(H-295) Web-based Algorithms for Clinical Vascular Risk Management

Introduction:: Algorithmic guidelines used for categorizing surgical risk have been established for aortic dissections and postoperative infection risk. The widespread clinical adoption of these algorithms is limited by ease-of-access and overall complexity. As a result, these guidelines are not readily available or clinically translatable. We introduce vasculink.net as a free-to-use repository of algorithms designed for vascular surgeons to make better informed clinical decisions.

Materials and Methods:: A website was created using the Flutter Software Development Kit to display three vascular guideline algorithms. Each algorithm was converted into an interactive form, allowing physicians to automatically input values and calculate predicted scores. The “Management of Aortic Graft Infection Collaboration” (MAGIC) algorithm classifies aortic graft infection by subdividing criteria into “major” or “minor” categories. The Surgical Site Infection (SSI) model stratifies patients based on their postoperative infection risk. Finally, the type B aortic dissection algorithm categorizes acuity of dissection by analyzing chronicity, aortic arch anatomy, and additional complications.

Results, Conclusions, and Discussions:: Vasculator.net hosts three risk prediction models. The first algorithm, Vasculink, was developed as an IOS application, which allowed users to download a mobile version. However, due to licensing limitations, the mobile app was converted to a web-application to increase the availability of the application. VascuM.A.G.I.C. was similarly developed as a web-application and utilizes flutter state-management techniques to help users predict risk of graft infection. Finally, VascuDissection incorporates a different state-management technique to provide live updates based on togglable risk inputs. Since the launch of the website, vascular surgeons have found these applications to be easy to use and effective as the web-based design allows users to access these algorithms on their devices such as phones, tablets, and laptop computers. Protected patient data is not used in any aspect of website or applications.

Vasculator’s suite of algorithm-based applications is freely available to the medical community based on current guidelines. The three applications serve as an aid for vascular surgeons and provide relevant clinical translations. Using key diagnostic criteria as inputs, the vasculator apps calculate predicted risk scores for aortic graft complication, postoperative infection, and aortic dissections. Although Vasculator has already proven useful in some practices, the predictive accuracy of the algorithms require further validation. Future steps include expanding the vasculator repository and introducing clinical decision making abilities.



Acknowledgements (Optional): :

References (Optional): :