Nurses turn to speech-recog­ni­tion soft­ware to speed doc­u­men­ta­tion

Modern Healthcare - - NEWS - By Joseph Conn

Rachel Mosci­cki is on the lead­ing edge of a move­ment for nurses to use speech recog­ni­tion and nat­u­ral-lan­guage-pro­cess­ing tech­nol­ogy to record their clin­i­cal doc­u­men­ta­tion, saving time and op­ti­miz­ing use of scarce staff.

Soft­ware that con­verts the hu­man voice to dig­i­tal text and ex­tracts mean­ing from those words pre­vi­ously was re­served for doc­tors. Now some health sys­tems are de­ploy­ing the tech­nol­ogy for nurses and an­cil­lary providers.

Mosci­cki, a car­di­ol­ogy nurse prac­ti­tioner at the Hud­son Val­ley Heart Cen­ter in Pough­keep­sie, N.Y., and some of her col­leagues switched in May to us­ing speech recog­ni­tion for real-time doc­u­men­ta­tion in the hos­pi­tal’s elec­tronic health-record sys­tem. The heart cen­ter is part of the three-hos­pi­tal Health Quest sys­tem, based in LaGrangeville, N.Y., which ini­tially rolled out speech recog­ni­tion for its physi­cians to re­place their tran­scribed dic­ta­tion as part of a doc­u­men­ta­tion-im­prove­ment ini­tia­tive.

The sys­tem is now making that same tech­nol­ogy avail­able to nurses and other hos­pi­tal staff, us­ing soft­ware from Nu­ance Com­mu­ni­ca­tions, a Burling­ton, Mass.-based de­vel­oper of speech recog­ni­tion and nat­u­ral-lan­guage-pro­cess­ing tech­nol­ogy.

Mosci­cki and her col­leagues use the new sys­tem for all their progress notes, ad­mis­sions, pa­tient his­to­ries, phys­i­cal exam re­sults and dis­charge sum­maries. Progress notes were pre­vi­ously recorded on pa­per. Mosci­cki said the soft­ware can trans­late her speech to text faster and more ac­cu­rately than she can type—and she can type fast. But, she said, “When you’re us­ing a dic­ta­tion sys­tem, you don’t have ty­pos.” And when she’s done, the hos­pi­tal’s EHR sys­tem, from Cerner Corp., is im­me­di­ately up­dated.

Im­prov­ing nurse ef­fi­ciency will be a ne­ces­sity given grow­ing health­care de­mands from baby boomers com­bined with a pre­dicted nurs­ing short­age. The U.S. Bureau of La­bor Sta­tis­tics projects a de­mand for 1.1 mil­lion new nurses over the next seven years. Half that num­ber will be needed to re­place nurses who will re­tire by 2020; the other half will be re­quired to fill an ex­pected 575,000 new po­si­tions. The squeeze is made more acute by a scarcity of qual­i­fied nurs­ing-school fac­ulty.

The in­tro­duc­tion of speechrecog­ni­tion tech­nol­ogy for nurses is sig­nif­i­cant be­cause of the amount of time they typ­i­cally spend on doc­u­men­ta­tion. A 2008 study pub­lished in the Per­ma­nente Jour­nal found nurses spent more time dur­ing their work­days on doc­u­men­ta­tion than on direct pa­tient care. The study as­sessed more than 700 med­i­cal­sur­gi­cal nurses at 36 hos­pi­tals.

Nurs­ing ex­perts say EHRs have not helped much with that time crunch, at least partly be­cause they weren’t de­signed with nurses in mind. A study pub­lished in the jour­nal CIN: Com­put­ers In­for­mat­ics Nurs­ing in 2012 found that nurses at 55 hos­pi­tals spent 19% of their time on doc­u­men­ta­tion, whether they used pa­per records or an EHR.

“It’s prob­a­bly worse now,” said Carol Bick­ford, a se­nior pol­icy fel­low at the Amer­i­can Nurses As­so­ci­a­tion. “The re­port­ing re­quire­ments are ter­ri­ble.” But, she added, “Nurses make it work. We find the work­arounds.”

A par­a­digm shift in nurs­ing doc­u­men­ta­tion is needed to im­prove care de­liv­ery, said Joyce Sens­meier, vice pres­i­dent of in­for­mat­ics for the Health­care In­for­ma­tion and Man­age­ment Sys­tems So­ci­ety. Speech recog­ni­tion could help with that. “We need the struc­tured data so we can use com­put­ers to ag­gre­gate, per­form data an­a­lyt­ics and look for treat­ment pat­terns to im­prove pa­tient care,” Sens­meier said. “But we also need some of that con­tex­tual in­for­ma­tion that’s in free text.” De­sign­ing sys­tems to do both “would ad­vance us much more rapidly,” she said.

