PREDICT-LBP · PRedictive Evidence Driven Intelligent Classification Tool for Low Back Pain

90% of people who have back pain are classified as having "non-specific" back pain - no specific cause can be found. This project aims to make non-specific back pain more specific, consider the multi-dimensional nature of pain and enable targeted therapies.

Current recommended treatments for low back pain are of limited efficacy and below the threshold of clinical significance. Approaches exist to classify and subgroup people with nonspecific low back pain, but existing classification approaches do not benefit patients more than simpler treatment approaches and do not consider the multifactorial nature of pain. Over 90% of cases of back pain are considered non-specific where no particular therapy is deemed better than another.

There is an urgent need to develop robust subgrouping of nonspecific low back pain, we argue for a return to basic research that takes into account the multidimensional nature of the pain experience and adopts an evidence- and data-science-driven approach. Our preliminary work shows that there is a significant knowledge gap for both work that considers the multiple causes (link) that contribute to nonspecific low back pain and data science (machine learning/artificial intelligence; link) approaches to this. Our pilot analyses using data from the UKBiobank (link, link) demonstrate our ability to perform the planned work. The results (link) demonstrated the potential presence of subgroups in non-specific low back pain, but also highlighted the need for further steps.

The aim of this project is to fill this evidence gap of the contribution of different pain systems to non-specific back pain and to develop a data-driven and evidence-based tool for classifying back pain. This project will result in a practical tool for clinicians to better diagnose non-specific back pain and to drive personalised management approaches.

Publications arising to date:

Belavy DL, Tagliaferri SD, Tegenthoff M, Enax-Krumova E, Schlaffke L, Bühring B, Schulte TL, Schmidt S, Wilke H-J, Angelova M, Trudel G, Ehrenbrusthoff K, Fitzgibbon B, Van Oosterwijck J, Miller CT, Owen PJ, Bowe S, Döding R, Kaczorowski S (2023): Evidence- and data-driven classification of low back pain via artificial intelligence: protocol of the PREDICT-LBP study. PLoS One 18(8):e0282346 http://dx.doi.org/10.1371/journal.pone.0282346 

Vickery S, Junker F, Döding R, Belavy DL, Angelova M, Chandan Karmakar C, Becker LA, Taheri N, Pumberger M, Reitmaier S, Schmidt H (2025): Integrating Multidimensional Data Analytics for Precision Diagnosis of Chronic Low Back Pain. Scientific Reports 15: 9675 https://doi.org/10.1038/s41598-025-93106-1