Method and system for predicting occurrence rate of allogeneic blood transfusion in joint replacement perioperative period of elderly patient

Abstract

The invention discloses a method and system for predicting the occurrence rate of allogeneic blood transfusion in the joint replacement perioperative period of an elderly patient, and belongs to the technical field of medical data processing, and the method comprises the steps: selecting potential factors from the historical clinical information of the patient; related factors are obtained by analyzing the influence degree of the potential factors on blood transfusion indications in the perioperative period of joint replacement; performing control variable analysis on the potential factors to screen out independent risk factors; in combination with related factors and independent risk factors, allogeneic blood transfusion evaluation indexes are obtained; and calculating an allogenic blood transfusion incidence prediction value of the target patient according to the allogenic blood transfusion evaluation index through a preset allogenic blood transfusion incidence prediction model. Therefore, by implementing the method and the device, the problem that in the prior art, other factors possibly causing the allogenic transfusion risk of the elderly patient are not considered, so that the calculated allogenic transfusion occurrence rate is inaccurate or the prediction time is too late, and reasonable blood use planning for the perioperative period of the patient cannot be realized can be solved.

Description

Technical Field

[0001] The present application belongs to the field of medical data processing technology, and specifically relates to a method and system for predicting the incidence of allogeneic blood transfusion during the perioperative period of joint replacement in elderly patients. Background Art

[0002] Total knee and total hip replacements often result in significant blood loss, accompanied by a decrease in hemoglobin, necessitating allogeneic blood transfusions in severe cases. Due to the limited availability of allogeneic blood resources and the risks of blood-borne diseases, infection, allergic reactions, fluid overload, prolonged hospitalization, and increased mortality, accurate prediction of the perioperative allogeneic transfusion probability could allow for early intervention in high-risk patients, significantly impacting the rational planning of blood use.

[0003] Preoperative anemia, intraoperative blood loss, and postoperative bleeding are the primary factors contributing to the high transfusion rate during joint replacement surgery. Therefore, traditional allogeneic blood transfusion rate prediction models typically estimate a patient's risk of perioperative allogeneic blood transfusion based solely on these three factors. However, for elderly patients, factors that induce massive blood loss during joint replacement surgery are more complex due to their advanced age and multiple comorbidities. Therefore, using only these three indicators to predict allogeneic blood transfusion rates in elderly patients is inaccurate and irrational, and cannot assist physicians in determining whether elderly patients require allogeneic blood transfusion during the perioperative period. Summary of the Invention

[0004] This application proposes a method and system for predicting the incidence of allogeneic blood transfusion during the perioperative period of joint replacement in elderly patients. It can solve the problem that the existing technology does not take into account other factors that may cause the risk of allogeneic blood transfusion in elderly patients, resulting in inaccurate calculated allogeneic blood transfusion incidence or too late prediction time, thereby failing to achieve reasonable planning of blood use for patients during the perioperative period.

[0005] The first aspect of the present application provides a method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients, the method comprising:

[0006] Several potential factors were selected from the patient's historical clinical information;

[0007] The influence of the potential factors on perioperative blood transfusion indications during joint replacement surgery was analyzed by univariate regression model, and several related factors were obtained.

[0008] The potential factors were analyzed by controlling variables through multivariate regression model to screen out several independent risk factors;

[0009] Combining the relevant factors and the independent risk factors, several allogeneic blood transfusion assessment indicators were obtained; wherein the allogeneic blood transfusion assessment indicators included the average pain score in the second 24 hours after surgery, patient age, comorbidity index, operation time, intraoperative blood loss, preoperative hemoglobin, and preoperative platelet count;

[0010] The predicted value of the allogeneic blood transfusion incidence of the target patient is calculated based on the allogeneic blood transfusion evaluation index using a preset allogeneic blood transfusion incidence prediction model.

[0011] The proposed approach first selects several indicators from historical clinical data as potential factors that may trigger a patient's perioperative allogeneic blood transfusion requirement. These potential factors are then analyzed using both univariate and multivariate regression. Univariate regression primarily analyzes the impact of a single potential factor on perioperative transfusion indications, while multivariate regression considers the combined effects of multiple potential factors. This approach is suitable for interpretation and prediction in complex scenarios, resulting in allogeneic transfusion assessment indicators that are closely correlated with perioperative transfusion indications for arthroplasty. Finally, a pre-defined allogeneic transfusion incidence prediction model is used to calculate each allogeneic transfusion assessment indicator in combination with the target patient's clinical data to derive a predicted allogeneic transfusion incidence for the target patient. Because the model also considers factors such as patient age and preexisting medical conditions in predicting allogeneic transfusion incidence, rather than solely focusing on preoperative anemia, intraoperative blood loss, and postoperative bleeding, it significantly improves the accuracy of predictions compared to traditional allogeneic transfusion incidence prediction models, providing data support for rationally planning blood use and accelerating postoperative recovery.

[0012] In a possible implementation method of the first aspect, several potential factors are selected from the patient's historical clinical information, specifically:

[0013] Obtain patients' historical clinical information from medical records;

[0014] The historical clinical information is screened based on preset factor screening criteria to obtain the potential factors; wherein the factor screening criteria include non-emergency surgery history and completeness of patient information.

[0015] The above scheme takes into account that the age and recent surgical history of different patients may induce massive blood loss during the perioperative period of joint replacement. Therefore, through the set factor screening criteria, potential factors that may be related to the perioperative blood transfusion indications of joint replacement are first selected from historical clinical information to provide data support for further analysis.

