Emergency Department Visits Among COPD Patients: An Example of Repeat Single and Recurrent-event Analysis

• Activity Number: 83
• Type: Contributed
• Date/Time: Monday, August 9, 2004 : 8:30 AM to 10:20 AM
• Sponsor: Section on Health Policy Statistics
• Author(s): Jianglin Xu*+ and Todd Lee and Whitney Witt and Ramon A. Durazo-Arvizu
• Companies: Northwestern University and Northwestern University and Northwestern University and Northwestern University
• Keywords: COPD ; ED ; VA ; HR

Abstract

There is increasing interest in applying survival analysis techniques to datasets with recurrent events in health services research. The objective of this analysis was to evaluate the association between follow-up office visits and repeat ED visits using single and recurrent event methods. A cohort of VA patients with a COPD-related ED visit was identified (n=27,416). Cox proportional hazard and Anderson-Gill models were used to compare the risk for repeat ED visits within 90 days of the index ED visit between patients with and without a follow-up physician visit within 30 days. Models were adjusted for age, gender, and health care utilization. When only the first repeat ED visit was considered (3,845 ED visits), patients with a follow-up office visit were at higher risk (HR=1.53; 95% CI, 1.42-1.63). When all repeat ED visits were considered (5,364 ED visits), the HR was 1.49 (95% CI, 1.40-1.58). The recurrent event model used more information; however, results were similar between models. Among COPD patients with an ED visit, those with a follow-up office visit within 30 days were more likely to have a repeat ED visit within 90 days.