ARTIFICIAL INTELLIGENCE IN REHABILITATION: FACILITATORS AND PERCEIVED BARRIERS AMONG PHYSICAL THERAPISTS

¹King Abdulaziz University, Physical Therapy Department, Jeddah, Saudi Arabia

Background: Artificial intelligence (AI) has been growing rapidly in the field of rehabilitation. Although numerous AI applications were studied in the literature, little is known about PTs’ knowledge, attitudes, and practices of AI in clinical practice. Nowadays, PTs need to be equipped with the latest AI-advanced technologies to improve patients' quality of care and outcomes.

Purpose: The purpose of the study was to understand PTs' views on AI and to identify the predictive factors of knowledge and attitudes toward AI based on multiple sociodemographic factors (gender, education qualification, years of experience, number of AI applications at work, and subspecialty). Also, the study aimed to identify barriers that might limit AI application in rehabilitation.

Methods: A total of 236 PTs responded to at least one of the qualitative questions. A self-administered survey consisting of close/open-ended questions investigating several aspects, including demographic, knowledge, uses, advantages, impacts, and barriers limiting AI utilization in rehabilitation. Concurrent mixed methods including quantitative and embedded qualitative design were used. The quantitative part of the study utilized a cross-sectional, predictive design with a set of multiple predictors to explain the current PTs' understanding and attitude toward AI applications, and it was analyzed using logistic regression. However, the qualitative part utilized open-ended questions to have an in-depth understanding of barriers that limit AI adoption, and content analysis was used to analyze the data.

Results: 152 (63.3%) PTs reported the absence of AI applications at work. Only 12 (5%) respondents have come across more than four AI applications in their clinical practice. The majority of PT respondents (75.0%, n=177) believed the clinicians’ opinions should be considered in case of conflict whereas 21.61% (n=51) thought that patients’ preferences should be prioritized over AI and clinicians’ judgments. The major factors influencing knowledge of AI among PTs were being a non-academic employee (OR= 1.77; 95% CI=1.01,3.12, P₌.04), being a senior PT (OR= 2.44; 95% CI₌1.40, 4.22, P₌ .002), having a post-graduate degree (OR= 1.97; 95% CI₌1.11, 3.50, P₌.02). However, cost, lack of knowledge, patient compliance, ethical issues, and technology trust were reported as potential barriers to implement AI in clinical practice.

Conclusions: Although technological innovation has an impact on improving the quality of healthcare, AI and advanced technology knowledge need to be transferred to PTs. Future studies should focus on improving knowledge of AI among PTs to bridge the gap between the ongoing research and current PT practices.

Implications: AI-based technologies can positively assist in the rehabilitation management of various clinical conditions. PTs should develop their knowledge of AI and be able to translate that knowledge into clinical practice. To cope with the newer rehabilitation technologies, PTs should be prepared with education and training. The academy must integrate AI knowledge into the PT curriculum and promote critical thinking skills needed for knowledge translation.

Funding acknowledgements: This research received no external funding.

Keywords:
Artificial intelligence
Physical Therapist
Barriers

Topics:
Innovative technology: information management, big data and artificial intelligence
Education: continuing professional development
Research methodology, knowledge translation & implementation science

Did this work require ethics approval? Yes

Institution: Center of Excellence in Genomic Medical Research

Committee: National Committee of Bioethics (KACST: HA-02-J-003)

Ethics number: 14-CEGMR-Bioeth-2021