[0016] In a possible implementation method of the first aspect, the influence of the potential factors on the perioperative blood transfusion indications of joint replacement surgery is analyzed by a univariate regression model to obtain several related factors, specifically:

[0017] Inputting the historical clinical information related to the potential factors into the univariate regression model, taking each potential factor as an independent variable and the perioperative blood transfusion indication for joint replacement as a dependent variable, fitting a straight line or a curve to describe the relationship between each independent variable and the dependent variable, and obtaining the degree of influence of each independent variable on the dependent variable;

[0018] The independent variable whose influence degree is greater than the first threshold is taken as the relevant factor.

[0019] The above scheme uses univariate regression analysis, with potential factors as independent variables and perioperative blood transfusion indications for joint replacement as dependent variables. By analyzing the impact of a single independent variable on the dependent variable, the possibility of a single potential factor inducing the patient to require allogeneic blood transfusion during the perioperative period is explained, thereby screening potential factors and obtaining relevant factors that can be used to more accurately predict the incidence of allogeneic blood transfusion.

[0020] In a possible implementation method of the first aspect, the relevant factors include: patient age, comorbidity index, operation time, intraoperative blood loss, preoperative hemoglobin, and preoperative platelet count;

[0021] The comorbidity index is a total assessment score obtained by quantitatively evaluating the patient's various existing diseases, and is used to predict the patient's overall health status.

[0022] The above scheme takes into account the impact of other diseases of the patient on perioperative bleeding volume, and proposes a comorbidity index to predict the patient's overall health status, providing data support for accurately predicting the incidence of allogeneic blood transfusion.

[0023] In a possible implementation method of the first aspect, a multivariate regression model is used to perform control variable analysis on the potential factors to screen out several independent risk factors, specifically:

[0024] Inputting the historical clinical information related to the potential factors into the multivariate regression model, with each potential factor as an independent variable and the perioperative blood transfusion indication during joint replacement surgery as a dependent variable;

[0025] By constructing a functional relationship between two or more independent variables and a dependent variable and performing a control variable analysis on the functional relationship, the independent risk factors are screened out from the potential factors.

[0026] The above scheme uses multivariate analysis to consider the combined effects of multiple potential factors on the perioperative transfusion indications of joint replacement surgery, which can more accurately predict the allogeneic blood transfusion rate of patients in complex situations.

[0027] In a possible implementation method of the first aspect, the independent risk factors include: average pain score in the second 24 hours after surgery, operation duration, intraoperative blood loss, preoperative hemoglobin, and preoperative platelet count;

[0028] The average pain score in the second 24 hours after surgery is the average pain score on the second day after the surgery, and the pain score is obtained from the patient's medical records.

[0029] The above scheme introduces the average pain score in the second 24 hours after surgery as one of the independent risk factors. By describing the patient's postoperative performance, it can accurately judge whether the patient is likely to bleed heavily after surgery, providing data support for accelerating patient recovery and reducing postoperative bleeding and oozing.

[0030] In a possible implementation method of the first aspect, the relevant factors and the independent risk factors are combined to obtain several allogeneic blood transfusion assessment indicators, specifically:

[0031] Taking the union of the relevant factors and the independent risk factors to obtain a union result;

[0032] Based on a preset coordinate system, the union results are integrated into the same plane to draw a nomogram, and the allogeneic blood transfusion evaluation index is obtained according to the nomogram.

[0033] The above scheme simplifies the complex model prediction results into an intuitive graphical display by drawing a nomogram, making the obtained allogeneic blood transfusion assessment indicators highly interpretable and helping users quickly calculate the incidence of allogeneic blood transfusion.

[0034] In a possible implementation method of the first aspect, a predicted value of the allogeneic blood transfusion incidence of a target patient is calculated based on the allogeneic blood transfusion assessment index using a preset allogeneic blood transfusion incidence prediction model, specifically:

[0035] Inputting the clinical information of the target patient into the allogeneic blood transfusion incidence prediction model, and calculating the index score corresponding to each of the allogeneic blood transfusion assessment indicators using a preset calculation formula;

[0036] The sum of all the index scores is calculated to obtain the predicted incidence of allogeneic blood transfusion in the target patient.

[0037] The above scheme selects relevant data from the clinical information of the target patient based on the allogeneic blood transfusion evaluation indicators, and then performs weighted calculation on these data through the allogeneic blood transfusion incidence prediction model to comprehensively derive the probability of allogeneic blood transfusion for the target patient during the perioperative period, providing data support for the rational planning of blood use for the target patient.

[0038] In a possible implementation method of the first aspect, the method further includes:

[0039] If the predicted value of the incidence of allogeneic blood transfusion is greater than or equal to a second threshold, an early warning signal for preparing a blood source is issued;

[0040] Based on the warning signals, the incidence of allogeneic blood transfusion in target patients can be reduced by improving anemia before surgery to increase hemoglobin levels, performing autologous blood transfusion during surgery, and using perioperative multimodal analgesia.

[0041] When the above scheme receives the early warning signal, the scores of allogeneic blood transfusion assessment indicators such as postoperative pain scores can be reduced through analgesic means to accelerate the patient's recovery, thereby reducing the probability of allogeneic blood transfusion.

