1. The novel severe coronary pneumonia prediction model is characterized in that: the prediction model is a calculation model of the progression risk score obtained according to lung diseases, the age, immunoglobulin M, CD16+/CD56+ NK cells and glutamic-oxaloacetic transaminase of a new coronary pneumonia patient so as to predict the risk of the progression from the mild/common type to the severe new coronary pneumonia.

2. The model for predicting the severity of coronary pneumonia according to claim 1, wherein: the progress risk score calculation model is as follows:

PAINT score ═ alpha₁*β₁+α₂*β₂+α₃*β₃+α₄*β₄+α₅*β₅

Wherein alpha is₁、α₂、α₃、α₄、α₅Are discrimination values, β, of five predictors respectively compared with corresponding thresholds₁、β₂、β₃、β₄、β₅The beta risk ratio values corresponding to the five prediction factors are obtained; IgM means immunoglobulin M and AST means glutamic-oxalacetic transaminase.

3. The model for predicting the severity of coronary pneumonia according to claim 2, wherein: beta is the same as₁＝2.4174，β₂＝1.3594，β₃＝1.8399，β₄＝1.2246，β₅＝1.5182。

4. The method for establishing a novel model for predicting severe coronary pneumonia according to claim 1, 2 or 3, wherein: the method comprises the following steps:

s1, collecting clinical relevant indexes of the new coronary pneumonia patient, including demographic information, basic disease history, clinical manifestations and biochemical indexes;

s2, statistical analysis, wherein the predictive ability of the population science, the basic disease history, the clinical manifestations and the biochemical indexes on the progress from the light type/the ordinary type to the heavy type is analyzed through a COX regression model, the predictive factors with high disease correlation degree are screened out, and the threshold value and the beta risk ratio value of each predictive factor are calculated;

and S3, obtaining a prediction model according to the prediction factor, the corresponding threshold value and the beta risk ratio value based on a preset formula.

5. The method for establishing a novel model for predicting severe coronary pneumonia according to claim 4, wherein: in step S1, the demographic information collected from the patient includes age, gender;

clinical manifestations include fever, cough, dyspnea, chest pain, angina, fatigue, myalgia, headache, vomiting, and diarrhea;

biochemical markers include leukocyte WBC, neutrophil count NEU, lymphocyte count LYM, hemoglobin HGB, platelet count PLT, prothrombin time PT, D-dimer, alanine ALT, aspartate aminotransferase AST, γ -glutamyl transpeptidase GGT, albumin ALB, total bilirubin TBlL, direct bilirubin DBlL, uric acid UA, creatinine Cr, creatine kinase CK, lactate dehydrogenase LDH, brain natriuretic peptide BNP, procalcitonin PCT, c-reactive protein CRP, neutrophil/lymphocyte ratio NLR, and SARS-CoV2 RNA tests.

6. The method for establishing a novel model for predicting severe coronary pneumonia according to claim 4, wherein: the step S2 includes:

s2.1, calculating the risk ratio of the clinically relevant indexes by using C0X single-factor regression to obtain an effect value of a single factor on the disease;

s2.2, bringing the factors with the P less than 0.05 in the single-factor regression result into a C0X multi-factor regression model, allowing the factors to form mutual correction, screening out prediction factors with high disease correlation, calculating a beta regression coefficient for each prediction factor, and taking the beta regression coefficient as a risk ratio value corresponding to the prediction factor;

the threshold for each predictor is calculated by the area under the curve ROC.

7. The method for establishing a novel model for predicting severe coronary pneumonia according to claim 4, wherein: the method also comprises the step of verifying the established prediction model.

8. The method for predicting the severe new coronary pneumonia is characterized by comprising the following steps: the method for predicting the severity of new coronary pneumonia, based on the model for predicting the severity of new coronary pneumonia according to any one of claims 1 to 3, comprising the steps of:

step 1, inputting the demographic information, the basic disease history, the clinical manifestations and the biochemical indexes of a patient with the new coronary pneumonia into the new coronary pneumonia severe prediction model;

and 2, predicting by using the new coronary pneumonia severe prediction model, and outputting the progression risk score of the tested patient from light/common type to heavy new coronary pneumonia.

9. The method for predicting the exacerbation of neocoronary pneumonia according to claim 8, further comprising: further comprising a step 3 of comparing the progression risk score obtained in the step 2 with a cut-off value and classifying the patients into a mild/normal type group or progression into a severe type group according to the comparison result.

