It is widely believed that emergency departments are being used for care that could better be provided in another, less expensive setting. Two Oregon Health & Science University researchers looked at this issue, one reviewing ED usage by Medicaid enrollees, the other reviewing 'nonemergency' ED visits. Their findings shed light on some commonly held misperceptions.
Dan Handel, M.D., M.P.H., found use of the emergency department by Oregon Medicaid enrollees to be much less than suspected. Anecdotal evidence had suggested use of the ED to be a substantial driver of Medicaid costs; however, Handel's study found ED usage accounted for only 6.8 percent of Medicaid spending.
Handel reviewed data from 589,903 individuals enrolled in Oregon's Medicaid program during 2002. He reviewed expenditures for hospital, physician and ancillary services associated with any ED visit not resulting in an inpatient admission. Results showed ED expenditures averaged $12.17 per Medicaid enrollee per month.
A small group of Medicaid enrollees accounted for a large portion of spending – 50 percent of all ED expenditures were attributed to 3 percent of enrollees.
ED expenses by Medicaid enrollees actually make up a small percentage of Medicaid spending. Assuming a very aggressive and successful policy to reduce ED expenditures by 25 percent, as has been previously suggested, Medicaid expenditures would be cut by less than 2 percent a year.
"A better plan of action would be to target the small group of frequent users to determine their barriers to accessing primary care," said Handel, a fellow in the OHSU Center for Policy and Research in Emergency Medicine.
Developers of a research tool called the Emergency Department Algorithm (EDA) claim the tool can evaluate the medical safety net because patients with worse access to care will use EDs for less urgent conditions. The algorithm assigns probabilities that each ED visit falls into each of four categories: nonemergency; primary care treatable emergency; preventable emergency; and non-preventable emergency. This algorithm has been widely adopted by health care policy-makers.
OHSU researchers evaluated whether the algorithm could detect changes in ED usage following cutbacks to the Oregon Health Plan in 2003 that affected access to care. Researchers reviewed all visits to 21 Oregon emergency departments during 2002 and 2004, comparing mean probabilities that ED visits in each payer category fell into each of the four categories before and after the Oregon Health Plan cuts.
Using the EDA, the largest change in mean probabilities was only 2 percent.
"What this shows us is that the EDA was less useful in demonstrating changes in access to care than were other, simpler measures," said Robert Lowe, M.D., M.P.H., director of the OHSU Center for Policy Research in Emergency Medicine. "While this is a promising methodology, it needs further refinement before being adopted widely."
Both Handel and Lowe will be presenting their research at the Society for Academic Emergency Medicine annual meeting in Chicago May 16-18.