Johns Hop­kins uses elec­tronic tool to im­prove triage of ED pa­tients

Modern Healthcare - - BEST PRACTICES - By Maria Castel­lucci

Triage nurses in the emer­gency depart­ment at Johns Hop­kins Hospi­tal in Bal­ti­more face a hard task. Just like in EDs across the coun­try, they must de­cide in just a few min­utes how crit­i­cal a pa­tient’s con­di­tion is and as­sign them a score that will de­ter­mine how quickly they are treated.

The nurses use the Emer­gency Sever­ity In­dex to help make their de­ci­sion. The ESI, a tool used widely in EDs across the U.S., is a way for care­givers to iden­tify pa­tients’ con­di­tions by as­sign­ing them to one of five groups, or lev­els. Level 1 in­di­cates the pa­tient needs im­me­di­ate at­ten­tion and is ex­pe­ri­enc­ing some­thing along the lines of car­diac ar­rest, while Level 5 means their needs aren’t ur­gent—a rash, for ex­am­ple.

That’s im­por­tant, ob­vi­ously, be­cause it gets the pa­tient the right treat­ment more quickly. The prob­lem is the ESI isn’t al­ways right and re­lies heav­ily on nurses’ sub­jec­tiv­ity, said Scott Levin, as­so­ci­ate pro­fes­sor of emer­gency medicine at the Johns Hop­kins Univer­sity School of Medicine.

Re­search on the ESI shows that about 70% of pa­tients are lumped into the medium cat­e­gory—Level 3—even though there can be wide vari­ance in the sever­ity of their symp­toms and ul­ti­mate di­ag­noses. “The ma­jor chal­lenge of the ESI is that it’s com­pletely sub­jec­tive,” Levin said. “When some­thing is com­pletely sub­jec­tive, there can be un­to­ward vari­abil­ity.”

In an at­tempt to make the triage process more ob­jec­tive, Levin and his col­leagues de­vel­oped an elec­tronic tool last year that is now used by triage nurses at Johns Hop­kins Hospi­tal.

The tool uses an al­go­rithm based on data from roughly 200,000 pa­tients treated at the six hospi­tals in the Johns Hop­kins sys­tem to pre­dict a pa­tient’s sever­ity of ill­ness. It takes into ac­count how pa­tients with the same symp­toms were treated and what their like­li­hood was for dy­ing, be­ing ad­mit­ted to the in­ten­sive-care unit or need­ing an emer­gency pro­ce­dure. The tool then as­signs the pa­tient a level score us­ing the ESI.

Nurses have been us­ing the tool since last De­cem­ber and find it’s help­ful to guide their clin­i­cal de­ci­sion­mak­ing.

But it took some time for staff to warm up to the tool, ad­mits Sophia Henry, a triage nurse in the Johns Hop­kins ED. Henry said she and other nurses were ini­tially wor­ried that the tool would take away their au­ton­omy or that they would be “re­placed by a com­puter.”

“We were very re­sis­tant at first be­cause for us, be­ing trained as a triage nurse is an honor. It shows clin­i­cal ex­cel­lence and that you un­der­stand clin­i­cal de­ci­sion­mak­ing,” she said.

Levin said he spent months with the triage nurses to en­sure they un­der­stood that the tool was not to re­place their clin­i­cal judg­ment but merely to sup­port them in their work. Levin said he tells nurses they should dis­agree with the tool when they think it is ap­pro­pri­ate. Af­ter all, the tool can’t in­ter­act with pa­tients the way a nurse can.

Levin, how­ever, said he’s con­fi­dent in the tool’s re­sults be­cause it’s tar­geted at Johns Hop­kins’ unique pa­tient pop­u­la­tion. He said health­care tools that rely on al­go­rithms aren’t usu­ally widely adopted be­cause providers don’t trust the data. But care­givers trust the ac­cu­racy of data used to de­velop this tool’s al­go­rithm.

The e-triage tool “is more mean­ing­ful to the peo­ple who are us­ing it,” Levin said. “Ev­ery ED is so dif­fer­ent— the pa­tient pop­u­la­tions they treat, the re­sources they bear and the care pro­cesses they use.”

The tool has been shown to work ef­fec­tively. A re­cent study led by Levin pub­lished in the An­nals of Emer­gency Medicine found that the tool iden­ti­fied 14,000 pa­tients, or 10%, triaged to ESI Level 3 who should have been cat­e­go­rized as a Level 1 or 2. The tool also in­creased the num­ber of pa­tients as­signed lower pri­or­ity lev­els like Lev­els 4 or 5.

Iden­ti­fy­ing pa­tients with less se­ri­ous con­di­tions sooner de­creases the time they have to spend in the ED, Levin said. EDs of­ten “fast track” pa­tients who aren’t ill enough to re­quire pro­longed care.

“If we put th­ese pa­tients in line with the very sick, they would never get out,” he said. “The hope is to not have them wait and get out quickly.”

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.