[0042] A second aspect of the present application provides a system for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients, the system comprising: a data selection module, a univariate analysis module, a multivariate analysis module, an evaluation index acquisition module, and a prediction value calculation module;

[0043] Among them, the data selection module is used to select several potential factors from the patient's historical clinical information;

[0044] The univariate analysis module is used to analyze the influence of the potential factors on the perioperative blood transfusion indications of joint replacement through a univariate regression model, and obtain several related factors;

[0045] The multivariate analysis module is used to perform control variable analysis on the potential factors through a multivariate regression model to screen out several independent risk factors;

[0046] The evaluation index acquisition module is used to combine the relevant factors and the independent risk factors to obtain a number of allogeneic blood transfusion evaluation indicators; wherein the allogeneic blood transfusion evaluation indicators include the average pain score in the second 24 hours after surgery, patient age, comorbidity index, operation time, intraoperative blood loss, preoperative hemoglobin and preoperative platelet count;

[0047] The predicted value calculation module is used to calculate the predicted value of the allogeneic blood transfusion incidence of the target patient according to the allogeneic blood transfusion evaluation index through a preset allogeneic blood transfusion incidence prediction model. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solution of the present application, the following is a brief introduction to the drawings required for use in the implementation. Obviously, the drawings described below are only some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0049] Figure 1 This is a schematic diagram of a specific process of a method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients provided by a certain embodiment of the present application;

[0050] Figure 2 This is the result of a univariate regression analysis of a method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients provided in a certain embodiment of the present application;

[0051] Figure 3 This is the result of a multivariate regression analysis of a method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients provided in a certain embodiment of the present application;

[0052] Figure 4 A nomogram of allogeneic blood transfusion assessment indicators for a method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients provided in a certain embodiment of the present application;

[0053] Figure 5 This is a comparison chart of ROC curves for a method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients, provided in one embodiment of the present application;

[0054] Figure 6 This is a flowchart of an allogeneic blood transfusion incidence prediction method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients provided in a certain embodiment of the present application;

[0055] Figure 7 This is a structural diagram of a system for predicting the incidence of perioperative allogeneic blood transfusion in elderly patients with joint replacement, provided in one embodiment of the present application. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0057] It should be understood that the step numbers used herein are only for convenience of description and are not intended to limit the order in which the steps are to be executed.

[0058] First embodiment

[0059] The perioperative period refers to a complete medical stage surrounding the entire surgical process, which includes three key periods: preoperative preparation, intraoperative operation, and postoperative recovery. During the perioperative period, it is necessary to monitor, intervene, and adjust the patient's health status throughout the process to accelerate the patient's recovery process and reduce the length of postoperative hospitalization. Joint replacement surgery usually causes massive bleeding, and the possibility of allogeneic blood transfusion is greatly increased. Due to the shortage of allogeneic blood resources and the high risk of allogeneic blood transfusion, if the probability of allogeneic blood transfusion in the perioperative period can be accurately predicted, early intervention can be performed on high-risk patients to reduce the possibility of allogeneic blood transfusion. The existing technology for predicting the perioperative allogeneic blood transfusion rate is mainly based on preoperative anemia, intraoperative blood loss, and postoperative bleeding. For elderly patients, factors such as age, other existing diseases, and recent use of anticoagulants may also lead to massive bleeding. Therefore, how to more accurately predict whether elderly patients need allogeneic blood transfusion in the perioperative period is of great significance for the rational planning of patient blood use.

[0060] like Figure 1 As shown, in order to solve the problem in the prior art that other factors that may cause the risk of allogeneic blood transfusion in elderly patients are not taken into account, resulting in inaccurate calculated allogeneic blood transfusion incidence, thereby failing to achieve reasonable planning of blood use for patients during the perioperative period, the first embodiment of the present application provides a specific flow chart of a method for predicting the incidence of allogeneic blood transfusion during joint replacement in elderly patients. The method for predicting the incidence of allogeneic blood transfusion during joint replacement in elderly patients in this embodiment includes steps S1 to S5, which are detailed as follows:

[0061] Step S1: Select several potential factors from the patient's historical clinical information.

[0062] In the embodiment of the present application, a large amount of historical clinical information of elderly patients who underwent unilateral hip or knee replacement surgery during hospitalization in the past few years is first obtained according to the data selection criteria, and then the historical clinical information is screened based on the factor screening criteria to select appropriate and complete data.

[0063] The data selection criteria are set as follows:

[0064] (1) Elderly patients undergoing elective hip replacement surgery;

[0065] (2) Patients aged 65 years or older, regardless of gender, and ASA grade ≤ IV. ASA grade is a classification of surgical risk based on the patient's physical condition before anesthesia. A higher grade indicates a higher perioperative mortality rate.

[0066] (3) No recent history of surgery.

[0067] The factor screening criteria are set as follows:

[0068] (1) Not used for emergency surgery

[0069] (2) If the data integrity is not high, it will not be selected.

[0070] The historical clinical information was screened based on the preset factor screening criteria to obtain the potential factors, including patient age, BMI, nutritional score, comorbidity index, preoperative hemoglobin, preoperative platelet count, albumin, albumin to lymphocyte / leukocyte ratio, prothrombin ratio, fibrinogen content, anesthesia method, ASA grade, operation time, intraoperative blood loss, intraoperative urine volume, average pain score in the second 24 hours after surgery, etc.

[0071] The comorbidity index, or the age-adjusted Charlson Comorbidity Index, is commonly used to assess the prognostic impact of a patient's comorbidities and predict their overall health. Comorbidities are assigned different scores based on severity to calculate the comorbidity index, which is then adjusted for age. In this application, comorbidities refer to other concurrent illnesses or health conditions that are not directly related to the underlying joint replacement. Because comorbidities may affect treatment, recovery, and prognosis, they need to be included in the assessment.

