@phdthesis{602, keywords = {Evidence-Based Medicine, Evidence Information, Clinical Practice Guidelines, Guideline Representation Languages, Guideline Representation Tools/ Decision-Making Process, Asbru, PROforma}, author = {Alime Öztürk}, title = {Embedding the evidence information in computer-supported guidelines into the decision-making process}, abstract = {

Clinical practice guidelines (CPGs) are widely used to support medical staff in treatment planning and decision-making during health care. To increase the effectiveness and the quality of health care, many research projects deal with computer-based representations and implementations of CPGs. Representation languages to formulate CPGs in a computer-interpretable way are complex, what makes the formulation process difficult and time-consuming. CPGs contain major recommendations about a certain disease that are usually based on clinical studies indicating the level of evidence and hence the strength of the recommendations. However, not all CPGs provide explicit information about the level of evidence or the strenght of recommendation (i.e., ungraded evidence information).

In this thesis we propose a methodology that supports guideline users during the decision-making process on the basis of a semi-formal representation of the evidence information that can be found in CPGs. A semi-formal representation is required to handle evidence information in computer-interpretable guideline representation languages. For this purpose, we have developed a meta schema that covers various kinds of grading systems including graded and ungraded evidence informaton. The classification of different recommendations in CPGs are one of the most important information sources to use. However, there is a lack of consensus amongst guideline developers, regarding those classification schemes. To address this problem, we mapped various kinds of grading systems into our meta schema. Furthermore, we extended two guideline representation languages (Asbru, PROforma) to model our meta schema.

Finally, we present the results of our qualitative study we performed with physicians, guideline developing organizations, and guideline representation language developers to examine the correctness, feasibility, and understandability of our meta schema and language extensions. The results of our evaluation indicate that using a semi-formal representation of the evidence information is of particular importance to facilitate the decision-making process.

}, year = {2007}, journal = {Institute of Visual Computing and Human-Centered Technology}, pages = {140}, publisher = {TU Wien}, address = {Vienna}, }