The European Health Data Space holds significant promise for transforming healthcare, research and innovation across Europe. The main objectives of the EHDS, which the European Commission launched in 2022, revolve around improving healthcare delivery, research, policy-making and patient empowerment.
One of its key promises is advancing research and public health decision-making by facilitating the creation of large-scale, high-quality health data sets, which are essential for developing new treatments and precision medicine.
For individuals, the EHDS enables digital access to their electronic health records (EHRs), prescriptions, medical images, lab results, and discharge reports across all of Europe.
The challenge of harmonising data for seamless reuse
The European Commission has set an ambitious goal: By 2030, all citizens across Europe will have full possession of their data.
However, despite its potential, the current EHDS framework presents two critical challenges:
- While it is designed to benefit the broader population, it does not adequately address the needs and control of individual patients over their own data.
- The process of generating secondary data sets remains complex and inefficient due to the heterogeneity and lack of interoperability within and between individual patient records.
To optimise secondary data uses, personal health data should be curated first rather than extracted in their source format and then curated for each use case. Currently, the curation and publishing process of data is done after extraction for secondary use and is costly, time-consuming and manageable only by experts. As a result, we are underutilising large quantities of data, and there are limited benefits for the patient, who remains with heterogeneous data with suboptimal quality.
To address those challenges, a consortium of 14 partners launched AIDAVA, a Horizon Europe-funded project. Started in 2022, the project aims to address how patient health data is managed, integrated, curated and utilised across Europe by developing an AI-driven virtual assistant. Through AIDAVA, the partners aim to provide patients with high-quality personal health records, offering personalized care while facilitating seamless and high-quality secondary data use.
Accessing all their personal health information is challenging for citizens and patients because it is scattered across different systems. Even healthcare professionals don’t always have access to that data or don’t archive it properly. In addition, health data is in heterogeneous format and is error-prone (up to 10% of records contain potentially life-threatening errors). AIDAVA is looking for AI virtual assistants to help it facilitate the integration and quality enhancement process.
The ‘how’ to data reusability
This four-year project has two main objectives:
- Maximise automation in the curation of personal health data to increase its reusability.
- Test the solution with three concrete actions:
- Establish an EU-wide breast cancer registry, integrating data from three federated centres across different countries and languages.
- Compute automatically a smart risk score to support follow-up of cardiovascular patients with a recent history of myocardial infarction.
- Automatically generate the Individual Patient Summary (IPS) in the EHDS-required European Electronic Health Record Exchange Format (EEHRxF) format, decreasing the burden on data holder organisations to generate these emerging standards.
The consortium is focusing its efforts on data curation by integrating all the data from multiple sources connected to each individual, then cleaning and homogenising it. The AI-driven virtual assistant will assist individuals in curating the data and flag issues that request human intervention either by the patient or their dedicated curator.
“Most healthcare initiatives, including EHDS, are focusing on population data. This does not benefit each individual patient with patient records that remain heterogeneous and error-prone; in addition recurrent curation of source data for secondary use at population level is not sustainable,” says Isabelle de Zegher, MD, MSc, the clinical coordinator of AIDAVA with 30+ years’ experience in digital health.
“AIDAVA proposes to maximise automation of the curation of individual health records and minimise the need for human intervention with AI technologies. Results from the first generation of the prototype – tested by patients in hospitals – are promising and demonstrate that this might be the correct direction to solve the long-standing problem of lack of interoperability of healthcare,” she adds.
Isabelle de Zegher will be at HIMSS Europe in Paris from 10-12 June, presenting the progress of the AIDAVA project and the initial results.