Back to search

HELSEVEL-Gode og effektive helse-, omsorgs- og velferdstjenester

AMK SIMULATOR: Et komplett verktøy for kompetansebygging for medisinsk nødmeldetjeneste, basert på simuleringstrening og kunstig intelligens

Alternative title: Dispatch simulator: A universal tool for training and competence improvement in the dispatch centre with help of artificial intelligence

Awarded: NOK 6.9 mill.

The Emergency Medical Communication Centre (EMCC) is an important link in the chain of survival for medical emergencies, such as cardiac arrest, stroke and trauma. The EMCC answers over 1 million calls for medical emergencies annually, and the number is rising. The EMCC is staffed by nurses and ambulance personnel with additional training as EMCC operators. Many calls are about critical and serious illness, and the EMCC operator must quickly determine the degree of the emergency, and which response unit should be sent. Furthermore, guidance in first aid must sometimes be given. All this happens over the phone, and preferably within seconds and minutes, as every second counts. The performance of the EMCC operator affects how fast the patient will receive help and correct treatment. It is an important, but demanding job. There is a lack of research and knowledge regarding the work that is done in the EMCCs. This is part of the reason that there is no standard education or validated training for EMCC operators. The degree of training the EMCC operators receive varies with the different EMCCs in Norway. Therefore, securing relevant education and training for EMCC operators could have a great impact on the chain of survival. The aim of the project is to develop a training simulator specifically designed for EMCC operators. The simulator will improve today`s standards for education and training, and give EMCC operators the possibility for flexible and frequent training on the job. Just like pilots, EMCC operators need to train for serious incidents before they "fly" in real life. EMCC Stavanger, SAFER (Stavanger Acute Medicine Foundation for Education and Research), NKLM (National Centre for Emergency Primary Health Care) and RAKOS (The Regional Centre for Emergency Medical Research and Development) are collaborating with the company SKLLS AS (formerly Headroom Lifescience) on the project. The aim is to develop a simulator where EMCC operators can train as realistically as possible. By developing technology such as artificial intelligence, EMCC operators will be called up by a computer, which is capable of talking to the operator and imitate several different callers with different injuries and illnesses. In this way the operators can prepare for complex, serious, and/or rare incidents by training in the simulator. By training through simulation, the operators will build experience that is similar to the experience gained through many years of clinical work. The aim is to improve the efficiency and quality of EMCC operators with positive effects on patient care. For operators, this can mean increased security and well-being at work, and for the patients with the most acute and severe incidents, frequent training of the operators can be the difference that saves vital seconds and minutes before treatment is given, thus saving lives. Since the start of the project in June 2020, technology partner SKLLS AS has worked closely with EMCC Stavanger and RAKOS on the development of an early prototype of the simulator. A reference group has previously been established with 1-2 people from each of the four EMCCs in Helse Vest. These have gained access to the script-driven version of the simulator, and the opportunity to test out and provide feedback on the simulator. The simulator has already been used at AMK courses in Stavanger in February and September 2021, and also in Bergen in June 2021. Prior to each of the courses, the simulator has been continuously improved and filled with more content. The script-driven version is planned to be used while AMK operators are at work around new year. It is now integrated with AMK's Integrated Communication and Control System (ICCS) so that it can be pushed out as a regular emergency call with clear marking that it is for training so that the call is not prioritized if an actual emergency call comes in at the same time. The project has gained access to audio recordings from real emergency calls in November 2021. Prior to the access, SKLLS AS has been able to incorporate their algorithms in the cloud service microsoft Azure administered through Helse Vest IKT. They have also improved their speech-to-text algorithms and created their own voices for the simulator by recording the voices of students reading out loud for several hours. Future development will include analysis of real emergency calls, in order to learn artificial intelligence how to imitate different callers and different medical scenarios. When the simulator is fully developed and completed, the effect of training in the simulator will be investigated and validated through prospective observational studies. As a part of this project, the project team will also work to further develop and adapt the simulator for use in the emergency primary health care sector.

Akuttmedisinsk kommunikasjonssentral (AMK) gir første hjelp ved livstruende skade eller sykdom utenfor sykehus. Tidlig diagnostikk og og behandling er essensielt for å redde liv. På kort tid tar AMK kritiske vurderinger, livsviktige beslutninger og veileder innringer i førstehjelp. Det er ca. 1 mio. nødsamtaler hvert år, og tallet øker. Det forventes at AMK sine oppgaver øker både i antall og kompleksitet. Jobben er faglig komplekst og krevende, og samhandling mellom innringer og AMK er avgjørende for utfallet. Likevel finnes ingen standardisert utdanning, opplæring, sertifisering eller vedlikeholdstrening for AMK-operatører. Det er behov for styrke kompetansen for å møte disse utfordringene, særlig omkring kommunikasjon, operativ psykologi, språk, og prioriteringskompetanse. Innovasjonspotentiale i prosjektet er at det utvikles en metode for dette. Prosjektet utvikler en fullskala simulator med 1:1 digitalt grensesnitt for alle funksjoner i AMK, og det utvikles virtuelle pasienter med kunstig intelligens (AI). 1:1 simulering betyr at brukerne kan trene i sitt vanlige arbeidsmiljø (in-situ). Dette er en helt ny løsning for kompetansebygging, og AMK kan trene kommunikasjon, klinisk og teknisk kompetanse i en virtuell setting som fremstår realistisk. Som bruker blir man ringt opp av en virtuell pasient, som simulerer realistiske sykdomsbilder og reagerer emosjonelt som en alminnelig innringer. Ved å lytte til lydlogger lærer AI å agere innringer, men også hva som er best praksis for AMK og kan gi tilbakemelding på brukers prestasjon. Læringen fra dialogen i lydloggene, bakgrunnstøy og andre metadata bruker AI også til å gi sanntid råd og anbefalinger, dvs. beslutningsstøtte. AMK-simulator valideres gjennom kvalitative og kvantitative deskriptive observasjonsstudier før og etter implementering av simulator og beslutningstøtte. Prosjektet vil også gjennomføre forundersøkelser og testing på legevakt mtp. å modifisere simulator til legevaktsentraloperatører.

Activity:

HELSEVEL-Gode og effektive helse-, omsorgs- og velferdstjenester