Intraoperative blood pressure management method based on individual risk stratification of patient

Abstract

The invention provides an intraoperative blood pressure management method based on individual risk stratification of patients, and belongs to the technical field of intraoperative blood pressure management. The continuous variables are subjected to five-quantile binning through evidence weight, original classification of the classification variables is reserved, and a preoperative variable data set after binning is obtained; screening variables, and performing minimum absolute value compression and selection operator regression dimensionality reduction to obtain a key preoperative variable set; establishing a perioperative complication score card model to divide the patients into low, medium and high risks; defining initial set threshold values of systolic pressure and average arterial pressure; calculating the correlation between the blood pressure threshold duration time proportion and perioperative complications, and selecting recommended blood pressure thresholds of each risk level; calculating a recommended duration threshold value corresponding to each blood pressure threshold value by adopting a restrictive cubic spline; and combining the recommended blood pressure thresholds and the recommended duration thresholds of different risk levels to form an intraoperative individualized blood pressure management strategy. The invention provides a management strategy for precise blood pressure regulation and control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intraoperative blood pressure management, and in particular to an intraoperative blood pressure management method based on patient individual risk stratification. BACKGROUND

[0002] It is estimated that the number of surgeries worldwide has exceeded 300 million per year, and this number is still increasing year by year. Hypotension is one of the most common risks for patients during the perioperative period. Studies have shown that up to 87% of patients experience one or more intraoperative hypotension events, and about 20% of surgical patients experience at least one postoperative hypotension. In addition, more and more evidence shows that hypotension is closely related to the incidence of perioperative organ damage (such as myocardial damage and acute kidney injury) and high perioperative mortality. Among them, postoperative acute kidney injury (PO-AKI) is the most common. Therefore, recent medical consensus emphasizes that it is crucial to shorten the duration and degree of perioperative hypotension in patients to improve patient outcomes.

[0003] Currently, optimizing intraoperative blood pressure management has become a new research hotspot. However, there is still some controversy about how to maintain intraoperative blood pressure within what range to effectively protect organ function. This field urgently needs more research to verify the effects of different blood pressure levels on organ function and the best blood pressure management strategies in specific patient populations.

[0004] Current intraoperative blood pressure management strategies lack precise solutions for different patient individuals: Risk assessment is blank: Traditional methods do not stratify the risk of perioperative complications (such as postoperative acute kidney injury PO-AKI) in patients, making it difficult to identify high-risk groups and implement differentiated interventions.

[0005] Blood pressure threshold is ambiguous: Existing guidelines do not clearly define the safe range of intraoperative blood pressure and the upper limit of duration for different risk patients, and there is a lack of quantitative basis in clinical practice, which may lead to over-treatment or insufficient organ protection.

[0006] Model lack of interpretability: Existing risk prediction models are mostly black box models, making it difficult for clinicians to understand the basis for decision-making, affecting the acceptance of application. SUMMARY

[0007] The present application provides an intraoperative blood pressure management method based on patient individual risk stratification to provide a precise blood pressure regulation management strategy.

[0008] To achieve the above-mentioned purpose, the present application adopts the following technical solutions: An intraoperative blood pressure management method based on patient individual risk stratification, comprising: S1. Collecting clinical data and intraoperative blood pressure data of patients undergoing general anesthesia non-cardiac surgery, establishing a disease database containing preoperative variables and intraoperative blood pressure monitoring values; S2. Five quantile bins are divided for continuous variables in the preoperative variables using evidence weight, and the original classification is retained for categorical variables, to obtain the preoperative variable data set after binning; S3. Based on the preoperative variable data set after binning, the information value of each variable is calculated, and the variables with information value greater than 0.02 are selected, and then the least absolute value compression and selector regression dimension reduction are used to obtain the key preoperative variable set; S4. A perioperative complication score card model is established according to the key preoperative variable set, and the risk score of the patient is calculated, The risk score corresponding to the specificity of 90% of the perioperative complication score card model is taken as the critical value between low-risk patients and medium-risk patients; the risk score corresponding to the sensitivity of 90% of the perioperative complication score card model is taken as the critical value between medium-risk patients and high-risk patients; S5. Defining the initial setting threshold of systolic blood pressure and mean arterial pressure with 5mmHg as a ladder, and calculating the blood pressure duration ratio under each initial setting threshold; S6. Single factor Logistic regression is used for patients in different risk levels respectively, the association between the initial setting blood pressure threshold duration ratio and perioperative complications is calculated, which is recorded as p value; the maximum initial setting blood pressure threshold with p<0.05 is selected as the recommended blood pressure threshold of each risk level; S7. Based on the blood pressure duration ratio, the recommended duration threshold corresponding to each initial setting blood pressure threshold is calculated by using the restricted cubic spline, and the intersection of the lower limit of the odds ratio confidence interval and the odds ratio=1 is taken as the recommended duration threshold; S8. Combining the recommended blood pressure threshold and the recommended duration threshold of different risk levels, the intraoperative individualized blood pressure management strategy is formed.

