Heart failure risk stratified evaluation system incorporating intestinal flora metabolites

Abstract

The invention provides a heart failure risk stratified evaluation system incorporating intestinal flora metabolites. Specifically, the present invention provides a heart failure risk stratified evaluation system incorporating intestinal flora metabolites, comprising: (a) an input module configured to input heart failure risk indicator data of an object to be tested; (b) a processing module which is configured to compare the input risk indicator data with a threshold value pre-stored in the equipment so as to obtain a comparison result; a risk score is obtained according to a comparison result; and obtaining an evaluation result according to the risk score. And (c) an output module configured to input the evaluation result. The system provided by the invention can more effectively and accurately carry out risk stratification on the heart failure patient.

Description

Technical Field

[0001] The present invention relates to the field of medical diagnosis, and more specifically to a heart failure risk stratification assessment system incorporating intestinal flora metabolites. Background Art

[0002] Heart failure (HF) is a complex pathophysiology that is influenced by multiple systems, including hemodynamics, neurohormones, metabolism, and kidneys. Recent studies have shown that HF ​​is also affected by the gut microbiota. The impact of gut microbiota-related metabolites on HF adds a new dimension to the understanding of the gut-heart axis in HF, especially certain gut microbiota metabolites have been shown to be closely related to the severity of HF.

[0003] Currently available clinical risk scores and risk stratification algorithms for heart failure are mainly based on traditional cardiovascular and hemodynamic factors (i.e., clinical parameters and circulating markers - mainly natriuretic peptides) and do not reflect the contribution of systemic factors (such as gut microbiota-associated metabolites) in heart failure pathology.

[0004] Therefore, there is an urgent need in this field to develop more effective and accurate prediction methods and assessment systems for heart failure risk, so as to achieve timely intervention treatment for heart failure. Summary of the invention

[0005] The purpose of the present invention is to provide a more effective and accurate prediction method and assessment system for heart failure risk.

[0006] In a first aspect of the present invention, a heart failure risk stratification assessment system incorporating intestinal flora metabolites is provided, the system comprising:

[0007] (a) an input module, the input module being configured to input heart failure risk indicator data of a subject to be tested;

[0008] The risk indicators of heart failure include: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure and intestinal flora metabolites; and the intestinal flora metabolites are: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide;

[0009] (b) a processing module, wherein the processing module is configured to perform the following functions:

[0010] i) comparing the input risk indicator data with a threshold value pre-stored in the device to obtain a comparison result; if a certain risk indicator data value input is greater than its corresponding threshold value, a value of 1 is assigned; if the risk indicator data value is less than its corresponding threshold value, a value of 0 is assigned;

[0011] ii) substituting the intestinal flora metabolite data comparison result into the first-level risk scoring formula to obtain its first-level risk score; and then obtaining the risk score value C of the intestinal flora metabolite according to its first-level risk score value; and

[0012] iii) Substituting the comparison result of the heart failure risk index data of the non-intestinal flora metabolites and the risk score value C into the second-level risk scoring formula for calculation, thereby obtaining the second-level risk score, and thus obtaining the evaluation result;

[0013] Wherein, when the second-tier risk score is below the risk cutoff value, it indicates that the subject has a low risk of heart failure; when the risk score is higher than the risk cutoff value, it indicates that the subject has a medium or high risk of heart failure;

[0014] Among them, the first-tier risk scoring formula is: The second-tier risk scoring formula is: Among them, the W i is the weight value of each risk indicator; i is the comparison result of each risk indicator data and its corresponding threshold value; said C is the risk score value of the intestinal flora metabolite obtained according to the result of S1; and

[0015] (c) An output module, wherein the output module is configured to input the evaluation result.

[0016] In another preferred embodiment, the heart failure risk indicator is selected from the following group: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure, diastolic blood pressure, hemoglobin, creatinine, intestinal flora metabolites, or a combination thereof.

[0017] In another preferred embodiment, the intestinal flora metabolite is selected from the following group: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide, or a combination thereof.

[0018] In another preferred embodiment, the risk indicators of heart failure include: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure and intestinal flora metabolites.

[0019] In another preferred embodiment, the intestinal flora metabolite is selected from the following group: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide, or a combination thereof.

[0020] In another preferred example, the risk cutoff value is 4.

[0021] In another preferred example, when the second-layer risk score is ≤4 points, the subject has a low risk of heart failure; when the second-layer risk score is 5-10 points, the subject has a medium risk of heart failure; when the second-layer risk score is >10, the subject has a high risk of heart failure.

[0022] In another preferred example, the scoring formula of the first-layer risk score is: S1=4*acetyl-L-carnitine+3*γ-butyl betaine+2*L-carnitine+4*trimethylamine oxide.

[0023] In another preferred example, when the first-layer risk score is ≤4, the risk score value C of the intestinal flora metabolites is 1; when the risk score is 5-10, the risk score value C of the intestinal flora metabolites is 2; when the risk score is >10, the risk score value C of the intestinal flora metabolites is 3.

[0024] In another preferred example, the scoring formula for the second-level risk score is: S2=3* previous history of heart failure + 3* NT-proBNP (or BNP) + 2* urea nitrogen + 2* age + 1* heart rate + 1* systolic blood pressure + C.

[0025] In another preferred example, the scoring formula of the first-layer risk score is: S1=4*acetyl-L-carnitine+3*γ-butylbetaine+2*L-carnitine+4*trimethylamine oxide, and the scoring formula of the second-layer risk score is: S2=3*previous history of heart failure+3*NT-proBNP (or BNP)+2*urea nitrogen+2*age+1*heart rate+1*systolic blood pressure+C.

[0026] In another preferred embodiment, the first-tier risk scoring formula and the second-tier risk scoring formula can be automatically calculated by a computer-assisted program.

[0027] In another preferred example, the system further includes (d) a storage module, which is configured to store data selected from the following group: various judgment or calculation result values, various risk indicator thresholds and risk cutoff values.

[0028] In another preferred example, the system further includes (e) a control module, wherein the control module is configured to control the operation of each module.

