Our new paper on using >2 million national symptom surveys to devise an algorithm for predicting the probability of an individual to test positive for COVID-19 based on answers to 9 simple questions (age, gender, and symptoms)
cell.com/med/fulltext/S…
cell.com/med/fulltext/S…
We validated our model on held-out individuals from Israel (auROC of 0.737) and on independently-collected dataset from the U.S., U.K. and Sweden by @timspector
We find that loss of taste and small is the most distinctive symptom for predicting a positive COVID-19 test
We find interactions between several symptoms and age, suggesting variation in the clinical manifestation of the disease in different age groups
We find interactions between several symptoms and age, suggesting variation in the clinical manifestation of the disease in different age groups
Our tool can be used online to directing the limited testing resources through prioritization of individuals for testing, thus increasing the rate at which positive individuals can be found
individuals at high risk for a positive test result can also be isolated prior to testing
individuals at high risk for a positive test result can also be isolated prior to testing
Great work by
Saar Shoer Tal Karady @AyyaKeshet @smadarshilo @H_Rossman @GavrieliAmir @tomer1812 Amit Lavon Dmitry Kolobkov @KalkaIris @Nastya08434121
And collaboration with @timspector @RanBalicer
Saar Shoer Tal Karady @AyyaKeshet @smadarshilo @H_Rossman @GavrieliAmir @tomer1812 Amit Lavon Dmitry Kolobkov @KalkaIris @Nastya08434121
And collaboration with @timspector @RanBalicer
You can check out the tool here:
hasadna.github.io/avid-covider-p…
hasadna.github.io/avid-covider-p…
• • •
Missing some Tweet in this thread? You can try to
force a refresh
