From register data to useful information: Framework for automating real-world evidence reporting

Authors

• Jani Miettinen Institute of Clinical Medicine, University of Eastern Finland https://orcid.org/0000-0003-2239-5376
• Reijo Sund Institute of Clinical Medicine, University of Eastern Finland, Kuopio; Clinical Research Centre, Kuopio University Hospital, Kuopio https://orcid.org/0000-0002-6268-8117

DOI:

https://doi.org/10.23996/fjhw.178833

Keywords:

data science, big data, medical informatics, open source software, automation, health care research

Abstract

Real-world data (RWD) are increasingly used in scientific research. However, transforming such data into structured and reproducible scientific evidence remains challenging. We present an automated analytical framework for transforming real-world data into standardized and reproducible real-world evidence (RWE) research reports. The novelty of this work lies in the integration of data harmonization, automated statistical modeling, and report generation within a lightweight graphical user interface supported by a simple common data model.

The framework standardizes key analytical components, including cohort construction, time-to-event modelling, and descriptive and summary reporting. As a proof of concept, the system is demonstrated using longitudinal data from OSTPRE cohort (Kuopio Osteoporosis Risk Factor and Prevention Study). The implementation illustrates how an automated workflow can efficiently generate transparent and consistent RWE outputs suitable both for research and healthcare system evaluation.

The framework is scalable to broader data ecosystems, such as regional hospital data repositories, enabling more detailed analyses and the production of robust real-world evidence. This approach supports more efficient utilization of real-world data in scientific and clinical decision-making.