SupportPrim is a research project financed by The Research Council of Norway through 2023 to improve primary health care management of common musculoskeletal pain disorders. The core research activity is to develop and apply artificial intelligence (AI) systems to identify the best treatment for patients with musculoskeletal pain.
The overall aim of the SupportPrim project is to improve current primary care management of common musculoskeletal disorders, including neck, shoulder, low back, hip, knee and multisite pain, expanding knowledge beyond evidence-based one-size-fits-all recommendations, via stratified care, to personalized care based on a clinical decision support system founded in AI.
A challenge for the society
Musculoskeletal pain complaints are common in the population, such as low back, neck, shoulder, hip, and knee pain, as well as multisite or complex pain, and musculoskeletal disorders are the number one cause of years lived with disability worldwide.
Most patients suffering from musculoskeletal pain have non-specific symptoms where studies show that common interventions either lack evidence to support their use or have modest or short-term effects.
Pain is however highly individual and influenced by many biological, psychological, and social factors, with large variation in symptoms and effect of treatment between patients even with the same pain complaint or diagnosis. In SupportPrim we believe an important reason that studies fail to document major effects is that they look at average effects among patients without regards for the heterogeneity of patients.
In the SupportPrim project we want to investigate if more individualized patient management can improve treatment effects and patient reported outcomes.
We will use artificial intelligence in terms of a method called Case-Based Reasoning (CBR) to build a computer-based decision support system (the SupportPrim-system) with a user-friendly interface (the SupportPrim-panel) that therapist and patient can use together to decide the best treatment based on knowledge of treatment in previous similar patients with successful outcome.
State-of-the-art personalized treatment plans are envisioned to benefit a much larger proportion of patients with musculoskeletal disorders than a "one-size-fits-all" approach.
To optimize person-centered care, the SupportPrim project will employ innovative methods from artificial intelligence using Case-Based Reasoning to build a clinical decision support system that uses patient data from previous patients in primary care physiotherapy.
Case-Based Reasoning aims to guide new treatment decisions based on the outcome in similar patients who received the treatment in the past. Previous cases with musculoskeletal disorders are used to help similar cases in the future, just as humans learn from their own experience.
The SupportPrim-system consist of a large database with systematic data of previous successful patients, their treatment and outcome, the CBR engine which is used to create the artificial intelligence-based decision support system, and the SupportPrim-panel interface where similar patients are matched and displayed. The system will be exposed to a cluster randomized controlled trial for assessing patient outcome and treatment efficacy compared to treatment as usual.
The AI decision-support system will be developed in Work Package (WP)1, led by Associate Professor Kerstin Bach, the interface decision-support dashboard will be developed in WP2, led by Researcher Ingebrigt Meisingset, and the system evaluated in a RCT study in primary care private physiotherapy practice in WP3, led by Professor Ottar Vasseljen.
Testing of the system
During 2021, 40 physiotherapists in Norway will be recruited to the project and randomized to either the intervention group or the control group. They will each recruit 18 of their own patients and each therapist serve as a cluster.
The patients in both groups will respond to questions about their conditions before their first consultation with their physiotherapist. The questions cover all aspects of the biopsychosocial domain. Only the intervention group will have access to the computer-based decision support system through the SupportPrim-panel. The panel interface graphically summarizes the current patient's profile, matches the current patient with previous successful patients and display information about their treatment as basis for deciding the best treatment options for the current patient. This is a co-decision process between the therapist and patient. Patients randomized to the control group will receive usual care.
Extending the decision support system to general practice
Improving musculoskeletal pain management in general practice will be addressed in two waves, firstly, by applying stratified care using the Keele STarT MSK Tool developed by Keele University (WP4), and secondly, by adapting the SupportPrim-system and -panel developed in WP1-3 to general practice (WP5).
The efficacy of the stratified care approach will be assessed in a cluster randomized controlled trial in general practice with similar selection criteria and design as in WP3. General practitioners within the PraksisNett network will be recruited to participate. GPs in the intervention group will be informed of their patient's risk group allocation (low, medium, high) based on the STarT MSK Tool instrument, and provided advice on management based on risk group. GPs in the control group will continue treatment as usual. For both groups, treatment and management procedures recorded by the GP will be collected by a small computer installed in the GP's office that fetch data from patient journals during night (Snow Health Alliance Box).
The RCT in General Practice will be done in WP4, led by Researcher Ingebrigt Meisingset, while the adaptation of the SupportPrim-system will be carried out in WP5, led by Associate Professor Kerstin Bach.
Bjarne Austad Associate Professor+47-73597528 +47-99029992 firstname.lastname@example.org Department of Public Health and Nursing
Kerstin Bach Associate Professor+47-73597410 +47-93032400 email@example.com Department of Computer Science
Anita Formo Bones Staff Engineer+47-45472843 firstname.lastname@example.org Department of Public Health and Nursing
Egil Andreas Fors Professor+47-73597581 +47-41236597 email@example.com Department of Public Health and Nursing
Fredrik Granviken PhD Candidate+47-93059497 firstname.lastname@example.org Department of Public Health and Nursing
Amar Jaiswal PhD Candidate+47-94080927 email@example.com Department of Computer Science
Pål Jørgensen Associate Professorpal.firstname.lastname@example.org Department of Public Health and Nursing
Lars Christian Naterstad Lervik Researcher+47-91853697 email@example.com Department of Public Health and Nursing
Jon Magnussen Vice Dean, Professor+47-93009681 firstname.lastname@example.org Department of Public Health and Nursing
Paola Marín Veites PhD Candidate+47-46954239 email@example.com Department of Computer Science
Ingebrigt Meisingset Researcher+47-73598901 +47-90066915 firstname.lastname@example.org Department of Public Health and Nursing
Paul Jarle Mork Professor+47-73590447 +47-90104615 email@example.com Department of Public Health and Nursing
Torbjørn Øien Professor+47-73597526 +47-95219502 firstname.lastname@example.org Department of Public Health and Nursing