Researchers at the Vrije Universiteit Brussel (VUB) have found a clever way to improve the computer calculations underpinning breast cancer screening programmes. Their new calculation method enables governments and policymakers to estimate the impact of changes to screening policy much more accurately. This helps to prevent policy decisions from being based on distorted model outcomes. The research, led by Dr Max Lelie, was recently published in the scientific journal *Applied Health Economics and Health Policy*.
Designing a population-based screening programme is always a complex balancing act. Screening everyone every month might seem ideal to ensure that no tumour is missed, but it would be practically impossible. Hospitals would become overburdened and the costs would skyrocket. Moreover, excessive testing leads to unnecessary treatments and anxiety among healthy women. Conversely, too little testing can result in cancers being detected too late. The government is therefore seeking the best combination of starting age, stopping age and testing frequency.
Until now, the mathematical models used to compare risk groups have encountered a curious problem. If the government relaxed the rules so that people were classified as high-risk more quickly, the computer became confused. The model would then conclude that the administrative change itself was causing more people to develop cancer. A government using such a model to test what would happen if a risk group were abolished would be told that there would suddenly be fewer cases of cancer. This could lead to misguided societal decisions.
Dr Max Lelie explains the aim of the research: “We are always looking for the optimal balance: how can we save as many lives as possible through policy, without costs rising unnecessarily? Investing extra to tackle a disease is certainly acceptable, but it must be done sensibly.”
To solve this problem, Dr Lelie adjusted the mathematics behind the model. In Flanders, around 13 per cent of women will develop breast cancer at some point in their lives. By setting that total number in the model in advance, the computer only needs to calculate when someone falls ill, rather than how many people in total will fall ill as a result of a policy change.
“In the old models, it seemed as though adjusting a category on paper changed the likelihood of developing cancer,” explains Dr Lelie. “The mathematical basis of the model always remains logically consistent with the known cancer statistics, however the government may reclassify the risk groups. In this way, the computer prevents policymakers from being led down the wrong track.”
Furthermore, the new calculation method is designed so that policymakers can easily feed it the most recent figures from the Cancer Register. As a result, the computer automatically adapts as the population becomes healthier or as lifestyles change.
“Our aim was to provide a tool that enables policymakers to safely run through various scenarios themselves using up-to-date figures,” concludes Dr Lelie.
The new model was extensively tested using Flemish data and proved to correspond very accurately with the actual annual diagnoses in Flanders. The method therefore provides a solid basis for planning future population-based screening programmes.
Reference
Lelie, M., VandenBulcke, B., Verhaeghe, N., Annemans, L., Simoens, S., & Putman, K. (2026) Ensuring Epidemiological Consistency in Risk-Stratified Cancer Screening Models: A Novel Approach Based on Flemish Breast Cancer Screening. Applied Health Economics and Health Policy. https://doi.org/10.1007/s40258-026-01029-3
Contact:
Dr Max Lelie: max.lelie@vub.be. Telephone number available via the press office.
Prof. Dr Koen Putman
Prof. Dr Steven Simoens