQR 1−4] to 1 [IQR 0–3], p = 0.002; post: 2 [IQR 1–3] to 1 [IQR 0–3], p = 0.004). The median number of anticholinergic medications remained 1 in both periods, although paired distributions shifted modestly downward (pre: p = 0.012; post: p = 0.005). The unadjusted mean difference in ACB change between periods was 0.025 points (bootstrap 95% CI −0.18 to 0.23); the Wilcoxon rank-sum p-value was 0.941. In the adjusted model, the post-intervention period was associated with a −0.064-point difference in discharge ACB (95% CI −0.265 to 0.137; p = 0.530). The between-period mean difference in anticholinergic-medication count change was −0.001 medications (bootstrap 95% CI −0.16 to 0.16); the Wilcoxon rank-sum p-value was 0.86. Conclusion Prescribed ACB decreased modestly and anticholinergic-medication count distributions shifted downward from admission to discharge in both periods, without evidence of additional adjusted between-period improvement. The uncontrolled design and incomplete implementation information preclude conclusions about the education program’s effectiveness. The findings can inform prospective evaluation of multimodal strategies.