Trajectories of Frequent Short-Term Emergency Department Visits Among Older Adults

Objectives Frequent emergency department (ED) use is typically defined over a one-year period, but short-term patterns of use among older adults remain poorly understood. We sought to identify distinct trajectories of ED use over a 90-day period and describe their associated patient and visit characteristics, with the goal of informing ED care for this population. Methods We conducted a retrospective population study in Quebec, Canada, using provincial administrative databases. Patients aged <i>≥</i>65 years with an index ED visit between July 2014 to December 2015 and three or more ED visits in the preceding 90 days were included. Group-based trajectory modeling was used to identify patient groups with distinct trajectories of ED visits; the patient and visit characteristics for each trajectory were compared. Results The 10,741 included patients were divided into two cohorts: those with all prior ED visits without admission (No Admission cohort) and those with at least one prior visit resulting in hospital admission (Admission cohort). In both cohorts, two conceptually similar patterns emerged - <i>Stable</i> (near-constant probability of ED visits) and <i>Increasing</i> (probability rising over time) - although the specific timing and magnitude of changes differed between cohorts. In the No Admission cohort, a third pattern, <i>Hyperacute</i> (rapidly rising and high probability of an ED visit near the index visit), was identified and was associated with shorter ED length of stay and fewer chronic conditions compared with other trajectories. The <i>Increasing</i> and <i>Stable</i> groups showed few differences in patient or visit characteristics apart from certain diagnoses. Conclusions This study demonstrates that older adults with frequent ED use can be characterized by distinct 90-day trajectories, which may represent clinically relevant subgroups. Incorporating trajectory-based approaches may enhance understanding of ED utilization patterns and inform strategies to optimize care delivery for this population.