Repeat patients to the emergency department in a Statewide database

Objectives: To describe the epidemiology of repeat users of the emergency department (ED) using a statewide database. Methods: Probabilistic linkage was used to convert three years of statewide ED visit data into a longitudinal, patient-based data set. Patients were classified as single, repeat (at least two visits within three years), or serial (four or more visits within a 365-day period) users of the ED. Serial patients were further stratified by the number of EDs attended. Descriptive statistics were used to assess differences between patient types. Results: There were 1,370,607 separate visits associated with 780,074 patients from 1996 to 1998. While repeat and serial patients represented 33% of the patients, they accounted for 62% of the ED visits during the study period. Repeat and serial patients were younger and had smaller median ED charges per visit than single-use patients. Serial patients attending five or more EDs were more likely to be coded as self-pay than other serial patients. Diagnosis codes relating to sprains, back problems, and headaches were prevalent among serial patients who visited five or more EDs. Approximately 30% of serial patients during the first year remained serial patients in the second year. Conclusions: Due to the high turnover in serial patients, control groups in future studies are necessary to evaluate interventions aimed at decreasing serial ED use. The likelihood of serial ED users to use multiple EDs indicates that those studying serial ED use should collect data from multiple EDs.