Evidence based decision support system for choosing appropriate radiologic examination

Data about the excess and inappropriate use of radiological teststriggered the systematic studiesthose discussthe reasons and solutions ofthe problem.Advances in information technologies influenced the complex decision making process in medical practice. Since radiology is one of the most rapidly developing and technologically advanced fields of medicine, it is expected to see more Decision Supportsystems(DSS) to be developed and used in future. Currently, there are only a few local DSS studies about radiological test appropriateness in the literature.In this thesis study, a DSS for assisting physicians about appropriate radiological test selection is developed, evaluated and discussed. Worldwide accepted American College of Radiology appropriateness criteria (ACRAC) was used as the base of the study. The scope covers the field of neuroradiology. An expert panel transformed the criteria into decision tree diagrams and the system was developed as an XML based web application. The system (RADDES) is designed to be a fast, basic tool that suits to common decision making processes in neuroradiology. Before testing RADDES, scenarios were given to the volunteer physicians and their radiological test decisions are collected for each scenario. The test decisions were compared with ACRAC criteria and order appropriateness was determined. In the second step, the physicians were asked to evaluate another scenario sets and they were advised to use RADDES in case they need assist during the decision making process. Orders from the first and second steps were then compared in terms of appropriateness. Statistical analysis showed that the decisions with RADDES support significantly increased the level of appropriateness. Moreover, RADDES was evaluated in terms of usability and the evaluation showed that the system has high usability.To our knowledge, this study is the first in the literature to compare the physicians? radiological test orders with evidence based guideline, ACRAC, and evaluate the appropriateness accordingly. It is also the first XML based DSS web application which supports mobile access for all mobile devices for order appropriateness. It covers the branch of neuroradiology as a whole and provides evaluation as well as usability analysis for such a DSS for first time in the literature. Moreover, RADDES is the first system that is developed in accordance with ACRAC criteria while using decision tree transformations as the base approach.In conclusion, radiological order appropriatenessis an important concept that needsto be improved in medical practice. DSSs which are based on ACRAC or other evidence based criteria will be beneficial in terms of improving the rate of appropriate radiological test orders. Decision making methodologiessuch as decision trees would be useful for developing such systems.Key Words: Order Appropriateness, Neuroradiology, Decision Support System