[0009] In this specification, the formula for five quantile bins using evidence weight is: ; In the formula, WOE i WOEi represents the WOE value corresponding to the ith bin, Good i Pi represents the good sample ratio of the ith bin, Good T P represents the good sample ratio in the total sample, Bad i Ni represents the bad sample ratio of the ith bin, Bad T N represents the bad sample ratio in the total sample.

[0010] In the specification, the calculation formula of information value is: ; ; In the formula, IV i represents the IV value corresponding to the i-th bin, i represents the i-th bin, IV represents the information value of the feature, n represents the total number of bins of the feature.

[0011] In the specification, the minimum absolute value compression and selection operator regression dimension reduction are performed, and 10-fold cross-validation evaluation is performed. The evaluation criteria include-2 times of natural logarithm of likelihood function value and area under curve. The λ corresponding to the highest AUC value in the range of 1 standard deviation is selected to obtain the reduced variables; wherein AUC is an index for evaluating the effect of the model; the formula is as follows: ; ; In the formula, argmin represents the minimum value solution of the objective function, n represents the total sample size, i represents the i-th sample, L represents the likelihood estimate of the sample point, y i represents the i-th sample outcome, x i represents the i-th feature variable, w represents the feature coefficient vector, f x i ; w represents the outcome index estimate value of the model, and λ represents the regularization coefficient, e represents the exponential function with natural constant e as the base, L y i , f x i ; w represents the likelihood estimate function of the sample point.

[0012] In the specification, the perioperative complication scoring card model is: ; Wherein, is the score; A is the score card model benchmark score coefficient, and B is the benchmark score coefficient; is the coefficient of the variable in the model, is the WOE value corresponding to the feature variable bin,​​​ is a characteristic variable, is a constant, the relationship between the key variables and the probability of complications is fitted by Logistic regression.

[0013] In the specification, the normality of the data in the database is evaluated by the Kolmogorov-Smirnov test, and the data with normal distribution is represented by mean ± standard deviation, and the data with non-normal distribution is represented by median and interquartile range.

[0014] In the specification, the initial setting of the systolic pressure threshold value includes less than 100 mmHg, 95 mmHg, …, 70 mmHg, the initial setting of the mean arterial pressure threshold value includes less than 80 mmHg, 75 mmHg, …, 55 mmHg, and the duration ratio is the ratio of the duration of blood pressure below the threshold value to the total duration of the operation.

[0015] In the specification, the calculation steps of the restricted cubic spline include: Converting the blood pressure duration ratio to a restricted cubic spline basis function; Fitting the spline curve to the odds ratio of the risk of complications; Taking the intersection of the lower limit of the confidence interval of the odds ratio and the odds ratio = 1 as the recommended duration threshold.

[0016] In the specification, the clinical data includes age, gender, ASA classification, history of diabetes, history of ischemic heart disease, serum cystatin C, creatinine, estimated glomerular filtration rate, operation site and type, and the intraoperative blood pressure data includes real-time monitoring values and duration of systolic pressure and mean arterial pressure.

[0017] In the specification, the current duration of the blood pressure threshold value is collected, the duration threshold value based on the blood pressure threshold value is calculated, the remaining duration corresponding to the systolic pressure threshold value and the mean arterial pressure threshold value of each risk level is calculated respectively, and the blood pressure threshold value, the duration threshold value, the duration ratio, the current duration and the remaining duration of the systolic pressure and the mean arterial pressure of different risk levels are combined to form an individualized intraoperative blood pressure management strategy.

[0018] In summary, the present application has at least the following beneficial effects: The present application fills the gap of individualized intraoperative blood pressure management through the complete chain of "risk stratification-threshold setting-system implementation", which not only solves the problem of traditional methods, but also enhances the clinical trust through an interpretable model, provides a precise blood pressure regulation management strategy, and provides an innovative solution for perioperative precision medicine. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 This is a schematic diagram of the intraoperative blood pressure management method based on individual patient risk stratification involved in the present invention.

[0021] Figure 2 This is a relationship diagram between the LASSO regression partial likelihood function deviation and Log (λ) of the PO-AKI score card model involved in the present invention.

[0022] Figure 3 Graph showing the relationship between the area under the LASSO curve and Log(λ) for the PO-AKI scoring card model involved in the present invention.

[0023] Figure 4 This is a receiver operating characteristic curve diagram of the PO-AKI score card model involved in the present invention.

