Leveraging Drools Workbench for Effective Clinical Decision Support System Design

This paper explores the conception, development, and real-world evaluation of a Clinical Decision Support System (CDSS) aimed at enhancing health-care decision-making. Utilizing the Drools Workbench for rule-based logic, a visual clinical flowchart, and seamless integration with Fast Healthcare Interoperability Resources (FHIR) input, the CDSS offers a robust solution for dynamic and evidence-based decision support. In this paper, we present a generalized method for creating complex CDSS using Drools Workbench. Our approach begins with original clinical guidelines, which are converted into clinical flowcharts and computer interpretable guidelines. We then perform a guideline reconciliation process on these computer interpretable guidelines to accommodate comorbidity requirements. Clinical concepts extracted from the reconciled guidelines are translated into Domain-Specific Language (DSL) files in Drools Workbench. Based on these DSL files and computer interpretable guidelines, we create a set of business processes that form a comprehensive CDSS for individual diseases. By running multiple CDSS within a single system, we develop a CDSS for comorbidities. Functional testing is conducted by both technical personnel and healthcare professionals to ensure that the output aligns with real-world practices. The innovation of this paper lies in presenting a more general method for creating CDSS with minimal programmer assistance, addressing not just a single disease but a range of diseases, with consideration of their comorbidities.