Based on the characteristics of symptoms, differences in physical and chemical examinations and tests, and characteristics of traditional Chinese medicine (TCM) disease differentiation and syndrome differentiation, the evaluation indicators for the clinical early warning model were screened, and the "three-level early warning mechanism for COVID-19 patients at high risks, with progression to severe cases and sequelae was constructed.

Adopt multi-level evidence base architecture, build "multi-source data analysis, feature extraction, multi-level evidence extraction, active The integrated Chinese and Western medicine big data platform integrates the original evidence processing into a multi-level evidence base, comprehensive analysis of The platform will generate a multi-level evidence base by processing the original evidence, analyze the evaluation indexes of New Coronary Pneumonia, select the features, build a deep neural network model, train and validate the model, and establish a high The model is trained and validated, and a three-level early warning model is established for people at high risk of new coronary pneumonia, those who turn serious, and those with sequelae.

Inclusion Criteria:

1. Meeting the diagnostic criteria of mild and moderate types of COVID-19 disease described in the "Diagnostic and Treatment Protocol for Novel Coronavirus Pneumonia (Trial version 9)" [9];
2. Age ≥18 years old and ≤90years old;
3. Not more than 5 days after the occurrence of the first symptom (or confirmed onset);
4. The subject agrees to participate the study and have signed the informed consent form by paper signature, electronic signature of mobile software or voice authorization.

Exclusion Criteria:

1. Pregnant or lactating women;
2. Patients complicated with severe heart, liver, kidney, blood system and other primary diseases and mental disorders.
3. Patients with disturbance of consciousness, aphasia, dementia and other conditions that prevent them from cooperating with the clinical investigators
4. Patients with incomplete study data.

COVID-19 patients