[0024] Figure 5 Schematic diagram of the PO-AKI scoring card involved in the present invention; in the figure: Elective / emergency, 1. Elective surgery, 2. Emergency surgery; Gender, 1: male, 2: female; Diabetes mellitus, 0: no history of diabetes, 1: non-insulin-dependent, 2: insulin-dependent; Exercise equivalent, 1: >6 metabolic equivalents, 2: 3-6 metabolic equivalents, 3: <3 metabolic equivalents; Hypertension, 0: no history of hypertension, 1: grade I, 2: grade II, 3: grade III; History of ischemic heart disease: 0: no chest pain or tightness after activity in the past 3 months; 1: history of non-myocardial infarction chest pain in the past 3 months; 2: acute myocardial infarction > 3 months; 3: acute myocardial infarction < 3 months; General condition, 1: good, 2: average, 3: poor; Consciousness state, 1: awake, 2: drowsy, 3: lethargic, 4: light coma, 5: deep coma; Surgical site: 1: limb orthopedic surgery, 2: spinal surgery, 3: body surface surgery, 4: cranial surgery, 5: head, face and neck surgery, 6: joint orthopedic surgery, 7: lower abdominal surgery, 8: thoracic surgery, 9: major vascular surgery, 10: upper abdominal surgery, 11: combined thoracoabdominal surgery, 12: total abdominal surgery; American Society of Anesthesiologists (ASA) classification, 1: grade I, 2: grade II, 3: grade III, 4: grade IV, 5: grade V.

[0025] Figure 6 Schematic diagram of the spline curve of the duration of systolic blood pressure <80 mmHg and postoperative acute kidney injury for the low-risk patients with postoperative acute kidney injury involved in the present application.

[0026] Figure 7 Schematic diagram of the intraoperative blood pressure threshold and duration threshold leading to an increased risk of postoperative acute kidney injury of patients involved in the present application.

[0027] Figure 8 Schematic diagram of the intraoperative individualized blood pressure management strategy for patients classified as high-risk for postoperative PO-AKI involved in the present application. DETAILED DESCRIPTION

[0028] In the following, only certain exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the embodiments of the present application. Therefore, the drawings and the description are considered to be exemplary in nature rather than limiting.

[0029] The following disclosure provides many different embodiments, or examples, for implementing different structures of the embodiments of the present application. For the purpose of simplifying the disclosure of the embodiments of the present application, the components and settings of specific examples are described in the following. Of course, they are only examples, and the purpose is not to limit the embodiments of the present application. In addition, the embodiments of the present application can repeatedly refer to numerals and / or reference letters in different examples, and such repetition is for the purpose of simplification and clarity, which itself does not indicate the relationship between the various embodiments and / or settings discussed.

[0030] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0031] As Figure 1 shown, the present embodiment provides an intraoperative blood pressure management method based on individual risk stratification of patients, comprising: S1. Collecting the clinical data and intraoperative blood pressure data of patients undergoing general anesthesia non-cardiac surgery, establishing a disease database containing preoperative variables and intraoperative blood pressure monitoring values; S2. Five quantile bins are used for the evidence weight of continuous variables in the preoperative variables, and the original classification is retained for the categorical variables, to obtain the binned preoperative variable data set; S3. Calculate the information value of each variable based on the preoperative variable dataset after binning, select the variables with information value greater than 0.02, and then reduce the dimension through the least absolute shrinkage and selection operator regression to obtain a key preoperative variable set; S4. Establish a perioperative complication scorecard model according to the key preoperative variable set, and calculate the risk score of the patient, The risk score corresponding to the specificity of 90% of the perioperative complication scorecard model is taken as the critical value between the low-risk patients and the medium-risk patients; and the risk score corresponding to the sensitivity of 90% of the perioperative complication scorecard model is taken as the critical value between the medium-risk patients and the high-risk patients; S5. Define the initial setting threshold of systolic pressure and mean arterial pressure with 5 mmHg as a ladder, and calculate the blood pressure duration ratio under each initial setting threshold; S6. For patients in different risk levels, a single factor Logistic regression is used to calculate the association between the initial setting blood pressure threshold duration ratio and the perioperative complications, and the association is denoted as p value; the maximum initial setting blood pressure threshold with p<0.05 is selected as the recommended blood pressure threshold of each risk level; S7. Based on the blood pressure duration ratio, a restricted cubic spline is used to calculate the recommended duration threshold corresponding to each initial setting blood pressure threshold, and the intersection of the lower limit of the odds ratio confidence interval and the odds ratio = 1 is taken as the recommended duration threshold; S8. Combine the recommended blood pressure threshold and the recommended duration threshold of different risk levels to form an intraoperative individualized blood pressure management strategy.

[0032] It should be noted that the blood pressure threshold is divided into two types, one is the initial setting blood pressure threshold, and the other is the recommended blood pressure threshold calculated. For example, the initial setting blood pressure threshold includes 100 mmHg, 95 mmHg, …, 75 mmHg, and 70 mmHg. Finally, the maximum initial setting blood pressure threshold with p<0.05 is calculated to be 90 mmHg, and the recommended blood pressure threshold is obtained as 90 mmHg.