10. The utility model provides a severe prediction system of new coronary pneumonia which characterized in that: the method comprises the following steps:

a prediction module comprising the new coronary pneumonia exacerbation prediction model of any one of claims 1-3;

the data input module is used for inputting the demographic information, clinical manifestations and biochemical indexes of the new coronary pneumonia patient into the new coronary pneumonia severe prediction model;

and the prediction output module is used for outputting the progression risk score of the tested patient from light/common type to heavy new coronary pneumonia according to the new coronary pneumonia severity prediction model.

Background

The etiology of the novel pneumonia is identified as a novel beta coronavirus, currently designated severe acute respiratory syndrome coronavirus 2(SARS-CoV-2), which has phylogenetic similarities with SARS-CoV. SARS-CoV-2 infection has been named as 2019 coronavirus disease (COVID-19) by the World Health Organization (WHO), namely: novel coronary pneumonia.

COVID-19 has now become a major public health problem worldwide. The overall lethality of COVID-19 was estimated to be about 1-2.3%, similar to Spanish influenza (2-3%), much higher than seasonal influenza (0.1%). According to the world health organization, the severity of COVID-19 is classified into mild, general, severe and critical. In 72314 case report studies at the Chinese center for disease prevention and control, approximately 81% of COVID-19 cases were defined as mild; in 14% of cases COVID-19, the case is severe, and in 5% of cases, the case is critical. The overall hospital mortality in hospitalized COVID-19 cases is about 15% to 20%, but in critically ill patients who need to be admitted to the ICU intensive care unit, the mortality rate is as high as 40-49%. Therefore, it is important to assess risk factors for disease progression in patients with COVID-19. Early detection of COVID-19 patients at risk of developing severe illness is particularly important for selecting optimal treatment regimens and reducing mortality.

Current studies have shown that certain abnormalities exist in the hematologic and immunological tests of patients with COVID-19. Studies suggest that older age, high Sequential Organ Failure Assessment (SOFA) scores, d-dimers greater than 1 μ g/mL, lymphopenia and coronary disease history increase the risk of death in patients with covi-19.

Currently, one of the major challenges in clinically treating patients with new coronary pneumonia (COVID-19) is predicting the severity of the disease.

Several predictive scoring models have been used to predict the prognosis of COVID-19 patients and other diseases. However, the associated risk factors that predict progression of COVID-19 patients from mild/general to severe remain limited and require further investigation.

Disclosure of Invention

In order to solve the technical problems, the application provides a new model and a system for predicting severe coronary pneumonia, and an establishment method and a prediction method thereof.

The application is realized by the following technical scheme:

a model for predicting severe neocoronary pneumonia, which is a model for calculating a risk score of progression according to lung diseases, ages, immunoglobulin M, CD16+/CD56+ NK cells and glutamic-oxalacetic transaminase of patients with the neocoronary pneumonia, so as to predict the risk of progression from light/common type to severe neocoronary pneumonia.

Further, the progress risk score calculation model is as follows:

the progress risk score calculation model is as follows:

PAINT score ═ α₁*β₁+α₂*β₂+α₃*β₃+α₄*β₄+α₅*β₅

Wherein alpha is₁、α₂、α₃、α₄、α₅Are discrimination values, β, of five predictors respectively compared with corresponding thresholds₁、β₂、β₃、β₄、β₅The risk ratio values corresponding to five predictors are IgM for immunoglobulin M and AST for glutamic-oxalacetic transaminase.

The method for establishing the new coronary pneumonia severe prediction model comprises the following steps:

s1, collecting clinical relevant indexes of the new coronary pneumonia patient, including demographic information, basic disease history, clinical manifestations and biochemical indexes;

s2, analyzing the prediction ability of the population science, the basic disease history, the clinical manifestations and the biochemical indexes on the progress from light type/common type to heavy type through a COX regression model, screening out prediction factors with high disease correlation degree, and calculating the threshold value and the risk ratio value of each prediction factor;

and S3, obtaining a prediction model according to the prediction factor, the corresponding threshold value and the risk ratio value based on a preset formula.

