Emergency Department Utilization Patterns and Subsequent Prescription Drug Overdose Death

The primary purpose of this dissertation was to understand if emergency department utilization patterns are associated with subsequent drug overdose death. Specifically, it was hypothesized that that increasing emergency department (ED) utilization (as measured by such indices as two or more visits in 72-hours, two or more visits in a 30-day period, four or more visits in a 365-day period) was associated with increasing risk of fatal unintentional drug overdose compared with patients without two or more visits in a given time frame. Using ED data from the New York State Department of Health's (NYSDOH) Statewide Planning and Research Cooperative System (SPARCS) for the years 2006-2010 linked with unintentional fatal prescription drug overdose data from death certificates and medical examiner case files from the New York City Department of Health and Mental Hygiene (NYC DOHMH) and the NYSDOH for the years 2006-2010, a retrospective dynamic cohort of ED encounter data was conducted. In the first study, the patient population consisted of 1,755,734 New York State residents who were 18-64 years of age and had selected diagnoses on their entry visit. Extended Cox proportional hazards regression models were conducted to estimate the association of ED utilization patterns and subsequent drug overdose death. Compared to time periods in which patients had no visits within a year, patients who had 3, 4-10, or > 10 visits in a year had elevated risks of prescription drug overdose death after adjustment for demographic characteristics: 3 visits (adjusted hazard ratio (aHR 4.77, 95% CI 3.60, 6.15)), 4 - 10 (aHR 7.39, 95%CI 5.81, 9.41), and > 10 ED (aHR 18.37, 95% CI 13.38, 25.23). ED utilization patterns are strong predictors of subsequent overdose death. Understanding the timing of overdose death in relation to ED utilization is essential to recognizing which patients to target with overdose prevention interventions. Identifying time-periods of increased risk may be used as an indicator for developing prediction tools to classify patients at increased risk for overdose.