AIR_PTE: AI-based risk prediction and treatment effect estimation based on health data

AIR_PTE aims to develop artificial intelligence (AI) based methods to improve and automate treatment effect estimation based on health insurance claims data in Canada and Germany. Both Canada and Germany can build upon large representative samples of long-term health claims data, well suited to obtain real world evidence and to exploit AI based methods.

The methods will be developed and compared on the example of the current therapeutic options to treat venous thromboembolism (VTE) on Canadian and German health care claims. The best methods will be generalized for treatments of other indications and population-based care models, to assess the robustness of the AI methods across study populations.

The project’s goal is to support multiple evaluations and therapy decisions with modern AI methods. The analysis platforms SAHRA (BMWK funding program "Smart Data"), EVA (ingef spectrumk) and Macadamian HealthConnect Platform™ will be used.

In the project, experience from the operation of digital health platforms, especially with regard to data protection and user acceptance, will be combined with experience from the application of innovative modeling methods to historical health insurance data. The result will provide reliable methods for treatment effect estimation and for personalized decision support at the point of care in Germany, Europe and Canada.

Partners:
• DCC Risikoanalytik GmbH (Project lead, Germany)
• Ingef- Institut für Angewandte Gesundheitsforschung (Germany)
• McGill University Montreal, Dep. of Medicine (Canada)
• Macadamian HealthConnect (Canada)

Project leader:
DCC Risikoanalytik GmbH

Project duration: Sept. 2020 - Aug. 2022

Total volume: 1.0 million euros

Funding volume: 0.7 million euros