[0029] In another preferred embodiment, the subject to be tested is a patient with heart failure.

[0030] In another preferred embodiment, the risk is the risk of re-hospitalization and / or death of a heart failure patient due to heart failure within one year.

[0031] In a second aspect of the present invention, a method for stratifying and assessing the risk of heart failure incorporating intestinal flora metabolites is provided, the method comprising:

[0032] (a) providing data, wherein the data includes heart failure risk indicator data of a subject to be tested;

[0033] The risk indicators of heart failure include: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure and intestinal flora metabolites; and the intestinal flora metabolites are: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide;

[0034] (b) Analytical evaluation, said evaluation comprising the following steps:

[0035] (s1) comparing the risk indicator data with its corresponding threshold value to obtain a comparison result; if a certain risk indicator data value is greater than its corresponding threshold value, assigning a value of 1; if the risk indicator data value is less than its corresponding threshold value, assigning a value of 0;

[0036] (s2) Substituting the comparison result of the intestinal flora metabolite data into the first-level risk scoring formula to obtain the first-level risk score; and then obtaining the risk score value C of the intestinal flora metabolite according to the first-level risk score;

[0037] (s3) substituting the comparison result of the heart failure risk index data of the non-intestinal flora metabolites and the risk score value C into the second-level risk scoring formula for calculation, thereby obtaining the second-level risk score; and

[0038] (s4) comparing the second-tier risk score with the risk cutoff value to obtain an assessment result;

[0039] Wherein, when the second-tier risk score is below the risk cutoff value, it indicates that the subject has a low risk of heart failure; when the risk score is higher than the risk cutoff value, it indicates that the subject has a medium or high risk of heart failure;

[0040] Among them, the first-tier risk scoring formula is:

[0041] The second-tier risk scoring formula is: Among them, the W i is the weight value of each risk indicator; i is the comparison result of each risk indicator data and its corresponding threshold value; said C is the risk score value of the intestinal flora metabolite obtained according to the result of S1; and

[0042] (c) Outputting the evaluation result.

[0043] In another preferred embodiment, the heart failure risk indicator is selected from the following group: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure, diastolic blood pressure, hemoglobin, creatinine, intestinal flora metabolites, or a combination thereof.

[0044] In another preferred embodiment, the intestinal flora metabolite is selected from the following group: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide, or a combination thereof.

[0045] In another preferred embodiment, the risk indicators of heart failure include: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure and intestinal flora metabolites.

[0046] In another preferred embodiment, the intestinal flora metabolite is selected from the following group: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide, or a combination thereof.

[0047] In another preferred example, the risk cutoff value is 4.

[0048] In another preferred example, the scoring formula of the first-layer risk score is: S=4*acetyl-L-carnitine+3*γ-butyl betaine+2*L-carnitine+4*trimethylamine oxide.

[0049] In another preferred example, when the first-layer risk score is ≤4, the risk score value C of the intestinal flora metabolites is 1; when the risk score is 5-10, the risk score value C of the intestinal flora metabolites is 2; when the risk score is >10, the risk score value C of the intestinal flora metabolites is 3.

[0050] In another preferred embodiment, the subject to be tested is a patient with heart failure.

[0051] In a third aspect of the present invention, a method for constructing a heart failure risk stratification prediction model is provided, comprising the steps of:

[0052] (S1) providing a first data set, wherein the first data set comprises heart failure risk indicator data, and the risk indicator data comprises the following data: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure, and intestinal flora metabolites; and the intestinal flora metabolites are: acetyl-L-carnitine, γ-butyl betaine, L-carnitine, and trimethylamine oxide; and

[0053] (S2) Based on the data information of the first data set, a first-level risk prediction model is constructed using logistic regression and Cox regression, and then a heart failure risk stratification model is constructed based on the first-level risk prediction model.

[0054] In another preferred example, the first-layer risk prediction model is a risk prediction model constructed based on intestinal flora metabolite index data.

[0055] In another preferred example, the input risk indicators of the first-level risk prediction model are: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide.

[0056] In another preferred embodiment, the scoring formula of the first-level risk prediction model is:

[0057] Among them, the W i is the weight value of each risk indicator; i It is the comparison result of each risk indicator data and its corresponding threshold value, and if the input risk indicator data value is greater than its corresponding threshold value, it is assigned a value of 1, and if the risk indicator data value is less than its corresponding threshold value, it is assigned a value of 0.

[0058] In another preferred example, the scoring formula of the first-level risk prediction model is: S1=4*acetyl-L-carnitine+3*γ-butyl betaine+2*L-carnitine+4*trimethylamine oxide.

[0059] In another preferred embodiment, the heart failure risk stratification model is a heart failure risk stratification model that incorporates intestinal flora metabolites.

[0060] In another preferred example, the risk indicators input into the heart failure risk stratification model are: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure and intestinal flora metabolites; and the intestinal flora metabolites are: acetyl-L-carnitine, γ-butylbetaine, L-carnitine and trimethylamine oxide.

[0061] In another preferred embodiment, the scoring formula of the heart failure risk stratification model is: Among them, the W i is the weight value of each risk indicator; i is the comparison result of each risk indicator data and its corresponding threshold value; said C is the risk score value of the intestinal flora metabolite obtained according to the result of said S1.

[0062] In another preferred example, when the first-layer risk score is ≤4, the risk score value C of the intestinal flora metabolites is 1; when the risk score is 5-10, the risk score value C of the intestinal flora metabolites is 2; when the risk score is >10, the risk score value C of the intestinal flora metabolites is 3.

[0063] In another preferred embodiment, the scoring formula of the heart failure risk stratification model is: S2=3* previous history of heart failure + 3* NT-proBNP (or BNP) + 2* urea nitrogen + 2* age + 1* heart rate + 1* systolic blood pressure + C.

[0064] In another preferred example, when the heart failure risk stratification model score is ≤4 points, the subject has a low risk of heart failure; when the heart failure risk stratification model score is 5-10 points, the subject has a medium risk of heart failure; when the heart failure risk stratification model score is >10, the subject has a high risk of heart failure.

