COVID-19 Risk in ONcology Evaluation Tool
CORONET is an online tool to support decisions regarding hospital admissions or discharge in cancer patients presenting with symptoms of COVID-19 and the likely severity of illness. It is based on real world patient data. This the 2nd version of the model. Detailed information as to how version 2 of the tool was created and details regarding the datasets can be found in the JCO clinical cancer informatics. Version 1 is available at medRxiv. We would be grateful for feedback as to utility of this tool (email the-christie.coronet@nhs.net or tweet @beckilee @CORONET). Version 2 uses data from many countries and new analysis has resulted in an update of the features important for the model and a validation on an external cohort of cancer patients. The model was recently validated on the Omicron variant (details in Cancers).
CORONET asks for some details about the patient, their cancer and blood test results on presentation to hospital with symptoms of COVID-19. It then uses data about the admission, requirement for oxygen and survival of similar patients in the past to show likely outcome of the patient.
CORONET is intended for use by health care professionals or by patients and their families in consultation with a medical professional.
CORONET is based on data from cancer patients presenting with COVID-19. It is designed to support decisions as to whether patients should be admitted due to COVID-19 but not for a cancer related problem. Of note it does not take into account new treatments for COVID-19, which may alter the clinical course - this will be a feature that will be added in future versions.
Results
STOP AND CHECK: Has the patient got an oncology problem that requires admission?
CORONET only advises if a patient should be admitted due to COVID-19; it is not designed to aid a decision as to whether to admit due to a cancer or cancer treatment related problem.
Action:
Predicted Score:
-
Information About The Plot
The plot shows all patients used for training of the CORONET model. Each dot represents an individual patient. The colour corresponds to their true outcome. The location on the X-axis is determined by the CORONET score based on the individual’s data.
All patients are sorted from left to right according to the predicted CORONET score. The plot allows you to locate your patient in the whole cohort, considering both true outcomes and CORONET recommendations.
-
Learn More About the Model’s Prediction
Learn More About Most Similar Patients
Most Similar Patients
Your Patient 1st Patient 2nd Patient 3rdPatient 4th Patient 5th Patient Outcome Admitted/Discharged Required O₂ Death Due to COVID-19 CORONET Score CORONET Recommendation Age Biological Sex Number of Comorbidities ECOG Performance Status Cancer Type Early/Advanced stage Haem/Solid Solid Cancer Stage Chemotherapy Immunotherapy Targeted Therapy Radiotherapy Treatment Intent C-Reactive Protein (mg/L) NEWS2 Albumin (g/L) Neutrophils (x10⁹/L) Platelets (x10⁹/L) Lymphocytes (x10⁹/L) Neutrophil : Lymphocyte Ratio Lactate Dehydrogenase (IU/L) Consciousness Respiratory Rate (breaths per minute) Oxygen Saturation (%) The table presents the 5 patients in the CORONET dataset who are most similar to your patient. The patients are presented from left to right in decreasing order of similarity, i.e., the 1st patient is the most similar one. The similarity was calculated using 10 features, weighted according to their contribution to the prediction for your patient (see Important Features Contributing to the Model Prediction). The more the feature contributes to the prediction, the more important it is when considering similarity.
Disclaimer: CORONET dataset is available upon request to the author. However it may not include all details due to information governance regulations. Any attempts to reproduce the dataset based on similar patients presented in the CORONET online tool are not permitted.