Osteoporosis must be detected earlier
With a DKK 10 million grant from NordForsk, researchers from the University of Southern Denmark (SDU) and Odense University Hospital (OUH) will use artificial intelligence to help general practitioners detect osteoporosis at an earlier stage. This could improve patients’ quality of life and reduce the costs associated with bone fractures.
An estimated 700,000 people in Denmark are living with osteoporosis. The condition weakens the bones, making them fragile and more likely to break – often resulting in painful, disabling, and in some cases even life-threatening fractures.
Hvad er knogleskørhed?
- Osteoporosis is a chronic condition where bone mass and strength are reduced, making bones more fragile and prone to fractures. According to the Danish Health Authority, around 700,000 people in Denmark are living with the disease.
- It affects nearly 1 in 3 women and 1 in 6 to 8 men over the age of 50.
- Fractures most commonly occur in the spine, hips or wrists.
- Risk factors include early menopause (before age 45), smoking, being underweight, lack of physical activity, and deficiencies in vitamin D and calcium. Certain medical conditions can also increase the risk.
- The condition is often not diagnosed until after the first fracture.
Source: sundhed.dk
Osteoporosis can have a serious impact not only on a person’s health and mobility, but also on their social life and the wellbeing of their loved ones. The condition also places a growing burden on society. And the longer it goes undiagnosed, the more costly and complex it becomes to treat.
– Osteoporosis is often not diagnosed until after a patient has suffered their first fracture, which means that opportunities for early detection and prevention are frequently missed, explains Katrine Hass Rubin, professor at the Department of Clinical Research at SDU and the research unit OPEN.
– We’ve developed a tool designed to help general pracitioners identify high-risk patients earlier. The next step is to test it and implement it in clinical practice.
The tool, called FREM_ML, is a decision support system that uses artificial intelligence to estimate the risk of osteoporosis-related fractures. Katrine Hass Rubin and her research team have recently been awarded a DKK 10 million grant from NordForsk to support the implementation of the tool.
FREM_ML
FREM_ML calculates the risk of osteoporosis-related fractures and alerts general practitioners to patients who may be at high risk. These patients can then be offered further evaluation or preventive care.
Uses existing data
FREM_ML can be integrated into the electronic medical record systems already in use by general practitioners. It helps general practitioners identify patients at high risk of osteoporosis based on data that is already available.
– All the information we need is already in the system, explains Professor Katrine Hass Rubin. The risk is calculated automatically – with no manual input required from the doctor.
The project was awarded an “Outstanding” rating by NordForsk – the highest possible score. Only five projects out of 74 applicants received funding.
Improving quality of life for older adults
Fractures caused by osteoporosis often require hospitalisation and, in severe cases, can result in permanent placement in a care home. Many patients suffer a significant loss of mobility and independence. Recovery is typically slow and incomplete – and life may never return to how it was before. The condition often demands long-term treatment and care.
By detecting osteoporosis before fractures occur, doctors can start treatment earlier – helping to assess risk and prevent injuries before they happen.
– We’ll begin initial testing in September 2025, says Katrine Hass Rubin. What’s groundbreaking is that the data analysis happens automatically – the general practitioner doesn’t have to enter anything, because the data is already in the system.
About the project:
- Project title: CHOICE – Continuous Healthcare in OsteoporosIs CasE-finding
- Funding: DKK 10 million from NordForsk under the call “Sustainable health and social care systems for the elderly”
- Principal investigator (PI): Professor Katrine Hass Rubin, OPEN Research Unit, Odense University Hospital and the Department of Clinical Research, University of Southern Denmark (SDU)
- Partners: The Research Unit for General Practice, the Maersk Mc-Kinney Møller Institute, and the OPEN Research Unit at SDU, in collaboration with partners from Norway and Sweden
- Project period: 2025–2029
FREM_ML is an AI-based model developed using national healthcare data and clinical data from patients who have previously suffered osteoporosis-related fractures. It has been trained to identify key risk factors and can estimate an individual’s risk of sustaining such a fracture within the next year.
The model takes into account a range of variables – including gender, age, medical diagnoses, and medication use over the past 15 years.
FREM_ML can be integrated into the electronic medical record systems already used by general practitioners. It draws on existing health data to detect patterns linked to risk factors and disease progression. In short, FREM_ML enables fast and automatic identification of high-risk patients – directly within the general practitioner’s system, using data that is already available.
The model now needs to be tested in real-world clinical settings so researchers can fine-tune its performance. Testing and validation will take place in collaboration with both primary and secondary healthcare sectors in Denmark, Norway and Sweden.
A user council has been established to contribute throughout the process and ensure that the perspectives of patients – and those potentially at risk of osteoporosis – are taken into account.
Read more:
Meet the researcher
Katrine Hass Rubin is professor of epidemiology at the OPEN research unit, Odense University Hospital and the Department of Clinical Research, SDU and OUH.