Establishment method for predicting death risk index of acute diquat pesticide poisoning

Abstract

The invention belongs to the field of pesticide poisoning prognosis, and particularly discloses a method for establishing an acute diquat pesticide poisoning prediction death risk index, which comprises the following steps: collecting data; establishing a death risk index model based on demographic characteristics and blood routine test results; and carrying out model training, internal verification and external verification. The death risk index provided by the invention comprehensively considers age, diquat blood concentration, leukocyte count and glutamic oxalacetic transaminase concentration, and aims to accurately evaluate the death risk of acute diquat pesticide poisoning patients in an index form. Through deep analysis of large-scale clinical data, the death risk index of the invention not only has scientificity and accuracy on the data level, but also can play an important role in clinical practice, can be applied to construction of acute diquat pesticide poisoning death risk prediction terminal equipment, helps medical teams to distribute medical resources more reasonably, and has a good application prospect. High-risk patients are preferentially processed, so that the overall treatment effect and the survival opportunity of the patients are improved.

Description

Technical Field

[0001] The invention belongs to the field of pesticide poisoning prognosis, and specifically discloses a method for establishing a mortality risk index for predicting acute diquat pesticide poisoning. Background Art

[0002] Poisoning is a global public health problem, and effective control of poisoning incidents is of great significance to reducing morbidity and mortality. The American Association of Poison Control Centers pointed out that there are more than 2 million poisoning cases reported in the United States each year, with a total incidence rate of 633 per 100,000 people. Data show that from 2009 to 2013, poisoning has been the fifth leading cause of death in China, mainly related to attempted suicide rather than accidental poisoning.

[0003] Paraquat and its new post-ban substitute, diquat, are toxic to organs such as the kidney and liver. Studies have found that the mortality rate of acute diquat poisoning is about 18.60% to 60.00%, and oral administration is the main route of poisoning. Due to the lack of effective antidotes, emergency departments mainly treat diquat poisoning patients with gastric lavage, adsorption, catharsis and whole-bowel lavage, but these measures may be difficult to implement in underdeveloped areas. The World Health Organization's International Programme on Chemical Safety is the only guideline that provides a lethal dose of diquat. Without considering other factors such as exposure time and age, it defines the lethal dose of diquat as 6 to 12 grams. Indeed, due to the lack of prognostic assessment of patients with diquat poisoning, clinical decision-making often makes it difficult to implement personalized treatment plans for patients. Summary of the invention

[0004] To solve the above problems, the present invention discloses an acute diquat pesticide poisoning prediction mortality risk index and an establishment method thereof, which provides scientific support for clinical decision-making based on demographic characteristics and blood routine test results. The index established based on this method can accurately estimate the mortality risk of patients with acute diquat poisoning.

[0005] The present invention includes the following technical solutions:

[0006] A method for establishing a mortality risk index for predicting acute diquat pesticide poisoning comprises the following steps:

[0007] Step 1: Collect data and build a queue

[0008] Data including patient demographic characteristics and routine blood test results were collected to construct a cohort of patients with acute diquat pesticide poisoning;

[0009] Step 2: Establish a mortality risk index based on demographic characteristics and blood routine test results

[0010] The collected demographic characteristics and blood routine test results were used as candidate predictive variables. The Boruta algorithm, univariate Cox model, and stepwise Cox regression algorithms were used to evaluate the feature importance. The variables that showed predictive value in all three algorithms were included in the death risk index, and a death risk index model based on demographic characteristics and blood routine test results was established.

[0011] Step 3: Model training, internal validation, and external validation

[0012] The cohort data established in step 1 was used as the development set, and the data was divided into a training set and a test set in a certain ratio. The mortality risk index model was trained and internally validated, and externally validated in the cohort data other than the development set.

[0013] Furthermore, in the above method for establishing a risk index for predicting mortality from acute diquat pesticide poisoning, in step 2:

[0014] The Boruta algorithm determines the important features by iteratively generating random forests and calculating the importance of new features using the important features determined in the previous round and their corresponding shadow features;

[0015] The univariate Cox model used a P value < 0.05 as the threshold for screening variables;

[0016] Stepwise Cox regression outputs the variable combination that achieves the minimum AIC.

