An elderly multiple chronic disease elderly critical patient prognosis poor risk auxiliary evaluation system

Abstract

The present application relates to the technical field of medical data processing and clinical risk auxiliary evaluation, and particularly relates to an elderly critical patient with multiple chronic diseases prognosis risk auxiliary evaluation system. In view of the problems that the existing evaluation method is difficult to fully integrate MEWS score, DAVROS score and key clinical indicators, and part of the high-risk patients is not fully identified and the risk result is not intuitive enough, the present application integrates APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level and GCS score, and uses nomogram model to calculate the risk, so as to realize the quantitative evaluation of patient prognosis risk, the probability prediction and the risk stratification output.

Description

Technical Field

[0001] This invention relates to the field of medical data processing technology, specifically to an auxiliary assessment system for poor prognosis risk in elderly patients with multiple chronic diseases and acute and severe illnesses. Background Technology

[0002] As my country's population ages, the demand for emergency and critical care for elderly patients is increasing significantly. Elderly patients experience a gradual decline in bodily functions, with immune, cardiopulmonary, renal, and neuroregulatory abilities all significantly reduced compared to younger patients. They often also suffer from two or more chronic diseases simultaneously, such as cardiovascular and cerebrovascular diseases, diabetes, chronic respiratory diseases, and chronic kidney disease. The long-term presence of multiple chronic diseases further weakens a patient's pathophysiological reserves, making them more susceptible to rapid disease progression, multiple organ dysfunction, and even death when faced with acute illnesses such as infections, respiratory failure, heart failure, cerebrovascular events, and metabolic disorders.

[0003] In the clinical practice of emergency and intensive care units, elderly patients with multiple chronic diseases and acute and critical illnesses are characterized by complex conditions, multiple underlying diseases, atypical clinical manifestations, rapid changes in condition, and high treatment risks. Failure to identify high-risk patients with poor prognosis in the early stages of admission may lead to insufficient monitoring, delayed intervention, or inefficient allocation of medical resources. Therefore, developing an auxiliary tool that can early, accurately, and intuitively assess the risk of poor prognosis in these patients is of great significance for optimizing clinical triage, developing individualized treatment plans, and improving the efficiency of emergency and critical care.

[0004] Currently, the Modified Early Warning Score (MEWS), commonly used in clinical practice, primarily stratifies patients' risk based on vital signs such as body temperature, heart rate, respiratory rate, systolic blood pressure, and level of consciousness. It offers advantages such as ease of operation, rapid data acquisition, and suitability for early emergency assessment. However, MEWS mainly reflects the degree of abnormality in a patient's current vital signs and cannot fully reflect the complexities of elderly patients, including the cumulative effects of chronic diseases, declining organ reserve, and multiple system dysfunctions. Therefore, it may lack specificity and fail to adequately identify some high-risk patients in elderly patients with multiple chronic diseases and acute and critical illnesses.

[0005] The DAVROS score is a risk assessment tool for emergency patients. It comprehensively evaluates the prognostic risk of emergency patients by integrating parameters such as age, physiological function indicators, and blood biochemistry test results. Compared with vital sign scores alone, the DAVROS score can incorporate more emergency risk-related information. However, when used alone, it should be used for critically ill patients who have been diagnosed and is not suitable for widespread screening.

[0006] In addition, indicators such as APACHE II score, GCS score, peripheral blood oxygen saturation (SpO2), and total bilirubin are closely related to the severity of illness and prognosis in critically ill patients. Among them, the APACHE II score reflects acute physiological disturbances and the burden of underlying diseases, the GCS score reflects the level of consciousness, SpO2 reflects oxygenation, and total bilirubin can reflect changes in liver function and overall condition to some extent. However, in current technologies, these indicators are often used individually or assessed separately, lacking an auxiliary assessment system that can integrate MEWS, DAVROS, and key clinical indicators to form a visualized risk score and risk stratification result. Summary of the Invention

[0007] In view of the aforementioned background art, the existing assessment methods are difficult to fully integrate MEWS scores, DAVROS scores, and key clinical indicators, and there are problems such as insufficient identification of some high-risk patients and insufficient intuitiveness of risk results. This application aims to provide an auxiliary assessment system for poor prognostic risk of elderly patients with multiple chronic diseases and acute and critical illnesses. By integrating APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score, and using a nomogram model for risk calculation, it can realize the quantitative assessment of the patient's poor prognostic risk, the prediction of the probability of risk occurrence, and the output of risk stratification.