[0065] It should be understood that within the scope of the present invention, the above-mentioned technical features of the present invention and the technical features specifically described below (such as embodiments) can be combined with each other to form a new or preferred technical solution. Due to space limitations, they will not be described one by one here. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 Kaplan-Meier survival curves are shown (gut flora metabolite risk score)

[0067] Figure 2 Forest plots showing the Cox proportional hazards regression models for readmission and / or death due to heart failure within 1 year: A. Model with risk score of gut microbiota metabolites; B. Model with risk score of gut microbiota metabolites added to ADHERE; C. Model with risk score of gut microbiota metabolites added to OPTIMIZE-HF; D. Model with risk score of gut microbiota metabolites added to GWTGHF.

[0068] Figure 3Kaplan-Meier survival curves (death and / or readmission due to heart failure within one year) (top) and ROC curves (bottom) are shown: (A) Modeling cohort (acute heart failure patients in a European hospital, threshold values ​​for acetyl-L-carnitine (12.1 μmol / L), L-carnitine (52.1 μmol / L), γ-butylbetaine (0.9 μmol / L), trimethylamine oxide (5.6 μmol / L), previous heart failure, NT-proBNP (2160 ng / L), urea nitrogen (8.9 mmol / L), age (78 years), heart rate (90 beats / min) and systolic blood pressure (133 mmHg) g)), (B) validation cohort (BIOSTAT-CHF, method 1, the threshold was the threshold of the modeling cohort), (C) validation cohort (BIOSTAT-CHF, method 2, the thresholds were acetyl-L-carnitine (8.2 μmol / L), L-carnitine (89.0 μmol / L), γ-butylbetaine (1.2 μmol / L), trimethylamine oxide (6.4 μmol / L), previous history of heart failure, NT-proBNP (376 ng / L), urea nitrogen (12.1 mmol / L), age (70 years), heart rate (79 beats / min) and systolic blood pressure (120 mmHg)).

[0069] Figure 4 Forest plots are shown showing the Cox proportional hazards regression model (death and / or readmission due to heart failure within one year): (A) modeling cohort (acute heart failure patients in a European hospital, threshold values ​​for acetyl-L-carnitine (12.1 μmol / L), L-carnitine (52.1 μmol / L), γ-butyl betaine (0.9 μmol / L), trimethylamine oxide (5.6 μmol / L), previous history of heart failure, NT-proBNP (2160 ng / L), urea nitrogen (8.9 mmol / L), age (78 years), heart rate (90 beats / min) and systolic blood pressure (133 mmHg)), (B ) validation cohort (BIOSTAT-CHF, method 1, the threshold was the threshold of the modeling cohort), (C) validation cohort (BIOSTAT-CHF, method 2, the thresholds were acetyl-L-carnitine (8.2 μmol / L), L-carnitine (89.0 μmol / L), γ-butylbetaine (1.2 μmol / L), trimethylamine oxide (6.4 μmol / L), previous history of heart failure, NT-proBNP (376 ng / L), urea nitrogen (12.1 mmol / L), age (70 years), heart rate (79 beats / min) and systolic blood pressure (120 mmHg)).

[0070] Figure 5 The calculation of 1-year HF survival rate using the gut microbiota metabolite heart failure risk stratification model (GuM-HF) is shown: (A) risk groups, (B) predicted survival rate after 1 year.

[0071] Figure 6 The ROC curves under different scoring conditions are shown (A: weighted score; B: average score; C: score of comparison model 1; D. score of comparison model 2; E. score of comparison model 3). DETAILED DESCRIPTION

[0072] After extensive and in-depth research, the inventors have developed for the first time a more effective and accurate method and evaluation system for predicting the risk of heart failure. Specifically, the inventors used specific heart failure risk indicator data, and the heart failure risk indicators included: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure and intestinal flora metabolites; and the intestinal flora metabolites were: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide; the risk of heart failure was predicted early using the constructed risk scoring model formula. The present invention was completed on this basis.

[0073] the term

[0074] The terms used in the present invention have the meanings commonly understood by those of ordinary skill in the relevant art. However, in order to better understand the present invention, some definitions and related terms are explained as follows:

[0075] As used herein, the terms "comprise", "include", and "contain" are used interchangeably and include not only closed definitions, but also semi-closed and open definitions. In other words, the terms include "consisting of", "consisting essentially of".

[0076] As used herein, the term "LC-MS" is an abbreviation for high performance liquid chromatography-mass spectrometry, that is, "LC-MS" and "high performance liquid chromatography-mass spectrometry" can be used interchangeably.

[0077] Acetyl-L-carnitine, L-carnitine, and trimethylamine oxide are used as examples for description.

[0078] As used herein, the term "acetyl-L-carnitine" is abbreviated as ALC, that is, "acetyl-L-carnitine" and "ALC" can be used interchangeably.

[0079] As used herein, the term "L-carnitine" is abbreviated as carnitine, that is, "L-carnitine" and carnitine can be used interchangeably.

[0080] As used herein, the term “trimethylamine oxide” is abbreviated as TMAO, that is, “trimethylamine oxide” and “TMAO” can be used interchangeably.

[0081] As used herein, the term "ultra-high performance liquid chromatography" is abbreviated as UPLC, that is, "ultra-high performance liquid chromatography" and "UPLC" can be used interchangeably.

[0082] As used herein, "mass spectrometry" (MS) refers to an analytical technique for identifying compounds by their mass. MS techniques generally include (1) ionizing compounds to form charged compounds; and (2) detecting the molecular weight of the charged compounds and calculating the mass-to-charge ratio (m / z). The compounds can be ionized and detected by any suitable means. A "mass spectrometer" generally includes an ionizer and an ion detector.

[0083] The term "about" as used herein in reference to a quantitative measurement means plus or minus 10% of the stated value.

[0084] According to the present invention, the term "biomarker panel" refers to one biomarker, or a combination of two or more biomarkers.

