Ebola RISK is a proof-of-concept app designed to illustrate how Machine Learning models for clinical prognosis could inform patient care and management. It offers treatment options compiled from World Health Organizations guidelines for hemorrhagic fever patients, prioritized by predicted mortality risk. The app and the underlying models are described in this paper:
http://biorxiv.org/cgi/content/short/294587v3
Use the predictions and clinical management recommendations from this app in combination with clinical judgement.
The app incorporates prognosis predictors for Ebola Virus Disease (EVD), trained on clinical data collected in 5 Emergency Treatment Units in Liberia and Sierra Leone, installed and operated by International Medical Corps during the 2014-14 outbreak. It offers to visualization options, one displaying the contributors of each predictor in the model to the overall risk, and the the other showing the most significant predictors and the treatment options available for each clinical sign/symptom. These options were compiled from the "Clinical management of patients with viral hemorrhagic fever" and "Manual for the care and management of patients in Ebola Care Units/Community Care Centers" guides from WHO.