Back

Triaging patients in the outbreak of the 2019 novel coronavirus

Huang, G.; Zeng, W.; Wang, W.; Song, Y.; Mo, X.; Li, J.; Wu, P.; Wang, R.; Zhou, F.; Wu, J.; Yi, B.; Xiong, Z.; Zhou, L.; Wang, F.; Tian, Y.; Hu, W.; Xu, X.; Zhai, R.; Yuan, K.; Li, X.; Qiu, X.; Qiu, J.; Wang, A.

2020-03-16 infectious diseases
10.1101/2020.03.13.20035212 medRxiv
Show abstract

In the end of 2019, the epidemic of a new coronavirus (SARS-CoV-2) occurred in Wuhan and spread rapidly. Changsha, a city located south to the epicenter, was soon impacted. To control the transmission of the coronavirus and avoid nosocomial infection, triage procedures based on epidemiology were implemented in a local hospital of the city. This retrospective study analyzed the data collected during the triage period and found that COVID-19 patients were enriched seven folds into the Section A designated for rapid detection and quarantine. On the other side, roughly triple amounts of visits were received at the Section B for patients without obvious epidemiological history. Eight COVID-19 cases were spotted out of 247 suspected patients. More than 50% of the suspected patients were submitted to multiple rounds of nucleic acid analysis for SARS-CoV-2 infection. Of the 239 patients who were diagnosed as negative of the virus infection,188 were successfully revisited and none was reported as a COVID-19 case. Of the eight COVID-19 patients, three were confirmed only after multiple rounds of nucleic acid analysis. Besides comorbidities, delayed sharing of epidemiological history added another layer of complexity to the diagnosis in practice. While SARS-CoV-2 epidemic is being alerted in many countries, our report will be helpful to other colleagues in rapid identification of COVID-19 cases and controlling the transmission of the disease.

Matching journals

The top 8 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.