[0017] Furthermore, in the above method for establishing a risk index for predicting mortality from acute diquat pesticide poisoning, the demographic characteristic is age.

[0018] Furthermore, in the above method for establishing a risk index for predicting death from acute diquat pesticide poisoning, the blood routine test results include initial diquat blood concentration, white blood cell count and aspartate aminotransferase concentration.

[0019] Furthermore, in the above method for establishing a risk index for predicting mortality from acute diquat pesticide poisoning, in step three, the data ratio of the training set to the test set is 7:3.

[0020] Furthermore, in the method for establishing a mortality risk index for predicting acute diquat pesticide poisoning, in step 2, age, initial diquat blood concentration, white blood cell count, and aspartate aminotransferase concentration are incorporated into the model to obtain a mortality risk index based on demographic characteristics and blood routine test results, as shown below:

[0021] Death Risk Index

[0022] =100-(98.2031318439685)*(3.8342×Age+0.0136601

[0023] *Concentration+6.7254×WBC+0.2535×AST)

[0024] Wherein, Age: age (years); Concentration: initial diquat blood concentration (ng / mL); WBC: white blood cell count (*10 9 / L); AST: aspartate aminotransferase concentration (U / L); * represents power operation, and × represents multiplication operation.

[0025] Furthermore, in the above-mentioned method for establishing a death risk index for predicting acute diquat pesticide poisoning, the higher the death risk index, the more serious the patient's condition and the greater the probability of death; the cutoff value for predicting the survival outcome based on the death risk index is 81, that is, if the calculated death risk index is ≤81, the patient's predicted outcome is survival, otherwise, if the calculated death risk index is >81, the patient's predicted outcome is death.

[0026] The present invention also discloses a mortality risk index for predicting acute diquat pesticide poisoning, which is established by any of the methods described above.

[0027] The present invention also discloses a device for predicting the risk of death due to acute diquat pesticide poisoning, comprising:

[0028] Data screening module: used to collect data including the patient's demographic characteristics and blood routine test results, and construct personal data of patients with acute diquat pesticide poisoning;

[0029] Scoring module: Use the above acute diquat pesticide poisoning prediction mortality risk index to score the patient's personal data;

[0030] Probability prediction module: compares the score in the scoring module with the cutoff value of the survival outcome prediction in the database and outputs the prediction result.

[0031] The present invention also discloses a terminal device for predicting the risk of death due to acute diquat pesticide poisoning, which is characterized in that it includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it calls the above-mentioned acute diquat pesticide poisoning prediction risk of death index.

[0032] The present invention has the following beneficial effects:

[0033] Based on demographic characteristics and routine blood test results, the present invention proposes an index model for predicting the risk of death in patients with acute diquat pesticide poisoning. The model comprehensively considers multiple key factors, including the patient's age, diquat blood concentration, white blood cell count and aspartate aminotransferase level, thereby providing an accurate prognosis prediction for poisoned patients and ensuring the reliability and accuracy of the judgment. Through in-depth analysis of large-scale clinical data, the death risk index of the present invention is not only scientific and accurate at the data level, but also can play an important role in clinical practice. Specifically, the index provides an important decision-making tool for medical staff during the treatment process, enabling them to formulate more personalized and effective treatment plans based on the specific conditions of the patient, especially the demographic characteristics and routine blood test results. At the same time, in emergency medical situations, the index can also help the medical team to allocate medical resources more reasonably and give priority to high-risk patients, thereby improving the overall treatment effect and the patient's chance of survival. In addition, the death risk index of the present invention is also highly adaptable and practical. It can not only be promoted and applied in different types of medical institutions, but also can be adjusted and optimized according to continuously updated clinical data to ensure its accuracy and effectiveness in different clinical scenarios. Through this index, clinicians can better understand the pathological process of diquat poisoning, optimize treatment strategies, and ultimately improve patients' prognosis and quality of life. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is the ROC curve in the embodiment of the present invention;

[0035] Figure 2 is the confusion matrix in the embodiment of the present invention;

[0036] Figure 3 It is a nomogram in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] The reagents or instruments used in the examples of the present invention without indicating the manufacturer are all conventional reagent products that can be obtained through commercial purchase.