[0033] In some embodiments, the formula for five quantile binning with evidence weight is: ; In the formula, WOE i WOEi represents the WOE value corresponding to the ith bin, Good i pi represents the good sample ratio of the ith bin, Good T pi represents the good sample ratio of the ith bin, Bad i ni represents the bad sample ratio of the ith bin, Bad TThe ratio of bad samples in total samples.

[0034] In some embodiments, the formula for calculating the information value of information is: ; ; In the formula, IV i The IV value corresponding to the i-th bin, i The i-th bin, IV The information value of the feature, n The total number of bins of the feature.

[0035] In some embodiments, the dimensionality reduction is performed by the least absolute value compression and the selection operator regression, and 10-fold cross-validation evaluation is performed. The evaluation criteria include -2 times of the natural logarithm of the likelihood function value and the area under the curve (AUC). The λ corresponding to the highest AUC in the range of 1 standard deviation is selected to obtain the reduced variables; wherein the AUC is an index for evaluating the effect of the model; the formula is as follows: ; ; In the formula, argmin The minimum value solution of the objective function, n The total sample size, i The i-th sample, L The likelihood estimate of the sample point, y i The i-th sample outcome, x i The i-th feature variable, w The feature coefficient vector, f ( x i ; w ) The outcome index estimate value of the model, λ represents the regularization coefficient, e The exponential function with the natural constant e as the base, L ( y i , f ( x i ; w )) The likelihood estimate function of the sample point.

[0036] In some embodiments, the perioperative complication scoring card model is: ; Wherein, A is the score card model base score coefficient, B is the base score coefficient; A is the coefficient of the variable in the model, A is the WOE value corresponding to the feature variable bin, A is the feature variable, A is a constant, and the relationship between the key variable and the probability of complications is fitted by Logistic regression.

[0037] In some embodiments, the risk score corresponding to the specificity of 90% of the perioperative complication score card model is taken as the critical value between low-risk patients and medium-risk patients; and the risk score corresponding to the sensitivity of 90% of the perioperative complication score card model is taken as the critical value between medium-risk patients and high-risk patients.

[0038] In some embodiments, the initial setting of the systolic blood pressure threshold value includes less than 100 mmHg, 95 mmHg, …, 70 mmHg, the initial setting of the mean arterial pressure threshold value includes less than 80 mmHg, 75 mmHg, …, 55 mmHg, and the duration ratio is the ratio of the duration of blood pressure below the threshold value to the total duration of the operation.

[0039] In some embodiments, the calculation step of the restrictive cubic spline includes: Converting the blood pressure duration ratio into a restrictive cubic spline base function; Fitting the ratio of the spline curve to the complication risk (Odds Ratio, OR); Taking the intersection of the lower limit of the odds ratio confidence interval and the odds ratio = 1 as the recommended duration threshold.

[0040] In some embodiments, the clinical data includes age, gender, ASA classification, history of diabetes, history of ischemic heart disease, serum cystatin C, creatinine, estimated glomerular filtration rate, operation site and type, and the intraoperative blood pressure data includes real-time monitoring values and durations of systolic blood pressure and mean arterial pressure.

[0041] In some embodiments, the normality of the data in the database is evaluated by the Kolmogorov-Smirnov test, the normally distributed data is represented by mean ± standard deviation, and the non-normally distributed data is represented by median and interquartile range. Data statistics is completed by Python and R language.

[0042] In some embodiments, the current duration of the blood pressure threshold value is collected, the remaining duration corresponding to the systolic blood pressure threshold value and the mean arterial pressure threshold value of each risk level is calculated based on the duration threshold value of the blood pressure threshold value, and the recommended blood pressure threshold value, the recommended duration threshold value, the duration ratio, the current duration and the remaining duration of the systolic blood pressure and the mean arterial pressure of different risk levels are combined to form an individualized intraoperative blood pressure management strategy (such asFigure 8 ).

[0043] In a specific embodiment, data from 34668 adult patients undergoing non-cardiac surgery under general anesthesia in a hospital were used as an example.

[0044] Postoperative acute kidney injury (PO-AKI) was used as the target perioperative complication, and the definition of PO-AKI was an increase of creatinine by ≥0.3mg / dl (≥26.5μmol / L) within 48 hours after surgery or an increase of creatinine by ≥1.5 times of preoperative creatinine within 7 days after surgery.

[0045] Clinical data and intraoperative blood pressure data of patients were collected to establish a database.

[0046] Preoperative variables were extracted from the database, including categorical variables and continuous variables, the original classification of the categorical variables was retained, and the continuous variables were divided into five quantiles using WOE.

[0047] After completing the WOE of the variables, the IV values of all variables were calculated, and a total of 104 variables with IV values greater than 0.02 were screened out.