[0008] In a first aspect, the present invention provides an auxiliary assessment system for the poor prognostic risk of elderly patients with multiple chronic diseases and acute and severe illnesses, comprising:

[0009] The data input module is used to acquire clinical indicator data of the patients to be evaluated. The clinical indicator data includes APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score.

[0010] The analysis module is used to input APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level and GCS score into a preset adverse prognostic risk prediction model, calculate the adverse prognostic risk score of the patient to be evaluated, and determine the corresponding adverse prognostic probability and risk stratification results based on the adverse prognostic risk score.

[0011] The results output module is used to output the patient's poor prognosis risk score, probability of poor prognosis, and risk stratification results.

[0012] The pre-defined adverse prognostic risk prediction model is a nomogram model, which converts the APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score into corresponding individual scores, and adds up the individual scores to obtain the total score of the nomogram model. The total score of the nomogram model is used to represent the adverse prognostic risk score of the patient to be evaluated.

[0013] Optionally, the risk stratification results determined based on the adverse prognostic risk score include:

[0014] Compare the adverse prognostic risk score with the preset risk cutoff value;

[0015] When the adverse prognostic risk score is greater than or equal to the preset risk cutoff value, the patient to be evaluated is identified as high-risk for adverse prognosis; when the adverse prognostic risk score is less than the preset risk cutoff value, the patient to be evaluated is identified as low-risk for adverse prognosis.

[0016] Optionally, the default risk cutoff value is 126 points.

[0017] Optionally, the nomogram model for predicting poor prognosis risk is constructed using the following method:

[0018] To obtain clinical sample data and follow-up prognostic results of elderly patients with multiple chronic diseases and acute and critical illnesses;

[0019] Univariate analysis was used to screen candidate variables from clinical sample data, and follow-up prognosis was used as the dependent variable, with death assigned a value of 1 and survival assigned a value of 0.

[0020] Candidate variables were incorporated into the LASSO regression model, and the penalty parameter λ was determined through 10-fold cross-validation. Candidate predictive variables were then selected.

[0021] Multicollinearity diagnosis was performed on the candidate predictor variables. When the variance inflation factor of each candidate predictor variable was less than 10, the candidate predictor variables were incorporated into a multivariate logistic regression model to construct a nomogram model for predicting adverse prognostic risk.

[0022] Optionally, candidate variables include GCS score, respiratory rate, heart rate, SpO2 level, body temperature, hemoglobin level, white blood cell count, platelet count, serum sodium level, serum potassium level, urea level, creatinine level, blood glucose level, total bilirubin level, mechanical ventilation time, total length of hospital stay, APACHE II score, MEWS score, and DAVROS score.

[0023] Optionally, candidate predictors include APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score.

[0024] Optionally, the nomogram model for predicting adverse prognostic risk includes nine rows of scale arranged from top to bottom:

[0025] The first line is a scoring scale, with a score range of 0 to 100, used to determine the individual score for each clinical indicator.

[0026] The second row is the APACHE II scoring scale, with the index value ranging from 7 to 15. As the APACHE II score increases, the corresponding individual item score also increases.

[0027] The third row is a scale for total bilirubin levels, in μmol / L, with values ​​ranging from 9 to 21. The individual scores increase as total bilirubin levels rise.

[0028] The fourth row is the MEWS scoring scale, with the index value ranging from 1 to 4. As the MEWS score increases, the corresponding individual score also increases.

[0029] The fifth row is the DAVROS rating scale, with the index value ranging from 6 to 21. As the DAVROS score increases, the corresponding individual item score also increases.

[0030] The sixth line is a SpO2 level scale, with the unit being %, and the index value ranges from 96 to 100. The individual score increases when the SpO2 level decreases.

[0031] The seventh row is the GCS scoring scale, with the index value ranging from 7 to 15. When the GCS score decreases, the corresponding individual item score increases.