[0072] For example, common comorbidities and their corresponding scores are: heart disease: 1 point, stroke: 1 point, peripheral vascular disease: 1 point, lung disease (such as chronic obstructive pulmonary disease): 1 point, diabetes (without complications): 1 point, diabetes (with complications): 2 points, kidney disease: 1 point, cancer (metastatic): 6 points, and AIDS / HIV: 6 points. Diseases are extracted from the medical record system and statistically analyzed. The score ranges from 0 to 15.

[0073] The average pain score in the second 24 hours after surgery is the average pain score on the second day after the surgery. The pain score is obtained from the patient's medical records and generally ranges from 0 to 10 points.

[0074] Step S2: Analyze the influence of the potential factors on the perioperative blood transfusion indications of joint replacement through a univariate regression model to obtain several related factors.

[0075] In the embodiment of the present application, a univariate regression model is used with potential factors as independent variables and perioperative blood transfusion indications for joint replacement as dependent variables. The influence of a single independent variable on the dependent variable is analyzed to explain the possibility that a single potential factor may induce a patient to require allogeneic blood transfusion during the perioperative period.

[0076] Specifically, a single-factor regression model is used to fit a straight line or curve to each independent variable and dependent variable to describe the linear or nonlinear mathematical relationship between each independent variable and the dependent variable. This is used to quantify the degree of influence of each independent variable on the dependent variable, and the p-value corresponding to each independent variable in the mathematical relationship is calculated. The p-value is used to indicate whether the influence of the independent variable on the dependent variable is statistically significant. If the p-value is small, the influence of the independent variable on the dependent variable is considered significant, indicating a large influence. Conversely, if the p-value is large, the influence is considered insignificant, indicating a small influence.

[0077] As shown in the figure, Figure 2 The data presented here is a partial list of univariate regression analysis results, showing the p-values ​​for seven potential factors: patient age, average pain score in the second 24 hours after surgery, comorbidity index, operation duration, intraoperative blood loss, preoperative hemoglobin, and preoperative platelet count. Except for the average pain score in the second 24 hours after surgery, the p-values ​​for the remaining potential factors were all less than 0.05. Therefore, this example selected patient age, comorbidity index, operation duration, intraoperative blood loss, preoperative hemoglobin, and preoperative platelet count as relevant factors, indicating their close correlation with perioperative blood transfusion.

[0078] Step S3: performing control variable analysis on the potential factors through a multivariate regression model to screen out several independent risk factors.

[0079] In an embodiment of the present application, the historical clinical information related to the potential factors is input into a preset multivariate regression model, and each potential factor is used as an independent variable, and the perioperative blood transfusion indication for joint replacement is used as a dependent variable. A functional relationship between two or more independent variables and dependent variables is constructed to simultaneously consider the comprehensive effects of multiple potential factors on the perioperative blood transfusion indication for joint replacement, which can more accurately predict the allogeneic blood transfusion rate of patients in complex situations.

[0080] Specifically, a multivariate regression model was used to evaluate the impact of multiple potential factors on perioperative transfusion indications for arthroplasty. This involved constructing a functional relationship between two or more independent variables and the dependent variable, performing a control variable analysis on this functional relationship, and calculating the p-value corresponding to each potential factor (this p-value served the same purpose as the p-value in the univariate regression model). The p-value threshold was set at 0.05, and potential factors with a p-value below 0.05 were selected as independent risk factors.

[0081] Finally, the independent risk factors obtained included the average pain score in the second 24 hours after surgery, operation time, intraoperative blood loss, preoperative hemoglobin and preoperative platelet count.

[0082] Figure 3This figure shows a subset of the results from the multivariate regression analysis, showing the p-values ​​for seven potential factors: patient age, average pain score in the second 24 hours after surgery, comorbidity index, operative duration, intraoperative blood loss, preoperative hemoglobin, and preoperative platelet count. Except for patient age and comorbidity index, which had p-values ​​greater than 0.05, all other potential factors were selected as independent risk factors.

[0083] Step S4: combining the relevant factors and the independent risk factors to obtain several allogeneic blood transfusion assessment indicators.

[0084] The obtained related factors and independent risk factors were combined and these factors were taken as a union.

[0085] Then the union results are processed by nomogram analysis, and the union results are integrated into the same plane based on the preset coordinate system and a nomogram is drawn, as shown in the following example. Figure 4 As shown in Figure 2, nomogram analysis is a visual prediction tool used to transform complex mathematical models into intuitive graphs. Variable scale lines are used to indicate the range of each variable and its contribution to the result.

[0086] Specifically, in Figure 4 The variable scale lines for all allogeneic blood transfusion assessment indicators are displayed in the table. The numerical range of the scale lines represents the value range of the allogeneic blood transfusion assessment indicators. Intraoperative blood loss refers to the total blood loss during surgery, including gauze, suction, and other blood loss, with a range of 0 to 1000 ml. Operation duration refers to the duration from skin incision to suture and dressing, with a range of 50 to 500 minutes. Preoperative hemoglobin refers to the hemoglobin concentration per unit volume of blood, with a normal range of 70 to 170 g / L. Preoperative platelet count refers to the number of platelets per unit volume of blood, with a normal range of 150 to 450 × 109 platelets per liter (L).

[0087] Through nomogram analysis, it can be found that incorporating the average pain score in the second 24 hours after surgery can improve the accuracy of predicting perioperative allogeneic blood transfusion. Therefore, the innovation of the embodiment of the present application is to consider the impact of the average pain score in the second 24 hours after surgery on the patient's perioperative period to predict the incidence of allogeneic blood transfusion.

