Back

Continuous monitoring of the use of molecular diagnostics in oncology: advancing towards a learning health system

Ebben, K.; Jacobs, F. J. C.; Pasman, R.; Louis - van den Broek, C.; de Kroon, C. D.; van der Wekken, A. J.; Seegers, P. A.; Voorham, Q.; Heidenrijk, E.; van der Werf, J.; van der Flier, S.; Barjesteh van Waalwijk van Doorn - Khosrovani, S.; Ligtenberg, M. J. L.; Kibbelaar, R. E.; de Leng, W. W. J.; Schuuring, E.

2025-09-12 health systems and quality improvement
10.1101/2025.09.07.25335287 medRxiv
Show abstract

Molecular Diagnostics (MDx) is integral to the personalization of cancer treatment, offering both prognostic and predictive insights that inform therapeutic decision-making. However, the rapid advancement of MDx in oncology comes with significant challenges for the development, implementation, and adherence to clinical guidelines. A systematic approach is required to monitor clinical practice, including the utilization of MDx and the evaluation of patient outcomes. Such an approach can facilitate evidence-based updates to clinical guidelines, enhance adherence, and increase transparency in healthcare decision-making. This study aimed to develop a proof-of-principle practice monitor for MDx using two showcases, non-small cell lung cancer (NSCLC) and endometrial cancer (EC). Adopting a socio-technical framework, stakeholders were engaged to ensure alignment of the monitors content with clinical needs. Critical clinical questions were identified, and computational decision algorithms were developed based on current guidelines and expert consensus. Nationwide real-world data registries were analyzed to assess the availability, quality, and completeness of data in relation to the clinical questions and decision algorithms. Variables included in decision algorithms were largely present for NSCLC and EC, confirming the feasibility of nationwide monitoring. While some clinical questions could be fully or partially addressed with the available data, many could not. This was due to fragmented data, inconsistent coding practices, and the use of non-standardized free-text entries. Accordingly, this initiative addresses key challenges related to data standardization, computational representation of guidelines, and variability in clinical reporting. With strong national databases and infrastructure already in place, successful implementation now hinges on focused multidisciplinary and multi-organizational collaboration and a learning community to sustain ongoing progress. This approach assessed the feasibility and challenges of continuous evaluation of MDx practice in a proof-of-principle practice monitor, as an important step towards a Learning Health System.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.