[0048] LASSO regression was performed on the variables screened out by IV value, and after 10-fold cross-validation, the parameter adjustment relationship of LASSO regression was as shown in Figure 2 and Figure 3 The λ corresponding to the maximum AUC in the range of 1 standard deviation was selected, and the best tolerable Log(λ) value was-6.063649, and a total of 27 candidate variables were screened out after LASSO regression dimension reduction.

[0049] The 27 variables after dimension reduction included American Society of Anesthesiologists classification, elective / emergency, surgical site, serum cystatin C, history of ischemic heart disease, creatinine, diastolic blood pressure, consciousness state, estimated glomerular filtration rate, general condition, age, albumin, red cell distribution width variation coefficient, exercise equivalent, glutamyl transpeptidase, systolic blood pressure, heart rate, absolute value of lymphocytes, red blood cell count, diabetes, prothrombin time, magnesium, albumin-globulin ratio, red cell distribution width standard deviation, aspartate aminotransferase, hypertension, and gender.

[0050] Based on the 27 variables screened out, a PO-AKI scorecard model was established by logistic regression method, and the model showed good prediction effect, with an average AUROC of 0.80 (95%CI: 0.78-0.82) ( Figure 4 ). For the convenience of clinical practice application, the scorecard score of all variables included in the model was calculated, and the PO-AKI scorecard was formed, seeFigure 5 .

[0051] Patients were stratified into 3 tiers according to the risk score corresponding to the model's high sensitivity (90%) and high specificity (90%): <15 points for PO-AKI low risk; 15-67 points for PO-AKI moderate risk; >67 points for PO-AKI high risk.

[0052] The number of low-risk patients was the largest, with 16586 patients, and the incidence of PO-AKI was also the lowest at 0.49%; The number of moderate-risk patients was 14380, and the incidence of PO-AKI was 2.48%, which was closest to the incidence of PO-AKI in the whole population (2.43%); The number of high-risk patients was the smallest, with 3702 patients, and the incidence of PO-AKI was also the highest at 10.99%.

[0053] The systolic pressure threshold was defined, with a 5mmHg step, the proportion of the duration of the patient's intraoperative systolic pressure below 100mmHg, 95mmHg···75mmHg and below 70mmHg.

[0054] The mean arterial pressure threshold was defined, with a 5mmHg step, the proportion of the duration of the patient's intraoperative mean arterial pressure below 80mmHg, 75mmHg···60mmHg and below 55mmHg.

[0055] The relationship between the proportion of the duration of exceeding different intraoperative blood pressure thresholds and PO-AKI was calculated by single factor Logistic regression, and this relationship was recorded as p value; The maximum value of the intraoperative blood pressure threshold with p value <0.05 in patients of different risk levels was selected as the recommended intraoperative blood pressure threshold (as shown in Table 1).

[0056]

[0057] Note: * represents p<0.05, ** represents p<0.01, and *** represents p<0.001.

[0058] The duration threshold corresponding to different intraoperative blood pressure thresholds was calculated by restricted cubic spline, and the intersection point of the lower limit of the confidence interval of the OR value and OR value = 1 was taken as the duration threshold corresponding to the intraoperative blood pressure threshold. As shown in Table 2, the duration threshold of systolic pressure <80mmHg for PO-AKI low-risk patients was 9 minutes and 51 seconds. Figure 6

[0059] Combination to form individualized intraoperative blood pressure management strategy, such as Figure 7 ​As shown, for PO-AKI low-risk patients, the systolic blood pressure threshold value is <80 mmHg, and the duration threshold value is 9 minutes 51 seconds; the mean arterial pressure threshold value is 55 mmHg, and the duration threshold value is 13 minutes 32 seconds. For PO-AKI medium-risk patients, the systolic blood pressure threshold value is <85 mmHg, and the duration threshold value is 13 minutes 6 seconds; the mean arterial pressure threshold value is 65 mmHg, and the duration threshold value is 17 minutes 44 seconds. For PO-AKI high-risk patients, the systolic blood pressure threshold value is <100 mmHg, and the duration threshold value is 49 minutes 10 seconds; the mean arterial pressure threshold value is 80 mmHg, and the duration threshold value is 107 minutes 32 seconds.

[0060] The intraoperative individualized blood pressure management strategy for patients classified as postoperative PO-AKI high-risk can also be as shown in FIG. 6. Figure 8 Figure 8 In the figure, red: represents that the duration threshold value tolerated by the patient has been exceeded, and the subsequent blood pressure below the blood pressure threshold value needs to be strictly avoided; yellow: represents that the remaining time to the duration threshold value tolerated by the patient is less than 10 minutes, and attention needs to be paid to avoid the subsequent blood pressure below the blood pressure threshold value as much as possible; blue: represents that the remaining time to the duration threshold value tolerated by the patient is more than 10 minutes, and other blood pressure threshold values with a remaining time of less than 10 minutes need to be focused on.