[0088] By drawing a nomogram of allogeneic blood transfusion assessment indicators to assist in model training of the allogeneic blood transfusion incidence prediction model, the accuracy of predicting the perioperative allogeneic blood transfusion incidence can be improved.

[0089] In order to better illustrate the impact of the average pain score in the second 24 hours after surgery on the accuracy of predicting the incidence of allogeneic blood transfusion, Figure 5Two ROC curves are provided: the red line represents the ROC curve for the training set, and the green line represents the ROC curve for the test set. The upper figure shows the performance of the allogeneic blood transfusion incidence prediction model when the nomogram for the average pain score in the second 24 hours after surgery is not plotted. The AUC for the ROC curve for the training set is 0.847, and the AUC for the test set is 0.729. The lower figure shows the performance of the allogeneic blood transfusion incidence prediction model when the nomogram for the average pain score in the second 24 hours after surgery is included and the corresponding nomogram is plotted. The AUC for the training set and the AUC for the test set are 0.858 and 0.757, respectively. This indicates that the inclusion of the average pain score in the second 24 hours after surgery in both the training and test sets results in closer agreement with the actual results, indicating better model performance and higher prediction accuracy. The AUC is the area under the ROC curve enclosed by the coordinate axes. A closer AUC is to 1, the more reliable the prediction.

[0090] Step S5: Calculate the predicted value of the target patient's allogeneic blood transfusion incidence rate according to the allogeneic blood transfusion evaluation index using a preset allogeneic blood transfusion incidence prediction model.

[0091] The data related to the allogeneic blood transfusion assessment index in the clinical data of the target patient is input into the trained allogeneic blood transfusion incidence prediction model for calculation. The allogeneic blood transfusion incidence prediction model has a calculation equation corresponding to each allogeneic blood transfusion assessment index, which is as follows:

[0092] Patient age score = 0*age^2+0.617908041*age+-40.164022641;

[0093] Comorbidity index score = 0*Charlson comorbidity index^2+3.300593204*age-adjusted Charlson comorbidity index+-6.601186407;

[0094] Preoperative hemoglobin score = 0*hemoglobin^3 + 0*hemoglobin^2 + -0.394518521*hemoglobin + 67.068148548";

[0095] Preoperative platelet count score = -0.091487958*platelet count + 50.318376939";

[0096] Operation time score = 0*operation time (mm)^2 + 0.222222222*operation time (mm) + -11.111111111";

[0097] Intraoperative blood loss = 0.036717727*blood loss + 0;

[0098] The average pain score in the second 24 hours after surgery = 0 * the average pain score in the second 24 hours after surgery^2 + 6.411594831 * the average pain score in the second 24 hours after surgery + -6.411594831;

[0099] The incidence of allogeneic blood transfusion for each variable = -6.34e-06*variable score^3+0.002152556*variable score^2+-0.224333147*variable score+7.507629815;

[0100] The total incidence of perioperative allogeneic blood transfusion in patients = the sum of the incidences of allogeneic blood transfusion of 7 variables.

[0101] The above formula can be used to obtain the predicted incidence of allogeneic blood transfusion in target patients.

[0102] In addition, if the predicted value of the incidence of allogeneic blood transfusion is greater than or equal to 50%, an early warning signal for preparing blood sources is issued.

[0103] Furthermore, after receiving the early warning signal, the patient's recovery can be accelerated through nerve blockade, local injection of "cocktail" into the joint capsule and periarticular soft tissue after osteotomy, improvement of anemia and increase of hemoglobin level before surgery, active autologous blood transfusion during surgery, and use of patient-controlled analgesia pump after surgery, so as to reduce postoperative bleeding and exudation, and reduce the incidence of perioperative allogeneic blood transfusion.

[0104] Figure 6 This example provides a flowchart for predicting the incidence of allogeneic blood transfusion during joint replacement surgery, using a small program (Figure (a)). Entering the patient's name or hospitalization number in the rectangular box (Figure (b)) will extract the corresponding clinical information. Then, entering the relevant clinical information according to the content shown in Figure (c) will yield the perioperative allogeneic blood transfusion rate (Figure (d)).

[0105] The implementation of the embodiments of the present application has the following beneficial effects:

[0106] This embodiment of the present application first selects several indicators from historical clinical information as potential factors that may cause a patient to require allogeneic blood transfusion during the perioperative period. These potential factors are then analyzed using univariate and multivariate regression. Univariate regression primarily analyzes the impact of a single potential factor on perioperative transfusion indications, while multivariate regression simultaneously considers the combined effects of multiple potential factors on perioperative transfusion indications, making it suitable for interpretation and prediction in complex scenarios. This results in allogeneic transfusion assessment indicators that are closely related to perioperative transfusion indications for joint replacement surgery. Finally, a pre-defined allogeneic transfusion incidence prediction model is used to calculate each allogeneic transfusion assessment indicator in combination with the target patient's clinical data to obtain a predicted allogeneic transfusion incidence value for the target patient. Because the model also considers factors such as patient age and other preexisting medical conditions when predicting allogeneic transfusion incidence, rather than simply considering preoperative anemia, intraoperative blood loss, and postoperative bleeding, it significantly improves the accuracy of prediction results compared to traditional allogeneic transfusion incidence prediction models, providing data support for rationally planning blood use and accelerating postoperative recovery.

