Data Spaces for the Statutory Health Insurance Companies for Data Analysis

Practical example and current challenges

• The statutory health insurers (SHI) carry out various forms of healthcare analytics. The available data from the routine procedures of the SHI (billing data from the outpatient and inpatient areas, pharmacy billing data and billing data from other service providers) are used and analyzed. The focus is on identifying specific care needs of a health insurance fund's community of insureds. One goal is to identify basic health care needs and to set up measures to meet these needs, for example, with the help of selective contracts. Machine learning is used to quantify specific risks for individual insureds in order to offer preventive measures to insureds with high risks in particular.
• These analyses are usually carried out by research institutes commissioned by the health insurers or as part of the insurers' own data analyses in the data warehouse systems of the health insurers. It must be noted that these data are highly sensitive health data that may only be accessed using solutions that meet the highest level of protection requirements. Data protection is therefore of particular importance in these processes. In addition, high regulatory requirements for data processing must be observed specifically for health insurance companies using cloud services.
• The use case is designed to combine the perspectives of health insurance, IT providers and data analytics processes into a homogeneous solution that complies with all existing data protection requirements. This is accomplished by setting up a Gaia-X-compliant data room that allows comprehensive analysis procedures with regard to the data stored in it. In addition, a secure transfer of the relevant data from the health insurance company to the data room and of the available results from the data room back to the health insurance company will be made possible.
• Health insurance companies are given the opportunity to enter standardized pseudonymized data into the data room. They can sovereignly decide for which legally legitimized processing purpose their data will be used.
• For health insurance companies, the main benefit is the secure and authorized use of cloud infrastructures in order to be able to tap into the advantages of these systems over on-premise solutions, even in sensitive data areas. Data can thus be shared with cooperation partners and also competitors in a protected space in a standardized and use-case-related manner, in compliance with data protection regulations, in order to be able to use larger data sets for training AI and machine learning. This leads to a significant improvement in quality and cost-effectiveness and contributes to a more efficient scaling of computing power compared to on-premise solutions. For the industrial companies to be involved, new business cases based on this will open up in the future.

What added value does the "Gaia-X project" offer?

• The Gaia-X architecture is to be used to establish secure data rooms within cloud infrastructures for sensitive data of the SHI. Among other things, the requirements of GDPR and BAS should be mentioned here. The pilot use case is intended to show that Gaia-X makes it easier to make cloud infrastructures in the public sector accessible for sensitive data.
• Gaia-X does not focus solely on cloud solutions, but its architecture also provides for edge solu-tions in which local services can be linked with cloud services to add value. This enables SHI providers to link IT services that were previously operated locally, such as core process-related procedures, with cloud systems, at least in part.
• Gaia-X will also act as a kind of "seal of approval" in the future. By proving that a company can operate services in compliance with the Gaia-X architecture, it will be possible to gain easier ac-cess to the market for SHI providers. Other expected added values are a more economical and productive use of AI and machine learning through larger data volumes and technical scalability, as well as the elimination of transaction costs and efforts in merging data.

Use Case Team

• André Czernia – BearingPoint
• Jan Steffen – BearingPoint
• Stefan Pechardscheck – BearingPoint