[0061] Using the intraoperative individualized blood pressure management strategy can improve the timeliness of intraoperative decision-making. The red warning directly prompts that the subsequent blood pressure below the threshold value needs to be strictly avoided, avoiding the delay of the doctor's judgment leading to related postoperative complications. The yellow warning helps the doctor to plan intervention measures (such as adjusting the dose of vasoactive drugs) in advance through the 10-minute countdown mechanism. The individualized blood pressure management strategy allows the doctor to make adjustments suitable for the patient according to the patient's condition, avoiding excessive intervention or insufficient intervention caused by one-size-fits-all blood pressure management. At the same time, the strategy integrates the systolic blood pressure and mean arterial pressure threshold values (such as high-risk patients needing to pay attention to SBP <100 mmHg and MAP <80 mmHg at the same time), helping the doctor to establish a multi-dimensional blood pressure management thinking and avoiding misjudgment and omission of a single parameter.

[0062] ​As shown in Tables 2 and 3, the intraoperative individualized blood pressure management strategy has a significant effect on controlling the incidence of PO-AKI. The data shows that the incidence of PO-AKI in patients who follow the strategy (not exceeding the recommended blood pressure threshold and recommended duration time threshold) in the whole population is 1.05%, while that of those who do not follow (exceeding the threshold) is 4.71%, with a risk increase of 4.49 times. In terms of stratification, the incidence of the strategy-compliant patients in the low-risk population is 0.45%, and that of the non-compliant patients is 0.91% (an increase of 2.02 times); in the medium-risk population, they are 1.97% and 2.81% (an increase of 1.43 times); and in the high-risk population, they are 6.35% and 12.19% (an increase of 1.92 times). Compared with the blood pressure threshold of <100 mmHg recommended by the current guidelines, the intraoperative individualized blood pressure management strategy has obvious advantages: the incidence of PO-AKI in the whole population not exceeding the threshold under the traditional threshold is 2.97%, and that of those exceeding the threshold is 2.42%, indicating that the one-size-fits-all blood pressure management strategy currently used in clinical practice is not suitable for all populations, and the blood pressure threshold and duration threshold need to be individually set according to the preoperative conditions of patients. The above data quantifies the clinical value of the intraoperative individualized blood pressure management strategy in reducing PO-AKI from the perspective of evidence-based medicine.

[0063]

[0064]

[0065] In summary, the present application significantly improves the scientificity and clinical practicability of perioperative blood pressure management by integrating risk stratification and individualized blood pressure management, and the specific effects are as follows: 1. Precise risk stratification optimizes resource allocation Interpretable scorecard: Based on WOE binning, IV value screening and LASSO regression, the scorecard model (such as the PO-AKI scorecard) converts 27 key variables before surgery (such as ASA classification, creatinine, and surgery type) into intuitive risk scores. Clinicians can manually calculate the scores to quickly identify low-risk (<15 points, PO-AKI incidence 0.49%), medium-risk (15-67 points, incidence 2.48%), and high-risk (>67 points, incidence 10.99%) patients, facilitating the development of individualized monitoring programs before surgery.

[0066] Stratified intervention basis: High-risk patients can be given priority for intensive blood pressure monitoring and intervention, while low-risk patients can be spared from excessive medical treatment, optimizing the allocation of medical resources.

[0067] 2. Quantitative blood pressure threshold for individualized management Stratified blood pressure threshold: Differentiated blood pressure thresholds are determined for patients of different risks through single-factor Logistic regression (such as a systolic blood pressure <80 mmHg for low-risk patients and <100 mmHg for high-risk patients), avoiding "one-size-fits-all" management.

[0068] Duration control: Restricted cubic splines are used to calculate the safety duration corresponding to each threshold (e.g., systolic blood pressure <80 mmHg for ≤9 minutes 51 seconds for low-risk patients, and <100 mmHg for ≤49 minutes 10 seconds for high-risk patients), providing clear time limits for real-time intraoperative control and reducing the risk of hypotension-related organ damage.

[0069] 3. Systematic data integration to improve management efficiency Full-process closed loop: From preoperative data collection and risk stratification to intraoperative blood pressure threshold calculation and strategy generation, a standardized process is formed to support the continuity of clinical decision-making.

[0070] Technical tool integration: Use Python and R languages ​​for data processing, modeling, and statistical analysis, combined with scorecard visualization and threshold charts to improve the convenience of doctors' operations.

[0071] 4. Significant clinical benefits and promotion of precision medicine Reduce complications: By identifying high-risk patients early and strictly controlling blood pressure fluctuations, the incidence of perioperative complications such as PO-AKI can be reduced and patient prognosis can be improved.

[0072] Improve the standardization of diagnosis and treatment: Provide a replicable risk stratification and blood pressure management framework to facilitate promotion in different medical institutions and promote the standardization and homogenization of perioperative blood pressure management.