Further, in step S1, the demographic information collected from the patient includes age, gender;

the basic disease history comprises previous hypertension, diabetes, cardiovascular diseases, lung diseases and liver diseases;

clinical manifestations include fever, cough, dyspnea, chest pain, angina, fatigue, myalgia, headache, vomiting, and diarrhea;

biochemical indices include leukocyte WBC, neutrophil count NEU, lymphocyte count LYM, hemoglobin HGB, platelet count PLT, prothrombin time PT, D-dimer, glutamic-pyruvic transaminase ALT, glutamic-oxalacetic transaminase AST, γ -glutamyl transpeptidase GGT, albumin ALB, total bilirubin TBIL, direct bilirubin DBIL, uric acid UA, creatinine Cr, creatine kinase CK, lactate dehydrogenase LDH, brain natriuretic peptide BNP, procalcitonin PCT, c-reactive protein CRP, neutrophil/lymphocyte ratio NLR, and SARS-CoV2 RNA tests.

Further, step S2 specifically includes:

s2.1, calculating the risk ratio of the clinically relevant indexes by using COX single-factor regression through single-factor COX risk model regression; clinically relevant indices include demographic information, basic disease history, clinical manifestations and biochemical indices;

and 2.2, by multi-factor COX risk model regression, bringing the factors with the P less than 0.05 in the single-factor COX regression result in the S2.1 into the COX multi-factor regression model, and then mutually correcting the factors. Through step S2, the predictor most relevant to the disease is screened out, the β regression coefficient is calculated and output for each predictor, and the β regression coefficient is used as the risk ratio value corresponding to the predictor, thereby establishing a prediction model including a plurality of predictors.

The result of the COX multifactor regression model indicates that after the model automatically corrects various factors, the most key five prediction factors are output, namely: lung disease, age, IgM, CD16+/CD56+ NK cells, AST; these five influencing factors have the best ability to predict the progression of COVID-19 patients from mild/normal to severe.

Further, the threshold (critical value) of each predictor, namely, the presence or absence of lung diseases, age of 75 years or more, IgM of 0.84 or less, CD16+/CD56+ NK cells of 116.5 or less and AST of 25 or more was obtained by ROC calculation of the area under the curve.

According to the results, the lung disease history is judged by using a threshold (critical value), the lung disease history is that the lung disease history is more than or equal to 75 years old, the lung disease history is less than or equal to 0.84 IgM, the lung disease history is less than or equal to 116.5 CD16+/CD56+ NK cells and more than or equal to 116.5 AST and more than or equal to 25, and beta regression coefficients (2.4174, 1.3594, 1.8399, 1.2246 and 1.5182 in sequence) corresponding to each risk factor are combined.

Preferably, the corresponding critical values of lung disease, age, IgM, CD16+/CD56+ NK cells, AST are in order: there was a history of pulmonary disease, age 75, 0.84, 116.5, 25.

The method for predicting the severity of new coronary pneumonia comprises the following steps based on the model for predicting the severity of new coronary pneumonia:

step 1, inputting the demographic information, the basic disease history, the clinical manifestations and the biochemical indexes of a patient with the new coronary pneumonia into the new coronary pneumonia severe prediction model;

and 2, predicting the new coronary pneumonia severe prediction model, and outputting the progression risk score of the tested patient from light/common type to heavy new coronary pneumonia.

Further, the method for predicting the exacerbation of new coronary pneumonia further comprises a step 3 of comparing the progression risk score obtained in the step 2 with a cut-off value, and classifying the patients into a light/normal type group or progression to an intensive type group according to the comparison result.

A system for predicting severe acute coronary pneumonia, comprising:

the prediction module comprises the new severe coronary pneumonia prediction model;

the data input module is used for inputting the demographic information, clinical manifestations and biochemical indexes of the new coronary pneumonia patient into the new coronary pneumonia severe prediction model;

and the prediction output module is used for outputting the progression risk score of the tested patient from light/common type to heavy new coronary pneumonia according to the new coronary pneumonia severity prediction model.

Compared with the prior art, the method has the following beneficial effects:

the invention is beneficial to identifying the light/common patients with high progression risk and is beneficial to improving the diagnosis and treatment success rate of severe patients.

Drawings

The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention.

