Clinical Characteristics of Severe COVID-19 Patients During Omicron Epidemic and a Nomogram Model Integrating

Yanfei Lu1,2, Wenying Xia1,2, Shuxian Miao1,2

  • 1Department of Laboratory Medicine, Jiangsu Province Hospital and Nanjing Medical University First Affiliated Hospital, Nanjing, People's Republic of China.

PubMed

Insights

Severe COVID-19 has a high mortality rate, with cell-free DNA (cfDNA) levels above 97.67 ng/mL significantly increasing risk. A nomogram incorporating age, intubation, shock, cfDNA, and BUN accurately predicts mortality in these patients.

Area of Science:

  • Infectious Diseases
  • Critical Care Medicine
  • Biomarkers

Background:

  • The Omicron variant surge presented challenges in managing severe COVID-19.
  • Understanding mortality risk factors and developing predictive tools are crucial for patient outcomes.

Purpose of the Study:

  • Investigate clinical characteristics and mortality risk factors in severe COVID-19 patients during the Omicron wave.
  • Evaluate the clinical utility of plasma cell-free DNA (cfDNA) as a mortality predictor.
  • Develop and validate a nomogram for predicting patient mortality.

Main Methods:

  • Retrospective analysis of 282 severe COVID-19 patients (December 2022-January 2023).
  • Comparison of clinical data, laboratory indicators, and cfDNA levels between survival and death groups.
  • Logistic regression for identifying independent risk factors and nomogram construction using R software.

Main Results:

  • Mortality rate was 55.7% in the severe COVID-19 cohort (median age 80).
  • Independent risk factors for death included age, tracheal intubation, shock, cfDNA, and blood urea nitrogen (BUN).
  • Plasma cfDNA demonstrated strong predictive value (AUC=0.805); the nomogram achieved high accuracy (AUC=0.856).

Conclusions:

  • Severe COVID-19 carries a high mortality risk.
  • Elevated cfDNA levels (≥97.67 ng/mL) are associated with increased mortality.
  • The developed nomogram integrating clinical and cfDNA data offers accurate and consistent mortality prediction.
Abstract