Drug safety risk evaluation index system construction method and system

Abstract

The invention discloses a drug safety risk evaluation index system construction method and system, belongs to the technical field of data processing, and aims to solve the technical problem of how to construct a comprehensive, accurate and effective drug safety risk evaluation system. Comprising the following steps: constructing a medicine safety evaluation index system for medicine safety evaluation, namely a medicine safety supervision capability index system, a medicine safety supervision social co-treatment index system and a medicine life cycle safety risk index system; for each drug safety evaluation index system, setting hierarchical indexes of the drug safety evaluation index system; the index weight of each hierarchical index is calculated based on an analytic hierarchy process, a Delphi method and a principal component analysis method.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for constructing a drug safety risk evaluation index system. Background Art

[0002] Drug safety research focuses on the following key areas: First, the drug development and approval process, focusing on the entire process from laboratory to market, including clinical trials, safety and efficacy evaluation and regulatory approval. Second, the quality control of drug production, involving raw material control, standardization of production processes and quality inspection. In addition, the study of drug regulatory policies and regulations is also very important, focusing on evaluating the effectiveness of drug regulatory policies and ensuring drug safety through legal means. Market supervision and risk management are also key points, including risk management measures such as monitoring of drugs on the market, adverse reaction reporting and drug recalls. Patient medication safety and education are also important parts of the research, focusing on improving public awareness of safe drug use. The issue of antimicrobial resistance, especially in light of the challenges posed by the abuse of antibiotics, has become an important research direction. Drug supply chain management has also received attention, covering the entire process from production to end users. Finally, international cooperation and global drug safety have become increasingly important in the context of globalization, involving the unification of transnational regulatory standards and international market monitoring. These research areas comprehensively cover all aspects of drug safety, and they are constantly developing and changing with technological advances and the development of globalization. The research on drug safety risk assessment index has the following main deficiencies:

[0003] (1) Insufficient multi-dimensional integration: Existing research often focuses on a single area, such as government supervision, enterprise production quality control, or public drug safety, but lacks discussion on how to effectively integrate risk factors in these different areas into a unified assessment framework. This results in the assessment of drug safety risks being often one-sided and unable to fully reflect the risks of the entire chain of drugs from research and development, production, circulation to use.

[0004] (2) Lack of dynamic monitoring mechanism: Drug safety risk is a dynamic process that requires real-time or regularly updated monitoring mechanisms to track and evaluate. In existing research, there is insufficient discussion on the construction of such a dynamic monitoring mechanism, which often remains at the static risk assessment level.

[0005] (3) Insufficient interdisciplinary integration: Drug safety risk assessment is a complex interdisciplinary issue involving multiple fields such as medicine, pharmacy, public administration, and economics. Current research often lacks an interdisciplinary perspective, and there is limited discussion on how to integrate theories and methods from different disciplines to more comprehensively assess drug safety risks.

[0006] (4) Data acquisition and application issues: In drug safety risk assessment, data acquisition and application are key. In current research, there is little discussion on how to obtain accurate and comprehensive data and how to use these data for effective analysis. Especially in the context of the big data era, how to use big data technology to improve the accuracy and efficiency of drug safety risk assessment is an issue worthy of in-depth study.

[0007] (5) Insufficient stakeholder participation: Drug safety risk assessment is not only a technical issue, but also a policy issue. Existing research has not explored how to better involve stakeholders such as the government, enterprises, and consumers in the drug safety risk assessment process, especially the lack of in-depth research on public participation mechanisms and consumer education.

[0008] (6) Insufficient international comparison and experience learning: In the context of globalization, drug safety is a transnational issue. However, existing research has not done enough to compare drug safety risk assessment systems in different countries and learn from international experience.

[0009] In summary, the current research on the drug safety risk assessment indicator system has deficiencies in terms of multi-dimensional integration, dynamic monitoring, interdisciplinary integration, data application, stakeholder participation, and international experience reference. In the face of these challenges, research needs to pay more attention to comprehensiveness, dynamism, interdisciplinaryness, and practicality, in order to build a more comprehensive, accurate, and effective drug safety risk assessment system.

[0010] How to build a comprehensive, accurate and effective drug safety risk assessment system is a technical problem that needs to be solved. Summary of the invention

[0011] The technical task of the present invention is to address the above shortcomings and provide a method and system for constructing a drug safety risk evaluation index system to solve the technical problem of how to construct a comprehensive, accurate and effective drug safety risk assessment system.