[0032] The eighth line is the total score scale, with a score range of 0 to 204, used to display the sum of the individual scores for the APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score;

[0033] The ninth line is a prognostic risk scale with a risk probability range of 0.10 to 0.95, used to determine the probability of adverse prognosis for the patient to be evaluated based on the total score.

[0034] Compared with existing technologies, the prognostic risk assessment system for elderly patients with multiple chronic diseases and acute and severe illnesses of the present invention brings the following significant effects:

[0035] 1. This invention incorporates APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score into a single adverse prognostic risk prediction model, simultaneously reflecting the severity of the patient's condition, liver function status, early warning risk, emergency risk, oxygenation status, and level of consciousness. Compared to methods relying on a single score or single clinical indicator, this invention improves the completeness of the assessment dimensions, helping to more comprehensively characterize the adverse prognostic risk of elderly patients with multiple chronic diseases and acute and severe illnesses.

[0036] 2. This invention employs a multivariate logistic regression model and the nomogram model derived from it to convert multiple clinical indicators into corresponding individual scores, and outputs a poor prognostic risk score or probability of occurrence in the form of a total score. This method enables complex statistical models to be presented in a visual and quantitative form, facilitating medical personnel to quickly obtain risk assessment results based on specific patient indicators, thereby improving the intuitiveness and operability of risk assessment results;

[0037] 3. This invention outputs a poor prognostic risk score, the probability of poor prognosis, and risk stratification results through a results output module. A score of 126 is used as the risk stratification cutoff value to classify patients under evaluation as either high-risk or low-risk for poor prognosis. This system can provide auxiliary reference for early clinical identification of high-risk patients, helping medical staff to assess monitoring intensity, allocate treatment resources, and make subsequent management decisions based on the patient's actual condition. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a graph showing the changes in the regression coefficients of each candidate variable as a function of Log(λ) in LASSO regression.

[0040] Figure 2 A variable selection plot for determining the optimal penalty parameter λ for LASSO regression based on 10-fold cross-validation;

[0041] Figure 3 This is a nodal chart model for intelligent prediction of adverse prognostic risk based on MEWS-DAVROS;

[0042] Figure 4 ROC curve for internal validation of prediction performance of the MEWS-DAVROS intelligent prediction nomogram model;

[0043] Figure 5 DCA decision curve for internal validation of clinical net benefit of MEWS-DAVROS intelligent prediction nomogram model;

[0044] Figure 6 Calibration curves for internal verification of prediction consistency in the MEWS-DAVROS intelligent prediction nomogram model;

[0045] Figure 7 ROC curve for external validation of the prediction performance of the MEWS-DAVROS intelligent prediction nomogram model;

[0046] Figure 8 Calibration curves for external validation of the prediction consistency of the MEWS-DAVROS intelligent prediction nomogram model. Detailed Implementation

[0047] To enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and embodiments. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various modifications or alterations to the present invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0048] Example 1: Construction of a MEWS-DAVROS-based Intelligent Predictive Nodal Chart Model

[0049] 1. Materials and Methods

[0050] 1.1 Case Data

[0051] Sixty elderly patients with multiple chronic diseases and acute and critical illnesses admitted to the Emergency Department and Intensive Care Unit of our hospital from April 2023 to April 2024 were selected as the study subjects. Inclusion criteria: ① Age 80 years and above; ② Diagnosed with at least two chronic diseases and currently in the acute exacerbation stage, requiring emergency intervention or ICU treatment; ③ Normal level of consciousness upon admission, able to understand and cooperate with the research process; ④ Compliance with the principle of informed consent; ⑤ Expected hospitalization time must last for 1 day or more. Exclusion criteria: ① Comorbid severe trauma or requiring surgery; ② Comorbid terminal diseases such as advanced tumors; ③ Participation in other interventional clinical studies before admission; ④ Missing or incomplete key diagnostic and treatment information. This invention was reviewed by the Medical Ethics Management Committee of the Affiliated Honghe Hospital of Kunming Medical University (ethics approval number: (2024) Yun Dian Nan Lun Shen No. (96)) and implemented in accordance with the principles of the Declaration of Helsinki.