[0039] The data used in this example are from the Jiangsu Medical Association Poisoning Branch cohort, including demographic characteristics and blood test results of patients obtained from the hospital electronic medical record system, as well as 28-day prospective follow-up outcomes. This project uses data from teaching hospitals of Nanjing Medical University for model development (training and internal validation), and uses patient data from other hospitals for external validation.

[0040] This study (ethics number: 2021-SR-394, ClinicalTrial.gov ID: NCT 05215457, October 7, 2017) was approved by the Medical Ethics Committee of the First Affiliated Hospital of Nanjing Medical University and was conducted in accordance with the principles of the Declaration of Helsinki. This study only involved the analysis of routinely collected data, so a written informed consent exemption was obtained from the institutional review board. To protect the privacy of the patients, all information was anonymized before analysis.

[0041] Inclusion criteria were as follows: (1) history of oral diquat exposure reported by the patient or his / her legal representative; (2) completion of toxicological tests immediately after admission; (3) complete blood routine test results; (4) the patient or his / her legal representative (in the case of unconsciousness) knew and agreed to the treatment plan. Exclusion criteria were as follows: (1) the patient had ingested other toxicants in addition to diquat; (2) no diquat pesticide residues were detected in blood or urine samples and no clinical symptoms occurred; (3) a history of major primary heart, lung, liver, kidney, brain, etc. diseases.

[0042] Finally, 104 patients from the teaching hospital of Nanjing Medical University were included in the development set, and data from 98 patients from the Jiangsu Medical Association Poisoning Branch cohort other than the development set were included in the external validation set.

[0043] Example

[0044] A method for establishing a mortality risk index for predicting acute diquat pesticide poisoning comprises the following steps:

[0045] Step 1: Collect data and build a queue

[0046] The collected data included demographic characteristics and blood test results to construct a cohort of patients with acute diquat pesticide poisoning. Demographic characteristics included age and gender; blood test results included initial diquat blood concentration, aspartate aminotransferase, platelet count, hemoglobin count, red blood cell count, white blood cell count, blood urea nitrogen, neutrophil count, monocyte count, serum creatinine concentration, lymphocyte count, basophils, and eosinophils.

[0047] Step 2: Establish a mortality risk index based on demographic characteristics and blood routine test results

[0048] The collected demographic characteristics and blood routine test results were used as candidate predictor variables, and the feature importance was evaluated using three algorithms: Boruta algorithm, univariate Cox model, and stepwise Cox regression.

[0049] First, the Boruta algorithm is used to screen these candidate variables. Boruta is a feature selection algorithm based on random forests. The key is that it generates multiple "shadow variables", which are meaningless data generated by randomizing the original data. The random forest model trains all candidate variables together with these shadow variables. The algorithm determines whether the variable is an important feature by comparing the importance score of each actual variable with the shadow variable. Specifically, the Boruta algorithm iteratively constructs multiple random forest models to gradually eliminate actual variables with scores lower than the shadow variables. Finally, those variables with scores significantly higher than the shadow variables are retained, and these variables are considered to have important contributions to survival outcomes. The advantage of this step is that it can capture the nonlinear relationship between complex data from multiple dimensions and screen out more comprehensive candidate variables.

[0050] Secondly, univariate Cox regression analysis was performed to evaluate the association between variables and survival outcomes. In this step, each variable was subjected to a separate Cox regression analysis to evaluate the strength of its association with prognostic events. By calculating the hazard ratio (HR) of each variable and its corresponding P value, it can be determined which variables are statistically significant in the univariate analysis. The P value was set to less than 0.05 as the screening threshold, that is, only those variables that were statistically significantly associated with prognosis were retained. This step helps to further narrow the range of candidate variables and remove those variables that do not have a significant effect on prognosis when viewed individually.

[0051] Finally, a stepwise Cox regression analysis was performed. This step uses the stepwise regression method to optimize the variable combination according to AIC (Akaike Information Criterion). AIC is a standard for evaluating model quality, which finds a balance between model fit and complexity. Stepwise Cox regression repeatedly adds and removes variables, constantly comparing the AIC values ​​of each model to find the model with the smallest AIC. The variable combination that is finally retained not only has strong explanatory power, but also avoids overfitting and ensures the generalization ability of the model. In the stepwise regression process, the introduction or elimination of variables is dynamic, and the algorithm will continue to adjust until the AIC value of the model reaches the optimal state. This step helps to build a concise and accurate model. Among the three algorithms, age, diquat blood concentration, white blood cell count, and aspartate aminotransferase level can reflect predictive value and be included in the mortality risk index;

