Knowledge engineering for adverse drug event prevention: On the design and development of a uniform, contextualized and sustainable knowledge-based framework

Abstract

The primary aim of this work was the development of a uniform, contextualized and sustainable knowledge-based framework to support adverse drug event (ADE) prevention via Clinical Decision Support Systems (CDSSs). In this regard, the employed methodology involved first the systematic analysis and formalization of the knowledge sources elaborated in the scope of this work, through which an application-specific knowledge model has been defined. The entire framework architecture has been then specified and implemented by adopting Computer Interpretable Guidelines (CIGs) as the knowledge engineering formalism for its construction. The framework integrates diverse and dynamic knowledge sources in the form of rule-based ADE signals, all under a uniform Knowledge Base (KB) structure, according to the defined knowledge model. Equally important, it employs the means to contextualize the encapsulated knowledge, in order to provide appropriate support considering the specific local environment (hospital, medical department, language, etc.), as well as the mechanisms for knowledge querying, inference, sharing, and management. In this paper, we present thoroughly the establishment of the proposed knowledge framework by presenting the employed methodology and the results obtained as regards implementation, performance and validation aspects that highlight its applicability and virtue in medication safety.

Evaluation of the Dynamic Construct Competition Miner for an eHealth System

“…ADRs expression has a different meaning than "side effect", since side effect might also imply that the effects can be beneficial. A more general term, Adverse Drug Event (ADE) [11,6], refers to any injury caused by the drug (whether drugs were used at normal dosage and/or due to overdose) and any harm associated in such case, i.e. ADRs are a special type of ADEs.…”

Leveraging Medication Safety through Mobile Computing: Decision Support and Guidance Services for Adverse Drug Event Prevention

“… Patients/citizens that are interested in accessing knowledge on potential ADEs according to their medical profile or health condition. This concerns information like diagnosis and laboratory examination results (if any), as well as current medication treatment for which ADE prevention rules may be applicable [4]. The user's interest may stem from checking simple drug-drug interactions to searching whether a drug could be appropriate for their medical condition.…”

“…In [4] we presented a knowledge engineering framework aiming to represent and manage various ADE patterns, with major focus on novel, rule-based signals obtained through knowledge discovery techniques applied on Electronic Health Records across Europe, and validated via a knowledge elicitation phase. The framework provides the basis for developing contextualized decision support services for ADE prevention through alerts and recommendations based on the assessment of the respective medical data [9].…”

“…In particular, a RESTful Web service was implemented in order to handle the requests from the mobile application. The service accepts the patient data along with contextual information (like filtering parameters and thresholds for rule firing [4]) in an XML-based format and queries the respective Clinical Decision Support System (CDSS). The CDSS response is also encoded in XML and returned to the corresponding application, which is hosted in our case in a mobile terminal.…”

“…In this paper, we present an approach towards ubiquitous access to contextualized medication-related guidance and decision support services targeting patients/citizens and healthcare professionals, respectively. Our work concerns the design and development of a "mobile component" aiming to exploit a novel decision support framework that was constructed in the scope of the PSIP EU-funded project [4].…”