[0012] In a first aspect, the present invention provides a method for constructing a drug safety risk assessment index system, comprising the following steps:

[0013] For drug safety evaluation, a drug safety evaluation index system is constructed, which includes the drug safety supervision capability index system, the drug safety supervision social co-governance index system, and the drug life cycle safety risk index system;

[0014] For each drug safety evaluation index system, set the hierarchical indicators of the drug safety evaluation index system;

[0015] The weights of indicators at each level are calculated based on the hierarchical analysis method, Delphi method and principal component analysis.

[0016] Preferably, each drug safety evaluation index system is set with multiple first-level indicators, each first-level indicator is set with multiple second-level indicators, and each second-level indicator is set with multiple third-level indicators.

[0017] As a preference, for the drug safety supervision capability index system and the drug safety supervision social co-governance index system, the indicator weights of each level indicator are calculated based on the hierarchical analysis method and the Delphi method, including the following steps:

[0018] M100, construct a judgment matrix, generate a regular questionnaire for the indicator system, and compare indicators at different levels in pairs using a scale of 1 to 9, as shown below:

[0019] 1: The i-th indicator is as important as the j-th indicator

[0020] 3: The i-th indicator is slightly more important than the j-th indicator

[0021] 5: The i-th index is greater than the j-th index.

[0022] 7: The i-th indicator is more important than the j-th indicator

[0023] 9: The i-th indicator is definitely more important than the j-th indicator

[0024] 1 / 3: The i-th indicator is slightly less important than the j-th indicator

[0025] 1 / 5: The i-th indicator is less important than the j-th indicator

[0026] 1 / 7: The i-th indicator is much less important than the j-th indicator

[0027] 1 / 9: The i-th indicator is definitely less important than the j-th indicator

[0028] Among them, 2, 4, 6, 8, 1 / 2, 1 / 4, 1 / 6, 1 / 8: indicate that the importance is between 1 and 3, 2 and 5, ...;

[0029] M200, send the questionnaire to the experts for filling;

[0030] M300, collect the questionnaire and import it into the AHP analysis software to calculate the importance ranking of indicators at each level. When calculating, find the eigenvector w corresponding to the maximum eigenvalue max according to the judgment matrix. The calculation formula is as follows:

[0031] P ω P ω =λ max ω,

[0032] M400, perform consistency check on the judgment matrix, the test formula is as follows:

[0033]

[0034] Among them, CR represents the random consistency ratio of the judgment matrix, CI represents the consistency index of the judgment matrix, and RI represents the average random consistency index of the judgment matrix;

[0035] M500, import the expert survey form into the AHP software to check the incompleteness and inconsistency of the matrix, and prompt the experts for abnormal situations, including incomplete and inconsistent matrix data. For missing items in the judgment matrix, communicate with the experts for confirmation and then correct them. For inconsistencies in the matrix, use the algorithm provided by the Maishi AHP software system to automatically adjust them to consistency when making decisions. The weights of experts are all set to 1;

[0036] M600, after generating the weights of each three-level indicator based on AHP software, the normalized scores of the corresponding three-level indicators are weighted and added and multiplied by 100 to obtain the final index, whose value is between 0-100.

[0037] Preferably, for the drug life cycle safety risk index system, the indicator weight of each level indicator is calculated based on the principal component analysis method, including the following steps:

[0038] L100 input: m samples, a data set with n features, that is, sample data X = [X1, X2, ..., X m ], the target dimension of dimension reduction is k, and the sample set is denoted as matrix X, which can be expressed as:

[0039]

[0040] L200 output: the sample set after dimensionality reduction Y = [y1, y2, ..., y m ].

[0041] L300 calculates the covariance matrix, including the following operations:

[0042] Step 1: Decentralize the matrix to obtain a new matrix X, that is, zero-mean each column:

[0043]

[0044] Step 2: Calculate the covariance moment of the decentralized matrix X That is, an n*n-order matrix;

[0045] Step 3: Perform eigendecomposition on the covariance matrix C to find the eigenvalues ​​and corresponding eigenvectors of the covariance matrix, i.e. cυ k =λ k v k ;

[0046] Step 4: Arrange the eigenvectors in descending order from left to right according to the corresponding eigenvalues, and take the first k columns to form the matrix W, that is, an n*k-order matrix;

[0047] Step 5: Calculate the sample features after dimension reduction to k dimensions through Y=XW, that is, an m*k-order matrix.

[0048] In a second aspect, the present invention provides a drug safety risk assessment index system construction system, including a system construction module, a hierarchical index setting module, and an index weight calculation module;

[0049] The system construction module is used to perform the following: for drug safety evaluation, construct a drug safety evaluation index system, which includes a drug safety supervision capability index system, a drug safety supervision social co-governance index system, and a drug life cycle safety risk index system;

[0050] The hierarchical index setting module is used to perform the following: for each drug safety evaluation index system, set the hierarchical index of the drug safety evaluation index system;

[0051] The indicator weight calculation module is used to perform the following: calculating the indicator weights of indicators at each level based on the hierarchical analysis method, the Delphi method and the principal component analysis method.

[0052] Preferably, each drug safety evaluation index system is set with multiple first-level indicators, each first-level indicator is set with multiple second-level indicators, and each second-level indicator is set with multiple third-level indicators.

[0053] Preferably, for the drug safety supervision capability index system and the drug safety supervision social co-governance index system, the indicator weight calculation module is used to calculate the indicator weight of each level indicator based on the hierarchical analysis method and the Delphi method, including the following steps:

[0054] M100, construct a judgment matrix, generate a regular questionnaire for the indicator system, and compare indicators at different levels in pairs using a scale of 1 to 9, as shown below:

[0055] 1: The i-th indicator is as important as the j-th indicator

[0056] 3: The i-th indicator is slightly more important than the j-th indicator

[0057] 5: The i-th index is greater than the j-th index.

[0058] 7: The i-th indicator is more important than the j-th indicator

[0059] 9: The i-th indicator is definitely more important than the j-th indicator

[0060] 1 / 3: The i-th indicator is slightly less important than the j-th indicator

[0061] 1 / 5: The i-th indicator is less important than the j-th indicator

[0062] 1 / 7: The i-th indicator is much less important than the j-th indicator

[0063] 1 / 9: The i-th indicator is definitely less important than the j-th indicator

[0064] Among them, 2, 4, 6, 8, 1 / 2, 1 / 4, 1 / 6, 1 / 8: indicate that the importance is between 1 and 3, 2 and 5, ...;

[0065] M200, send the questionnaire to the experts for filling;

[0066] M300, collect the questionnaire and import it into the AHP analysis software to calculate the importance ranking of indicators at each level. When calculating, find the eigenvector w corresponding to the maximum eigenvalue max according to the judgment matrix. The calculation formula is as follows:

[0067] P ω P ω =λ max ω,

[0068] M400, perform consistency check on the judgment matrix, the test formula is as follows:

[0069]

[0070] Among them, CR represents the random consistency ratio of the judgment matrix, CI represents the consistency index of the judgment matrix, and RI represents the average random consistency index of the judgment matrix;

[0071] M500, import the expert survey form into the AHP software to check the incompleteness and inconsistency of the matrix, and prompt the experts for abnormal situations, including incomplete and inconsistent matrix data. For missing items in the judgment matrix, communicate with the experts for confirmation and then correct them. For inconsistencies in the matrix, use the algorithm provided by the Maishi AHP software system to automatically adjust them to consistency when making decisions. The weights of experts are all set to 1;

[0072] M600, after generating the weights of each three-level indicator based on AHP software, the normalized scores of the corresponding three-level indicators are weighted and added and multiplied by 100 to obtain the final index, whose value is between 0-100.

[0073] Preferably, for the drug life cycle safety risk index system, the indicator weight calculation module is used to calculate the indicator weight of each level indicator based on the principal component analysis method, including the following steps:

[0074] L100 input: m samples, a data set with n features, that is, sample data X = [X1, X2, ..., Xm ], the target dimension of dimension reduction is k, and the sample set is denoted as matrix X, which can be expressed as:

[0075]

[0076] L200 output: the sample set after dimensionality reduction Y = [y1, y2, ..., y m ].

[0077] L300 calculates the covariance matrix, including the following operations:

[0078] Step 1: Decentralize the matrix to obtain a new matrix X, that is, zero-mean each column:

[0079]

[0080] Step 2: Calculate the covariance moment of the decentralized matrix X That is, an n*n-order matrix;

[0081] Step 3: Perform eigendecomposition on the covariance matrix C to find the eigenvalues ​​and corresponding eigenvectors of the covariance matrix, i.e. cυ k =λ k v k ;

[0082] Step 4: Arrange the eigenvectors in descending order from left to right according to the corresponding eigenvalues, and take the first k columns to form the matrix W, that is, an n*k-order matrix;

[0083] Step 5: Calculate the sample features after dimension reduction to k dimensions through Y=XW, that is, an m*k-order matrix.

[0084] The method and system for constructing a drug safety risk assessment index system of the present invention have the following advantages: relying on the data resource system of the drug supervision big data center, using drug, medical device and cosmetic supervision-related data, processing data through comprehensive processing, superposition, statistics and other models, performing data sampling, weighting and calculation, establishing a drug safety assessment index system that truly reflects the drug safety, comprehensively analyzing the drug safety situation, and being able to display the drug safety assessment index status by region, variety and theme, providing an important basis for precise drug safety supervision. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0086] The present invention is further described below in conjunction with the accompanying drawings.

[0087] Figure 1 This is a flowchart of a method for constructing a drug safety risk assessment index system in Example 1. DETAILED DESCRIPTION

[0088] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it. However, the embodiments are not intended to limit the present invention. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments may be combined with each other.

[0089] The embodiments of the present invention provide a method and system for constructing a drug safety risk evaluation index system, which are used to solve the technical problem of how to construct a comprehensive, accurate and effective drug safety risk evaluation system.

[0090] Embodiment 1:

[0091] The present invention provides a method for constructing a drug safety risk assessment index system, comprising the following steps:

[0092] (1) For drug safety evaluation, a drug safety evaluation index system is constructed, including the drug safety supervision capacity index system, the drug safety supervision social co-governance index system, and the drug life cycle safety risk index system;

[0093] (2) for each drug safety evaluation index system, setting the hierarchical indicators of the drug safety evaluation index system;

[0094] (3) The weights of indicators at each level are calculated based on the hierarchical analysis method, Delphi method and principal component analysis method.

[0095] In this embodiment, the drug regulatory capacity evaluation index is divided into seven categories of first-level indicators, namely, basic capacity, ability to implement supervision and management responsibilities, ability to perform duties, organizational leadership capacity, supervision capacity, decision-making and innovation capacity, and food and drug safety committee reference base, as well as several second-level indicators.

[0096] The drug safety supervision capability index system consists of seven first-level indicators: basic capabilities, job performance capabilities, organizational leadership capabilities, supervision and management responsibility implementation capabilities, supervision capacity building, decision-making and innovation capabilities, and the food and drug safety committee reference base. Each first-level indicator has multiple second-level indicators, and each second-level indicator has multiple third-level indicators, forming a multi-level indicator system.

[0097] The drug production safety supervision index consists of six first-level indicators: government information disclosure quality, supervision information, government service quality, social entity participation satisfaction, public opinion supervision quality, and food and drug safety committee reference base. Each first-level indicator has multiple second-level indicators, and each second-level indicator has multiple third-level indicators, forming a multi-level indicator system.

[0098] The drug life cycle safety risk index system consists of seven first-level indicators: drug production registration and review, drug production supervision and inspection, drug production supervision sampling, drug production administrative law enforcement, drug production complaints and reports, drug production adverse reactions, and drug production Food Safety Committee scoring. Each first-level indicator has multiple second-level indicators, and each second-level indicator has multiple third-level indicators, forming a multi-level indicator system.

[0099] In this embodiment, for the drug safety supervision capability index system and the drug safety supervision social co-governance index system, the index weight of each level index is calculated based on the hierarchical analysis method and the Delphi method, including the following steps:

[0100] M100, construct a judgment matrix, generate a regular questionnaire for the indicator system, and compare indicators at different levels in pairs using a scale of 1 to 9, as shown below:

[0101] 1: The i-th indicator is as important as the j-th indicator

[0102] 3: The i-th indicator is slightly more important than the j-th indicator

[0103] 5: The i-th index is greater than the j-th index.

[0104] 7: The i-th indicator is more important than the j-th indicator

[0105] 9: The i-th indicator is definitely more important than the j-th indicator

[0106] 1 / 3: The i-th indicator is slightly less important than the j-th indicator

[0107] 1 / 5: The i-th indicator is less important than the j-th indicator

[0108] 1 / 7: The i-th indicator is much less important than the j-th indicator

[0109] 1 / 9: The i-th indicator is definitely less important than the j-th indicator

[0110] Among them, 2, 4, 6, 8, 1 / 2, 1 / 4, 1 / 6, 1 / 8: indicate that the importance is between 1 and 3, 2 and 5, ...;

[0111] M200, send the questionnaire to the experts for filling;

[0112] M300, collect the questionnaire and import it into the AHP analysis software to calculate the importance ranking of indicators at each level. When calculating, find the eigenvector w corresponding to the maximum eigenvalue max according to the judgment matrix. The calculation formula is as follows:

[0113] P ω P ω =λ max ω,

[0114] M400, perform consistency check on the judgment matrix, the test formula is as follows:

[0115]

[0116] Among them, CR represents the random consistency ratio of the judgment matrix, CI represents the consistency index of the judgment matrix, and RI represents the average random consistency index of the judgment matrix;

[0117] M500, import the expert survey form into the AHP software to check the incompleteness and inconsistency of the matrix, and prompt the experts for abnormal situations, including incomplete and inconsistent matrix data. For missing items in the judgment matrix, communicate with the experts for confirmation and then correct them. For inconsistencies in the matrix, use the algorithm provided by the Maishi AHP software system to automatically adjust them to consistency when making decisions. The weights of experts are all set to 1;

[0118] M600, after generating the weights of each three-level indicator based on AHP software, the normalized scores of the corresponding three-level indicators are weighted and added and multiplied by 100 to obtain the final index, whose value is between 0-100.

[0119] For the drug life cycle safety risk index system, the indicator weights of each level indicator are calculated based on the principal component analysis method, including the following steps:

[0120] L100 input: m samples, a data set with n features, that is, sample data X = [X1, X2, ..., X m ], the target dimension of dimension reduction is k, and the sample set is denoted as matrix X, which can be expressed as:

[0121]

[0122] L200 output: the sample set after dimensionality reduction Y = [y1, y2, ..., y m ].

[0123] L300 calculates the covariance matrix, including the following operations:

[0124] Step 1: Decentralize the matrix to obtain a new matrix X, that is, zero-mean each column:

[0125]

[0126] Step 2: Calculate the covariance moment of the decentralized matrix X That is, an n*n-order matrix;

[0127] Step 3: Perform eigendecomposition on the covariance matrix C to find the eigenvalues ​​and corresponding eigenvectors of the covariance matrix, i.e. cυ k =λ k v k ;

[0128] Step 4: Arrange the eigenvectors in descending order from left to right according to the corresponding eigenvalues, and take the first k columns to form the matrix W, that is, an n*k-order matrix;

[0129] Step 5: Calculate the sample features after dimension reduction to k dimensions through Y=XW, that is, an m*k-order matrix.

[0130] The current research on the drug safety risk assessment indicator system has deficiencies in terms of multi-dimensional integration, dynamic monitoring, interdisciplinary integration, data application, stakeholder participation, and international experience reference. In the face of these challenges, research needs to pay more attention to comprehensiveness, dynamism, interdisciplinaryness, and practicality, in order to build a more comprehensive, accurate, and effective drug safety risk assessment system.

[0131] Embodiment 2:

[0132] The present invention provides a drug safety risk evaluation index system construction system, which includes a system construction module, a hierarchical index setting module and an index weight calculation module.

[0133] The system construction module is used to perform the following: For drug safety evaluation, a drug safety evaluation index system is constructed, which includes the drug safety supervision capability index system, the drug safety supervision social co-governance index system and the drug life cycle safety risk index system.

[0134] The hierarchical index setting module is used to perform the following: for each drug safety evaluation index system, set the hierarchical index of the drug safety evaluation index system.

[0135] The indicator weight calculation module is used to perform the following: calculate the indicator weights of indicators at each level based on the hierarchical analysis method, the Delphi method and the principal component analysis method.

[0136] In this embodiment, the drug regulatory capacity evaluation index is divided into seven categories of first-level indicators, namely, basic capacity, ability to implement supervision and management responsibilities, ability to perform duties, organizational leadership capacity, supervision capacity, decision-making and innovation capacity, and food and drug safety committee reference base, as well as several second-level indicators.

[0137] The drug safety supervision capability index system consists of seven first-level indicators: basic capabilities, job performance capabilities, organizational leadership capabilities, supervision and management responsibility implementation capabilities, supervision capacity building, decision-making and innovation capabilities, and the food and drug safety committee reference base. Each first-level indicator has multiple second-level indicators, and each second-level indicator has multiple third-level indicators, forming a multi-level indicator system.

[0138] The drug production safety supervision index consists of six first-level indicators: government information disclosure quality, supervision information, government service quality, social entity participation satisfaction, public opinion supervision quality, and food and drug safety committee reference base. Each first-level indicator has multiple second-level indicators, and each second-level indicator has multiple third-level indicators, forming a multi-level indicator system.

[0139] The drug life cycle safety risk index system consists of seven first-level indicators: drug production registration and review, drug production supervision and inspection, drug production supervision sampling, drug production administrative law enforcement, drug production complaints and reports, drug production adverse reactions, and drug production Food Safety Committee scoring. Each first-level indicator has multiple second-level indicators, and each second-level indicator has multiple third-level indicators, forming a multi-level indicator system.

[0140] In this embodiment, for the drug safety supervision capability index system and the drug safety supervision social co-governance index system, the index weight of each level index is calculated based on the hierarchical analysis method and the Delphi method, including the following steps:

[0141] M100, construct a judgment matrix, generate a regular questionnaire for the indicator system, and compare indicators at different levels in pairs using a scale of 1 to 9, as shown below:

[0142] 1: The i-th indicator is as important as the j-th indicator

[0143] 3: The i-th indicator is slightly more important than the j-th indicator

[0144] 5: The i-th index is greater than the j-th index.

[0145] 7: The i-th indicator is more important than the j-th indicator

[0146] 9: The i-th indicator is definitely more important than the j-th indicator

[0147] 1 / 3: The i-th indicator is slightly less important than the j-th indicator

[0148] 1 / 5: The i-th indicator is less important than the j-th indicator

[0149] 1 / 7: The i-th indicator is much less important than the j-th indicator

[0150] 1 / 9: The i-th indicator is definitely less important than the j-th indicator

[0151] Among them, 2, 4, 6, 8, 1 / 2, 1 / 4, 1 / 6, 1 / 8: indicate that the importance is between 1 and 3, 2 and 5, ...;

[0152] M200, send the questionnaire to the experts for filling;

[0153] M300, collect the questionnaire and import it into the AHP analysis software to calculate the importance ranking of indicators at each level. When calculating, find the eigenvector w corresponding to the maximum eigenvalue max according to the judgment matrix. The calculation formula is as follows:

[0154] P ω P ω =λ max ω,

[0155] M400, perform consistency check on the judgment matrix, the test formula is as follows:

[0156]

[0157] Among them, CR represents the random consistency ratio of the judgment matrix, CI represents the consistency index of the judgment matrix, and RI represents the average random consistency index of the judgment matrix;

[0158] M500, import the expert survey form into the AHP software to check the incompleteness and inconsistency of the matrix, and prompt the experts for abnormal situations, including incomplete and inconsistent matrix data. For missing items in the judgment matrix, communicate with the experts for confirmation and then correct them. For inconsistencies in the matrix, use the algorithm provided by the Maishi AHP software system to automatically adjust them to consistency when making decisions. The weights of experts are all set to 1;

[0159] M600, after generating the weights of each three-level indicator based on AHP software, the normalized scores of the corresponding three-level indicators are weighted and added and multiplied by 100 to obtain the final index, whose value is between 0-100.

[0160] For the drug life cycle safety risk index system, the indicator weights of each level indicator are calculated based on the principal component analysis method, including the following steps:

[0161] L100 input: m samples, a data set with n features, that is, sample data X = [X1, X2, ..., X m ], the target dimension of dimension reduction is k, and the sample set is denoted as matrix X, which can be expressed as:

[0162]

[0163] L200 output: the sample set after dimensionality reduction Y = [y1, y2, ..., y m ].

[0164] L300 calculates the covariance matrix, including the following operations:

[0165] Step 1: Decentralize the matrix to obtain a new matrix X, that is, zero-mean each column:

[0166]

[0167] Step 2: Calculate the covariance moment of the decentralized matrix X That is, an n*n-order matrix;

[0168] Step 3: Perform eigendecomposition on the covariance matrix C to find the eigenvalues ​​and corresponding eigenvectors of the covariance matrix, i.e. cυ k =λ k v k ;

[0169] Step 4: Arrange the eigenvectors in descending order from left to right according to the corresponding eigenvalues, and take the first k columns to form the matrix W, that is, an n*k-order matrix;

[0170] Step 5: Calculate the sample features after dimension reduction to k dimensions through Y=XW, that is, an m*k-order matrix.

[0171] The system of this embodiment can execute the method disclosed in Example 1 to realize the construction of a drug safety risk evaluation index system.

[0172] The present invention has been shown and described in detail above through the accompanying drawings and preferred embodiments. However, the present invention is not limited to these disclosed embodiments. Based on the above multiple embodiments, those skilled in the art can know that the means in the above different embodiments can be combined to obtain more embodiments of the present invention, and these embodiments are also within the protection scope of the present invention.

Claims

1. A method for constructing a drug safety risk assessment index system, characterized in that: The steps include: For drug safety evaluation, a drug safety evaluation index system is constructed, which includes the drug safety supervision capability index system, the drug safety supervision social co-governance index system, and the drug life cycle safety risk index system; For each drug safety evaluation index system, set the hierarchical indicators of the drug safety evaluation index system; The weights of indicators at each level are calculated based on the hierarchical analysis method, Delphi method and principal component analysis.

2. The method for constructing a drug safety risk assessment index system according to claim 1, characterized in that: Each drug safety evaluation index system is set with multiple first-level indicators, each first-level indicator is set with multiple second-level indicators, and each second-level indicator is set with multiple third-level indicators.

3. The method for constructing a drug safety risk assessment index system according to claim 1, characterized in that: For the drug safety supervision capability index system and the drug safety supervision social co-governance index system, the indicator weights of each level indicator are calculated based on the hierarchical analysis method and the Delphi method, including the following steps: M100, construct a judgment matrix, generate a regular questionnaire for the indicator system, and compare indicators at different levels in pairs using a scale of 1 to 9, as shown below: 1: The i-th indicator is as important as the j-th indicator 3: The i-th indicator is slightly more important than the j-th indicator 5: The i-th index is greater than the j-th index. 7: The i-th indicator is more important than the j-th indicator 9: The i-th indicator is definitely more important than the j-th indicator 1 / 3: The i-th indicator is slightly less important than the j-th indicator 1 / 5: The i-th indicator is less important than the j-th indicator 1 / 7: The i-th indicator is much less important than the j-th indicator 1 / 9: The i-th indicator is definitely less important than the j-th indicator Among them, 2, 4, 6, 8, 1 / 2, 1 / 4, 1 / 6, 1 / 8: indicate that the importance is between 1 and 3, 2 and 5, ...; M200, send the questionnaire to the experts for filling; M300, collect the questionnaire and import it into the AHP analysis software to calculate the importance ranking of indicators at each level. When calculating, find the eigenvector w corresponding to the maximum eigenvalue max according to the judgment matrix. The calculation formula is as follows: P ω P ω =λ max oh, M400, perform consistency check on the judgment matrix, the test formula is as follows: Among them, CR represents the random consistency ratio of the judgment matrix, CI represents the consistency index of the judgment matrix, and RI represents the average random consistency index of the judgment matrix; M500, import the expert survey form into the AHP software to check the incompleteness and inconsistency of the matrix, and prompt the experts for abnormal situations, including incomplete and inconsistent matrix data. For missing items in the judgment matrix, communicate with the experts for confirmation and then correct them. For inconsistencies in the matrix, use the algorithm provided by the Maishi AHP software system to automatically adjust them to consistency when making decisions. The weights of experts are all set to 1; M600, after generating the weights of each three-level indicator based on AHP software, the normalized scores of the corresponding three-level indicators are weighted and added and multiplied by 100 to obtain the final index, whose value is between 0-100.

4. The method for constructing a drug safety risk assessment index system according to claim 1, characterized in that: For the drug life cycle safety risk index system, the indicator weights of each level indicator are calculated based on the principal component analysis method, including the following steps: L100 input: m samples, a data set with n features, that is, sample data X = [X1, X2, ..., X m ], the target dimension of dimension reduction is k, and the sample set is denoted as matrix X, which can be expressed as: L200 output: the sample set after dimensionality reduction Y = [y1, y2, ..., y m ]. L300 calculates the covariance matrix, including the following operations: Step 1: Decentralize the matrix to obtain a new matrix X, that is, zero-mean each column: Step 2: Calculate the covariance moment of the decentralized matrix X That is, an n*n-order matrix; Step 3: Perform eigendecomposition on the covariance matrix C to find the eigenvalues ​​and corresponding eigenvectors of the covariance matrix, i.e. cυ k =λ k ν k ; Step 4: Arrange the eigenvectors in descending order from left to right according to the corresponding eigenvalues, and take the first k columns to form the matrix W, that is, an n*k-order matrix; Step 5: Calculate the sample features after dimension reduction to k dimensions through Y=XW, that is, an m*k-order matrix.

5. A drug safety risk assessment index system construction system, characterized in that: It includes system construction module, hierarchical indicator setting module and indicator weight calculation module; The system construction module is used to perform the following: for drug safety evaluation, construct a drug safety evaluation index system, which includes a drug safety supervision capability index system, a drug safety supervision social co-governance index system, and a drug life cycle safety risk index system; The hierarchical index setting module is used to perform the following: for each drug safety evaluation index system, set the hierarchical index of the drug safety evaluation index system; The indicator weight calculation module is used to perform the following: calculating the indicator weights of indicators at each level based on the hierarchical analysis method, the Delphi method and the principal component analysis method.

6. The drug safety risk assessment index system construction system according to claim 5, characterized in that: Each drug safety evaluation index system is set with multiple first-level indicators, each first-level indicator is set with multiple second-level indicators, and each second-level indicator is set with multiple third-level indicators.

7. The drug safety risk assessment index system construction system according to claim 5, characterized in that: For the drug safety supervision capability index system and the drug safety supervision social co-governance index system, the indicator weight calculation module is used to calculate the indicator weight of each level indicator based on the hierarchical analysis method and the Delphi method, including the following steps: M100, construct a judgment matrix, generate a regular questionnaire for the indicator system, and compare indicators at different levels in pairs using a scale of 1 to 9, as shown below: 1: The i-th indicator is as important as the j-th indicator 3: The i-th indicator is slightly more important than the j-th indicator 5: The i-th index is greater than the j-th index. 7: The i-th indicator is more important than the j-th indicator 9: The i-th indicator is definitely more important than the j-th indicator 1 / 3: The i-th indicator is slightly less important than the j-th indicator 1 / 5: The i-th indicator is less important than the j-th indicator 1 / 7: The i-th indicator is much less important than the j-th indicator 1 / 9: The i-th indicator is definitely less important than the j-th indicator Among them, 2, 4, 6, 8, 1 / 2, 1 / 4, 1 / 6, 1 / 8: indicate that the importance is between 1 and 3, 2 and 5, ...; M200, send the questionnaire to the experts for filling; M300, collect the questionnaire and import it into the AHP analysis software to calculate the importance ranking of indicators at each level. When calculating, find the eigenvector w corresponding to the maximum eigenvalue max according to the judgment matrix. The calculation formula is as follows: P ω P ω =λ max oh, M400, perform consistency check on the judgment matrix, the test formula is as follows: Among them, CR represents the random consistency ratio of the judgment matrix, CI represents the consistency index of the judgment matrix, and RI represents the average random consistency index of the judgment matrix; M500, import the expert survey form into the AHP software to check the incompleteness and inconsistency of the matrix, and prompt the experts for abnormal situations, including incomplete and inconsistent matrix data. For missing items in the judgment matrix, communicate with the experts for confirmation and then correct them. For inconsistencies in the matrix, use the algorithm provided by the Maishi AHP software system to automatically adjust them to consistency when making decisions. The weights of experts are all set to 1; M600, after generating the weights of each three-level indicator based on AHP software, the normalized scores of the corresponding three-level indicators are weighted and added and multiplied by 100 to obtain the final index, whose value is between 0-100.

8. The drug safety risk assessment index system construction system according to claim 5, characterized in that: For the drug life cycle safety risk index system, the indicator weight calculation module is used to calculate the indicator weight of each level indicator based on the principal component analysis method, including the following steps: L100 input: m samples, a data set with n features, that is, sample data X = [X1, X2, ..., X m ], the target dimension of dimension reduction is k, and the sample set is denoted as matrix X, which can be expressed as: L200 output: the sample set after dimensionality reduction Y = [y1, y2, ..., y m ]. L300 calculates the covariance matrix, including the following operations: Step 1: Decentralize the matrix to obtain a new matrix X, that is, zero-mean each column: Step 2: Calculate the covariance moment of the decentralized matrix X That is, an n*n-order matrix; Step 3: Perform eigendecomposition on the covariance matrix C to find the eigenvalues ​​and corresponding eigenvectors of the covariance matrix, i.e. cυ k =λ k v k ; Step 4: Arrange the eigenvectors in descending order from left to right according to the corresponding eigenvalues, and take the first k columns to form the matrix W, that is, an n*k-order matrix; Step 5: Calculate the sample features after dimension reduction to k dimensions through Y=XW, that is, an m*k-order matrix.

Images