[0085] According to the present invention, the term "intestinal flora metabolites" refers to metabolites produced by intestinal flora in an organism, such as choline, betaine aldehyde, betaine, L-carnitine, croton betaine, γ-butyl betaine, acetyl-L-carnitine, trimethyl lysine, etc. According to the present invention, the term "risk" refers to the risk of re-hospitalization and / or death due to heart failure in a patient with heart failure within one year after the diagnosis of heart failure.

[0086] According to the present invention, the first level of risk score is the risk score of intestinal flora metabolites.

[0087] According to the present invention, the second-level risk score is a heart failure risk score after incorporating intestinal flora metabolites into other internationally accepted scoring systems, such as ADHERE (Acute Decompensated Heart Failure National Registry), OPTIMIZE-HF (Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure) and GWTG-HF (GWTG-HF: Get With The Guidelines-Heart Failure, American Heart Association (AHA) "Follow the Guidelines-Heart Failure" risk scoring system, etc.

[0088] As used herein, B-type natriuretic peptide (BNP) and N-terminal pro-B-type natriuretic peptide (NT-proBNP) are both commonly used markers of heart failure, and the two can be converted to each other. For example, the NT-proBNP threshold used in this article is 2160 ng / L, which is equivalent to the BNP threshold of 480 ng / L. (ISHIHARA S et al., (2022). New Conversion Formula Between B-TypeNatriuretic Peptide and N-Terminal-Pro-B-Type Natriuretic Peptide. Circulation Journal, 2022, 1-9 (doi: 10.1253 / circj.CJ-22-0032)

[0089] According to the present invention, the content of intestinal flora metabolites is indicated by the mass spectrometry signal area normalized value.

[0090] According to the present invention, it is known from the prior art that the training set and the validation set have the same meaning. In one embodiment of the present invention, the training set refers to a modeling cohort, which is a set of the content of intestinal flora metabolites in biological samples of patients with heart failure. In one embodiment of the present invention, the validation set refers to a data set used to test the performance of the training set. In one embodiment of the present invention, the content of intestinal flora metabolites can be represented as an absolute value or a relative value according to the method of determination. For example, when mass spectrometry is used to determine the level (such as content) of intestinal flora metabolites, the intensity or area of ​​the peak can represent the level of intestinal flora metabolites, which is a relative value level; when PCR is used to determine the level of biomarkers, the number of copies of genes or the number of copies of gene fragments can represent the level of biomarkers.

[0091] In one embodiment of the present invention, the reference value refers to the reference value or normal value of a healthy control. It is clear to those skilled in the art that, when the number of samples is sufficient, the range of normal values ​​(absolute values) of each intestinal flora metabolite can be obtained by inspection and calculation methods. Therefore, when other methods other than mass spectrometry are used to detect the level of biomarkers, the absolute values ​​of these biomarker levels can be directly compared with normal values ​​to evaluate the diagnosis of heart failure or early diagnosis of heart failure. In the present invention, statistical methods can also be used.

[0092] According to the present invention, the term "biomarker", also known as "biological marker", refers to a measurable indicator of a biological state of an individual. Such biomarkers can be any substance in an individual as long as they are related to a specific biological state (e.g., disease) of the individual being tested, for example, nucleic acid markers (e.g., DNA), protein markers, cytokine markers, chemokine markers, carbohydrate markers, antigen markers, antibody markers, species markers (species / genus markers) and functional markers (KO / OG markers), etc. Biomarkers are measured and evaluated, often to examine normal biological processes, pathogenic processes, or pharmacological responses to therapeutic interventions, and are useful in many scientific fields.

[0093] According to the present invention, the term "subject" refers to an animal, particularly a mammal, such as a primate, preferably a human.

[0094] According to the present invention, the term "plasma" refers to the liquid component of whole blood. Depending on the separation method used, plasma may be completely free of cellular components or may contain varying amounts of platelets and / or small amounts of other cellular components.

[0095] According to the present invention, terms such as "a", "an" and "the" refer not only to the singular individual but include the general class that may be used to describe a particular embodiment.

[0096] It should be noted that the explanations of the terms provided herein are only for enabling those skilled in the art to better understand the present invention and are not intended to limit the present invention.

[0097] System of the present invention

[0098] The present invention provides a heart failure risk stratification assessment system incorporating intestinal flora metabolites, the system comprising:

[0099] (a) an input module, the input module being configured to input heart failure risk indicator data of a subject to be tested;

[0100] The risk indicators of heart failure include: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure and intestinal flora metabolites; and the intestinal flora metabolites are: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide;

[0101] (b) a processing module, wherein the processing module is configured to perform the following functions:

[0102] i) comparing the input risk indicator data with a threshold value pre-stored in the device to obtain a comparison result; if a certain risk indicator data value input is greater than its corresponding threshold value, a value of 1 is assigned; if the risk indicator data value is less than its corresponding threshold value, a value of 0 is assigned;

[0103] ii) substituting the intestinal flora metabolite data comparison result into the first-level risk scoring formula to obtain its first-level risk score; and then obtaining the risk score value C of the intestinal flora metabolite according to its first-level risk score value; and

[0104] iii) Substituting the comparison result of the heart failure risk index data of the non-intestinal flora metabolites and the risk score value C into the second-level risk scoring formula for calculation, thereby obtaining the second-level risk score, and thus obtaining the evaluation result;

[0105] Wherein, when the second-tier risk score is below the risk cutoff value, it indicates that the subject has a low risk of heart failure; when the risk score is higher than the risk cutoff value, it indicates that the subject has a medium or high risk of heart failure;

[0106] Among them, the first-tier risk scoring formula is: The second-tier risk scoring formula is: Among them, the W i is the weight value of each risk indicator; i is the comparison result of each risk indicator data and its corresponding threshold value; said C is the risk score value of the intestinal flora metabolite obtained according to the result of S1; and

[0107] (c) An output module, wherein the output module is configured to input the evaluation result.

[0108] In another preferred example, the system further includes (d) a storage module, which is configured to store data selected from the following group: various judgment or calculation result values, various risk indicator thresholds and risk cutoff values.

