Back

Data-driven prognosis for COVID-19 patients based on symptoms and age

Paul, S.; Lorin, E.

2021-11-09 infectious diseases
10.1101/2021.11.08.21266091 medRxiv
Show abstract

In this article, we develop an algorithm and a computational code to extract, analyze and compress the relevant information from the publicly available database of Canadian COVID-19 patients. We digitize the symptoms, that is, we assign a label/code as an integer variable for all possible combinations of various symptoms. We introduce a digital code for individual patient and divide all patients into a myriad of groups based on symptoms and age. In addition, we develop an electronic application (app) that allows for a rapid digital prognosis of COVID-19 patients, and provides individual patient prognosis on chance of recovery, average recovery period, etc. using the information, extracted from the database. This tool is aimed to assist health specialists in their decision regarding COVID-19 patients, based on symptoms and age of the patient. This novel approach can be used to develop similar applications for other diseases.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.