[0107] Second embodiment

[0108] Furthermore, in order to implement the above method embodiment corresponding to the elderly patients' joint replacement perioperative allogeneic blood transfusion incidence prediction system to achieve the corresponding functions and technical effects, Figure 7 A structural diagram of a system for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients is provided. For ease of illustration, only the parts relevant to this embodiment are shown. The system for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients provided by this embodiment of the application includes:

[0109] The data selection module 201 is used to select several potential factors from the patient's historical clinical information.

[0110] In an embodiment of the present application, a large amount of historical clinical information of elderly patients who underwent unilateral hip or knee replacement surgery during their hospitalization in the past few years is first obtained according to data selection criteria, and then the historical clinical information is screened based on factor screening criteria to select appropriate and complete data.

[0111] The data selection criteria are set as follows:

[0112] (1) Elderly patients undergoing elective hip replacement surgery;

[0113] (2) Patients aged 65 years or older, regardless of gender, and ASA grade ≤ IV. ASA grade is a classification of surgical risk based on the patient's physical condition before anesthesia. A higher grade indicates a higher perioperative mortality rate.

[0114] (3) No recent history of surgery.

[0115] The factor screening criteria are set as follows:

[0116] (1) Not used for emergency surgery

[0117] (2) If the data integrity is not high, it will not be selected.

[0118] The historical clinical information was screened based on the preset factor screening criteria to obtain the potential factors, including patient age, BMI, nutritional score, comorbidity index, preoperative hemoglobin, preoperative platelet count, albumin, albumin to lymphocyte / leukocyte ratio, prothrombin ratio, fibrinogen content, anesthesia method, ASA grade, operation time, intraoperative blood loss, intraoperative urine volume, average pain score in the second 24 hours after surgery, etc.

[0119] The comorbidity index, or the age-adjusted Charlson Comorbidity Index, is commonly used to assess the prognostic impact of a patient's comorbidities and predict their overall health. Comorbidities are assigned different scores based on severity to calculate the comorbidity index, which is then adjusted for age. In this application, comorbidities refer to other concurrent illnesses or health conditions that are not directly related to the underlying joint replacement. Because comorbidities may affect treatment, recovery, and prognosis, they need to be included in the assessment.

[0120] For example, common comorbidities and their corresponding scores are: heart disease: 1 point, stroke: 1 point, peripheral vascular disease: 1 point, lung disease (such as chronic obstructive pulmonary disease): 1 point, diabetes (without complications): 1 point, diabetes (with complications): 2 points, kidney disease: 1 point, cancer (metastatic): 6 points, and AIDS / HIV: 6 points. Diseases are extracted from the medical record system and statistically analyzed. The score ranges from 0 to 15.

[0121] The average pain score in the second 24 hours after surgery is the average pain score on the second day after the surgery. The pain score is obtained from the patient's medical records and generally ranges from 0 to 10 points.

[0122] The univariate analysis module 202 is used to analyze the influence of the potential factors on the perioperative blood transfusion indications of joint replacement surgery through a univariate regression model to obtain several related factors.

[0123] In the embodiment of the present application, a univariate regression model is used with potential factors as independent variables and perioperative blood transfusion indications for joint replacement as dependent variables. The influence of a single independent variable on the dependent variable is analyzed to explain the possibility that a single potential factor may induce a patient to require allogeneic blood transfusion during the perioperative period.

[0124] Specifically, a single-factor regression model is used to fit a straight line or curve to each independent variable and dependent variable to describe the linear or nonlinear mathematical relationship between each independent variable and the dependent variable. This is used to quantify the degree of influence of each independent variable on the dependent variable, and the p-value corresponding to each independent variable in the mathematical relationship is calculated. The p-value is used to indicate whether the influence of the independent variable on the dependent variable is statistically significant. If the p-value is small, the influence of the independent variable on the dependent variable is considered significant, indicating a large influence. Conversely, if the p-value is large, the influence is considered insignificant, indicating a small influence.

[0125] Therefore, in the embodiment of the present application, the p-value of each independent variable is compared with the size of the set threshold to measure the degree of influence of the independent variable on the dependent variable, so as to select the corresponding relevant factors.

[0126] For example, when the p-value is less than or equal to 0.05, it indicates that the independent variable has a greater impact on the dependent variable, and the corresponding independent variable can be selected as a related factor, which includes patient age, comorbidity index, operation time, intraoperative blood loss, preoperative hemoglobin and preoperative platelet count.

[0127] The multi-factor analysis module 203 is used to perform control variable analysis on the potential factors through a multi-factor regression model to screen out a number of independent risk factors.

[0128] In an embodiment of the present application, the historical clinical information related to the potential factors is input into a preset multivariate regression model, and each potential factor is used as an independent variable, and the perioperative blood transfusion indication for joint replacement is used as a dependent variable. A functional relationship between two or more independent variables and dependent variables is constructed to simultaneously consider the comprehensive effects of multiple potential factors on the perioperative blood transfusion indication for joint replacement, which can more accurately predict the allogeneic blood transfusion rate of patients in complex situations.

[0129] Specifically, a multivariate regression model was used to evaluate the impact of multiple potential factors on perioperative transfusion indications for arthroplasty. This involved constructing a functional relationship between two or more independent variables and the dependent variable, performing a control variable analysis on this functional relationship, and calculating the p-value corresponding to each potential factor (this p-value served the same purpose as the p-value in the univariate regression model). The p-value threshold was set at 0.05, and potential factors with a p-value below 0.05 were selected as independent risk factors.

[0130] Finally, the independent risk factors obtained included the average pain score in the second 24 hours after surgery, operation time, intraoperative blood loss, preoperative hemoglobin and preoperative platelet count.

