Related Experiment Video
Updated: Oct 17, 2025
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
Early predictors and screening tool developing for severe patients with COVID-19
Le Fang1, Huashan Xie2, Lingyun Liu2
1Zhejiang Provincial Center for Disease Control and Prevention, 3399 Binsheng Road, Binjiang District, Hangzhou, Zhejiang, China.
Insights
A new screening tool identifies severe COVID-19 cases early using easily obtainable predictors like age and chronic kidney disease. This aids in reducing mortality during health service overload.
Area of Science:
- Infectious Diseases
- Public Health
- Epidemiology
Background:
- Coronavirus disease 2019 (COVID-19) pandemic necessitates early identification of high-risk individuals to reduce mortality.
- Accessible screening tools are crucial, as not all predictors are universally available in healthcare settings.
Purpose of the Study:
- To develop and validate a simple screening tool for early prediction of severe COVID-19 cases.
- To identify easily obtainable predictors for risk stratification in COVID-19 patients.
Main Methods:
- Retrospective study of 813 confirmed COVID-19 cases.
- Logistic regression analysis to select key predictors.
- Development and ROC curve analysis of a novel screening tool.
Main Results:
- Seven early predictors identified: chronic kidney disease, age >60, low lymphocyte count, high Neutrophil to Lymphocyte Ratio, high fever, male sex, and cardiovascular disease.
- The screening tool achieved an Area Under the ROC Curve of 0.798, with 72.0% sensitivity and 75.3% specificity at a cut-off >4.5.
Conclusions:
- The developed screening tool offers a practical approach for early prediction of severe COVID-19.
- It is particularly valuable for healthcare systems facing overload, enabling timely alerts for at-risk patients.
Background:
Coronavirus disease 2019 (COVID-19) is a declared global pandemic, causing a lot of death. How to quickly screen risk population for severe patients is essential for decreasing the mortality. Many of the predictors might not be available in all hospitals, so it is necessary to develop a simpler screening tool with predictors which can be easily obtained for wide wise.
Methods:
This retrospective study included all the 813 confirmed cases diagnosed with COVID-19 before March 2nd, 2020 in a city of Hubei Province in China. Data of the COVID-19 patients including clinical and epidemiological features were collected through Chinese Disease Control and Prevention Information System. Predictors were selected by logistic regression, and then categorized to four different level risk factors. A screening tool for severe patient with COVID-19 was developed and tested by ROC curve.
Results:
Seven early predictors for severe patients with COVID-19 were selected, including chronic kidney disease (OR 14.7), age above 60 (OR 5.6), lymphocyte count less than < 0.8 × 109 per L (OR 2.5), Neutrophil to Lymphocyte Ratio larger than 4.7 (OR 2.2), high fever with temperature ≥ 38.5℃ (OR 2.2), male (OR 2.2), cardiovascular related diseases (OR 2.0). The Area Under the ROC Curve of the screening tool developed by above seven predictors was 0.798 (95% CI 0.747-0.849), and its best cut-off value is > 4.5, with sensitivity 72.0% and specificity 75.3%.
Conclusions:
This newly developed screening tool can be a good choice for early prediction and alert for severe case especially in the condition of overload health service.
Related Concept Videos
01:27Chronic Obstructive Pulmonary Disease-IV: Assessement and Diagnostic Studies
Medical History
01:30Pneumonia III: Complications and Assessment
01:29Pulmonary Embolism II: Diagnostic Studies and Interprofessional Care