[0052] 1.2 Methods

[0053] 1.2.1 Clinical Data Collection

[0054] This invention system collects patients' clinical data, including basic information (name, gender, age), medical information (chief physician, past medical history, length of hospital stay, and outcome), and immediate indicators in the resuscitation room: vital signs (body temperature, blood pressure, heart rate, respiratory rate, peripheral oxygen saturation (SpO2), Glasgow Coma Scale (GCS), etc.) and laboratory tests (complete blood count, electrolytes, renal function, blood glucose); as well as medical history (circulatory system diseases, respiratory system diseases, digestive system diseases, neurological system diseases, etc.). All data are collected and stored in a standardized manner.

[0055] 1.2.2 MEWS, DAVROS, and Apache II scores

[0056] Data on patient condition severity were collected upon admission. The DAVROS score, which integrates multiple parameters including age, physiological function indicators, and blood biochemistry results, constructs a comprehensive risk assessment model. The score is significantly positively correlated with patient prognosis; higher scores indicate more severe conditions. The MEWS score monitors five key clinical indicators: heart rate, respiratory rate, body temperature, systolic blood pressure, and level of consciousness. The score ranges from 0 to 15, with higher scores indicating worsening clinical condition. The APACHE II score includes 12 core physiological parameters, with a total score range of 0 to 71. This system comprehensively reflects the degree of acute physiological dysfunction and the burden of underlying diseases; higher scores indicate increased severity of illness.

[0057] 1.2.3 Follow-up and grouping

[0058] Follow-up will be conducted for six months via telephone, outpatient visits, or readmissions, and detailed statistics will be compiled on patient prognosis, including deaths from any cause. Based on the follow-up results, patients who died will be classified as the death group (13 cases), and those who did not die will be classified as the survival group (47 cases).

[0059] 1.3 Quality Control

[0060] Regarding data integrity, the missing rate of key variables was strictly controlled to be below 5%. For the unavoidable data missing issues in the study, multiple imputation was used for data processing. In the data entry stage, a dual-person independent entry method was adopted, and the Kappa test was used to ensure that the consistency of entry reached above 0.85. At the same time, 10% of the cases were randomly selected for verification. To control research bias, the research team developed a unified operation manual to clearly define the standards of each variable, and implemented blinding in the data analysis stage to ensure that the data analysts were unaware of the case grouping information.

[0061] 1.4 Statistical Analysis

[0062] This invention uses SPSS 27.0 and R4.1.1 statistical software for data analysis. Categorical variables are described in the form of [n (%)], and chi-square tests are used for comparisons between groups; continuous variables are described in the form of (n(%)). (±s) indicates the mean, and the independent samples t-test was used to compare differences between groups. Binary logistic regression analysis was performed on factors influencing poor prognosis in elderly patients with multiple chronic diseases and acute and severe illnesses. A nomogram model was constructed using R software, and the receiver operating characteristic (ROC) was used to analyze the predictive power of the model. All statistical tests were two-tailed, with P < 0.05 as the cutoff criterion.

[0063] 2 Results

[0064] 2.1 Clinical Data

[0065] The follow-up results of this invention showed that 13 patients died and 47 survived. Analysis of the clinical data of the two groups showed that the survival group had significantly lower respiratory rate, heart rate, body temperature, white blood cell count, platelet count, potassium, urea, creatinine, blood glucose, total bilirubin, mechanical ventilation time, APACHE II score, MEWS, DAVROS score, and total length of hospital stay than the death group; while the survival group had significantly higher GCS score, systolic blood pressure, SpO2, hemoglobin, and sodium (P<0.05), as shown in Table 1.

[0066] Table 1 Comparison of clinical data [n (%) / ( ±s)]

[0067]

[0068]

[0069] Note: "-" indicates that Fisher's exact probability method was used for calculation.