[0131] The evaluation index acquisition module 204 is used to combine the relevant factors and the independent risk factors to obtain a number of allogeneic blood transfusion evaluation indicators; wherein the allogeneic blood transfusion evaluation indicators include the average pain score in the second 24 hours after surgery, patient age, comorbidity index, operation time, intraoperative blood loss, preoperative hemoglobin and preoperative platelet count.

[0132] The obtained related factors and independent risk factors were combined and these factors were taken as a union.

[0133] The union results are then processed using nomogram analysis, which integrates them onto a single plane based on a pre-set coordinate system and plots a nomogram. Nomogram analysis is a visual forecasting tool used to transform complex mathematical models into intuitive graphs, using variable scale lines to indicate the range of each variable and its contribution to the result.

[0134] Specifically, according to the nomogram, the value range of each allogeneic blood transfusion assessment indicator can be known. Intraoperative blood loss refers to the total amount of blood lost by the patient during the operation, including all blood lost by gauze, suction device, etc., with a value range of 0 to 1000 ml; operation duration refers to the time from skin incision to suture and bandaging, with a value range of 50 to 500 minutes; preoperative hemoglobin refers to the concentration of hemoglobin per unit volume of blood, with a value range of 70 to 170 g / L; preoperative platelet count refers to the number of platelets per unit volume of blood, with a normal value range of 150 to 450×10^9 platelets per liter (L).

[0135] Through nomogram analysis, it can be found that incorporating the average pain score in the second 24 hours after surgery can improve the accuracy of predicting perioperative allogeneic blood transfusion. Therefore, the innovation of the embodiment of the present application is to consider the impact of the average pain score in the second 24 hours after surgery on the patient's perioperative period to predict the incidence of allogeneic blood transfusion.

[0136] The prediction value calculation module 205 is used to calculate the predicted value of the allogeneic blood transfusion incidence of the target patient according to the allogeneic blood transfusion evaluation index using a preset allogeneic blood transfusion incidence prediction model.

[0137] The data related to the allogeneic blood transfusion assessment index in the clinical data of the target patient is input into the trained allogeneic blood transfusion incidence prediction model for calculation. The allogeneic blood transfusion incidence prediction model has a calculation equation corresponding to each allogeneic blood transfusion assessment index, which is as follows:

[0138] Patient age score = 0*age^2+0.617908041*age+-40.164022641;

[0139] Comorbidity index score = 0*Charlson comorbidity index^2+3.300593204*age-adjusted Charlson comorbidity index+-6.601186407;

[0140] Preoperative hemoglobin score = 0*hemoglobin^3 + 0*hemoglobin^2 + -0.394518521*hemoglobin + 67.068148548";

[0141] Preoperative platelet count score = -0.091487958*platelet count + 50.318376939";

[0142] Operation time score = 0*operation time (mm)^2 + 0.222222222*operation time (mm) + -11.111111111";

[0143] Intraoperative blood loss = 0.036717727*blood loss + 0;

[0144] The average pain score in the second 24 hours after surgery = 0 * the average pain score in the second 24 hours after surgery^2 + 6.411594831 * the average pain score in the second 24 hours after surgery + -6.411594831;

[0145] The incidence of allogeneic blood transfusion for each variable = -6.34e-06*variable score^3+0.002152556*variable score^2+-0.224333147*variable score+7.507629815;

[0146] The total incidence of perioperative allogeneic blood transfusion in patients = the sum of the incidence of allogeneic blood transfusion of 7 variables.

[0147] The above formula can be used to obtain the predicted incidence of allogeneic blood transfusion in target patients.

[0148] In addition, if the predicted value of the incidence of allogeneic blood transfusion is greater than or equal to 50%, an early warning signal for preparing blood sources is issued.

[0149] Furthermore, after receiving the early warning signal, the patient's recovery can be accelerated through nerve blockade, local injection of "cocktail" into the joint capsule and periarticular soft tissue after osteotomy, and the use of patient-controlled analgesia pump after surgery, so as to reduce postoperative bleeding and exudation of blood, and reduce the incidence of perioperative allogeneic blood transfusion.

[0150] The implementation of the embodiments of the present application has the following beneficial effects:

[0151] This embodiment of the present application first selects several indicators from historical clinical information as potential factors that may cause a patient to require allogeneic blood transfusion during the perioperative period. These potential factors are then analyzed using univariate and multivariate regression. Univariate regression primarily analyzes the impact of a single potential factor on perioperative transfusion indications, while multivariate regression simultaneously considers the combined effects of multiple potential factors on perioperative transfusion indications, making it suitable for interpretation and prediction in complex scenarios. This results in allogeneic transfusion assessment indicators that are closely related to perioperative transfusion indications for joint replacement surgery. Finally, a pre-defined allogeneic transfusion incidence prediction model is used to calculate each allogeneic transfusion assessment indicator in combination with the target patient's clinical data to obtain a predicted allogeneic transfusion incidence value for the target patient. Because the model also considers factors such as patient age and other preexisting medical conditions when predicting allogeneic transfusion incidence, rather than simply considering preoperative anemia, intraoperative blood loss, and postoperative bleeding, it significantly improves the accuracy of prediction results compared to traditional allogeneic transfusion incidence prediction models, providing data support for rationally planning blood use and accelerating postoperative recovery.

[0152] The specific embodiments described above further illustrate the purpose, technical solutions, and beneficial effects of this application. It should be understood that the above description is merely a specific embodiment of this application and is not intended to limit the scope of protection of this application. In particular, it should be noted that for those skilled in the art, any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of this application should be included in the scope of protection of this application.

