Back

Features of C-reactive protein in COVID-19 patients within various period: a cohort study

Huang, G.; Qu, G.; Yu, H.; Zhang, M.; Song, X.; Chen, L.; Zhu, H.; Wang, Y.; Pei, B.

2020-10-27 infectious diseases
10.1101/2020.10.26.20219360 medRxiv
Show abstract

BACKGROUNDCoronavirus disease 2019 (COVID-19) has been declared as a threat to the global. Due to the lack of efficient treatments, indicators were urgently needed during the evolvement of disease to analyze the illness and prognosis and prevent the aggravation of COVID-19. METHODSAll laboratory confirmed COVID-19 patients hospitalized in Xiangyang No.1 Peoples Hospital were included. Patients general information, clinical type, CRP value and outcome were collected. CRP values of all patients during disease course from different initial time were analyzed. RESULTSThe 131 enrolled patients were 50.13{+/-}17.13 years old. All cases underwent 724 tests of CRP since symptom onset, 53.18% of the test results were abnormal and the median value was 9.52(2.63-34.10) mg/L. The first median value on the day 8 from exposure onset was 39.08(11.92-47.89) mg/L then fluctuated around it until the day 28. The CRP median increased from 15.93 mg/L to 41.44 mg/L and then decreased to 18.26 mg/L before transformation of severe type, and then increased to 62.25 mg/L on the transforming date. Conversely, the CRP median increased from 56.17 mg/L 102.75 mg/L before transformation of critical type but decreased to 68.68 mg/L on the transforming date. The changes of CRP median over time before death ranged from 77.77 mg/L to 133.52 mg/L. CONCLUSIONSCRP increased before symptom onset and substantially increased during the early-to-mid stage (especially early stage), which was different from other virus-infected diseases. The changes of CRP before the transformation of clinical type was inconsistent with the aggravating of illness. And the CRP maintained over 100.00 mg/L prompted poor prognosis.

Matching journals

The top 10 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.