Applying Artificial Intelligence in Developing Personalized and Sustainable Healthcare for Spinal Disorders

Overview

The primary objective is to use machine learning methods on large survey and health register data to identify participants with different treatment trajectories and health outcomes after surgical and/or conservative treatment for spinal disorders. Secondary objectives are to 1) conduct external validation of the prediction models, and 2) explore how the prediction models can be implemented into AI-based clinical co-decision tools and interventions.

Three work packages are conducted. In the first, the investigators will use data from three general population surveys in Norway (HUNT, Tromsø, and Ullensaker) linked to administrative health registry data (Norwegian Patient Registry (for secondary care) and Norwegian Registry for Primary Health Care) and clinical registers on spinal disorders (the Norwegian registry for spine surgery, NorSpine, and the Norwegian registry for neck and back pain) to explore treatment trajectories and health outcomes following an episode of back and/or neck pain. The investigators will use different combinations of these data sets to assess the impact of a wide range of risk/ prognostic factors and to develop prognostic models for different health and welfare outcomes. Four major outcomes will be adressed; a) unfavourable outcomes, b) use of prescribed medication, c) use of sickness absence and other disability benefits, and d) patient-reported outcomes. In the second work package, the investigators will conduct external validation studies of the prediction models by using Danish and Swedish data. There is a large overlap and similarities in health and welfare registers across the Nordic countries. In the third work package the investigators will first conduct a feasibility study in a secondary care hospital setting in which surgeons examine and assess referred patients with disc herniation and spinal stenosis for surgical treatment (or not). Qualitative interviews will be used to gain a better understanding of today's clinical decision-making process.

Conditions