Abstract
Background: Frequent presenters to emergency departments (EDs) pose many challenges around care delivery and health service management. The aim of this study was to investigate the presentation patterns of people with 5 or more ED visits in any calendar month (5+ frequent presenter [FP5+]) to develop a useful methodological framework on which the real impact of interventions may be assessed. Methods: This study is a retrospective analysis of de-identified frequent ED presentation data using segmented regression analysis of an interrupted time series (ITS). Results: A total of 82 FP5+ to this single ED were identified in a year. Of these presenters, 77% had 10 or more presentations in a year. The total FP5+ presentations in the 12 months preceding and after each participant's ≥5 presentations in 1 month (the trigger month for inclusion in the study) accounted for 1,064 and 1,606 visits, respectively. ITS analysis of frequent ED presentations did not show a significant level change or trend change during the data collection period. Monthly review of people who frequently present to a single ED showed that presentations typically occurred in bouts that may span calendar years. Presentation bouts then typically slow, potentially distorting evaluation of the effects of interventions. Conclusion: Rolling monthly examination of presentation data may facilitate timely case review and care delivery, as well as provide a holistic picture of the impacts of interventions targeting patient care needs. This unique analysis demands a reconsideration of the typical before-and-after analysis of interventions for this vulnerable and high-cost group of patients.