[0109] In another preferred example, the system further includes (e) a control module, wherein the control module is configured to control the operation of each module.

[0110] Detection Methods

[0111] According to the present invention, mass spectrometry (MS) can be divided into ion trap mass spectrometry, quadrupole mass spectrometry, orbital trap mass spectrometry and time-of-flight mass spectrometry, with deviations of 0.2amu, 0.4amu, 3ppm and 5ppm respectively. In the present invention, tandem quadrupole mass spectrometry is used to obtain MS data.

[0112] ROC-AUC

[0113] ROC-AUC is a method to evaluate the accuracy of a model. The ROC curve is the receiver operating characteristic curve, which is a coordinate graph composed of the false positive rate as the horizontal axis and the true positive rate as the vertical axis. It is a comprehensive indicator reflecting the sensitivity and specificity of continuous variables. AUC is the area under the ROC curve. The ROC-AUC value is between 1.0 and 0.5. The closer it is to 1, the better the prediction effect. It has lower accuracy when it is 0.5-0.7, and has a certain accuracy when it is 0.7-0.9. When AUC is above 0.9, it has higher accuracy. When AUC = 0.5, it means that the prediction method does not work at all. AUC < 0.5 does not conform to the actual situation and rarely occurs in practice.

[0114] The main advantages of the present invention are:

[0115] 1. The present invention provides an effective and accurate method for predicting the risk of heart failure.

[0116] 2. The present invention performs weighted scoring on the metabolites of different intestinal flora, thereby improving the predictive ability of the model.

[0117] 3. This invention incorporates the metabolites of intestinal flora into the heart failure risk stratification model for the first time; adding the metabolites of intestinal flora to the traditional risk scoring model can also improve the predictive ability of the traditional model.

[0118] 4. The present invention demonstrates that both clinical variables and elevated concentrations of intestinal metabolites have the ability to distinguish risk groups.

[0119] The present invention is further described below in conjunction with specific examples. It should be understood that these examples are only used to illustrate the present invention and are not intended to limit the scope of the present invention. The experimental methods in the following examples that do not specify specific conditions are usually carried out under conventional conditions, such as the conditions described in CLSI C62-A Clinical Mass Spectrometry New Standard | Liquid Chromatography-Mass Spectrometry Methods-Approved Guideline, or according to the conditions recommended by the manufacturer. Unless otherwise stated, percentages and parts are weight percentages and weight parts.

[0120] Example 1. Screening of indicators and construction of risk model

[0121] 1.1 Experimental subjects

[0122] The risk model was established using clinical variables and biomarker data from 806 patients admitted to a European hospital for acute heart failure, and the optimal indicators were obtained. The model was validated using data from 1,265 patients with new or progressive heart failure from multiple European centers. During at least one year of follow-up, each patient provided a blood sample and written informed consent for risk assessment.

[0123] Table 1 Patient statistics of the modeling cohort (a European hospital) and the validation cohort (multi-center European center)

[0124]

[0125]

[0126] Note: (1) *Only 597 patients

[0127] (2) For continuous variables, data are median (interquartile range); for categorical variables, data are %.

[0128] (3) NYHA: New York Heart Association; eGFR: estimated glomerular filtration rate; NT-proBNP: N-terminal pro-B-type natriuretic peptide; ADHERE: Acute Decompensated Heart Failure National Registry; OPTIMIZE-HF: Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure; GWTG-HF: Get With The Guidelines-Heart Failure.

[0129] 1.2 Metabolites of intestinal flora

[0130] Logistic regression and Cox regression were used to establish a risk stratification model for death and / or readmission of heart failure patients within one year. Multivariate logistic regression and Cox proportional hazard regression were used to analyze the relationship between carnitine-trimethylamine oxide pathway metabolites and death and / or readmission of heart failure within one year. The odds ratio (OR) (logistic regression) and hazard ratio (HR) (Cox proportional hazard regression) are shown in Table 2.

[0131] Table 2 Multivariate logistic regression and Cox proportional hazards regression analysis of the relationship between carnitine-trimethylamine oxide pathway metabolites and one-year heart failure death and / or readmission

[0132]

[0133]

[0134] 1.3 Construction and verification of the first-tier risk model

[0135] The contribution of intestinal flora metabolites was evaluated by odds ratio (OR) (logistic regression) and hazard ratio (HR) (Cox regression) data in Table 2, and a scoring system was established to obtain the risk scores of each intestinal flora metabolite as shown in Table 3.

[0136] Table 3 Risk ranking and risk score of each intestinal flora metabolite

[0137] Metabolites OR / HR ranking Risk score (weight) Acetyl L-Carnitine ＝1 4 γ-Butyl Betaine 4 3 L-Carnitine 3 2 Trimethylamine oxide ＝1 4 total 13

[0138] The highest OR / HRs were for acetyl-L-carnitine (OR, 2.12; HR, 2.03) and trimethylamine oxide (TMAO) (OR, 2.31; HR, 1.94), both of which were rated the highest 4 points, followed by γ-butyl betaine (3 points) and L-carnitine (2 points). Scores were given based on levels above the median (threshold) of each gut microbiota metabolite and then summed up to a maximum of 13 points.

[0139] The weight scores and thresholds of each intestinal flora metabolite are shown in Table 4.

[0140] Table 4 Weight scores and thresholds of each intestinal flora metabolite

[0141]

[0142] In addition, the present invention divides patients into three groups according to the scores of the metabolites of each intestinal flora of the patients: low, medium and high risk groups. Patients with a score of ≤4 are in the low risk group, patients with a score of 5-10 are in the medium risk group, and patients with a score of >10 are in the high risk group, as shown in Table 5 and Figure 1 shown.

[0143] Table 5

[0144]

[0145] For example, in a clinical sample, the concentration of acetyl-L-carnitine is 14.65 μmol / L (greater than the threshold, assigned a value of 1, see Table 4 for the threshold), the concentration of γ-butyl betaine is 1.76 μmol / L (greater than the threshold, assigned a value of 1), the concentration of L-carnitine is 51.58 μmol / L (less than the threshold, assigned a value of 0), and the concentration of trimethylamine oxide is 7.67 μmol / L (greater than the threshold, assigned a value of 1). According to the scoring formula of the first-level risk score: S=4*acetyl-L-carnitine+3*γ-butyl betaine+2*L-carnitine+4*trimethylamine oxide=4*1+3*1+2*0+4*1=4+3+0+4=11, it belongs to the high-risk group.