FIG. 1 is a flowchart of a method for establishing a model for predicting severe coronary pneumonia;

FIG. 2 is a flow chart of a statistical analysis established by the PAINT scoring model;

FIG. 3 is a graph of risk ratio forests for 5 preferred independent risk factors;

in FIG. 4, a is a result of model stability test for the presence or absence of lung disease index, b is a result of model stability test for the age index, c is a result of model stability test for the IgM index, d is a result of model stability test for the CD16+/CD56+ NK cell index, and e is a result of model stability test for the AST index;

FIG. 5 is a graph of survival analysis of predictor lung disease, where a is a graph of survival analysis, b is a graph of analysis of the number of at-risk persons at the time points corresponding to graph a, and c is a graph of analysis of the number of missing persons at the time points corresponding to graph a;

FIG. 6 is a graph of survival analysis of age as a predictor, where a is a graph of survival analysis, b is a graph of analysis of the number of people at risk at the time points corresponding to the graph a, and c is a graph of analysis of the number of people lost at the time points corresponding to the graph a;

FIG. 7 is a graph showing the survival analysis of the prediction factor IgM, wherein a is a graph showing the survival rate analysis, b is a graph showing the number of persons at risk at the time points corresponding to the a-graph, and c is a graph showing the number of persons deleted at the time points corresponding to the a-graph;

FIG. 8 is a graph showing the survival analysis of the predictor CD16+/CD56+ NK cells, wherein a is a graph showing the survival analysis, b is a graph showing the analysis of the number of persons at risk at the time points corresponding to the graph a, and c is a graph showing the analysis of the number of persons lost at the time points corresponding to the graph a;

FIG. 9 is a graph of survival analysis of the predictor AST, where a is a graph of survival analysis, b is a graph of analysis of the number of persons at risk at the time points corresponding to the a graph, and c is a graph of analysis of the number of persons lost at the time points corresponding to the a graph;

FIG. 10a is a weight plot for Principal Component Analysis (PCA);

in fig. 10b is a vector calculation diagram of Principal Component Analysis (PCA);

FIG. 11 is a graph of survival analysis of the PAINT score, wherein a is a graph of survival analysis, b is a graph of analysis of the number of people at risk at the time points corresponding to the a graph, and c is a graph of analysis of the number of people lost at the time points corresponding to the a graph;

FIG. 12 is a graph comparing PAINT scores to existing qSOFA scores and CURB-65;

fig. 13 a is a plot of the consensus index for PAINT scores and b is a 1000-time internal sample validation plot for PAINT scores.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments. It is to be understood that the described embodiments are only a few embodiments of the present invention, and not all embodiments.

Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.

In addition, the embodiments of the present invention and the features of the embodiments may be combined with each other without conflict. It should be noted that, in the present specification, the embodiments are all described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments may be referred to each other. For the device-like embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.

Example 1

Natural Killer (NK) cells are innate lymphocytes that respond to viral infection and may be associated with the severity of COVID-19 disease. Therefore, this example establishes a predictive model of the progression of mild/normal type COVID-19 to severe disease, hereinafter referred to as PAINT scoring model, based on the clinical manifestations of NK cells and other four preferred relevant factors. In total 239 CODV-19 hospitalizations from two medical centers in China were included in this example. Prognostic effects of variables, including clinical data and laboratory test results of electronic medical records in various hospitals, were analyzed using a Cox proportional hazards model and the Kaplan-Meier method. The PAINT scoring model established herein to help predict risk of progression from light/normal to heavy COVID-19. After adjusting for various confounding factors, the five preferred factors of age > 75 years, IgM < 0.84, CD16+/CD56+ NK cells < 116.5, glutamic-oxaloacetic transaminase < 25 > and having a history of pulmonary diseases are independent influencing factors progressing to severe COVID-19. Based on these five factors, a predictive scoring model is established. By verification, the prediction scoring model shows a higher predicted value (C index is 0.91, 0.902 ± 0.021, p < 0.001). The predictive scoring model of the invention helps to identify light/common type COVID-19 patients with high progression risk, and improves the diagnosis and treatment success rate of high risk severe patients.

As shown in fig. 1 and 2, the method for establishing the prediction scoring model of the present invention includes the steps of:

and S1, collecting the demographic information, the basic disease history, the clinical manifestations and the biochemical indexes of the new coronary pneumonia patients.

This example recruited COVID-19 patients from two medical centers in China for study, and included 239 COVID-19 patients in total. In this cohort, 216 (90.38%) light/normal cases and 23 (9.62%) heavy cases were included. Table 1 lists the demographic and clinical baseline characteristics of the study population, with a median age of 58 years (range: 26-90 years) and 58.20% male (139/239 cases). The median maximum temperature was 38 deg.C (range 36.0-41.0).

