Background and importance Readmission of elderly patients is an issue of concern for both healthcare professionals and health authorities. The observed rate of unscheduled 30 day readmission is up to 14% in patients aged 75 years or over. Moreover, the proportion of readmissions deemed avoidable is estimated at 23%. In elderly patients, falls are frequent and can lead to consultations in the emergency department (ED) or even hospitalisation. The proportion of people hospitalised after visiting the ED for a fall increases with age: from 25% at 65 years to almost half at 90 years. At the end of the index hospital stay, readmission of older fallers are thus challenging for the healthcare system.
Aim and objectives To describe older patients hospitalised for falls and identify the risk of readmission in that population.
Material and methods We conducted an observational, single centre, prospective study (from April to June 2019). Inclusion criteria were: patients aged 75 and over, admitted to the ED for falls and consenting to the study. For patients subsequently hospitalised, geriatric scores were determined (risk of readmission (ISAR score), state of frailty, degree of autonomy (Katz score)), and when appropriate, medication treatments were listed and compliance of patients was assessed (Girerd score).
Results During this 3 month study, 154 patients were included (median age 86 years (min 75–max 103), sex ratio 0.44), of whom 73 patients were hospitalised. Among these patients, 63% presented a high risk of readmission; 45.2% were at risk of frailty; 72.6% were dependent. Finally, 53 of the 73 patients (72.6%) had medications in the primary care setting and presented a 71.7% non-compliance or low compliance rate. 58 patients (79.5%) had at least one drug that can cause falls (min 1–max 7).
Conclusion and relevance Older patients presenting at hospital with a fall were often likely to become frail and the majority were dependent. More importantly, this population was at high risk of readmission. Therefore, future studies are now needed to test interventions aimed at reducing this risk.
Conflict of interest No conflict of interest