THE ANALYTICAL FRAMEWORK OF CLINICAL TRIALS EVALUATING CLINICAL OUTCOMES OF ARTIFICIAL INTELLIGENCE-BASED DIGITAL HEALTH INTERVENTIONS FOR MUSCULOSKELETAL DISORDERS: A SYSTEMATIC LITERATURE REVIEW
Author(s)
Dimitrije Grbic, PhD (c), Filip Stanicic, PhD (c), Vlad Zah, PhD;
ZRx Outcomes Research, Inc., Mississauga, ON, Canada
ZRx Outcomes Research, Inc., Mississauga, ON, Canada
OBJECTIVES: The systematic literature review (SLR) aimed to establish an analytical framework for clinical trials evaluating outcomes of artificial intelligence-based digital health interventions (AI-DHI) among patients with musculoskeletal disorders.
METHODS: The SLR was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines. Literature search was performed in PubMed and Embase databases, with additional hand-search on Google Scholar platform. Population included patients with musculoskeletal disorders using AI-DHI. Only clinical trials that evaluated clinical outcomes and written in English were considered. The National Institute for Health and Care Excellence quality appraisal checklist was used to assess quality of included studies.
RESULTS: After the review process, the initial number of 2,247 studies was decreased to 17 studies in the final sample. Only one study (5.9%) was published before 2020. The most common indications were back pain (58.8%) and rheumatic disorders (26.4%). Most studies were conducted in Denmark and Norway (29.4%). Majority of studies (70.6%) were controlled, parallel-group clinical trials with at least two arms. Standard of care was the most common comparator (66.7%) in parallel-group clinical trials. Despite blinding should be applied both for investigators and participants, only one study (5.9%) was single-blinded, 82.3% of trials were open-label, while 11.8% of trials did not report the blinding level. There were 23.5% of studies conducted at multiple sites across different countries. Power analysis was conducted to determine sample size in 58.8% of trials. Dropout rates in clinical trials should be <20% at all endpoints, with 23.5% of studies reported higher values. Statistical tests were used according to the outcome type. Main outcomes reflecting clinical efficacy were pain intensity (70.6%) and physical activity scores (47.1%).
CONCLUSIONS: The SLR results reflected on the characteristics of clinical trials evaluating AI-DHI clinical value among patients with musculoskeletal disorders and emphasized current methodological gaps.
METHODS: The SLR was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines. Literature search was performed in PubMed and Embase databases, with additional hand-search on Google Scholar platform. Population included patients with musculoskeletal disorders using AI-DHI. Only clinical trials that evaluated clinical outcomes and written in English were considered. The National Institute for Health and Care Excellence quality appraisal checklist was used to assess quality of included studies.
RESULTS: After the review process, the initial number of 2,247 studies was decreased to 17 studies in the final sample. Only one study (5.9%) was published before 2020. The most common indications were back pain (58.8%) and rheumatic disorders (26.4%). Most studies were conducted in Denmark and Norway (29.4%). Majority of studies (70.6%) were controlled, parallel-group clinical trials with at least two arms. Standard of care was the most common comparator (66.7%) in parallel-group clinical trials. Despite blinding should be applied both for investigators and participants, only one study (5.9%) was single-blinded, 82.3% of trials were open-label, while 11.8% of trials did not report the blinding level. There were 23.5% of studies conducted at multiple sites across different countries. Power analysis was conducted to determine sample size in 58.8% of trials. Dropout rates in clinical trials should be <20% at all endpoints, with 23.5% of studies reported higher values. Statistical tests were used according to the outcome type. Main outcomes reflecting clinical efficacy were pain intensity (70.6%) and physical activity scores (47.1%).
CONCLUSIONS: The SLR results reflected on the characteristics of clinical trials evaluating AI-DHI clinical value among patients with musculoskeletal disorders and emphasized current methodological gaps.
Conference/Value in Health Info
2026-05, ISPOR 2026, Philadelphia, PA, USA
Value in Health, Volume 29, Issue S6
Code
MT33
Topic
Medical Technologies
Topic Subcategory
Digital Health
Disease
SDC: Musculoskeletal Disorders (Arthritis, Bone Disorders, Osteoporosis, Other Musculoskeletal)