Self-anamnesis in Physiotherapy Practice: Collecting patient data with a virtual assistant
Translated title
Selv-anamnese i fysioterapeutisk praksis: Indsamling af patientdata med en virtuel assistent
Authors
Tengberg, Fie Louise ; Chaika, Alona
Term
4. Term
Publication year
2024
Submitted on
2024-05-31
Pages
54
Abstract
Complex patient cases in physiotherapy often take the most time, which paradoxically leaves less time for documentation and treatment. We explored whether a virtual assistant could help by collecting patients’ medical histories (anamnesis) before the consultation. Our aim was to reduce the documentation burden on physiotherapists and enable a faster start to tests and treatment. We built a self-anamnesis prototype that used speech recognition (the system listens) and speech synthesis (the system speaks) so patients could answer by voice. In our evaluation, potential patients interacted with the system and then shared their views in a survey and semi-structured interviews. Next, a group of seven physiotherapists were shown selected answers from traditional anamnesis recordings and from the system, and asked to tell them apart. Overall, the physiotherapists could not distinguish between information gathered by the virtual assistant and by conventional methods, suggesting that patients do not respond differently when questioned by a purpose-built system. Both physiotherapists and potential patients gave generally positive feedback, while noting drawbacks. The main concerns were limited customization and practical, logistical challenges that could hinder successful use in everyday practice.
Komplekse patientforløb i fysioterapi tager ofte mest tid, hvilket paradoksalt nok efterlader mindre tid til både dokumentation og behandling. Vi undersøgte, om en virtuel assistent kan hjælpe ved at indsamle patientens sygehistorie (anamnesen) før konsultationen. Målet var at lette fysioterapeuters dokumentationsarbejde og gøre det muligt hurtigere at starte relevante undersøgelser og behandling. Vi udviklede en selv-anamnese-prototype med talegenkendelse (systemet lytter) og talesyntese (systemet taler), så patienter kunne besvare spørgsmål med stemmen. I evalueringen afprøvede potentielle patienter systemet og gav deres vurderinger i et spørgeskema og semistrukturerede interviews. Derefter blev en gruppe på syv fysioterapeuter præsenteret for udvalgte svar fra traditionelle anamneseinterview og fra systemet og blev bedt om at skelne mellem dem. Overordnet kunne fysioterapeuterne ikke skelne mellem data indsamlet af den virtuelle assistent og data fra den konventionelle metode, hvilket tyder på, at patienter ikke svarer anderledes, når spørgsmålene stilles af et formålsspecifikt system. Både fysioterapeuter og potentielle patienter gav overvejende positiv feedback, men pegede også på ulemper. De vigtigste bekymringer var manglende mulighed for tilpasning samt praktiske og logistiske forhold, der kan gøre implementering vanskelig i praksis.
[This apstract has been rewritten with the help of AI based on the project's original abstract]
Keywords
fysioterapi ; anamnese ; virtuel assistent ; AI ; data ; NLP