Joe Petro, se­nior vice pres­i­dent of health­care re­search and de­vel­op­ment at Nu­ance Com­mu­ni­ca­tions, said us­ing speech-recog­ni­tion tech­nol­ogy to re­duce the time nurses spend on doc­u­men­ta­tion frees up more time for pa­tient care. “Even if we could im­prove that with a mod­est goal of 10 min­utes a day, that’s an ad­di­tional 40 hours over the

“Just the speed to cap­ture things in real time on the mo­bile de­vice will be ex­tremely help­ful.”

Stephanie John­son Clin­i­cal in­for­ma­tion spe­cial­ist Epic Sys­tems Corp.

course of a year we could move to­ward pa­tient care,” he said.

“There are op­por­tu­ni­ties with voice recog­ni­tion,” ac­knowl­edged Ann Shep­ard, chief nurs­ing in­for­mat­ics of­fi­cer at Catholic Health Ini­tia­tives. The En­gle­wood, Colo.-based sys­tem cur­rently uses speech-recog­ni­tion sys­tems as a doc­u­men­ta­tion aid for physi­cians and ther­a­pists.

Shep­ard fore­sees CHI nurses com­plet­ing some of their doc­u­men­ta­tion with voice recog­ni­tion in­stead of typ­ing. One pos­si­ble tar­get for that shift is hos­pi­tal nurses’ end-of-shift re­ports, which cover pa­tient sit­u­a­tions, back­grounds, as­sess­ments and rec­om­men­da­tions (SBAR). A typ­i­cal SBAR re­port in­cludes both data, such as blood pres­sure and other vi­tal signs, and a text nar­ra­tion, in­clud­ing the nurse’s pa­tient-care plan.

But Bick­ford ques­tions the vi­a­bil­ity of speech-recog­ni­tion soft­ware for doc­u­men­ta­tion, par­tic­u­larly in busy, noisy hos­pi­tal en­vi­ron­ments. “How can you have a con­ver­sa­tion about pro­tected in­for­ma­tion in a hall­way?” she asked.

Epic Sys­tems Corp., a lead­ing EHR de­vel­oper, is ready­ing a mo­bile ap­pli­ca­tion for nurses and ther­a­pists that will con­vert speech to text and insert it into the pa­tient’s EHR, said Stephanie John­son, a clin­i­cal in­for­ma­tion spe­cial­ist at the Verona, Wis.-based com­pany. Epic aims to of­fer the app as part of its 2016 up­grade next fall. “I think there is a mar­ket for it,” she said. “Just the speed to cap­ture things in real time on the mo­bile de­vice will be ex­tremely help­ful.”

Epic is eval­u­at­ing Florence, Nu­ance’s voice-driven dig­i­tal as­sis­tant with ar­ti­fi­cial in­tel­li­gence. That project is cur­rently in the pro­to­type stage with no set launch date. “I could say blood pres­sure is 120 over 80 and that would go right into our nurs­ing flow sheet,” John­son said, ex­plain­ing a po­ten­tial use.

Last month at Health Quest, a dozen speech ther­a­pists, many with nurs­ing back­grounds, re­ceived train­ing in the Nu­ance speech-recog­ni­tion sys­tem and be­came its lat­est users.

Health Quest has bought an en­ter­prisewide li­cense to im­ple­ment the tech­nol­ogy across its whole sys­tem, en­abling any staff mem­ber who wants to use speech recog­ni­tion to do so. “That’s definitely the way to go,” said Adem Arslani, a nurse-in­for­mat­ics con­sul­tant who’s work­ing with Health Quest on the project. “You don’t want to cre­ate the (tech­nol­ogy) haves and have-nots.”

Don Bar­barino, a Health Quest speech pathol­o­gist who was part of last month’s train­ing group, ex­pressed sat­is­fac­tion with his ex­pe­ri­ence. “We started the first day and have been us­ing it ever since,” he said. “It’s cut my doc­u­men­ta­tion at least in half.”

That time sav­ings has en­abled him to see one or two more pa­tients a day. As a side ben­e­fit, he said he ex­pe­ri­ences less eye strain from star­ing at his com­puter screen while en­ter­ing data.

Nurse prac­ti­tioner Rachel Mosci­cki this spring be­gan us­ing speech-recog­ni­tion soft­ware for her doc­u­men­ta­tion, with help from con­sult­ing nurse-in­for­mati­cist Adem Arslani.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.