[0146] Univariate Cox proportional hazards regression analysis showed that the composite endpoint (death and / or readmission within one year) was significantly associated with the risk score when the low-risk group (reference group) was compared with the intermediate-risk group (HR 1.36 (95% CI 1.04-1.77) p = 0.023) and the high-risk group (HR 2.42 (95% CI 1.81-3.23) p < 0.001). Compared with the low-risk group, the high-risk group had increased levels of intestinal flora metabolites, and the survival rate of patients was almost half of that of the former.

[0147] When the scores of gut microbiota metabolites were added to the current risk score model, significant differences were found between the high-risk group and the low-risk group (ADHERE, HR 1.77 95% CI 1.26-2.49 p < 0.001; GWTG-HF HR 1.65 95% CI 1.18-2.32 p = 0.003; and OPTIMIZE-HF HR 1.78 95% CI 1.21-2.64 p = 0.004).

[0148] The above data results show that adding intestinal flora metabolites to the current risk model has an additional effect, that is, it has excellent and specific discrimination ability for certain heart failure patients and can detect heart failure risks that cannot be detected by some existing risk models, such as Figure 2 As shown in AD.

[0149] 1.4 Other indicators

[0150] Univariate Cox proportional hazards regression was used to further analyze clinical and biochemical variables. Ten variables associated with HF risk stratification were identified, including age, previous history of HF, heart rate, systolic blood pressure, diastolic blood pressure, hemoglobin, urea nitrogen, creatinine, eGFR, and N-terminal pro-B-type natriuretic peptide (NT-proBNP) (p≤0.010), as shown in Table 6.

[0151] Table 6 Univariate Cox proportional hazards regression analysis of the relationship between clinical and biochemical variables (p < 0.1) and outcomes (death and / or readmission within 1 year)

[0152] HR 95% CI P-value age 1.24 1.13-1.03 <0.001 Previous history of heart failure 1.64 1.30-2.05 <0.001 Heart rate 0.99 0.99-1.00 0.002 Systolic blood pressure 0.99 0.99-0.99 <0.001 Diastolic blood pressure 0.98 0.98-0.99 <0.001 Hemoglobin 0.93 0.88-0.98 0.010 Urea nitrogen 1.06 1.04-1.07 <0.001 Creatinine 1.01 1.00-1.01 <0.001 <![CDATA[eGFR(mL / min / 1.73m 2 )]]> 0.98 0.97-0.99 <0.001 NT-proBNP (ng / L) 1.81 1.39-2.35 <0.001

[0153] 1.5 Construction of a heart failure risk stratification model incorporating gut microbiota metabolites (GuM-HF model) The 10 variables in 1.4 were used for reverse proportional hazards regression and reverse logistic regression analysis to identify multivariate model factors associated with the risk of heart failure (death and / or readmission due to heart failure within one year).

[0154] Finally, a model including 7 variables (age, previous history of heart failure, heart rate, systolic blood pressure, urea nitrogen, NT-proBNP (or BNP) and intestinal flora metabolites (acetyl-L-carnitine, γ-butylbetaine, L-carnitine and trimethylamine oxide) was established.

[0155] The area under the receiver operating characteristic (ROC) curve / C statistic of the model was 0.70 (95% CI 0.66-0.74, p<0.001), which was significantly higher than the ADHERE scoring model 0.64 (p<0.001), the GWTG-HF scoring model 0.64 (p=0.002) and the OPTIMIZE-HF scoring model 0.65 (p=0.003). The practicability and effectiveness of the model of the present invention (comprising acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide) were demonstrated.

[0156] In order to further establish a practical risk scoring system, the above constructed model was optimized. First, the variables were dichotomized and ranked according to the hazard ratio of reverse cox regression.

[0157] Under this method, those with a history of heart failure and NT-proBNP>2160ng / L (or BNP>480ng / L) get 3 points, otherwise 0 points; those with urea nitrogen>8.9mmol / L and age>78 years old get 2 points, otherwise 0 points; those with heart rate>90 beats / min and systolic blood pressure>133mmHg get 1 point, otherwise 0 points. These scores are added to the scoring system of intestinal flora metabolites (low, medium and high risk, those judged as high risk by the first-level risk model get 3 points, those with medium risk get 2 points, and those with low risk get 1 point), and then the sum is calculated to get a score range of 1-15 points.

[0158] Finally, the risk model scores and risk situations were divided into three groups: ≤4 (low risk), 5-10 (medium risk) and >10 (high risk), as shown in Table 7.

[0159] Table 7

[0160]

[0161] Figure 3 Kaplan-Meier analysis in A showed significant differences in survival associated with increased risk groups (Chi-square test 13.713-38.436, p<0.001) and increased event rates (Group 1 (low risk) = 21%, Group 2 (intermediate risk) = 40%, Group 3 (high risk) = 57%).

[0162] Likewise, Figure 4 As shown, Cox proportional hazards regression showed that the intermediate-risk group and the high-risk group had an increased correlation with the prognostic outcome compared with the low-risk group (reference) (HR 2.10-3.70 (95% CI 1.40-5.73) p<0.001).

[0163] Taken together, these results demonstrate the ability of clinical variables and elevated gut microbiota metabolites to discriminate between risk groups.

[0164] 1.6 Validation of the GuM-HF risk model

[0165] The risk model was validated using the BIOSTAT-CHF cohort, and two methods (method 1 and method 2) were used for validation. The median value of the intestinal flora metabolites in the modeling cohort was used as the threshold for method 1, and the median value of the intestinal flora metabolites in the validation cohort (BIOSTAT-CHF) was used as the threshold for method 2. Survival analysis of the GuM-HF model in the validation cohort showed that the survival rate significantly deteriorated with the increase in the risk group (p<0.001).

