Clinical Decision Support for Prosthodontic Diagnostic Index

Objectives: The Prosthodontic Diagnostic Index (PDI) is a classification tool for completely dentate, partially, and completely edentulous conditions. It provides a framework for organizing clinical observations and classifying dental prosthodontic patients based on complexity and helps clinicians make treatment and referral decisions. PDI is increasingly being adopted and there is a need for computerized Clinical Decision Support (CDS) tools that will automate the decision-making process for diagnostic classification. Such CDS tools can help in screening patients and can help simplify treatment planning and referral decision-making.
Methods: Here we report the development of CDS tools for completely dentate, partially, and completely dentate conditions. PROforma was the modeling language used to develop the models. The criteria defined in the PDI were converted to data elements and arguments for making recommendations. PROformajs was used to create and execute the models. PROformajs generates a web-based interface that allows a user to enter clinical observations. The CDS tools provide a diagnostic classification and a rationale based on the data entered. Scenario-based testing using synthetic observational data was performed to evaluate the ability of the CDS tools to generate a diagnostic classification (5 scenarios per condition).
Results: Modeling in PROforma identified 14 out of 53 criteria within all 3 PDI conditions that can benefit from further definition. The results of the testing showed that the CDS tools were able to provide an accurate classification of all (n=15) scenarios.
Conclusions: Defining terms like ‘oral manifestations of systemic diseases’ using coded terminology can help retrieve the data from electronic health records and further simplify the process of data collection. Further studies are needed to validate the tools and compare their performance to experts. We also need to investigate if CDS recommendations can change decision-making in real-world scenarios and improve health outcomes. Our initial results show that PROforma-based CDS tools are a viable approach for PDI classification and have potential applications in dental education and clinical practice.