Developing a clinical decision support system for prediction of warfarin dosage based on computer interpretable guideline

Anticoagulation therapy is one of the most important strategies for preventing clot formation and subsequent stroke, with warfarin representing the most widely used oral anticoagulant. However, predicting the outcomes of warfarin administration poses a major challenge for physicians because of the narrow boundary between the therapeutic and toxic levels of warfarin. Clinical decision support systems (CDSSs) can be used as a tool to improve both adherence to clinical guidelines and transfer of evidence-based knowledge to daily clinical practice for dose adjustment, thereby helping reduce medical errors. The aim of this study was to develop a prototype CDSS for predicting warfarin doses according to computerized clinical guidelines.

This applied developmental study involving a qualitative design was conducted in two major steps.First, computer-interpretable guidelines were extracted from existing clinical guidelines as a workflow diagram for warfarin therapy and were subsequently evaluated by an expert panel. Second, a prototype CDSS was designed with PHP programming language and SQL database, and Nielson’s heuristic evaluation checklist was used for usability testing.

In the first step, the findings were presented in two main workflows and two sub-workflows. In the second step, a prototype CDSS was designed. The overall usability of the prototype was found to be at a "relatively acceptable" level, with a rating percentage of 92.09.

The usability evaluation results suggest that CDSSs similar to the one presented herein could serve as valuable clinical decision support tools for estimating warfarin dosage. These promising results call for further research aimed at exploring the feasibility of implementing such systems in clinical settings.