[0070] 2.2 LASSO regression and multivariate logistic regression analysis

[0071] See Figure 1 , Figure 2Using the prognostic outcomes of elderly patients with multiple chronic diseases and acute and severe illnesses as the dependent variable (mortise = 1, survival = 0), candidate variables with P < 0.05 in the univariate analysis in Table 1 were included in LASSO regression analysis. During the screening process, the mean squared error as a function of log(λ) was plotted using the 10-fold cross-validation method. Finally, six influencing factors were selected: APACHE II score, total bilirubin, MEWS score, DAVROS score, SpO2 level, and GCS score. All were substituted with real values. Multicollinearity diagnosis showed that the VIF among the independent variables was < 10. Further multivariate logistic regression analysis (Table 2) showed that the total bilirubin level (OR = 1) was the most significant factor. The following risk factors were identified during follow-up after treatment: GCS score (OR=0.522, 95%CI: 1.521-1.944), APAPACHE II score (OR=2.053, 95%CI: 1.212-2.917), MEWS score (OR=2.156, 95%CI: 1.478-2.824), and DAVROS score (OR=1.679, 95%CI: 1.035-2.264). The following protective factors were identified: GCS score (OR=0.520, 95%CI: 0.119-0.922) and SpO2 level (OR=0.591, 95%CI: 0.172-0.915) (P<0.05).

[0072] Table 2 Multivariate Logistic Regression Analysis

[0073]

[0074] 2.3 Construction of MEWS-DAVROS-based Intelligent Predictive Nodal Chart Model

[0075] This application uses R software to construct a MEWS-DAVROS-based intelligent prediction nomogram based on APACHE II score, total bilirubin, MEWS score, DAVROS score, SpO2, and GCS score. The highest individual scores for each indicator are as follows: APACHE II score 39 points, total bilirubin 26 points, MEWS score 22 points, DAVROS score 46 points, SpO2 41 points, and GCS score 30 points, for a total score of 204 points. Higher scores indicate a higher risk of poor prognosis in elderly patients with multiple chronic diseases and acute or severe illnesses. (See [link to relevant documentation]). Figure 3 .

[0076] The nomogram model includes 9 rows of rulers arranged from top to bottom:

[0077] The first line is a scoring scale, with a score range of 0 to 100, used to determine the individual score for each clinical indicator.

[0078] The second row is the APACHE II scoring scale, with the index value ranging from 7 to 15. As the APACHE II score increases, the corresponding individual item score also increases.

[0079] The third row is a scale for total bilirubin levels, in μmol / L, with values ​​ranging from 9 to 21. The individual scores increase as total bilirubin levels rise.

[0080] The fourth row is the MEWS scoring scale, with the index value ranging from 1 to 4. As the MEWS score increases, the corresponding individual score also increases.

[0081] The fifth row is the DAVROS rating scale, with the index value ranging from 6 to 21. As the DAVROS score increases, the corresponding individual item score also increases.

[0082] The sixth line is a SpO2 level scale, with the unit being %, and the index value ranges from 96 to 100. The individual score increases when the SpO2 level decreases.

[0083] The seventh row is the GCS scoring scale, with the index value ranging from 7 to 15. When the GCS score decreases, the corresponding individual item score increases.

[0084] The eighth line is the total score scale, with a score range of 0 to 204, used to display the sum of the individual scores for the APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score;

[0085] The ninth line is a prognostic risk scale with a risk probability range of 0.10 to 0.95, used to determine the probability of adverse prognosis for the patient to be evaluated based on the total score.

[0086] Example 2: Performance Evaluation of the MEWS-DAVROS Intelligent Predictive Nodal Chart Model

[0087] Data from 60 patients described in Example 1 were used, and internal validation was performed by repeated sampling 1000 times using the Bootstrap method. ROC analysis showed that the MEWS-DAVROS intelligent predictive nomogram model had high predictive efficacy for poor prognosis in elderly patients with multiple chronic diseases and acute and severe illnesses, with AUC=0.863 (95%CI: 0.782–0.944, P < 0.001, optimal cutoff point was 126 points, Youden index was 0.632, sensitivity was 82.6%, and specificity was 80.6%). Figure 4 The prediction performance of the MEWS-DAVROS intelligent prediction model was validated using DCA curves, revealing that the prediction probability exceeded 75%, demonstrating positive net benefits. (See attached image.) Figure 5The calibration curve closely approximates the ideal curve, indicating that the nomogram model's prediction of poor patient prognosis is consistent with actual observations (P < 0.05). Figure 6 .

