Study design and participants
We conducted a retrospective cohort to study involving COVID-19 patients, and characterized their epidemiological, clinical, and laboratory data between January 25, 2020, and March 31, 2020, at the JinYinTan infectious disease Hospital. We included 186 mild and moderate patients and 113 severe and critical COVID-19 patients. We then performed a small sample study including 17 severe patients who rapidly developed critical illness, but eventually recovered, and 15 severe illness patients who did not recover, which included patients who had blood oxygen saturation score of 4 points or above.
Data collection
Nose and throat swab specimens or nasopharyngeal aspirates collected from patients were tested for COVID-19 using positive high-throughput sequencing or real-time reverse‐transcriptase polymerase chain reaction (RT‐PCR).
Epidemiological information, medical history, and clinical symptoms were recorded at admission. Imaging results, laboratory findings such as partial arterial oxygen pressure, oxygen saturation, complete blood count (white blood cell, lymphocyte, monocyte count, eosinophil count, neutrophil to lymphocyte ratio, hemoglobin levels, and platelet counts), serum biochemical tests (liver and kidney function and lactate dehydrogenase) as well as coagulation indicator (D-dimer levels, prothrombin time, and activated partial thromboplastin time) were routinely conducted for all the patients. Epidemiological characteristics such as age, sex were also recorded. Besides, variables of medical history included cardiovascular disease, digestive system disease, endocrine system disease, malignant tumor, nervous system disease, or respiratory system disease were also evaluated. Clinical symptoms included: temperature, respiratory rate, heartbeat rate, cardiac arrhythmia, blood pressure, fever, chest pain, palpitation, dyspnea, cough, and stridor, fatigue, xerostomia, nausea and vomiting, diarrhea, and anorexia.
All the data were extracted from hospital information systems (HISs) and laboratory information systems (LISs) using a standardized data collection form. A group of experienced physicians determined the frequency of examinations, reviewed and cross-checked the data. Each data was double-blind input by two physicians and were scrutinized by other 2 independent physicians.
Definition
An axillary temperature of at least 37 °C was considered as fever.
The clinical spectrum of COVID-19 pneumonia ranged from mild-type group, moderate-type group, severe-type group to critical-type group, according to the Chinese management guideline for COVID-19; new coronavirus pneumonia diagnosis and treatment plan (trial version 7) developed by the National Health Committee of the People's Republic of China (http://www.nhc.gov.cn/). The clinical spectrum used was as follows: (1) mild, with mild clinical symptoms with imaging reporting negative for pneumonia. (2) moderate, with respiratory symptoms and fever, with imaging showing pneumonia. (3) severe, meeting any of the following: a) shortness of breath, respiratory rate ≥ 30 beats/min; b) during resting-state, oxygen saturation without oxygen uptake ≤ 93%; c) arterial partial pressure of oxygen (PaO2) / oxygen concentration ≤ 300 mmHg (1 mmHg = 0.133 kPa); d) progressive worsening of clinical symptoms, lung imaging findings showed rapid progress ༞50% during the past 24–48 hours. 4) critical illness: meeting any of the following: a) respiratory failure, and need for mechanical ventilation; b) shock; c) multisystem organ failure, and urgent admission to the intensive care unit.
Blood oxygen saturation score: (1) 6 points: death; (2) 5 points: need for invasive ventilator; (3) 4 points: high flow humidification oxygen/high frequency oxygen/non-invasive ventilator are needed; (4) 3 points: low flow (1 liters − 5 liters per minutes) face mask/nasal cannula for oxygen inhalation; (5) 2 point: oxygen saturation of above 93 without oxygen inhalation; (6) 1 point: discharge[6].
Chest X-Ray and Chest Computed Tomography abnormality (CXR): 0: Bilateral pneumonia with no abnormal lesions (Unilateral or Bilateral pneumonia); 1: Unilateral pneumonia with multiple mottling and ground-glass opacity; 2: Bilateral pneumonia with multiple mottling and ground-glass opacity.
Variable Selection and Score Construction
We initially enrolled all 302 the COVID-19 patients at the JinYinTan hospital for the selection of variables and development of risk score system. Categorical and continuous variables were presented by number (%) and by median (interquartile range), respectively. We applied the Chi-square test or Wilcoxon rank sum test to compare the differences between the two groups.
LASSO regression and 3-fold cross-validation was used to minimize the potential collinearity of variables while multiple logistic regression was applied to select the independent risk factors which influence the results. Then those statistically significant variables were used to establish a nomogram. In case of a P < 0.05 for non-linearity, the restricted cubic spline (RCS) was drawn to demonstrate the relationship between the variable and the risk.
We then enrolled 17 patients who recovered from the suddenly worsened condition and 16 patients who did not recover. We used the logistic regression analysis to select predictive factors which would be protective from sudden development of worsened illness.
Assessment of Accuracy
Receiver Operating Characteristic Curve (ROC) was drawn to measure the optimal clinical value of the prediction probability. The accuracy of COVID-19 risk factors were assessed with the area under the ROC (AUC), Calibration, Decision Curve, Net Reclassification Improvement (NRI), and Clinical Impact Curve. An AUC < 0.5 indicated meaningless prediction, an AUC of 0.5 ~ 0.7 indicated lower accuracy prediction, while an AUC > 0.7 indicated higher prediction accuracy. Integrated Discrimination Improvement (IDI) was applied to evaluate the improvement of standard CURB-65 in our model. Analysis was performed with SPSS 22.0 and R software (version 3.6.2), and a P-value < 0.05 was considered statistically significant.
Score Validation
For internal validation of the accuracy of the estimates, we used the training set in the 3-fold cross-validation, while we used the validation set for external validation.