[0073] 5. The model has reliable performance, accurate patient stratification, and supports dynamic adjustment High predictive efficacy: The PO-AKI scorecard model achieved an average AUC of 0.80 (95% CI: 0.78-0.82), ensuring accurate risk stratification.

[0074] Accurate patient stratification: The incidence of PO-AKI in patients of different risk stratifications was significantly different, including low risk (0.49%), medium risk (2.43%), and high risk (10.99%).

[0075] Dynamic adaptability: Intraoperatively, real-time blood pressure data and risk scores can be combined to update management strategies to respond to changes in patient status and improve program flexibility.

[0076] The above embodiments are intended to illustrate the present invention, not to limit the present invention. Therefore, changes in illustrative values ​​or substitutions of equivalent components should still fall within the scope of the present invention.

[0077] From the above detailed description, it will be clear to those skilled in the art that the present invention can indeed achieve the aforementioned objectives and is in compliance with the provisions of the Patent Law.

[0078] Although preferred embodiments of the application have been described herein, those skilled in the art will appreciate that various modifications and changes can be made to the described embodiments without departing from the spirit and scope of the application. Accordingly, it is intended that all claims be interpreted to include the preferred embodiments and all modifications and changes equivalent to those described above. The above description is intended to be illustrative only and is not intended to limit the scope of the application, which is set forth with the following claims. Any modification, equivalent replacement or improvement made within the spirit and principle of the application should be included in the scope of the application.

[0079] It should be noted that the above description of the flow is merely for example and illustration, and does not limit the scope of the present specification. Those skilled in the art can make various modifications and changes to the flow under the guidance of the present specification. However, these modifications and changes are still within the scope of the present specification.

[0080] The above has described the basic concept, and it is obvious that the above application disclosure is only as an example and does not constitute a limitation on the present application for those skilled in the art after reading this application. Although it is not explicitly stated here, those skilled in the art can make various modifications, improvements and modifications to the present application. Such modifications, improvements and modifications are suggested in the present application, so such modifications, improvements and modifications are still within the spirit and scope of the exemplary embodiments of the present application.

[0081] Meanwhile, specific words are used in the present application to describe the embodiments of the present application. For example, "one embodiment", "an embodiment", and / or "some embodiments" means a certain feature, structure or characteristic related to at least one embodiment of the present application. Therefore, it should be emphasized and noted that the "an embodiment" or "one embodiment" or "an alternative embodiment" mentioned in different places in the specification does not necessarily refer to the same embodiment. In addition, some features, structures or characteristics in one or more embodiments of the present application can be properly combined.

[0082] In addition, unless the claim explicitly states otherwise, the order of the processing elements and sequences described in the present application, the use of numerals and letters, or the use of other names, is not intended to limit the order of the processes and methods of the present application. Although some currently considered useful embodiments of the application are discussed in the above disclosure through various examples, it should be understood that such details are only for the purpose of illustration, and the additional claims are not limited to the disclosed embodiments, on the contrary, the claims are intended to cover all modifications and equivalent combinations within the spirit and scope of the embodiments of the present application. For example, although the implementation of the above various components can be realized in a hardware device, it can also be realized as a pure software solution, for example, installation on an existing server or mobile device.

[0083] For similar reasons, it is to be appreciated that the teachings of the present application provided herein can be applied to any embodiment of the present application, and that actual details thereof can depart from those set forth in a given embodiment, or that even methods, or implementations, which do not provide all of the benefits and features set forth herein, are within the scope of the present application.

Claims

1. A method for intraoperative blood pressure management based on individual patient risk stratification, characterized in that: include: S1. Collect clinical data and intraoperative blood pressure data of patients undergoing non-cardiac surgery under general anesthesia and establish a disease database containing preoperative variables and intraoperative blood pressure monitoring values; S2. Bin the continuous variables in the preoperative variables into quintiles using weight of evidence, and retain the original classification of categorical variables to obtain the binned preoperative variable dataset; S3. Based on the binned preoperative variable dataset, calculate the information value of each variable, select variables with an information value greater than 0.02, and then perform dimensionality reduction using minimum absolute value compression and selection operator regression to obtain the key preoperative variable set. S4. Develop a perioperative complication scorecard model based on the key preoperative variable set and calculate the patient risk score. The risk score corresponding to a 90% specificity of the perioperative complication scorecard model is used as the cutoff between low-risk and moderate-risk patients. The risk score corresponding to a 90% sensitivity of the perioperative complication scorecard model is used as the cutoff between moderate-risk and high-risk patients. S5. Define initial thresholds for systolic and mean arterial pressure using 5 mmHg steps, and calculate the proportion of blood pressure duration at each initial threshold. S6. Use univariate logistic regression to calculate the association between the proportion of duration of the initial blood pressure threshold and perioperative complications in patients across different risk strata. This association is recorded as the p-value. The maximum initial blood pressure threshold with a p < 0.05 was selected as the recommended blood pressure threshold for each risk stratum. S7. Based on the blood pressure duration ratio, use restricted cubic splines to calculate the recommended duration threshold corresponding to each initially set blood pressure threshold, and take the intersection of the lower limit of the odds ratio confidence interval and the odds ratio = 1 as the recommended duration threshold; S8. Combine the recommended blood pressure thresholds and recommended duration thresholds for different risk levels to form an individualized blood pressure management strategy during surgery.