[0088] External validation was conducted on 33 elderly patients with multiple chronic diseases and acute and critical illnesses admitted to the Emergency Department and Intensive Care Unit of our hospital from May to December 2024. Inclusion / exclusion criteria were the same as in Example 1. Follow-up results showed that 27 patients survived and 6 died. External validation results showed that the MEWS-DAVROS intelligent predictive nomogram model had an AUC of 0.833 (95% CI: 0.776–0.892, P < 0.001) for predicting poor prognosis in elderly patients with multiple chronic diseases and acute and critical illnesses, with a Youden index of 0.611, a sensitivity of 83.3%, and a specificity of 77.8%. (See [link to relevant documentation]). Figure 7 The calibration curve closely approximates the ideal curve, indicating that the nomogram model's prediction of poor patient prognosis risk is consistent with actual observations. (See...) Figure 8 .

[0089] This embodiment demonstrates that the constructed MEWS-DAVROS intelligent predictive nomogram model can assist in assessing the risk of poor prognosis in elderly patients with multiple chronic diseases and acute and severe illnesses, and has high discrimination ability, good net clinical benefit and good consistency.

[0090] Example 3: An Auxiliary Evaluation System Based on the MEWS-DAVROS Intelligent Prediction Novologram Model

[0091] This embodiment provides an auxiliary assessment system for the poor prognostic risk of elderly patients with multiple chronic diseases and acute and severe illnesses. The auxiliary assessment system includes a data input module, an analysis module, and a result output module.

[0092] 1. Data Input Module

[0093] The data input module is used to acquire clinical indicator data of the patients to be evaluated. The clinical indicator data includes APACHE II score, total bilirubin level, MEWS score, DAVROS score, peripheral blood oxygen saturation SpO2 level, and GCS score.

[0094] The data input module can be a manual input interface, an electronic medical record data interface, an examination and testing system data interface, or a data table import interface. Through the data input module, medical personnel or the system can input the above-mentioned clinical indicators of the patient to be evaluated into the auxiliary assessment system.

[0095] 2. Analysis Module

[0096] The analysis module is used to input APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level and GCS score into the adverse prognosis risk prediction model, calculate the adverse prognosis risk score of the patient to be evaluated, and determine the corresponding adverse prognosis probability and risk stratification results based on the adverse prognosis risk score.

[0097] In one implementation, the adverse prognostic risk prediction model includes a multivariate logistic regression model and a nomogram model derived from the multivariate logistic regression model.

[0098] In one implementation, the nomogram model is used to convert the APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score into corresponding individual scores, and then sums the individual scores to obtain the total nomogram score, which is used to characterize the risk of poor prognosis for the patient to be evaluated.

[0099] In one implementation, the analysis module uses 126 points as the risk stratification cutoff value. When the total score of the nomogram is greater than or equal to 126 points, the patient to be evaluated is classified into the high-risk group with poor prognosis; when the total score of the nomogram is less than 126 points, the patient to be evaluated is classified into the low-risk group with poor prognosis.

[0100] 3. Result Output Module

[0101] The results output module is used to output the patient's poor prognostic risk score, probability of poor prognosis, and risk stratification results.

[0102] In one implementation, the output module outputs the following: a patient's adverse prognostic risk score, the probability of adverse prognosis, and risk stratification results.

[0103] In one implementation, the results output module outputs the results in the form of text, tables, graphical interfaces, or reports.

[0104] The auxiliary assessment system in this embodiment is used to assist medical staff in assessing the risk of poor prognosis for elderly patients with multiple chronic diseases and acute and severe illnesses, and does not directly replace doctors in making disease diagnosis or treatment decisions.

[0105] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention, enabling those skilled in the art to understand and apply it. However, it should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, several simple deductions or substitutions can be made without departing from the inventive concept, without requiring creative effort. Therefore, any simple improvements made to the present invention by those skilled in the art based on the disclosure of this invention should be within the scope of protection of this invention.

