Partnerships serving patients

Heidelberg University Hospital has focused on building partnerships for over 10 years. It is therefore important to have an IT structure that advances those partnerships and ensures a smooth exchange of data. The first pilot projects were successful.

“We wanted to keep complexity to a minimum in the PEPA project.”

Oliver Reinhard
Function: Head of the Department of Medical Information Systems
Organization: University Hospital Heidelberg

Heidelberg has the highest physician density in Baden-Württemberg and over 30 hospitals. Residents and patients from all over the world value the excellent medical care there. Some of the credit also goes to the University of Heidelberg.

We believe it is important to pool our efforts in defined treatment processes for the benefit of patients, through coordinated partner profiling, staff sharing and rotation models, and cooperative work, and not to weaken one another by competing. This is in keeping with our role as a tertiary-care hospital with a long track record in research. At the same time, we can retain our unique features and our top-ranking position both within Germany and internationally. Our partner institutions handle the majority of the primary care and standard treatment in the region, but have quick access to the University Hospital’s expertise when needed; as teaching hospitals, they also facilitate medical staff training. This networked approach with cross-sector collaborations also extends to ambulatory facilities, such as specialized practices, primary care doctors and specialists.

“The patients play a vital role, so we developed a plan for a personalized, cross-enterprise health and patient record (PEPA) early on.”

Prof. Dr. med. Björn Bergh
Funktion: Director Department of Medical Information Systems
Organization: University Hospital Heidelberg


Achieving long-term viability together with innovative information technology
Because of the classical IT structures and legal requirements for data protection, electronic patient records (EPAs) are still very limited and facility-specific. This is an obstacle to cross-enterprise and cross-sectional cooperation that benefits patients. Thus the acquisition and delivery of information remains a time-intensive process, and unfortunately duplicate requests and tests are still an everyday occurrence.

We made a start with the introduction of teleradiology in 2003, which was an early and significant step toward cross-enterprise IT in the region. We then moved ahead with this development process in other specialized fields, such as teleneurology. In teleradiology, the emphasis is on consulting activity in the field of medical imaging. This does a great deal to facilitate collaboration, but many other areas with the potential for cooperation are not covered.

This is why we have spent over 10 years working on the question of how we can design a cross-enterprise electronic patient record that is useful to everyone. Prof. Bergh’s article in this issue compares the various records systems. The PEPA architecture allows patients to access their cross-enterprise health and patient records personally via a patient portal and to use permissions to control healthcare providers’ access. A PEPA becomes fully functional once it is linked to the respective facility-based records (EPAs), such as hospital information systems and medical practice systems, and when —with the patient’s consent— the relevant patient information, documents, and image data are transferred from the EPAs to the PEPA, preferably automatically. When the PEPA concept was developed, we asked ourselves how we could put it into practice. The following factors played a key role:

  • Deciding to use a standards-based architecture (IHE)
  • Choosing suitable industry partners (ICW, CHILI GmbH)
  • Deep integration into the primary systems
  • Launching a cross-enterprise EPA (eEPA) with the prospect of developing it into a PEPA


Standards-based architecture (IHE): A cross-enterprise patient record has to be linked to various primary systems from a wide variety of manufacturers; at the same time, it must be able to communicate with other regional records systems (national and international) in the future. Sustainability can only be achieved with standards-based communication. IHE-based approaches have since proven successful at the national and international level.

Industry partners: For component-based implementation, we chose ICW to handle the MPI and record components (XDS registry and repository, professional portal), and CHILI GmbH for the image components (image registration and display). Even then we were impressed by their expertise in the IHE field and their willingness to make the PEPA journey with us.

Deep integration into primary systems: The interplay of the systems should appear as integrated as possible for the clinical user. Consent management is dealt with immediately during the admission process and documented in the hospital’s primary system. If the patient consents, data transfer to the PEPA starts once the admission is registered, and access is immediately enabled for doctors in that particular facility. Doctors have context-based access to records in the primary system. The departments decide which document types (e.g. discharge letters or reports) will be automatically transferred to the PEPA once consent has been given and documents are released. Doctors can send specific documents of other types from the primary system to the PEPA. All image data for a given patient will also be sent to the PEPA if consent has been given.

“In the INFOPAT project, we developed a patient portal, a medication module and a linked research platform for quality assurance.”

Dr. Oliver Heinze
Function: Project Manager Infopat-Project
Organization: University Hospital Heidelberg


Multi-stage plan for implementation and launch:
The full PEPA architecture requires numerous interfaces, components (e.g. a patient portal), and carefully coordinated processes for all the parties involved. At that time, no one had any experience with patient involvement yet. To keep complexity to a minimum, we decided we would first set up an eEPA for everyday use in the hospitals, and then integrate private-practice physicians and patients in subsequent stages.

Current status and experiences
To date the PEPA system has been adopted in selected areas of Heidelberg University Hospital and the whole of the Thoraxklinik (Center for Thoracic Surgery), and it currently contains some 26,000 records. Random sampling at the Thoraxklinik showed a consent rate of about 95 percent. As the records become more comprehensive, the physicians involved are more motivated to use them. Many phone calls are no longer necessary because doctors can easily pull documents from partner institutions into their own records.

Data protection-compliant consent management is an administrative challenge. The first step is to lay the groundwork for a secure network architecture. The steps toward IHE-compliant implementation must then be defined, concerning issues like setting up an OID structure or using IHE XDS metadata. Although a great deal of effort is involved at first, it is worthwhile. Many industry partners are now very open to the idea of integration via IHE. For example, the in-house archive viewer (Agfa HYDMedia) is already able to display external patient documents from the PEPA in addition to its own archived documents in a test installation. The first proof of concept to integrate physician information systems from CompuGroup Medical using IHE was performed successfully. And the initial results from the INFOPAT study, which examines selected patients’ use of PEPAs via a patient portal, are a cause for optimism.

Nevertheless, linking partners to achieve deep integration is a challenging and complex process. We need to continue working with our industry partners to simplify the process. There will also be further technical, administrative, and cultural challenges to deal with in the nationwide integration of private-practice physicians and patients.

Looking ahead:
The PEPA project’s goal for 2017 is to integrate another partner facility as well as physician information systems. A proof of concept will also be conducted for patient apps.