OntoHealth - problem-aware Semantic eHealth Services

An ever-increasing number of patients that comes along with a permanent increase of techniques and medical knowledge for diagnosis and treatment put high pressure on the healthcare system. In order to keep costs low and guarantee respectively increase quality of treatment there is, among other things, the need for information systems that are highly flexible and guarantee a high level of functional quality along the process of medical treatment. The concept of an Electronic Health Record (EHR) sets out to close the gap between institution-specific patient data and a comprehensive, longitudinal collection of a patient`s health and healthcare data that supports information exchange between various healthcare providers. Currently many countries are actually working on the implementation of national Electronic Health Records, aiming at the establishment of a basic EHR infrastructure. A problem, which is not tackled sufficiently by the majority of the current efforts, is the functional level of Electronic Health Records. Electronic Health Records are still very often regarded solely as records allowing basic access to data such as to view and retrieve existing documents and not as integrated, complex systems of data and functionality. Although EHRs are regarded as distributed, inter-organizational information systems from a data perspective, they are often (implicitly) regarded as monolithic software systems from a functional perspective. Even if projects use e.g. web service as their technological base, the general idea of dynamically orchestrated services (functions) or the idea of service (function) orchestration according to certain clinical problems as far as to the level of the user is not widely recognized. The projects pursues the following goals: Conceptualize and implement a workflow-enabled IHE-based trans-institutional information system architecture Determine and categorize functional respectively informational needs as well as non-functional requirements of health care professionals Design and implementation of a semantic service grid for Electronic Health Records based on the information system architecture Identify and prototypically implement selected common clinical services for the semantic service grid Evaluate all concepts with regard to their usefulness and ability to support health professionals The project results in several new and innovative concepts and solutions, including: an IHE-based information system that allows native, dynamic workflow support regardless of the type of the clinical workflow; a user-centered, workflow-oriented and (semi-) formal model to describe functional / informational needs as well as additional non-functional requirements to support automated, semantics-based service orchestration in a clinical context; a semantic-services-enabled EHR (semantic services grid) framework that allows for the (semi-) automatic, dynamic orchestration of services along clinical workflows and problems; selected common services in the context of diabetes.

Electronic documentation is seen as an essential part of a comprehensive healthcare system. Medical progress as well as technology developments is generating continuous amounts of patient data. Current systems for trans-institutional exchange of diagnoses, therapies, examination results should realize a meaningful use of healthcare data. With the advent of electronic health records (EHR) a document-based, inter organizational access of a lifelong medical history of all patients got achieved. Although existing standards and best practices used for the realization of an EHR ensure high syntactic and semantic interoperability, the functional focus of EHR is limited to the basic exchange of healthcare documents. This means, the end users of EHRs (physicians) can only query documents (e.g. reports, laboratory results, etc.) based on the document's metadata. The project OntoHealth extends the existing concept of an EHR with a functional layer in order to provide individual functions, which are able to process, transform and extract accessible healthcare data on a user- and context specific manner. The goal of the project is to investigate whether shared, semantic eHealth web services can be (semi-)automated selected and executed by individually set requirements. An ontology got developed that was used as a semantic description for both, the user requirements and the eHealth web services. All developed concepts are initially focused on diabetes Mellitus. Further, based on the service oriented architecture, even non-functional parameter (NFP) were integrated in the service delivery process: The end user describes the required needs or defines the problem (e.g. display a graphical trend but also of all measured HbA1c value of a patient considering the past 24 months or calculate the cardiovascular risk) but also includes desired NFP (e.g. performance, costs, availability). The system then tries to use this information in order to find suitable web services or if necessary compose new web services based on atomic services which are executed along the intended workflow. Essentially the idea of OntoHealth is to utilize web service delivery to answer the defined needs under consideration of NFP.