Document Type


Publication Date



Recurrent utilization of emergency medical services by certain patient populations is a challenge at the University of New Mexico Emergency Department (UNM ED) and at other institutions. Data suggest that a significant percentage of patients treated in the UNM ED could be treated in other, less costly outpatient settings. Existing data were used to analyze trends in UNM ED utilization among four distinct populations — managed uninsured patients (UNM Care), unmanaged uninsured patients (Self-pay), managed Medicaid patients (Salud), and unmanaged Medicaid patients (Medicaid). We hypothesized that those who previously had high usage of the Emergency Department (ED) in the index year would continue high use of the ED in the next year. In addition, we hypothesized that patients in managed programs, UNM Care and Salud, would likely have a lower recurrent use of the ED than patients in unmanaged programs, Self-pay and Medicaid, patients. Our data included 19,461 adult patients, with 1,104 having \u22653 visits in Fiscal Year 2004 (FY04). Of those patients, 145 (13.1%) patients had \u22653 visits in Fiscal Year 2005 (FY05). Unmanaged Medicaid patients who had \u22653 visits in FY04 were 17.80 times as likely to have \u22653 visits in FY05 (p<0.001) compared to unmanaged Medicaid patients with 1-2 visits in FY04. The next highest recurrent high utilization pattern was the Self-pay group with \u22653 visits in FY04, who were 9.37 times as likely to have \u22653 visits in 2005 as the Self-pay patients with 1-2 visits in FY04. The managed groups also showed that patients with high utilization in the index year continued high use in the next year, but to a lesser extent than the unmanaged groups. Salud patients with \u22653 visits in 2004 were 6.69 times more likely to have \u22653 visits in 2005, while the UNM Care group shows significantly less recurrent ED use at 4.28 times as likely to have \u22653 visits in 2005, compared to patients from the respective groups with 1-2 visits in FY04, which is also lower than the other three groups (p<0.05).