A Conversational AI Tool for ICF-Informed Physiotherapy: A Worked Example of a Formative Mixed-Method Evaluation from Patient and Physiotherapist Perspectives

Abstract Background: Implementing the biopsychosocial perspective and the International Classification of Functioning, Disability and Health in routine physiotherapy remains challenging because comprehensive assessment requires time, clinical expertise, and integration within existing workflows. We developed a conversational prototype to address these barriers. This study evaluated its user experience, usability, and acceptability from patient and physiotherapist perspectives and provides a worked example of a multi-actor formative mixed-methods evaluation. Methods: A single-centre formative mixed-methods study was conducted in a Swiss physiotherapy clinic. Thirty patients completed a conversational interview before a simulated initial consultation; ten physiotherapists each reviewed three AI-generated reports. Patient outcomes included interview completion, time, study-specific ISO 9241-11 usability indices, User Experience Questionnaire scores, acceptability, and open-ended feedback. Physiotherapists completed the System Usability Scale, a context-adapted Technology Acceptance Model questionnaire, and open-ended questions. Quantitative findings were summarised descriptively. Inductive thematic analysis captured concerns; deductive framework analysis explored findings prioritised using established benchmarks where applicable and study-specific criteria otherwise. Joint displays integrated findings into refinement proposals. Results: ISO 9241-11 usability indices and objective measures were reported as moderate. Patients reported favourable perspicuity and novelty, while efficiency and stimulation were below average and attractiveness and dependability were rated as bad against User Experience Questionnaire benchmarks. Twenty-two patients (73.3\%) supported future implementation. Physiotherapists reported moderate usability (mean System Usability Scale score 66.0, SD 13.3), perceived ease of use of 3.50/5, perceived usefulness of 2.83/5, and behavioural intention to use of 3.25/5. Qualitative findings indicated that the interview could support patient reflection, disclosure, and consultation preparation. However, repetitive or insufficiently tailored questions, conversational reliability problems, and incomplete or insufficiently prioritised reports reduced usability and perceived clinical value. Integration identified discrepancies between successful completion and interaction burden, and between basic learnability and uncertain clinical usefulness. Conclusions: The prototype showed potential as an acceptable and usable support tool for ICF-informed physiotherapy but requires further refinement before routine implementation. Proposals include adaptive questioning, improved conversational reliability, concise and clinically prioritised reports, traceability to patient responses, and explicit physiotherapist verification of AI-generated content. The multi-actor mixed-methods approach enabled the identification of actionable refinements that would not have been otherwise apparent.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-16
DOI
https://doi.org/10.21203/rs.3.rs-11470004/v1
Primary Topic
Artificial Intelligence in Healthcare and Education
Type
preprint

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

A Conversational AI Tool for ICF-Informed Physiotherapy: A Worked Example of a Formative Mixed-Method Evaluation from Patient and Physiotherapist Perspectives

Marco Barbero, Alessandro Schneebeli, Andrea Piña, Andrea Franchini et al.
2 links
Research Square
Artificial Intelligence in Healthcare and Education
preprint

A Conversational AI Tool for ICF-Informed Physiotherapy: A Worked Example of a Formative Mixed-Method Evaluation from Patient and Physiotherapist Perspectives

Marco Barbero, Alessandro Schneebeli, Andrea Piña, Andrea Franchini, Marco Griseri, Tiziano Negrisolo, Fabio Rinaldi
preprint en

Abstract

Abstract Background: Implementing the biopsychosocial perspective and the International Classification of Functioning, Disability and Health in routine physiotherapy remains challenging because comprehensive assessment requires time, clinical expertise, and integration within existing workflows. We developed a conversational prototype to address these barriers. This study evaluated its user experience, usability, and acceptability from patient and physiotherapist perspectives and provides a worked example of a multi-actor formative mixed-methods evaluation. Methods: A single-centre formative mixed-methods study was conducted in a Swiss physiotherapy clinic. Thirty patients completed a conversational interview before a simulated initial consultation; ten physiotherapists each reviewed three AI-generated reports. Patient outcomes included interview completion, time, study-specific ISO 9241-11 usability indices, User Experience Questionnaire scores, acceptability, and open-ended feedback. Physiotherapists completed the System Usability Scale, a context-adapted Technology Acceptance Model questionnaire, and open-ended questions. Quantitative findings were summarised descriptively. Inductive thematic analysis captured concerns; deductive framework analysis explored findings prioritised using established benchmarks where applicable and study-specific criteria otherwise. Joint displays integrated findings into refinement proposals. Results: ISO 9241-11 usability indices and objective measures were reported as moderate. Patients reported favourable perspicuity and novelty, while efficiency and stimulation were below average and attractiveness and dependability were rated as bad against User Experience Questionnaire benchmarks. Twenty-two patients (73.3\%) supported future implementation. Physiotherapists reported moderate usability (mean System Usability Scale score 66.0, SD 13.3), perceived ease of use of 3.50/5, perceived usefulness of 2.83/5, and behavioural intention to use of 3.25/5. Qualitative findings indicated that the interview could support patient reflection, disclosure, and consultation preparation. However, repetitive or insufficiently tailored questions, conversational reliability problems, and incomplete or insufficiently prioritised reports reduced usability and perceived clinical value. Integration identified discrepancies between successful completion and interaction burden, and between basic learnability and uncertain clinical usefulness. Conclusions: The prototype showed potential as an acceptable and usable support tool for ICF-informed physiotherapy but requires further refinement before routine implementation. Proposals include adaptive questioning, improved conversational reliability, concise and clinically prioritised reports, traceability to patient responses, and explicit physiotherapist verification of AI-generated content. The multi-actor mixed-methods approach enabled the identification of actionable refinements that would not have been otherwise apparent.

Research Square
University of Applied Sciences and Arts of Southern Switzerland (CH), Università della Svizzera italiana (CH)
Innosuisse - Schweizerische Agentur für Innovationsförderung
Good health and well-being
Artificial Intelligence in Healthcare and Education
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon