A medication case management system and method for chronic disease patients

Abstract

The application relates to the technical field of data storage management, and discloses a medication case management system and method for chronic disease patients, which comprises the following steps: acquiring a case set, extracting a key information field to construct a feature sequence, performing effectiveness verification and a case screening experiment, calculating time-dependent representation parameters and combination-dependent representation parameters to determine a case-dependent feature value, and destroying invalid cases based on sensitive case types. The application can efficiently manage medication cases of chronic disease patients, reduce the storage space occupied by invalid cases, avoid data redundancy and resource waste, and improve the accuracy and efficiency of case management.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data storage management, in particular to a medication case management system and method for chronic disease patients. BACKGROUND

[0002] In the field of chronic disease management, the medication records of patients are usually stored in the form of case data, which includes medication time, dosage, treatment plan and basic information of patients, etc. With the increase in the number of chronic disease patients, the amount of case data is showing a rapid growth trend. However, the existing case management system mainly focuses on simple storage and query, lacks in-depth analysis and regular mining of case characteristics, resulting in the existence of repeated, redundant and invalid case information, which affects the accurate extraction and analysis of medication rules.

[0003] At present, in the process of case storage and updating, the existing method usually does not fully consider the time dependence and combination dependence between cases, and lacks system identification of case type sensitivity. This deficiency makes it difficult to accurately judge the importance and potential value of the case when screening, processing or analyzing the case, thereby affecting the precision and scientificity of case management. In addition, some chronic disease cases may appear unresolvable or invalid during long-term management, and the existing technology lacks a dynamic identification and destruction mechanism for invalid cases, which cannot effectively reduce the occupation of useless data to the system, increasing the complexity and uncertainty of data processing.

[0004] Therefore, a management method capable of validity verification, dependency analysis and sensitive type identification of cases is needed to realize efficient, accurate and dynamic management of chronic disease medication cases. SUMMARY

[0005] In view of this, the present application provides a medication case management system and method for chronic disease patients, aiming to solve the problem that the existing chronic disease medication case management method in the current technology lacks system identification of the time dependence, combination dependence and sensitive type of cases, and cannot dynamically identify and process unresolvable or invalid cases, resulting in low case management precision, high data redundancy and high processing complexity.

[0006] The present application provides a medication case management method for chronic disease patients, comprising:

[0007] Obtaining a set of chronic disease patient medication cases within a preset time range, extracting key information fields of each case, constructing a case feature sequence based on the key information fields, recording the number of occurrences of the same case feature sequence to analyze the historical use rules of the case, and verifying the validity of the case;

[0008] Based on the validity verification result, a case screening experiment is performed, a predetermined proportion of cases in the case set is randomly selected, the processing time deviation corresponding to each case is recorded, a neighboring case group is screened, and the time-dependent representation parameter is determined based on the analysis result of the neighboring case group;

[0009] The cases subjected to the case screening experiment are analyzed, the number of unanalyzable cases is determined, and the combination-dependent representation parameter is obtained;

[0010] Based on the time-dependent representation parameter and the combination-dependent representation parameter, the case-dependent feature value is calculated, and the case-dependent category is divided;

[0011] Based on the division result of the case-dependent category, the storage and update process of the case is controlled, including monitoring the result of the case screening experiment;

[0012] Based on the result of the case screening experiment, the arrival distribution value and the analysis representation value of the cases corresponding to each case type are analyzed, and the case type sensitivity parameter is determined according to the arrival distribution value and the analysis representation value of the cases corresponding to each case type, so as to locate the sensitive case type;

[0013] According to the sensitive case type, the invalid cases that need to be destroyed are determined and the destruction operation is performed.

[0014] Further, when verifying the validity of the case, the historical use rule of the case is analyzed, including:

[0015] The number of times of similar use behaviors of the case in a plurality of historical use processes is recorded;

[0016] The ratio of the number of times of similar use behaviors to the total number of uses is determined to obtain a similar use behavior ratio;

[0017] If the similar use behavior ratio is greater than a predetermined similar use behavior ratio threshold, the case is determined to be valid;

[0018] The similar use behavior needs to satisfy that the case feature sequence corresponding to each case in the historical use process is consistent with the case feature sequence corresponding to each case in any other historical use process.

[0019] Further, when performing the case screening experiment based on the validity verification result, including:

[0020] If the case is valid, the case screening experiment needs to be performed, a predetermined proportion of cases in the case set is randomly selected, the processing time deviation corresponding to each case is recorded, a neighboring case group is screened, and the time-dependent representation parameter is determined based on the analysis result of the neighboring case group;

[0021] If the case is not valid, the case screening experiment does not need to be performed.

[0022] Further, in response to the determination result of the case dependency category, the storage and update of the case are controlled, and the method comprises the steps that:

[0023] A processing time deviation corresponding to each case is determined.

[0024] If there is a case satisfying the deviation condition, the case is combined with the adjacent case group to form an adjacent case group.

[0025] It is determined whether each case in the adjacent case group can be completely analyzed to determine a time-dependent case group.

[0026] The ratio of the number of time-dependent case groups to the total number of cases is determined as a time-dependent characterization parameter.

[0027] The deviation condition is that the deviation time corresponding to the case is greater than the predetermined time deviation, and the complete analysis is that each case in the adjacent case group can be completely analyzed.

[0028] Further, in determining the combination-dependent characterization parameter, the method comprises the steps that:

[0029] The difference between the number of unanalyzable cases and the number of damaged cases is determined.

[0030] The ratio of the difference value to the number of unanalyzable cases is determined as a combination-dependent characterization parameter.

[0031] Further, in calculating the case-dependent feature value based on the time-dependent characterization parameter and the combination-dependent characterization parameter, the method comprises the steps that:

[0032] The ratio of the time-dependent characterization parameter to the reference time-dependent characterization parameter is determined as a time-dependent influence factor.

[0033] The ratio of the combination-dependent characterization parameter to the reference combination-dependent characterization parameter is determined as a combination-dependent influence factor.

[0034] The weighted sum of the time-dependent influence factor and the combination-dependent influence factor is determined as the case-dependent feature value.

[0035] Further, in response to the determination result of the case dependency category, the storage and update of the case are controlled, and the method comprises the steps that:

[0036] If the case-dependent feature value is greater than the case-dependent feature value threshold, the case dependency category is divided into a first category, the results of the case screening experiment are monitored, the arrival distribution value and the analysis characterization value of each case type corresponding to the case are analyzed based on the results of the case screening experiment, the case type sensitive parameter is calculated to locate the sensitive case type, and the invalid case needing to be destroyed is selected based on the location result and the destruction operation is performed.

[0037] If the case-dependent feature value is less than or equal to the case-dependent feature value threshold, the case dependency category is divided into a second category.

[0038] Further, based on the result analysis of the case screening experiment, the arrival distribution value and the analytical characterization value of each case type corresponding to the case are analyzed, and the case type sensitive parameter is determined, so as to locate the sensitive case type, including:

[0039] The number of analyzable cases of each case type corresponding to the case is determined as the arrival distribution value;

[0040] The ratio of the number of analyzable cases to the total number of cases of the same case type is determined as the analytical characterization value;

[0041] The ratio of the reference arrival distribution value corresponding to the case type to the arrival distribution value is determined as the distribution value influence factor;

[0042] The ratio of the reference analytical characterization value corresponding to the case type to the analytical characterization value is determined as the analytical characterization value influence factor;

[0043] The weighted sum of the distribution value influence factor and the analytical characterization value influence factor is determined as the case type sensitive parameter.

[0044] Further, when locating the sensitive case type, selecting the invalid case to be destroyed based on the positioning result and performing the destruction operation, including:

[0045] If the case type sensitive parameter is greater than the preset sensitive parameter threshold, the case type is determined as a sensitive case type;

[0046] Selecting the case corresponding to the sensitive case type from the stored cases as the invalid case;

[0047] Performing the destruction operation on the invalid case;

[0048] Wherein, the destruction operation is to generate a destruction instruction based on the identification information of the invalid case, and to trigger the case management system to delete the invalid case through the destruction instruction.

[0049] Compared with the prior art, the application has the beneficial effects that: by acquiring a set of medication cases of chronic disease patients within a preset time range, and extracting key information and constructing feature sequences of the cases, the analysis of the case history use rule and the effectiveness verification are realized. This step can eliminate data that is not representative or has obvious abnormalities in advance, ensuring the accuracy and reliability of subsequent management from the source, and avoiding the analysis deviation caused by invalid data in the traditional method. On this basis, by introducing a case screening experiment and the construction of a neighboring case group, the time-dependent representation parameters are determined by using the processing time deviation and the analysis result, further revealing the dependence of the case in the time dimension. This design can dynamically depict the time-sensitive features of the case, providing quantitative indicators for the rule identification and time management in the medication process of chronic diseases, and improving the adaptability of the system to complex case changes. At the same time, by analyzing the number of unanalyzable cases in the screening experiment, combined with the number of damaged cases, the combined dependence representation parameters are calculated, realizing the dependence measurement of the case in the combination dimension. This mechanism effectively avoids the one-sidedness caused by relying on a single dimension, and can fully reflect the characteristics of the case in terms of data integrity and combination analyzability, thereby providing a more solid foundation for the classification and management of the case. Further, by weighting and fusing the time-dependent representation parameters and the combined dependence representation parameters, the case dependence feature value is calculated, and the case dependence category is divided accordingly. With the help of this division mechanism, the system can realize the differentiated storage and update management of the case: for the high-dependence category case, dynamic management is carried out; for the low-dependence category case, a simplified storage strategy is adopted. This hierarchical management effectively improves the utilization rate and update efficiency of data storage, avoiding resource waste. Finally, by calculating the case type sensitivity parameter based on the case screening experiment result, the sensitive case type is located, and invalid cases are automatically identified and destroyed based on the positioning result. This step realizes the dynamic cleaning of invalid data, ensures the high quality and high purity of the core database, and reduces the interference of redundant data on system operation and subsequent analysis. Through the automatic destruction mechanism, manual intervention and potential errors are reduced, making the case management process more efficient and intelligent.

[0050] In another aspect, the application also provides a medication case management system for chronic disease patients, comprising:

[0051] A case acquisition module is configured to acquire a set of medication cases of chronic disease patients within a preset time range, extract key information fields of each case, and construct a case feature sequence based on the key information fields;

[0052] A case analysis module is electrically connected to the case acquisition module, and is configured to analyze the case and determine the number of unanalyzable cases;

[0053] The case screening module is electrically connected with the case analysis module. The case screening module is used for performing a case screening experiment, recording a processing time deviation corresponding to each case, and screening a group of adjacent cases.

[0054] The case dependency calculation module is electrically connected with the case screening module. The case dependency calculation module is used for calculating a case dependency characteristic value based on the time dependency representation parameter and the combination dependency representation parameter.

[0055] The case destruction module is electrically connected with the case dependency calculation module. The case destruction module is used for positioning a sensitive case type based on a case type sensitive parameter, and selecting an invalid case to be destroyed.

[0056] The case storage module is electrically connected with the case destruction module. The case storage module is used for storing an effective case and updating a case library.

[0057] It can be understood that the medication case management system and method for chronic disease patients in each of the above embodiments have the same beneficial effects, which will not be repeated. BRIEF DESCRIPTION OF DRAWINGS

[0058] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments and are not intended to limit the scope of the application. Moreover, the same reference numerals are used throughout the same figures. In the drawings:

[0059] Figure 1 A flowchart of a medication case management method for chronic disease patients is provided for the embodiments of the present application;

[0060] Figure 2 A function block diagram of a medication case management system for chronic disease patients is provided for the embodiments of the present application. DETAILED DESCRIPTION

[0061] Exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be accurately conveyed to those skilled in the art. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0062] As Figure 1 shown in some embodiments of the present application, the present embodiments provide a medication case management method for chronic disease patients, comprising:

[0063] Step S100, obtaining a chronic disease patient medication case set in a preset time range, extracting key information fields of each case, constructing a case feature sequence based on the key information fields, recording the number of occurrences of the same case feature sequence to analyze the historical use rules of the case, and verifying the effectiveness of the case.

[0064] Specifically, when analyzing the historical use rules of the case and verifying the effectiveness of the case, the following steps are included: recording the number of times of similar use behaviors of the case in several historical use processes; determining the ratio of the number of similar use behaviors to the total number of uses to obtain a similar use behavior ratio; if the similar use behavior ratio is greater than a predetermined similar use behavior ratio threshold, the case is determined to be valid; wherein the similar use behavior needs to satisfy that the case feature sequence corresponding to each case in the historical use process is consistent with the case feature sequence corresponding to each case in any other historical use process.

[0065] It can be understood that by obtaining a patient medication case set from a preset time range and extracting key information fields of each case, such as medication time, dose, medication scheme, etc., a case feature sequence is constructed, thereby converting complex case data into a computable feature representation. After constructing the case feature sequence, the historical use rules of the case are analyzed by counting the number of occurrences of the same case feature sequence in the historical use process. This process records the number of times of similar use behaviors of the case in multiple historical uses, calculates the proportion of the number of similar use behaviors in the total number of uses, and obtains a similar use behavior ratio, thereby quantifying the stability and regularity of the case. Based on the above analysis results, further effectiveness verification is performed: when the similar use behavior ratio exceeds a predetermined threshold, the case is determined to be valid. The so-called similar use behavior refers to the case feature sequence of any one use being completely consistent with the feature sequence in other historical uses in the historical use process. Through this technical means, cases with stable regularity and reference value in medication management can be objectively identified, providing a reliable data basis for subsequent case screening, dependency analysis, and sensitive type positioning.

[0066] In the specific embodiments in the present application, the above steps realize the scenario in the following manner: in a certain chronic disease management system, a set of medication cases of a patient in the past year is collected, including the key information fields of the dose of each medication, the medication time, the medication scheme, etc. The system extracts and encodes each case information to construct the corresponding case feature sequence, for example, encoding "morning medication 10 mg, evening medication 5 mg, scheme A" into a specific sequence to facilitate unified analysis and comparison. Then, the system statistically analyzes the historical use data to record the number of times each case feature sequence appears in the historical use process. For example, a certain patient has 4 times of medication feature sequences that are completely consistent with other uses in the past 6 times of medication, and the similar use behavior times are 4, the total use times are 6, and the similar use behavior ratio is 4 / 6≈0.67. Finally, the system determines the case effectiveness according to the preset threshold (such as 0.6). Since the similar use behavior ratio of the case is 0.67, which is greater than the threshold 0.6, it is determined that the case is an effective case. The example shows the specific operation process of realizing the effectiveness verification of the medication cases of the chronic disease patient by constructing the case feature sequence, statistically analyzing the historical use rules, and calculating the similar use behavior ratio.

[0067] The above scenario is only a preferred embodiment of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0068] Step S200, based on the effectiveness verification result, performing case screening experiment, randomly extracting a predetermined proportion of cases in the case set, recording the processing time deviation corresponding to each case, screening the adjacent case group, and determining the time-dependent representation parameter based on the analysis result of the adjacent case group.

[0069] Specifically, based on the effectiveness verification result, when performing the case screening experiment, if the case has effectiveness, the case screening experiment needs to be performed, a predetermined proportion of cases in the case set is randomly extracted, the processing time deviation corresponding to each case is recorded, the adjacent case group is screened, and the time-dependent representation parameter is determined based on the analysis result of the adjacent case group; if the case does not have effectiveness, the case screening experiment does not need to be performed.

[0070] Specifically, when screening the adjacent case group and determining the time-dependent representation parameter based on the analysis result of the adjacent case group, it includes: determining the processing time deviation corresponding to each case; if there is a case that meets the deviation condition, the case is combined with the adjacent case group to form an adjacent case group; determining whether each case in the adjacent case group can be completely analyzed to determine the time-dependent case group; the ratio of the number of the time-dependent case group to the total number of cases is determined as the time-dependent representation parameter; wherein the deviation condition is that the deviation time corresponding to the case is greater than the predetermined time deviation, and the complete analysis is that each case in the adjacent case group can be completely analyzed.

[0071] It can be understood that, by screening the cases of chronic disease drugs after validity verification, the regularity of case processing time is analyzed, so as to quantify the time-dependent characteristics. First, according to the validity verification result, the cases with validity are executed for screening experiment, that is, a predetermined proportion of cases are randomly selected from the case set for analysis, and the time deviation of each case in the processing process is recorded to obtain the time performance information of the case in the actual application. Then, the time-dependent characteristics are determined by screening the adjacent case groups. Specifically, the processing time deviation of each case is calculated, and the cases meeting the condition are screened out according to the deviation condition (such as the deviation time being greater than a predetermined threshold), and combined with the cases similar in time or characteristics to form an adjacent case group. Then, each case in the adjacent case group is analyzed to determine whether it can be completely analyzed to determine the time-dependent case group. Finally, by counting the proportion of the number of time-dependent case groups to the total number of cases, the time-dependent characteristic parameter is calculated. Through the analysis of the processing time deviation and the combination of the adjacent case groups, the time regularity of the case is quantified as a representable parameter, which provides basic data for subsequent case-dependent relationship analysis and sensitive type identification, so as to realize the scientific quantification and dynamic evaluation of the time dependence of the case in the management process of the chronic disease patients.

[0072] In the specific embodiments in the present application, the above steps realize the scene in the following manner: in the chronic disease drug management system, a plurality of cases with validity are screened out by validity verification. For example, 50% of the cases are randomly selected from the total case set as experimental objects, and the deviation of the actual processing time and the expected processing time of each case is recorded. Assuming that the processing time of a case is 15 minutes more than the expected time, the deviation is recorded as 15 minutes. Then, according to a preset time deviation threshold (such as 10 minutes), the cases whose deviation exceeds the threshold are screened out, and these cases are combined with the cases similar in time or characteristics to form an adjacent case group. For example, if the processing time deviations of two other cases are 12 minutes and 18 minutes respectively, they meet the deviation condition, and the three cases form an adjacent case group. Finally, each case in the adjacent case group is analyzed to determine whether it can be completely analyzed, that is, all case data are complete and the characteristic information is matchable. If the three cases in the adjacent case group can be completely analyzed, a time-dependent case group is formed. Assuming that a total of 4 time-dependent case groups are formed in the experiment, and the total number of experimental cases is 20, the time-dependent characteristic parameter is 4 / 20=0.2. This example shows the specific operation process of quantifying the time-dependent characteristics of the case by random selection, time deviation screening and adjacent case group analysis.

[0073] The above scenarios are only preferred embodiments of the present application and are not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0074] Step S300, analyzing the cases subjected to the case screening experiment, determining the number of unanalyzable cases, and obtaining the combination dependency characterization parameter.

[0075] Specifically, when determining the combination dependency characterization parameter, the following steps are included: determining the difference between the number of unanalyzable cases and the number of broken cases; and determining the ratio of the difference to the number of unanalyzable cases as the combination dependency characterization parameter.

[0076] It can be understood that by analyzing the cases subjected to the case screening experiment, the unanalyzable cases, i.e. those cases that cannot be effectively analyzed due to data missing, incomplete features, or inability to match with neighboring case groups, are identified. Subsequently, the combination dependency characterization parameter is calculated according to the analysis result. Specifically, the system first determines the number of unanalyzable cases and compares it with the number of broken cases (e.g. the number of cases caused by data abnormality or missing), and obtains the difference between the two. The difference reflects the relationship between the number of cases that cannot form effective combinations in the combination analysis process and the broken cases. Finally, the combination dependency characterization parameter is calculated by the ratio of the difference to the number of unanalyzable cases, which can quantify the dependency and completeness degree of the cases in the combination analysis process. Through this technical means, the stability and reliability of the case combination can be scientifically evaluated, which provides basic data support for case dependency classification and sensitive type identification, and realizes the combination dependency feature quantification analysis in the management of chronic disease patient medication cases.

[0077] In the specific embodiments in the present application, the above steps realize the scene in the following manner: in the chronic disease drug management system, a case set for performing a case screening experiment is parsed. In the parsing process, the system finds that part of the cases cannot be matched with the adjacent case groups due to data missing or incomplete key feature information, for example, a certain case lacks a drug dosage field or the drug use time record is incomplete, and such cases are identified as unresolvable cases. It is assumed that there are 10 unresolvable cases in the experiment. Then, the unresolvable cases are compared with the broken cases. It is assumed that the broken cases are 5 cases caused by system data abnormalities or uploading errors. The system calculates the difference between the number of unresolvable cases and the number of broken cases, that is, 10-5=5, and the difference reflects the number of cases that cannot form an effective combination due to non-broken reasons in the combination parsing process. Finally, the ratio of the difference value to the number of unresolvable cases is calculated to obtain the combination dependency representation parameter, that is, 5 / 10=0.5. This parameter quantifies the dependency and completeness degree of the cases in the combination parsing process, and the higher the numerical value, the greater the influence of unresolvable cases on the combination parsing. Through this example, it can be intuitively shown how to calculate the combination dependency representation parameter using the parsing result and the broken case data, thereby providing an actual reference basis for case dependency category division and sensitive type identification.

[0078] The above scene is only a preferred embodiment of the present application and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0079] Step S400, calculating a case dependency feature value based on the time dependency representation parameter and the combination dependency representation parameter, and dividing the case dependency category.

[0080] Specifically, when calculating the case dependency feature value based on the time dependency representation parameter and the combination dependency representation parameter, it includes: determining the ratio of the time dependency representation parameter to the reference time dependency representation parameter as a time dependency influence factor; determining the ratio of the combination dependency representation parameter to the reference combination dependency representation parameter as a combination dependency influence factor; and determining the weighted sum value of the time dependency influence factor and the combination dependency influence factor as the case dependency feature value.

[0081] It can be understood that by quantitatively analyzing the time dependence and combination dependence of the case, the case dependence characteristic value is calculated, and the scientific division of the case dependence category is realized. First, based on the time dependence characteristic parameters and the combination dependence characteristic parameters obtained in the foregoing steps, they are compared with the corresponding benchmark parameters. Specifically, the system calculates the ratio of the time dependence characteristic parameters to the benchmark time dependence characteristic parameters to obtain the time dependence influence factor, which is used to quantify the dependence degree of the case in the time dimension. At the same time, the ratio of the combination dependence characteristic parameters to the benchmark combination dependence characteristic parameters is calculated as the combination dependence influence factor, which is used to reflect the dependence and integrity degree of the case in the combination analysis process. By quantitatively analyzing the time dependence and the combination dependence, the dependence characteristics of the case can be comprehensively evaluated from two key dimensions. Finally, the time dependence influence factor and the combination dependence influence factor are weighted and summed to obtain the case dependence characteristic value of each case. The characteristic value comprehensively reflects the dependence of the case in the time and combination dimensions, and can be used to divide the case into different dependence categories. Through this principle, the dependence analysis of the chronic disease patient medication case can be scientifically and quantitatively performed, and reliable data support can be provided for the subsequent sensitive case type identification, storage optimization and invalid case destruction.

[0082] In the specific embodiments in the present application, the above steps realize the scene in the following manner: in the chronic disease medication management system, the time dependence characteristic parameters of a case obtained through the foregoing steps are 0.3, the combination dependence characteristic parameters are 0.5, the benchmark time dependence characteristic parameters set by the system are 0.2, and the benchmark combination dependence characteristic parameters are 0.4. The system first calculates the time dependence influence factor, that is, 0.3 / 0.2=1.5, which is used to quantify the dependence degree of the case in the time dimension. Subsequently, the system calculates the combination dependence influence factor, that is, 0.5 / 0.4=1.25, which is used to reflect the dependence and integrity degree of the case on other cases in the combination analysis process. Through the two influence factors, the system can comprehensively evaluate the dependence characteristics of the case from the time and combination two key dimensions. Finally, the system weights and sums the time dependence influence factor and the combination dependence influence factor (assuming that the weights are both 0.5) to obtain the case dependence characteristic value: 0.5*1.5+0.5*1.25=1.375. According to the set dependence characteristic value division threshold, for example, the cases with a characteristic value greater than 1.2 are classified into the high dependence category, and the case is classified into the high dependence category. Through the example, how to calculate the case dependence characteristic value based on the time dependence and the combination dependence characteristic parameters can be intuitively displayed, and the dependence category division of the chronic disease patient medication case is realized.

[0083] The above scenarios are only preferred embodiments of the present application and are not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0084] In step S500, based on the division result of the case dependency category, the storage and update process of the case is controlled, including monitoring the result of the case screening experiment.

[0085] Specifically, in response to the division result of the case dependency category, the storage and update of the case are controlled, including: if the case dependency feature value is greater than the case dependency feature value threshold, the case dependency category is divided into the first category, the result of the case screening experiment is monitored, the arrival distribution value and the analytical representation value of the case corresponding to each case type are analyzed based on the result of the case screening experiment, the case type sensitive parameter is calculated to locate the sensitive case type, and the invalid case needing to be destroyed is selected based on the positioning result and a destruction operation is performed; if the case dependency feature value is less than or equal to the case dependency feature value threshold, the case dependency category is divided into the second category.

[0086] It can be understood that the cases are divided into different dependency categories by the case dependency feature value calculated according to the foregoing. For example, when the case dependency feature value is greater than a set threshold, it is divided into the first category; when the feature value is less than or equal to the threshold, it is divided into the second category. This division reflects the comprehensive dependency of the case in the time and combination dimensions. Subsequently, for the first category case of high dependency, the system dynamically monitors it, obtains the arrival distribution value and the analytical representation value of each category case by analyzing the result of the case screening experiment. Based on these data, the system calculates the case type sensitive parameter for quantifying the importance and potential impact of each category case in the management process, so as to scientifically identify the sensitive case type. Finally, according to the result of the positioning of the sensitive case type, the invalid case needing to be destroyed is selected and a destruction operation is performed to clean up redundant or invalid data, optimize the case storage structure and update efficiency. Through this technical principle, the system realizes the dynamic management of the case based on the dependency feature, which not only guarantees the scientific monitoring of the high dependency case, but also can efficiently identify and process the invalid case, improve the accuracy of the chronic disease drug case management and the system running efficiency.

[0087] Specifically, the numerical determination process of the case-dependent characteristic value can be divided into the following stages: First, in the case data collection stage, based on the historical records, follow-up data, drug regimen and experimental test results of the case, a number of characteristic variables related to case dependence are extracted. These characteristic variables include but are not limited to case arrival interval, case recurrence frequency, treatment response fluctuation amplitude, drug adjustment times and complication occurrence probability, etc. Through standardization and normalization processing of these variables, the dimensional difference is eliminated, and the characteristic matrix under unified scale is obtained, laying a foundation for subsequent calculation of case-dependent characteristic value. Secondly, in the feature aggregation stage, a multi-dimensional weighted aggregation model is introduced, with time-dependent factor, combination-dependent factor and result-dependent factor as the core calculation elements. Among them, the time-dependent factor is used to reflect the continuity and stability of the case in the time series, the combination-dependent factor reflects the cross-correlation strength between cases, and the result-dependent factor is used to depict the influence strength of the case on the specific treatment outcome. Weight coefficients are set for each factor, and weight training is usually carried out through least squares fitting or mutual information maximization algorithm, so that the overall dependence characteristic value can maximize the representation of the comprehensive dependence relationship of the case in the multi-dimensional feature space. Thirdly, in the determination process of the case-dependent characteristic value threshold, a combination of large sample statistical analysis and experimental verification is used. Specifically, a representative case set is selected from the sample library, and the dependence characteristic value of the case set is analyzed by distribution fitting. The probability density curve of the dependence characteristic value is fitted by using normal distribution or lognormal distribution model. Subsequently, the quantile (such as median or 75% quantile) of the characteristic value in the distribution is calculated as the threshold reference. In order to further optimize the classification accuracy, combined with the results of the case screening experiment, the optimal segmentation point is calculated based on the sensitivity-specificity curve (ROC curve) to determine the final case-dependent characteristic value threshold. Finally, in the sensitive parameter calculation stage, the case arrival distribution value and analytical representation value obtained by the screening experiment are used to perform secondary fitting analysis on the high-dependence category cases. By calculating the gradient change rate and variance contribution of the case type in the dependence feature space, the case type sensitive parameter is obtained. The value of this parameter reflects the sensitivity and importance of the case type in stability, which is used to further guide the identification and destruction strategy of invalid cases.

[0088] In the specific embodiments in the present application, the above steps realize the scenario in the following manner: in the chronic disease drug management system, the dependence characteristic value of a certain case is 1.5, and the system sets the dependence characteristic value threshold to be 1.2. Since the dependence characteristic value of the case is greater than the threshold, the system classifies it into the first category, i.e., the high dependence category, and starts to dynamically monitor the case. Correspondingly, if the dependence characteristic value of another case is 0.9, it is classified into the second category and does not enter the dynamic monitoring process. Subsequently, the high dependence cases in the first category are monitored and analyzed. Assuming that in the case screening experiment, the arrival distribution value of the cases in this category is 0.6, and the analytical representation value is 0.8. The system calculates the case type sensitivity parameter based on these data, for example, obtains the sensitivity parameter 0.7 by weighted summation, so as to identify the most critical sensitive case types in this category for management and analysis. Finally, based on the positioning result of the sensitive case types, some invalid cases are identified, for example, some cases with missing data or abnormal characteristics, and a destruction operation is performed. Assuming that in the experiment, 3 invalid cases are destroyed, thereby releasing the storage space and reducing the system redundancy burden. Through this example, it can be intuitively shown how to realize the whole process operation of high dependence case management and invalid case cleaning based on case dependence category division, dynamic monitoring and sensitive parameter analysis.

[0089] The above scenario is only a preferred embodiment of the present application and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0090] Step S600: based on the result of the case screening experiment, analyze the arrival distribution value and the analytical representation value of the cases corresponding to each case type, and determine the case type sensitivity parameter according to the arrival distribution value and the analytical representation value of the cases corresponding to each case type, so as to locate the sensitive case types.

[0091] Specifically, when the arrival distribution value and the analytical representation value of the cases corresponding to each case type are analyzed based on the result of the case screening experiment, the case type sensitivity parameter is determined, and the sensitive case types are located, it includes: determining the number of analyzable cases corresponding to each case type as the arrival distribution value; determining the ratio of the number of analyzable cases to the total number of cases of the same case type as the analytical representation value; determining the ratio of the reference arrival distribution value corresponding to the case type to the arrival distribution value as the distribution value influence factor; determining the ratio of the reference analytical representation value corresponding to the case type to the analytical representation value as the analytical representation value influence factor; and determining the weighted summation value of the distribution value influence factor and the analytical representation value influence factor as the case type sensitivity parameter.

[0092] It can be understood that through the results of the case screening experiment, key indicators reflecting the stability and reliability of different case types are extracted and further quantified as sensitive parameters for positioning sensitive case types. The principle is to establish an index chain from case data to sensitivity determination, and to convert complex case distribution characteristics and analytical characteristics into calculable numerical expressions. First, in units of case types, the number of analyzable cases corresponding to each type is counted, and the number of analyzable cases is taken as the arrival distribution value of the case type. This distribution value directly reflects the degree of effective analysis of a certain type of case in the experiment, and is an important basic index for measuring case quality. Then, the ratio of the number of analyzable cases to the total number of cases of the same type is defined as the analytical representation value, which is used to reflect the availability and integrity of the case type as a whole. On this basis, the concept of benchmark value is introduced, that is, the benchmark arrival distribution value and the benchmark analytical representation value are set for each case type. By comparing the actual arrival distribution value with the benchmark distribution value, the distribution value influence factor is obtained; then by comparing the analytical representation value with the benchmark analytical representation value, the analytical representation value influence factor is obtained. The technical principle of this design is to compare the performance of each case type with the expected or standard, so as to reveal the deviation. Finally, the case type sensitivity parameter is obtained by weighted sum of the distribution value influence factor and the analytical representation value influence factor. The sensitive parameter is essentially a fusion index, which integrates the influence of case distribution and analytical characteristics on the system, and can quantitatively reflect whether a certain case type has abnormal sensitivity or potential risk. Based on this parameter, sensitive case types can be located and managed or removed, thereby improving the robustness and reliability of the case storage and analysis process.

[0093] In the specific embodiments in the present application, the above steps realize the scene in the following manner: in a certain chronic disease management system, the system needs to periodically screen and update different types of cases. It is assumed that there are three types of cases in the system: hypertension cases, diabetes cases, and cardiovascular comprehensive cases. In a case screening experiment, the system statistics found that there were 1000 cases in the diabetes cases, of which 820 could be completely parsed, and the number of parseable cases obtained was the arrival distribution value 820. At the same time, the ratio of 820 to the total number of cases 1000 obtained the parsing representation value 0.82, indicating that the cases of this type have high overall parseability. Subsequently, the system compares these data with the preset benchmark value. It is assumed that the benchmark arrival distribution value is set to 900, and the benchmark parsing representation value is set to 0.90. Through comparison, the distribution value influence factor of the diabetes case is 900 / 820≈1.10, and the parsing representation value influence factor is 0.90 / 0.82≈1.098. The system calculates the sensitivity parameter by weighting the two factors, which is about 1.099. This indicates that there is a certain difference between the distribution and parsing characteristics of this type of case and the expectation, which belongs to the sensitive case type and needs to be further monitored. For example, in the cardiovascular comprehensive case, the total number of cases is 600, but only 300 can be completely parsed. At this time, the arrival distribution value is 300, and the parsing representation value is 0.5. Assuming that the benchmark values are 500 and 0.85, the distribution value influence factor is 500 / 300≈1.67, and the parsing representation value influence factor is 0.85 / 0.5=1.70. The weighted sensitivity parameter is about 1.685, which is significantly higher than the threshold value. The system therefore locates the cardiovascular comprehensive case as a high-sensitivity category and eliminates some invalid cases in subsequent processing to avoid data redundancy affecting system performance.

[0094] It can be seen that in actual case management, dynamic filtering and monitoring can be achieved. The case type with a high sensitivity parameter will be identified and subjected to more stringent parsing and screening, ultimately realizing fine management of case data storage and updating. This not only guarantees the quality of the case library, but also improves the reliability and operating efficiency in chronic disease drug management.

[0095] The above scene is only a preferred embodiment of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0096] Step S700, according to the sensitive case type, determine the invalid cases that need to be destroyed and perform the destruction operation.

[0097] Specifically, the positioning sensitive case type, selecting invalid cases requiring destruction based on the positioning results and performing the destruction operation, includes: if the case type sensitivity parameter is greater than the preset sensitivity parameter threshold, determining that the case type is a sensitive case type; selecting a case corresponding to the sensitive case type in the un-stored cases as an invalid case; performing a destruction operation on the invalid case; wherein the destruction operation is generating a destruction instruction based on the identification information of the invalid case, and triggering the case management system to delete the invalid case through the destruction instruction.

[0098] It can be understood that the case type sensitivity parameter calculated according to the foregoing is compared with the preset sensitivity parameter threshold. If the sensitivity parameter is greater than the threshold, it means that the case type has a large deviation in the resolvability or data distribution, and it is extremely likely to produce redundant, error or unusable data in subsequent management, so it is determined as a sensitive case type. After determining the sensitive case type, the specific case corresponding to it is further located. In this process, only the un-stored cases are screened, because these cases often have not entered the core data warehouse, and direct processing can avoid pollution of invalid cases to the database. For the un-stored cases belonging to the sensitive case type, they are marked as invalid cases. In this way, a mapping link from "case type-sensitive category-invalid case" is formed, ensuring that the cleaning target is targeted and accurate. Finally, a destruction instruction is generated based on the identification information of the invalid case, and a deletion operation is performed through the case management. The destruction operation is not only the physical deletion of data, but also contains the logical level of data identification update and call control, so as to ensure that the invalid case is completely eliminated and will not be called by subsequent processes. The technical principle is to use the automatic instruction generation and triggering mechanism to realize the quick cleaning of invalid cases and avoid the omission or error caused by manual intervention.

[0099] In the specific embodiments in the present application, the above steps realize the scene in the following manner: in chronic disease management, screening experiments are regularly conducted on case data. Through analysis, it is found that the sensitive parameters of some follow-up cases are significantly higher than the preset sensitive parameter threshold. Further examination finds that these cases have multiple problems, such as some cases lacking key detection indicators, some cases repeatedly submitting data, and some cases not matching the standard in record format, resulting in a much lower number of analyzable cases than the total number of cases of this type. Based on the judgment of sensitive parameters, these cases are divided into sensitive case types. Subsequently, only cases not stored in the core database are located, and cases belonging to the sensitive type are identified and automatically marked as invalid cases. To ensure efficient and reliable cleaning, destruction instructions are generated based on the unique identification information of these invalid cases, and deletion operations are triggered through case management. Finally, these invalid cases are completely cleaned up and cannot enter the core database, thereby avoiding the interference of redundant, erroneous, or unusable data on subsequent chronic disease medication dependence analysis and efficacy evaluation. This example shows that through the automatic link of sensitive parameter judgment, invalid case identification, and destruction, dynamic optimization management of case data can be achieved, ensuring the accuracy of case storage and the stability of operation.

[0100] The above scene is only a preferred embodiment of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0101] In the above embodiment, by acquiring the medication case set of chronic disease patients within a preset time range, and extracting key information and constructing feature sequence of the case, the analysis of case history use rule and effectiveness verification are realized. This step can eliminate non-representative or obviously abnormal data in advance, ensuring the accuracy and reliability of subsequent management from the source, avoiding the analysis deviation caused by invalid data in traditional methods. On this basis, by introducing case screening experiment and construction of adjacent case group, using processing time deviation and analysis result to determine time-dependent representation parameter, the dependence of the case in time dimension is further revealed. This design can dynamically depict the time-sensitive features of the case, providing quantitative indicators for rule identification and time management in the medication process of chronic diseases, and improving the adaptability of the system to complex case changes. At the same time, by analyzing the number of unanalyzable cases in the screening experiment, combining with the number of damaged cases, the combined dependence representation parameter is calculated to realize the dependence measurement of the case in the combination dimension. This mechanism effectively avoids the one-sidedness caused by relying on a single dimension, and can fully reflect the characteristics of the case in data integrity and combination analyzability, thereby providing a more solid foundation for case classification and management. Further, by weighting fusion of time-dependent representation parameter and combined dependence representation parameter, the case dependence feature value is calculated, and the case dependence category is divided accordingly. With this division mechanism, the system can realize differentiated storage and update management of cases: for high dependence category cases, dynamic management and monitoring are carried out; for low dependence category cases, simplified storage strategy is adopted. This hierarchical management effectively improves the utilization rate and update efficiency of data storage, avoiding resource waste. Finally, by calculating the case type sensitivity parameter based on the case screening experiment result, the sensitive case type is located, and invalid cases are automatically identified and destroyed based on the positioning result. This step realizes the dynamic cleaning of invalid data, ensures the high quality and high purity of the core database, and reduces the interference of redundant data on system operation and subsequent analysis. Through the automatic destruction mechanism, manual intervention and potential errors are reduced, making the case management process more efficient and intelligent.

[0102] In another preferred mode based on the above embodiment, as shown in Figure 2 the present embodiment provides a medication case management system for chronic disease patients, comprising a case acquisition module, a case analysis module, a case screening module, a case dependence calculation module, a case destruction module and a case storage module.

[0103] In particular, the case acquisition module is configured to acquire a set of medication cases of chronic disease patients within a preset time range, extract key information fields of each case, and construct a case feature sequence based on the key information fields; the case analysis module is configured to analyze the cases and determine the number of unanalyzable cases; the case analysis module is electrically connected to the case acquisition module, and the case analysis module is configured to analyze the cases and determine the number of unanalyzable cases; the case screening module is electrically connected to the case analysis module, and the case screening module is configured to perform a case screening experiment, record a processing time deviation corresponding to each case, and screen a neighboring case group; the case dependency calculation module is electrically connected to the case screening module, and the case dependency calculation module is configured to calculate a case dependency feature value based on a time dependency representation parameter and a combination dependency representation parameter; the case destruction module is electrically connected to the case dependency calculation module, and the case destruction module is configured to locate a sensitive case type based on a case type sensitivity parameter and select an invalid case that needs to be destroyed; and the case storage module is electrically connected to the case destruction module, and the case storage module is configured to store valid cases and update a case library.

[0104] It can be understood that the medication case management system and method for chronic disease patients in each of the above embodiments have the same beneficial effects, which will not be described again.

[0105] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can adopt a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program codes.

[0106] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems) and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The device that implements the functions specified in one flow or multiple flows and / or blocks. Figure 1 The device that implements the functions specified in one flow or multiple flows and / or blocks.

[0107] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks or multiple blocks.

[0108] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks or multiple blocks.

[0109] Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, rather than limiting the same. Even though the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced equivalently, and any modification or replacement without departing from the spirit and scope of the present application should be covered within the protection scope of the claims of the present application.

