Back

A Multi-center Study of COVID-19 with Multivariate Prognostic Analysis

Zeng, W.; Feng, X.; Huang, J.; Du, C.; Qu, D.; Zhang, X.; Zhang, j.

2020-09-28 respiratory medicine
10.1101/2020.09.26.20202234 medRxiv
Show abstract

PurposeCoronavirus disease (COVID-19) pandemic is now a global health concern. However, there is no detailed analysis of the factors related to patients improvement. Patients and methodsWe compared the clinical characteristics, laboratory findings, CT images, and treatment of COVID-19 patients from two different cities in China. One hundred and sixty-nine patients were recruited from January 27 to March 17, 2020 at five hospitals in Hubei and Guangxi. They were divided into four groups according to age and into two groups according to presence of comorbidities. Multivariate statistical analyses were performed for the prognosis of the disease. ResultsFifty-two patients (30.8%) had comorbidities, and the percentage of critical COVID-19was higher in the comorbidities group (11.6%vs.0.9%, p<0.05). Older patients had higher proportion of severe or critical disease. The results showed that lymphocyte count was significantly associated with the number of days from positive COVID-19 nucleic acid test to negative test; number of days from onset of symptoms to confirmation of diagnosis was significantly associated with the time it took for symptoms to improve; and number of days from onset of symptoms to confirmation of diagnosis and disease severity were significantly associated with chest computed tomography improvement. ConclusionsAge, comorbidities, lymphocyte count, and SpO2 may predict the risk of severity of COVID-19. Early isolation, early diagnosis, and early initiation of management can slow down the progression and spread of COVID-19. Key PointsAge and comorbidities can predict the risk of severity of COVID-19, Lymphocyte count and SpO2 may predict the risk of severity of COVID-19. Early isolation, Early diagnosis can slow down the progression of COVID-19

Matching journals

The top 4 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.