Table 1: demographic and clinical baseline profiles for study population

S2, analyzing the prediction ability of the population science, the basic disease history, the clinical manifestations and the biochemical indexes on the progress from the light type/the ordinary type to the heavy type through a COX regression model, screening out the prediction factors with high disease correlation, and calculating the risk ratio value of each prediction factor. The method comprises the following specific steps:

when comparing demographic data at admission, male patients and older and respiratory distress patients (all with P <0.05, table 1) progressed to a severe grade. The clinical manifestations of all patients are mainly as follows: fever 77.8% (186/239), cough 60.3% (144/239), sputum 23.8% (57/239), respiratory distress 10.5% (25/239), chest pain 4.6% (11/239), palpitations 4.2% (11/239), fatigue 28.5% (68/239), muscle pain 9.2% (22/239), headache 5.0% (12/239), vomiting 1.7% (4/239) and diarrhea 18.0% (43/239).

Then, the clinical characteristics of the study population were further summarized as shown in table 2. When comparing the biochemical indices of covi-19 between the advanced and non-advanced severe cases, significant statistical differences were found for neutrophils, lymphocytes, neutrophil/lymphocyte ratios, glutamic-oxaloacetic transaminase, total bilirubin, direct bilirubin, creatinine, blood glucose, blood sodium, clotting time, myoglobin, CD3+ T cells, CD4+ T cells, CD8+ T cells, CD19+ T cells, CD16+/CD56+ NK cells and IgM (all P < 0.05).

Table 2: laboratory test and clinical baseline profiles

In Table 2, fast organ failure score, qSOFA; IgA, immunoglobulin a; IgG, immunoglobulin G; IgM, immunoglobulin M; IgE, immunoglobulin E.

As shown in fig. 2, this example compares the demographics, basic disease history, clinical data and biochemical indices between light/normal and heavy cases in order to explore the risk factors for progressing from light/normal to heavy COVID-19 cases. Further, the direct occurrence probability multiple relationship between each factor and the disease was predicted using the results of the output risk ratio (ORs) by the risk correction between each factor using univariate and multivariate COX regression models, as shown in fig. 3. The results show that five factors of age > 75 years (ORs ═ 3.92, 95% CI 1.61-9.73, P ═ 0.003), existing history of pulmonary disease (ORs ═ 11.20, 95% CI 2.50-49.70, P ═ 0.001), IgM (ORs ═ 6.31, 95% CI 1.99-19.60, P ═ 0.002), CD16+/CD56+ NK cells (ORs ═ 3.40, 95% CI 1.31-9.13, P ═ 0.014) and AST (ORs ═ 4.60, 95% CI 1.31-16.00, P ═ 0.018) are significantly different, see table 3. This example uses these 5 independent risk factors as predictors for predicting the progression from mild/normal to severe cases, as shown in fig. 3. Meanwhile, COX multifactor regression analysis results in the study population showed that the preferred 5 factors were significantly associated with the progression of COVID-19 disease.

Table 3: single factor multifactor COX regression analysis results table

On this basis, as shown in fig. 2, the present example also uses the Cox diagnostic bias and Cox proportional hazards modeling of the global Schoenfeld test to assess the predictive power of these five independent risk factors leading to progression from mild/normal to severe cases. As shown in fig. 4, each red dot in fig. 4 is arranged in two dashed lines, close to the central solid line, indicating good stability of the diagnostic model. The results show that these 5 risk factors lead to a good prediction of progression from mild/normal to severe cases.

The threshold (critical value) of each factor, namely whether the lung disease exists or not, is obtained by calculating the area ROC under the curve, the age is more than or equal to 75 years, the IgM is less than or equal to 0.84, the CD16+/CD56+ NK cells are less than or equal to 116.5, and the AST is more than or equal to 25. Further, the analysis of the Kaplan-Meier survival curves showed significant differences in the survival curves of COVID-19 patients aged 75 years or older, IgM 0.84 or less, CD16+/CD56+ NK cells 116.5 or less and AST 25 or more, respectively, according to the history of pulmonary disease, as shown in FIGS. 5-9.

And S3, constructing a PAINT scoring model from light/common type to heavy type based on a preset formula.

Wherein, the preset formula is as follows:

in the above formula, α_iTaking 0 or 1, beta as the discrimination value of the comparison between the prediction factor and the corresponding critical value_iN is a natural number.

Based on statistical analysis, this example involved risk factors including lung disease, age, IgM, CD16+/CD56+ NK cells and AST in the generation of predictive scores for COVID-19 patients progressing from mild/moderate to severe cases, and constructed a model of PAINT scores for light/normal to heavy progression as follows:

PAINT score ═ α₁*2.4174+α₂*1.3594+α₃*1.8399+α₄*1.2246+α₅*1.5182

As shown in fig. 10a, the present embodiment further employs PCA principal component analysis to explore the proportion of 5 predictors in the model, thereby helping the clinical decision maker to determine the weights of the factors. As shown in fig. 10b, the weight contributed by each predictor is shown, suggesting that IgM and AST abnormalities are the worst for the prognosis of the patient.