[0052] The predicted mortality risk index for acute diquat pesticide poisoning is as follows:

[0053] Death Risk Index

[0054] =100-(98.2031318439685)*(3.8342×Age+0.0136601

[0055] *Concentration+6.7254×WBC+0.2535×AST)

[0056] Wherein, Age: age (years); Concentration: initial diquat blood concentration (ng / mL); WBC: white blood cell count (*10 9 / L); AST: aspartate aminotransferase concentration (U / L); * represents power operation, and × represents multiplication operation.

[0057] Step 3: Model training, internal validation, and external validation

[0058] A Cox regression model was trained on the development set, which was divided into training and internal validation datasets in a ratio of 7:3 to construct a risk index for predicting mortality in acute diquat pesticide poisoning. The consistency index (C-index) and area under the curve (AUC) were used to evaluate model fit, with 0.5 indicating random prediction and 1.0 indicating perfect accuracy. The model performance was further evaluated using receiver operating characteristic (ROC) curves, positive predictive value (PPV), negative positive value (NPV), accuracy, and confusion matrix. All statistical analyses were performed using R software (version 4.1.2). Statistical tests were two-tailed, with an α = 0.05 statistical significance level.

[0059] As shown in Table 1 , the development group included 104 patients with a mortality rate of 36.5%. 56% were male. Exposure history (self-reported diquat exposure and exposure time), initial diquat blood concentration, white blood cell count, monocyte count, basophil, red blood cell count, hemoglobin, alanine aminotransferase, aspartate aminotransferase concentration, blood urea nitrogen, serum creatinine concentration, and neutrophil-to-lymphocyte ratio were significantly different between the survivors and nonsurvivors in the development group. All parameters were comparable between the training and test sets (Table S2). The external validation set included 98 patients, including 69 survivors and 29 nonsurvivors. Most of them were female (61.2%). Survivors (median: 20 years) were younger than nonsurvivors (median: 30 years). Exposure history (self-reported diquat exposure and exposure time), initial diquat blood concentration, white blood cell count, monocyte count, red blood cell count, hemoglobin, alanine aminotransferase, aspartate aminotransferase concentrations, and serum creatinine concentrations were significantly different between surviving and nonsurviving patients.

[0060] Table 1 Comparison of demographic characteristics, exposure history, and biomarkers between patients who survived and those who died of acute diquat poisoning.

[0061]

[0062] The risk of death at 28 days increased with age (hazard ratio: 1.04 [1.02, 1.08]), with initial diquat blood concentration (hazard ratio: 1.14 [1.06, 1.23]), with white blood cell count (hazard ratio: 1.07 [1.03, 1.11]), and with aspartate aminotransferase level (hazard ratio: 1.002 [1.001, 1.004]). Figure 1 As shown in the figure, the consistency index of the mortality risk index for acute diquat pesticide poisoning was 0.86±0.06 (training set), 0.92±0.07 (internal validation set), and 0.87±0.07 (external validation set); the area under the ROC curve was 0.92 [0.85, 0.98] (training set), 0.93 [0.84, 1] (internal validation set), and 0.90 [0.83, 0.97] (external validation set). Figure 2 As shown, the PPV of the three data sets was 1, and there was no false positive (predicting the survival of patients as death). The NPV of the training, test, and external validation sets were 0.75, 0.77, and 0.77, respectively. In addition, the training accuracy of the model was 0.81, the test accuracy was 0.78, and the external validation accuracy was 0.79.

[0063] After training, the mature model can be used to estimate the risk of death in patients poisoned by diquat. Figure 3 As shown, when using it, first find the numerical position corresponding to each variable in the nomogram according to the patient's specific data (including age, initial diquat blood concentration, white blood cell count, and aspartate aminotransferase concentration), and draw a vertical line downward to intersect the "score" axis. After reading the score of each variable on the "score" axis, add up the scores to get the total score. Then draw a vertical line downward at the corresponding position of the total score on the "total score" axis to find the probability of death on the "Acute diquat pesticide poisoning predicted death risk index" axis. The death risk index uses 81 points as the dividing line for predicting death risk: when the total score is greater than or equal to 81 points, the patient is predicted to be at high risk and the probability of death increases significantly; when the total score is less than 81 points, the patient is predicted to be at low risk and the probability of death is low. This risk value can help clinicians more accurately assess the patient's prognosis and provide guidance in the formulation of treatment plans.