Claims

1. A method for medication case management for a patient with a chronic disease, characterized by, The method comprises the following steps: acquiring a set of chronic disease patient medication cases within a preset time range, extracting key information fields of each case, constructing a case feature sequence based on the key information fields, recording the number of occurrences of the same case feature sequence to analyze the historical use rules of the cases, and verifying the effectiveness of the cases; based on the effectiveness verification result, performing a case screening experiment, randomly selecting a predetermined proportion of cases in the case set, recording the processing time deviation of each case, screening a group of adjacent cases, and determining a time-dependent representation parameter based on the analysis result of the group of adjacent cases; analyzing the cases that have undergone the case screening experiment, determining the number of unanalyzable cases, and obtaining a combination-dependent representation parameter; calculating a case-dependent feature value based on the time-dependent representation parameter and the combination-dependent representation parameter, and dividing the case-dependent categories; based on the division result of the case-dependent categories, controlling the storage and update process of the cases, including monitoring the results of the case screening experiment; based on the results of the case screening experiment, analyzing the arrival distribution value and the analysis representation value of the cases corresponding to each case type, and determining a case type sensitivity parameter according to the arrival distribution value and the analysis representation value of the cases corresponding to each case type to locate sensitive case types; according to the sensitive case types, determining invalid cases that need to be destroyed and performing a destruction operation; wherein: the time-dependent representation parameter is specifically determined by the ratio of the number of time-dependent case groups to the total number of cases, and the time-dependent case group refers to a group of adjacent cases that can be completely analyzed; the combination-dependent representation parameter is specifically the difference between the number of unanalyzable cases and the number of damaged cases, and the ratio of the number of unanalyzable cases; the case-dependent feature value is specifically the weighted sum of the time-dependent influence factor and the combination-dependent influence factor; the arrival distribution value is specifically the number of analyzable cases corresponding to each case type; the analysis representation value is specifically the ratio of the number of analyzable cases to the total number of cases of the same case type; the case type sensitivity parameter is specifically the weighted sum of the distribution value influence factor and the analysis representation value influence factor; the processing time deviation is specifically the deviation between the actual processing time and the expected processing time of each case in the processing process; the adjacent case group is specifically a combination of cases that meet the processing time deviation condition and are close in time or similar in characteristics.

2. The medication case management method for chronic disease patients according to claim 1, wherein When analyzing the historical use rules of the cases and verifying the effectiveness of the cases, the following steps are included: record the number of similar use behaviors in the historical use process of the cases; determine the ratio of the number of similar use behaviors to the total number of uses to obtain the similar use behavior ratio; if the similar use behavior ratio is greater than a predetermined similar use behavior ratio threshold, the case is determined to be valid; wherein, the similar use behavior needs to meet that the case feature sequence corresponding to each case in the historical use process is consistent with the case feature sequence corresponding to each case in any other historical use process.

3. The medication case management method for chronic disease patients according to claim 2, wherein, When performing a case screening experiment based on the effectiveness verification result, the following steps are included: If the case is valid, a case screening experiment needs to be performed, a predetermined proportion of cases in the case set is randomly selected, the processing time deviation corresponding to each case is recorded, a neighboring case group is screened, and the time-dependent characterization parameter is determined based on the analysis result of the neighboring case group; If the case is not valid, the case screening experiment does not need to be performed.

4. The medication case management method for chronic patients according to claim 3, wherein When screening the neighboring case group and determining the time-dependent characterization parameter based on the analysis result of the neighboring case group, the following steps are included: Determine the processing time deviation corresponding to each case; If there is a case that meets the deviation condition, combine the case with the neighboring case to form a neighboring case group; Determine whether each case in the neighboring case group can be completely analyzed to determine the time-dependent case group; Determine the ratio of the number of time-dependent case groups to the total number of cases as the time-dependent characterization parameter; Wherein, the deviation condition is that the deviation time corresponding to the case is greater than the predetermined time deviation, and the complete analysis is that each case in the neighboring case group can be completely analyzed.

5. The medication case management method for chronic disease patients according to claim 4, wherein, When calculating the case-dependent feature value based on the time-dependent characterization parameter and the combination-dependent characterization parameter, the following steps are included: Determine the ratio of the time-dependent characterization parameter to the baseline time-dependent characterization parameter as the time-dependent influence factor; Determine the ratio of the combination-dependent characterization parameter to the baseline combination-dependent characterization parameter as the combination-dependent influence factor; Determine the weighted sum of the time-dependent influence factor and the combination-dependent influence factor as the case-dependent feature value.

6. The medication case management method for chronic disease patients according to claim 5, wherein, In response to the division result of the case dependency category, the storage and update of the case are controlled, including: If the case-dependent feature value is greater than the case-dependent feature value threshold, the case dependency category is divided into the first category, the results of the case screening experiment are monitored, the arrival distribution value and the analysis characterization value of the cases corresponding to each case type are analyzed based on the results of the case screening experiment, the case type sensitive parameter is calculated to locate the sensitive case type, the invalid case that needs to be destroyed is selected based on the positioning result, and the destruction operation is performed; If the case-dependent feature value is less than or equal to the case-dependent feature value threshold, the case dependency category is divided into the second category.

7. The medication case management method for chronic disease patients according to claim 6, wherein, When determining the case type sensitive parameter based on the arrival distribution value and the analysis characterization value of the cases corresponding to each case type based on the results of the case screening experiment, the following steps are included: Determine the number of analyzable cases corresponding to each case type as the arrival distribution value; Determine the ratio of the number of analyzable cases to the total number of cases of the same case type as the analysis characterization value; Determine the ratio of the baseline arrival distribution value corresponding to the case type to the arrival distribution value as the distribution value influence factor; Determine the ratio of the baseline analysis characterization value corresponding to the case type to the analysis characterization value as the analysis characterization value influence factor; Determine the weighted sum of the distribution value influence factor and the analysis characterization value influence factor as the case type sensitive parameter.

8. The medication case management method for chronic patients according to claim 7, wherein, When locating the sensitive case type and performing the destruction operation on the invalid case selected based on the positioning result, the following steps are included: If the case type sensitive parameter is greater than the preset sensitive parameter threshold, the case type is determined to be a sensitive case type; Select the case corresponding to the sensitive case type from the stored cases as the invalid case; Perform the destruction operation on the invalid case; The destroying operation is generating a destroying instruction based on the identification information of the invalid case, and triggering the case management system to delete the invalid case through the destroying instruction.

9. A medication case management system for a chronic disease patient, which is adapted to a medication case management method for a chronic disease patient according to any one of claims 1 to 8, characterized in that, Comprise: The case acquisition module is used for acquiring a set of chronic disease patient medication cases within a preset time range, extracting key information fields of each case, and constructing a case feature sequence based on the key information fields; The case analysis module is used for analyzing the case and determining the number of unanalyzable cases; The case screening module is used for performing a case screening experiment, recording the processing time deviation corresponding to each case, and screening adjacent case groups; The case dependence calculation module is used for calculating case dependence feature values based on time dependence representation parameters and combination dependence representation parameters; The case destroying module is used for positioning sensitive case types based on case type sensitivity parameters, and selecting invalid cases that need to be destroyed; The case storage module is used for storing valid cases and updating the case library.

Images