gms | German Medical Science

Introduction: The AKTIN Emergency Department (ED) Data Registry [1], a federated and distributed research infrastructure, faces challenges due to the increasing number of institutions and data requests. From 2018 to 2023, ED nodes grew from 16 to 62, and annual data requests rose from 25 to 629. The current manual request creation and management processes are time-consuming and lack a centralized tracking system, causing delays in request fulfillment and reliance on experts. Such administrative burdens are common in European health information systems [2], highlighting the need for technical solutions. To address these challenges, we are developing a software solution with user-friendly interfaces to streamline request creation and tracking, governance, and provenance management for expert and non-expert personnel. The objective of this work is to design and implement a prototype tailored to the needs of AKTIN Registry personnel.

State of the art: In recent years, health data networks like PopMedNet [3] have emerged to enhance EHR access. PopMedNet enables researchers to deposit, discover, and manage health data within a distributed network. Its menu-driven query builder simplifies creating standardized queries. However, customized research queries require manual data extraction and coordination across institutions, as each organization remains responsible for data management, provenance, and governance.

Concept: Employing Design Thinking [4], we interviewed AKTIN Registry personnel (n=4) about dissatisfactions with the current system and created personas based on operational roles. We converted interview insights into user stories (n=60), from which we extracted functional and non-functional requirements. In brainstorming sessions we generated solution concepts, leading to low-fidelity user-interface (UI) wireframes and backend architecture concepts. We refined these blueprints iteratively through user interviews and reviews by experienced software engineers. Following evolutionary prototyping principles, we implemented an initial software prototype.

Implementation: User interviews highlighted dissatisfaction with dispersed information and a need for automated serial query creation. The current system lacks query syntax validation, data provenance, and auditability, and assigns identical permissions to all registry personnel. Our backend concepts address these issues by centralizing request information and enabling automated creation and updates of serial queries, along with secure, hash-verified download of query results. A visual query builder simplifies the creation and validation of technical queries. All user actions are logged, and a role-based system ensures appropriate access for different personnel. We implemented a prototype web app using our institute-standards Java 17 with Spring Boot and Vue.js 3.4, featuring a complete UI based on the wireframes with static demo data, simulating the final application. Backend implementation is pending.

Lessons learned: We developed a software prototype to address operational challenges in a rapidly expanding health data network. While tailored to our specific use case, our solution may offer insights for similar efforts. Existing solutions like PopMedNet [3] were unsuitable due to our project environment and a design requiring technical expertise. PopMedNet lacks centralized data provenance, a feature AKTIN personnel requires, and we have implemented. Comprehensive evaluation of existing networks awaits full implementation, with rigorous software validation planned. The iterative development process will incorporate user feedback to ensure robustness and optimize usability.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.