To demonstrate the ability of the PAINT scoring model to find more severe patients for early clinical treatment, this example also used the ROC curve to find the optimal cut-off value. The cut-off value (threshold) for the score obtained was 14.687 points, and when the PAINT score was compared in size with the cut-off value, patients were classified into light/normal non-progression group and progression critical group, and the results were confirmed by Kaplan-Meier survival curve analysis (P0.001, as shown in fig. 11).

Patients with a PAINT score less than the cutoff score scored the light/general group, while patients with a PAINT score greater than or equal to the cutoff score scored the progression to the severe group. According to different groups, the specific treatment is carried out, and the diagnosis and treatment success rate of severe patients is improved.

This example further employed a C-Index (C-Index) analysis to evaluate the efficacy of the PAINT scoring model described above in predicting the progression of COVID-19 patients from mild/moderate to severe cases. The PAINT score was compared to the qSOFA and CURB-65 (confusion, uremia, respiratory rate, BP, age > 65) scores. As shown in fig. 12, the C-index (0.902 ± 0.021) of the PAINT scoring model used to predict progression from mild/moderate to severe cases was significantly better than qsfa (0.534 ± 0.027) and currb-65 (0.561 ± 0.058).

In addition, this example compares the PAINT scoring model described above with 5 independent risk factors (lung disease, age, IgM, CD16+/CD56+ NK cells and AST). The results indicate that the PANT score may be suitable for predicting progression from mild/normal to severe, with a significant improvement over the PANT score with a history of pulmonary disease (0.543 + -0.034), age > 75 years (0.639 + -0.052), IgM < 0.84 < 0.683 + -0.044, CD16+/CD56+ NK cells < 116.5(0.647 + -0.050), and glutamic oxaloacetic transaminase > 25(0.716 + -0.036), as shown in Table 4.

Table 4: c-index judges the predicted effect: from mild to severe

Variables of	Prediction ability determination (C-Index)	P value
			PAINT score	0.902±0.021
qSOFA score	0.534±0.027	p<0.001
			CURB-65 score	0.561±0.058	p<0.001
Pulmonary diseases (of)	0.543±0.034	p<0.001
			Age (age)	0.639±0.052	p<0.001
IgM antibodies	0.683±0.044	p<0.001
			CD16+/CD56+NK cell	0.647±0.050	p<0.001
Glutamic-oxalacetic transaminase	0.716±0.036	p<0.001

In table 4, qSOFA is the rapid organ failure score.

In addition, to internally validate the capabilities of the PAINT scoring model of the present invention, the present example also performed a consistency index analysis to evaluate the discrimination of the PAINT score, which was observed to be better than the qSOFA and CURB-65 scores. In addition, 1000 bootstrap internal verifications were performed in this example, and the results of the verification showed that the predicted PAINT score of the present invention also showed better discrimination, as shown in fig. 13.

Based on the model for predicting the severe new coronary pneumonia, the invention discloses a method for predicting the severe new coronary pneumonia, which comprises the following steps:

step 1, inputting the demographic information, the basic disease history, the clinical manifestations and the biochemical indexes of a patient with the new coronary pneumonia into the new coronary pneumonia severe prediction model;

and 2, predicting the new coronary pneumonia severe prediction model, and outputting the progression risk score of the tested patient from light/common type to heavy new coronary pneumonia.

And 3, comparing the progression risk score obtained in the step 2 with a critical value, and classifying the patients into a light/normal type group or progression into a severe type group according to the comparison result.

Based on the prediction model and the prediction method, the invention also discloses a new coronary pneumonia severe prediction system, which comprises the following steps:

the prediction module comprises the new severe coronary pneumonia prediction model;

the data input module is used for inputting the demographic information, clinical manifestations and biochemical indexes of the new coronary pneumonia patient into the new coronary pneumonia severe prediction model;

and the prediction output module is used for outputting the progression risk score of the tested patient from light/common type to heavy new coronary pneumonia according to the new coronary pneumonia severity prediction model.

The above-mentioned embodiments, objects, technical solutions and advantages of the present invention are further described in detail, it should be understood that the above-mentioned embodiments are only exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.