2. The intraoperative blood pressure management method based on individual patient risk stratification according to claim 1, characterized in that: The formula for quintile binning using weight of evidence is: ； Where, WOE i Indicates the WOE value corresponding to the i-th bin, Good i represents the proportion of good samples in the i-th bin, Good T represents the proportion of good samples in the total sample, Bad i Indicates the proportion of bad samples in the i-th bin, Bad T Indicates the proportion of bad samples in the total sample.

3. The intraoperative blood pressure management method based on individual patient risk stratification according to claim 2, characterized in that: The formula for calculating information value is: ； ； Where, IV i Indicates the IV value corresponding to the i-th bin, i represents the i-th bin, IV Indicates the information value of the feature, n Indicates the total number of bins for this feature.

4. The intraoperative blood pressure management method based on individual patient risk stratification according to claim 3, characterized in that: We used minimum absolute value compression and operator selection regression to reduce dimensionality, and performed a 10-fold cross-validation evaluation. The evaluation criteria included -2 ​​times the natural logarithm of the likelihood function value and the area under the curve. We selected the lambda corresponding to the highest AUC value within the range of 1 standard deviation to obtain the reduced-dimensional variable. The AUC is an indicator for evaluating the model's effectiveness; the formula is as follows: ； ； Where, argmin Indicates the minimum solution of the objective function. n represents the overall sample size, i represents the i-th sample, L represents the likelihood estimate of the sample point, y i represents the outcome of the i-th sample, x i represents the i-th characteristic variable, w represents the characteristic coefficient vector, f ( x i ; w ) represents the estimated value of the model outcome indicator, λ represents the regularization coefficient, e represents the exponential function with the natural constant e as the base, L ( y i , f ( x i ; w )) represents the likelihood estimation function of the sample point.

5. The intraoperative blood pressure management method based on individual patient risk stratification according to claim 4, characterized in that: The perioperative complication scorecard model is: ； in, is the scoring score; A is the benchmark score coefficient of the scoring card model, and B is the benchmark score coefficient; are the coefficients of the variables in the model, is the WOE value corresponding to the feature variable binning, is the characteristic variable, The relationship between key variables and the probability of complications was fitted by Logistic regression.

6. The intraoperative blood pressure management method based on individual patient risk stratification according to claim 1, characterized in that: The Kolmogorov-Smirnov test was used to assess the normality of the data in the database. Normally distributed data were expressed as mean ± standard deviation, and non-normally distributed data were expressed as median and interquartile range.

7. The intraoperative blood pressure management method based on individual patient risk stratification according to claim 1, characterized in that: The initial set systolic blood pressure thresholds include below 100 mmHg, 95 mmHg, ..., 70 mmHg, and the initial set mean arterial pressure thresholds include below 80 mmHg, 75 mmHg, ..., 55 mmHg. The duration ratio is the ratio of the time the blood pressure is below the threshold to the total duration of the operation.

8. The intraoperative blood pressure management method based on individual patient risk stratification according to claim 1, characterized in that: The computational steps for restricted cubic splines include: The blood pressure duration ratio was converted to a restricted cubic spline basis function; Fitted spline curves were used to calculate odds ratios for the risk of complications; The intersection of the lower limit of the confidence interval of the odds ratio and the odds ratio = 1 was used as the recommended duration threshold.

9. The intraoperative blood pressure management method based on individual patient risk stratification according to claim 1, characterized in that: The clinical data include age, gender, ASA classification, history of diabetes, history of ischemic heart disease, serum cystatin C, creatinine, estimated glomerular filtration rate, surgical site and type; the intraoperative blood pressure data include real-time monitoring values ​​and duration of systolic blood pressure and mean arterial pressure.

10. The intraoperative blood pressure management method based on individual patient risk stratification according to claim 1, characterized in that: The current duration of the blood pressure threshold is collected, and based on the duration threshold of the blood pressure threshold, the remaining duration corresponding to the systolic blood pressure threshold and mean arterial pressure threshold of each risk level is calculated respectively. The recommended blood pressure threshold, recommended duration threshold, duration ratio, current duration and remaining duration of systolic blood pressure and mean arterial pressure of different risk levels are combined to form an individualized blood pressure management strategy during surgery.

Images