Claims

1. A method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients, characterized by: include: Several potential factors were selected from the patient's historical clinical information; The influence of the potential factors on perioperative blood transfusion indications during joint replacement surgery was analyzed by univariate regression model, and several related factors were obtained. The potential factors were analyzed by controlling variables through multivariate regression model to screen out several independent risk factors; Combining the relevant factors and the independent risk factors, several allogeneic blood transfusion assessment indicators were obtained; wherein the allogeneic blood transfusion assessment indicators included the average pain score in the second 24 hours after surgery, patient age, comorbidity index, operation time, intraoperative blood loss, preoperative hemoglobin, and preoperative platelet count; The predicted value of the allogeneic blood transfusion incidence of the target patient is calculated based on the allogeneic blood transfusion evaluation index using a preset allogeneic blood transfusion incidence prediction model.

2. The method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients according to claim 1, characterized in that: The method selects several potential factors from the patient's historical clinical information, specifically: Obtain patients' historical clinical information from medical records; The historical clinical information is screened based on preset factor screening criteria to obtain the potential factors; wherein the factor screening criteria include non-emergency surgery history and completeness of patient information.

3. The method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients according to claim 1, characterized in that: The influence of the potential factors on the perioperative blood transfusion indications of joint replacement surgery was analyzed by univariate regression model, and several related factors were obtained, specifically: Inputting the historical clinical information related to the potential factors into the univariate regression model, taking each potential factor as an independent variable and the perioperative blood transfusion indication for joint replacement as a dependent variable, fitting a straight line or a curve to describe the relationship between each independent variable and the dependent variable, and obtaining the degree of influence of each independent variable on the dependent variable; The independent variable whose influence degree is greater than the first threshold is taken as the relevant factor.

4. The method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients according to claim 3, characterized in that: The related factors mentioned include: patient age, comorbidity index, operation duration, intraoperative blood loss, preoperative hemoglobin, and preoperative platelet count; The comorbidity index is a total assessment score obtained by quantitatively evaluating the patient's various existing diseases, and is used to predict the patient's overall health status.

5. The method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients according to claim 1, characterized in that: The multivariate regression model was used to analyze the potential factors and identify several independent risk factors, specifically: Inputting the historical clinical information related to the potential factors into the multivariate regression model, with each potential factor as an independent variable and the perioperative blood transfusion indication during joint replacement surgery as a dependent variable; By constructing a functional relationship between two or more independent variables and a dependent variable and performing a control variable analysis on the functional relationship, the independent risk factors are screened out from the potential factors.

6. The method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients according to claim 5, characterized in that: The independent risk factors include: average pain score in the second 24 hours after surgery, operation time, intraoperative blood loss, preoperative hemoglobin and preoperative platelet count; The average pain score in the second 24 hours after surgery is the average pain score on the second day after the surgery, and the pain score is obtained from the patient's medical records.

7. The method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients according to claim 1, characterized in that: The combination of the relevant factors and the independent risk factors yields several allogeneic blood transfusion assessment indicators, specifically: Taking the union of the relevant factors and the independent risk factors to obtain a union result; Based on a preset coordinate system, the union results are integrated into the same plane to draw a nomogram, and the allogeneic blood transfusion evaluation index is obtained according to the nomogram.

8. The method for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients according to claim 1, characterized in that: The predicted value of the allogeneic blood transfusion incidence rate of the target patient is calculated based on the allogeneic blood transfusion assessment index using the preset allogeneic blood transfusion incidence prediction model, specifically: Inputting the clinical information of the target patient into the allogeneic blood transfusion incidence prediction model, and calculating the index score corresponding to each of the allogeneic blood transfusion assessment indicators using a preset calculation formula; The sum of all the index scores is calculated to obtain the predicted incidence of allogeneic blood transfusion in the target patient.

9. The method for predicting the incidence of perioperative allogeneic blood transfusion in elderly patients undergoing joint replacement surgery according to any one of claims 1 to 8, wherein: Also includes: If the predicted value of the incidence of allogeneic blood transfusion is greater than or equal to a second threshold, an early warning signal for preparing a blood source is issued; Based on the warning signals, the incidence of allogeneic blood transfusion in target patients can be reduced by improving anemia before surgery to increase hemoglobin levels, performing autologous blood transfusion during surgery, and using perioperative multimodal analgesia.

10. A system for predicting the incidence of allogeneic blood transfusion during perioperative joint replacement in elderly patients, characterized by: include: Data selection module, single factor analysis module, multi-factor analysis module, evaluation index acquisition module and prediction value calculation module; Among them, the data selection module is used to select several potential factors from the patient's historical clinical information; The univariate analysis module is used to analyze the influence of the potential factors on the perioperative blood transfusion indications of joint replacement through a univariate regression model, and obtain several related factors; The multivariate analysis module is used to perform control variable analysis on the potential factors through a multivariate regression model to screen out several independent risk factors; The evaluation index acquisition module is used to combine the relevant factors and the independent risk factors to obtain a number of allogeneic blood transfusion evaluation indicators; wherein the allogeneic blood transfusion evaluation indicators include the average pain score in the second 24 hours after surgery, patient age, comorbidity index, operation time, intraoperative blood loss, preoperative hemoglobin and preoperative platelet count; The predicted value calculation module is used to calculate the predicted value of the allogeneic blood transfusion incidence of the target patient according to the allogeneic blood transfusion evaluation index through a preset allogeneic blood transfusion incidence prediction model.

Images