Claims

1. A system for auxiliary assessment of poor prognostic risk in elderly patients with multiple chronic diseases and acute and severe illnesses, characterized in that, include: The data input module is used to acquire clinical indicator data of the patients to be evaluated. The clinical indicator data includes APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score. The analysis module is used to input APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level and GCS score into a preset adverse prognostic risk prediction model, calculate the adverse prognostic risk score of the patient to be evaluated, and determine the corresponding adverse prognostic probability and risk stratification results based on the adverse prognostic risk score. The results output module is used to output the patient's poor prognosis risk score, probability of poor prognosis, and risk stratification results. The pre-defined adverse prognostic risk prediction model is a nomogram model, which converts the APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score into corresponding individual scores, and adds up the individual scores to obtain the total score of the nomogram model. The total score of the nomogram model is used to represent the adverse prognostic risk score of the patient to be evaluated.

2. The auxiliary evaluation system according to claim 1, characterized in that, The risk stratification results determined based on the adverse prognostic risk score include: Compare the adverse prognostic risk score with the preset risk cutoff value; When the adverse prognostic risk score is greater than or equal to the preset risk cutoff value, the patient to be evaluated is identified as high-risk for adverse prognosis; when the adverse prognostic risk score is less than the preset risk cutoff value, the patient to be evaluated is identified as low-risk for adverse prognosis.

3. The auxiliary evaluation system according to claim 2, characterized in that, The preset risk cutoff value is 126 points.

4. The auxiliary evaluation system according to claim 1, characterized in that, The nomogram model for predicting adverse prognostic risk is constructed using the following method: To obtain clinical sample data and follow-up prognostic results of elderly patients with multiple chronic diseases and acute and critical illnesses; Univariate analysis was used to screen candidate variables from clinical sample data, and follow-up prognosis was used as the dependent variable, with death assigned a value of 1 and survival assigned a value of 0. Candidate variables were incorporated into the LASSO regression model, and the penalty parameter λ was determined through 10-fold cross-validation. Candidate predictive variables were then selected. Multicollinearity diagnosis was performed on the candidate predictor variables. When the variance inflation factor of each candidate predictor variable was less than 10, the candidate predictor variables were incorporated into a multivariate logistic regression model to construct a nomogram model for predicting adverse prognostic risk.

5. The auxiliary evaluation system according to claim 4, characterized in that, Candidate variables included GCS score, respiratory rate, heart rate, SpO2 level, body temperature, hemoglobin level, white blood cell count, platelet count, serum sodium level, serum potassium level, urea level, creatinine level, blood glucose level, total bilirubin level, mechanical ventilation time, total length of hospital stay, APACHE II score, MEWS score, and DAVROS score.

6. The auxiliary evaluation system according to claim 5, characterized in that, Candidate predictors include APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score.

7. The auxiliary evaluation system according to claim 6, characterized in that, The prognostic adverse risk prediction nomogram model includes 9 rows of scales arranged from top to bottom: The first line is a scoring scale, with a score range of 0 to 100, used to determine the individual score for each clinical indicator. The second row is the APACHE II scoring scale, with the index value ranging from 7 to 15. As the APACHE II score increases, the corresponding individual item score also increases. The third row is a scale for total bilirubin levels, in μmol / L, with values ​​ranging from 9 to 21. The individual scores increase as total bilirubin levels rise. The fourth row is the MEWS scoring scale, with the index value ranging from 1 to 4. As the MEWS score increases, the corresponding individual score also increases. The fifth row is the DAVROS rating scale, with the index value ranging from 6 to 21. As the DAVROS score increases, the corresponding individual item score also increases. The sixth line is a SpO2 level scale, with the unit being %, and the index value ranges from 96 to 100. The individual score increases when the SpO2 level decreases. The seventh row is the GCS scoring scale, with the index value ranging from 7 to 15. When the GCS score decreases, the corresponding individual item score increases. The eighth line is the total score scale, with a score range of 0 to 204, used to display the sum of the individual scores for the APACHE II score, total bilirubin level, MEWS score, DAVROS score, SpO2 level, and GCS score; The ninth line is a prognostic risk scale with a risk probability range of 0.10 to 0.95, used to determine the probability of adverse prognosis for the patient to be evaluated based on the total score.

Images