Emergency room symptoms may not predict health care needs: study
NEW YORK (Reuters Health) - The symptoms people come to the emergency room with may not predict the actual diagnosis they're given when they're released from the hospital, according to a new study.
Researchers found about 6 percent of 35,000 patients who visited ERs in 2009 did not need immediate care, but that could not be predicted when they first got to the emergency department.
"Until we evaluate them in the emergency department, we don't know if their condition requires emergency treatment or not," said Dr. Maria Raven, the study's lead author from the University of California, San Francisco.
The findings may not be just academic. Some state Medicaid programs - government-run health insurance plans for the poor - have proposed discouraging unnecessary ER care by withholding hospital and doctor payments for problems that family doctors can manage.
"The concern is that the policies are based on the discharge diagnosis, and this is only arrived at when the entire emergency department visit already happened," Raven said.
For example, under the policy, a doctor would have to immediately determine if a man with chest pain has merely a simple problem, or something more serious - such as a heart attack. If not, the doctor and hospital run the risk of not being paid by Medicaid when the patient is released.
For the new study, published in JAMA on Tuesday, the researchers applied an algorithm that's popular with policymakers to 34,942 ER visits from 2009. Regulators hope the algorithm can predict which visits require immediate care based on patients' first symptoms.
Overall, 6.3 percent of the ER patients were released from the hospital with a diagnosis that could have been managed by a primary care doctor, according to the algorithm.
However, the researchers found those patients' first symptoms were not unique. In fact, 88.7 percent of all patients first came to the ER with similar complaints.
The researchers write that if a nurse were to turn away patients with those symptoms from the ER, 93 percent would have a condition that needed immediate attention.
"At a minimum, if any insurer is implementing this sort of policy, we'd really encourage them to use our methodology to determine whether they have findings different than ours, because I don't know if they will," Raven said.
WASHINGTON STATE'S EXPERIENCE
Carol Wagner, senior vice president of patient safety at the Washington State Hospital Association in Seattle, told Reuters Health that their state's Medicaid program considered a similar program to cut expenses.
"The problem with that plan is that patients going to the ER don't realize what's wrong with them. They may have pain in their chest and they don't know that it's not indigestion," said Wagner, who was not involved with the new research.
Instead, the state's doctors worked with policymakers to develop a program that streamlined several systems for Medicaid patients, including access to their healthcare records.
The changes are expected to save over $30 million during the fiscal year and already reduced the number of visits for Medicaid patients who made frequent visits to the ER, according to Washington's Health Care Authority.
In an editorial accompanying the new study, Dr. James Adams, from the Northwestern University Feinberg School of Medicine in Chicago, cited Washington's efforts.
"Attention should be redirected away from penalizing patients, physicians, or hospitals when a condition turns out to be minor. Instead, the emphasis should be on integration across sites of care, especially for the most complex and most expensive patients," he wrote.
SOURCE: bit.ly/JOTmp1 JAMA, online March 19, 2013.