[0166] There were significant differences between the high-risk group and the low-risk group (verification method 1, chi-square test 55.637, p<0.001; verification method 2, chi-square test 80.633, p<0.001) and the medium-risk group (verification method 1, chi-square test 27.945, p<0.001; verification method 2, chi-square test 40.443, p<0.001). In both verification methods, patients in the high-risk group showed a lower survival rate.

[0167] There are two verification methods, such as Figure 3 As shown in BC, there are also significant differences between the medium-risk group and the high-risk group (validation method 1, chi-square test 18.384, p < 0.001; validation method 2, chi-square test 22.466, p < 0.001), and the survival rate of the high-risk group is lower. For validation method 1 and validation method 2, the AUC of the risk model using ROC curve analysis was 0.72 and 0.71, respectively ((95% CI, 0.68-0.75) p < 0.001).

[0168] like Figure 5As shown, Cox proportional hazards regression analysis showed that the intermediate-risk group and high-risk group had an increased correlation with prognosis (death and / or readmission due to heart failure within one year) compared with the low-risk group (reference) (Validation 1 HR 2.24-4.69 (95% CI 1.64-7.22) p<0.001; Validation 2 HR 2.49-5.26 (95% CI 1.86-7.80) p<0.001).

[0169] Example 2. Comparison of the first-tier risk model with other similar risk models

[0170] Sample Information

[0171] After obtaining the patient's consent, this example included 30 plasma samples from patients with heart failure, and the blood was collected in the early morning on an empty stomach. The clinical information of the 30 patients with heart failure is as follows:

[0172] ① Samples 1-13: were readmitted to the hospital due to heart failure within one year after being diagnosed with heart failure; ② Samples 14-20: died of heart failure within one year after being diagnosed with heart failure; ③ Samples 21-30: were not readmitted to the hospital or died due to heart failure within one year after being diagnosed with heart failure.

[0173] After preliminary processing, the collected plasma samples were used to determine the content of intestinal flora metabolites using liquid chromatography-mass spectrometry (LC-MS).

[0174] Table 8 shows the concentration values ​​(μmol / L) of various intestinal flora metabolites in samples from 30 patients with heart failure.

[0175] Table 8 Concentration values ​​of each metabolite in the samples (unit: μmol / L)

[0176]

[0177]

[0178] Acquisition of different risk models

[0179] The inventors constructed a comparative risk model based on the average score (not the weight score of the present invention) and the scores of the odds ratio or risk ratio of the same intestinal flora metabolites in other studies, and normalized the sum of the metabolite weights to the same weight of 13 points as the present invention. As shown in Table 9.

[0180] Table 9 Score values ​​of each metabolite in different models

[0181]

[0182]

[0183] Comparative model 1: indicators include acetyl-L-carnitine, betaine, choline, γ-butyl betaine, L-carnitine, trimethylamine oxide and TFF-3 (trefoil factor 3);

[0184] Comparison model 2: indicators included those of comparison model 1, as well as age, previous history of heart failure hospitalization, peripheral edema, systolic blood pressure, BNP, hemoglobin, and serum sodium;

[0185] Comparison model 3: indicators included those of comparison model 2, plus current smoking, chronic obstructive pulmonary disease, and estimated glomerular filtration rate;

[0186] Evaluate risk value for different samples

[0187] According to the risk score (weight) of the metabolites obtained in Example 1 and the weight of each comparison model in Table 9, the concentration threshold of each metabolite in Table 4, and the concentration value of each metabolite in Table 8, each metabolite in each sample is scored.

[0188] When the concentration of metabolites in the sample is higher than the threshold, a corresponding score is obtained, and when it is lower than the threshold, zero score is obtained, and then the final total score of each sample is calculated. For example, when the model of the present invention is used for scoring, the concentration of acetyl-L-carnitine in sample 1 (14.65 μmol / L) is higher than the threshold (12.1 μmol / L), and 4 points are obtained; the concentration of acetyl-L-carnitine in sample 3 (6.77 μmol / L) is lower than the threshold, and 0 points are obtained. The same is true when other comparison models are used for scoring. The final scoring results are shown in Tables 10-12.

[0189] Table 10

[0190]

[0191]

[0192] Table 11

[0193]

[0194]

[0195] Table 12

[0196]

[0197]

[0198] Data analysis

[0199] IBM SPSS Statistics was used for data analysis: the samples of heart failure patients were divided into two groups, sample 1 to sample 20 (re-hospitalized and / or died due to heart failure within one year after diagnosis of heart failure) = 1; sample 21 to 30 (not re-hospitalized or died due to heart failure within one year after diagnosis of heart failure) = 0. The final grouping of the 30 samples of heart failure patients is shown in Table 13.

[0200] Table 13

[0201]

[0202]

[0203] Draw the ROC curve under different scoring conditions and calculate the area under the curve (AUC). The results are as follows Figure 6 shown. Figure 6 A and Figure 6 B are ROC curves under weighted scoring (the present invention) and average scoring conditions, and the calculated AUC values ​​are 0.870 and 0.758, respectively. Figure 6 C. Figure 6 D and Figure 6 E are the scores of comparison model 1, comparison model 2, and comparison model 3, and the calculated AUC values ​​are 0.660, 0.645, and 0.690, respectively.

[0204] The ROC curve is the receiver operating characteristic curve, which is drawn according to the sensitivity and specificity at different thresholds. The area under the ROC curve (AUC) can evaluate the accuracy of the test. AUC < 0.5 is considered to be inaccurate; the test has a lower accuracy when it is 0.5-0.7; the test has a certain accuracy when it is 0.7-0.9; and the test has a higher accuracy when it is > 0.9. The larger the AUC value, the higher the prediction accuracy. Therefore, compared with the average score, weighted scoring of intestinal flora metabolites can help improve the accuracy of the heart failure risk prediction model.