[0064] The above embodiments show and describe the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the present invention in any form, and any technical solution obtained by equivalent replacement or equivalent transformation falls within the protection scope of the present invention.

Claims

1. A method for establishing a risk index for predicting mortality from acute diquat pesticide poisoning, characterized in that: The steps include: Step 1: Collect data and build a queue Data including patient demographic characteristics and routine blood test results were collected to construct a cohort of patients with acute diquat pesticide poisoning; Step 2: Establish a mortality risk index based on demographic characteristics and blood routine test results The collected demographic characteristics and blood routine test results were used as candidate predictive variables. The Boruta algorithm, univariate Cox model, and stepwise Cox regression algorithms were used to evaluate the feature importance. The variables that showed predictive value in all three algorithms were included in the death risk index, and a death risk index model based on demographic characteristics and blood routine test results was established. Step 3: Model training, internal validation, and external validation The cohort data established in step 1 was used as the development set, and the data was divided into a training set and a test set in a certain ratio. The mortality risk index model was trained and internally validated, and externally validated in the cohort data other than the development set.

2. The method for establishing a risk index for predicting death from acute diquat pesticide poisoning according to claim 1, characterized in that: In the step 2: The Boruta algorithm determines the important features by iteratively generating random forests and calculating the importance of new features using the important features determined in the previous round and their corresponding shadow features; The univariate Cox model used a P value < 0.05 as the threshold for screening variables; Stepwise Cox regression outputs the variable combination that achieves the minimum AIC.

3. The method for establishing a risk index for predicting death from acute diquat pesticide poisoning according to claim 1, characterized in that: The demographic characteristic was age.

4. The method for establishing a risk index for predicting death from acute diquat pesticide poisoning according to claim 1, characterized in that: The routine blood test results include initial diquat blood concentration, white blood cell count and aspartate aminotransferase concentration.

5. The method for establishing a risk index for predicting death from acute diquat pesticide poisoning according to claim 1, characterized in that: In step 3, the data ratio of the training set to the test set is 7:

3.

6. The method for establishing a risk index for predicting death from acute diquat pesticide poisoning according to claim 1, characterized in that: In step 2, age, initial diquat blood concentration, white blood cell count, and aspartate aminotransferase concentration are incorporated into the model to obtain a mortality risk index based on demographic characteristics and blood routine test results, as shown below: Death RiskIndex =100-(98.2031318439685)*(3.8342×Age+0.0136601 *Concentration+6.7254×WBC+0.2535×AST) Wherein, Age: age (years); Concentration: initial diquat blood concentration (ng / mL); WBC: white blood cell count (*10 9 / L); AST: aspartate aminotransferase concentration (U / L); * represents power operation, and × represents multiplication operation.

7. The method for establishing a risk index for predicting death from acute diquat pesticide poisoning according to claim 6, characterized in that: The higher the death risk index is, the more serious the patient's condition is and the greater the probability of death is. The cutoff value of the survival outcome prediction based on the death risk index is 81, that is, if the calculated death risk index is ≤81, the patient's predicted outcome is survival, otherwise, if the calculated death risk index is >81, the patient's predicted outcome is death.

8. An acute diquat pesticide poisoning prediction mortality risk index, characterized in that: The method is established by any one of claims 1 to 7.

9. A device for predicting the risk of death from acute diquat pesticide poisoning, characterized in that: include: Data screening module: used to collect data including the patient's demographic characteristics and blood routine test results, and construct personal data of patients with acute diquat pesticide poisoning; Scoring module: using the acute diquat pesticide poisoning prediction death risk index as described in claim 8 to score the patient's personal data; Probability prediction module: compares the score in the scoring module with the cutoff value of the survival outcome prediction in the database and outputs the prediction result.

10. A terminal device for predicting the risk of death from acute diquat pesticide poisoning, characterized in that: The system comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, wherein the processor calls the acute diquat pesticide poisoning prediction death risk index as claimed in claim 8 when executing the computer program.

Images