[0205] According to the HR values ​​of various indicators in other studies, the AUC values ​​obtained were scored ( Figure 6 CE) is between 0.645 and 0.690, which is less than 0.870 ( Figure 6 A, weighted score); and the accuracy of the weighted score of the present invention is also significantly higher than the average weighted score of the present invention (ie, when the weights of the metabolites of various intestinal flora are the same).

[0206] Therefore, the risk prediction model of the present invention adopts a method of weighted scoring of intestinal flora metabolites to construct a heart failure risk prediction model, which has high accuracy.

[0207] All documents mentioned in the present invention are cited as references in this application, just as each document is cited as reference individually. In addition, it should be understood that after reading the above teachings of the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the claims attached to this application.

Claims

1. A heart failure risk stratification assessment system incorporating intestinal flora metabolites, characterized in that: The system comprises: (a) an input module, the input module being configured to input heart failure risk indicator data of a subject to be tested; The risk indicators of heart failure include: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure and intestinal flora metabolites; and the intestinal flora metabolites are: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide; (b) a processing module, wherein the processing module is configured to perform the following functions: i) comparing the input risk indicator data with a threshold value pre-stored in the device to obtain a comparison result; if a certain risk indicator data value input is greater than its corresponding threshold value, a value of 1 is assigned; if the risk indicator data value is less than its corresponding threshold value, a value of 0 is assigned; ii) substituting the intestinal flora metabolite data comparison result into the first-level risk scoring formula to obtain its first-level risk score; and then obtaining the risk score value C of the intestinal flora metabolite according to its first-level risk score value; and iii) Substituting the comparison result of the heart failure risk index data of the non-intestinal flora metabolites and the risk score value C into the second-level risk scoring formula for calculation, thereby obtaining the second-level risk score, and thus obtaining the evaluation result; Wherein, when the second-tier risk score is below the risk cutoff value, it indicates that the subject has a low risk of heart failure; when the risk score is higher than the risk cutoff value, it indicates that the subject has a medium or high risk of heart failure; Among them, the first-tier risk scoring formula is: The second-tier risk scoring formula is: Among them, the W i is the weight value of each risk indicator; i is the comparison result of each risk indicator data and its corresponding threshold value; said C is the risk score value of the intestinal flora metabolite obtained according to the result of S1; and (c) An output module, wherein the output module is configured to input the evaluation result.

2. The system according to claim 1, characterized in that The risk is the risk of a heart failure patient being readmitted to the hospital for heart failure and / or dying within one year.

3. The system according to claim 1, characterized in that When the first-level risk score is ≤4, the risk score value C of the intestinal flora metabolites is 1; when the risk score is 5-10, the risk score value C of the intestinal flora metabolites is 2; when the risk score is >10, the risk score value C of the intestinal flora metabolites is 3.

4. The system according to claim 1, characterized in that When the second-layer risk score score is ≤4 points, the subject has a low risk of heart failure; when the second-layer risk score score is 5-10 points, the subject has a moderate risk of heart failure; when the second-layer risk score score is >10, the subject has a high risk of heart failure.

5. The system according to claim 1, wherein: The subject to be tested is a patient with heart failure.

6. The system according to claim 2, characterized in that The scoring formula for the first-level risk score is: S1=4*acetyl-L-carnitine+3*γ-butylbetaine+2*L-carnitine+4*trimethylamine oxide, and the scoring formula for the second-level risk score is: S2=3*previous history of heart failure+3*NT-proBNP (or BNP)+2*urea nitrogen+2*age+1*heart rate+1*systolic blood pressure+C.

7. The system according to claim 1, characterized in that The system further includes (d) a storage module configured to store data selected from the following group: various judgment or calculation result values, various risk indicator thresholds and risk cutoff values.

8. The system of claim 1, further comprising (e) a control module configured to control the operation of each module.

9. A method for stratifying and assessing the risk of heart failure by incorporating intestinal flora metabolites, characterized in that: The method comprises: (a) providing data, wherein the data includes heart failure risk indicator data of a subject to be tested; Among them, the risk indicators of heart failure include: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure and intestinal flora metabolites; and the intestinal flora metabolites are: acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide; (b) Analytical evaluation, said evaluation comprising the following steps: (s1) comparing the risk indicator data with its corresponding threshold value to obtain a comparison result; if a certain risk indicator data value is greater than its corresponding threshold value, assigning a value of 1; if the risk indicator data value is less than its corresponding threshold value, assigning a value of 0; (s2) Substituting the comparison result of the intestinal flora metabolite data into the first-level risk scoring formula to obtain the first-level risk score; and then obtaining the risk score value C of the intestinal flora metabolite according to the first-level risk score; (s3) substituting the comparison result of the heart failure risk index data of the non-intestinal flora metabolites and the risk score value C into the second-level risk scoring formula for calculation, thereby obtaining the second-level risk score; and (s4) comparing the second-tier risk score with the risk cutoff value to obtain an assessment result; Wherein, when the second-tier risk score is below the risk cutoff value, it indicates that the subject has a low risk of heart failure; when the risk score is higher than the risk cutoff value, it indicates that the subject has a medium or high risk of heart failure; Among them, the first-tier risk scoring formula is: The second-tier risk scoring formula is: Among them, the W i is the weight value of each risk indicator; i is the comparison result of each risk indicator data and its corresponding threshold value; said C is the risk score value of the intestinal flora metabolite obtained according to the result of S1; and (c) Outputting the evaluation result.

10. A method for constructing a heart failure risk stratification prediction model, characterized in that: Includes steps: (S1) providing a first data set, wherein the first data set comprises heart failure risk indicator data, and the risk indicator data comprises the following group of data: previous history of heart failure, N-terminal pro-B-type natriuretic peptide (NT-proBNP) or B-type natriuretic peptide (BNP), urea nitrogen, age, heart rate, systolic blood pressure, acetyl-L-carnitine, γ-butyl betaine, L-carnitine and trimethylamine oxide; and (S2) Based on the data information of the first data set, logistic regression and Cox regression are used to construct a first-level risk prediction model based on intestinal flora metabolites, and then a heart failure risk stratification model is constructed based on the first-level risk prediction model.

Images