Research method for influence factors of medical security system

Abstract

The invention discloses a medical security system influence factor research method, and relates to the technical field of medical security system research. According to the method, the multi-level medical security development index is constructed, the barrier that the comprehensive security effect is difficult to carry out regional comparison under a multi-security system is broken through, the regional comparability of the overall level of the multi-level medical security system is realized, and the multi-level medical security development index is established from the perspective of the multi-level medical security system. Influence effects of digital general benefits finance on different guarantee levels such as basic medical insurance, serious illness insurance, medical assistance and commercial health insurance and structure effects and heterogeneity effects of the digital general benefits finance are systematically investigated, the blank of the research field is filled, and research contents and methods are enriched; a causal forest algorithm is used for carrying out heterogeneity analysis, the reliability and universality of research results are verified, possible interference factors and deviations are eliminated, large-scale, high-dimensional, nonlinear and non-stationary data can be effectively processed, and potential rules and characteristics of the data can be mined.

Description

Technical Field

[0001] The present invention relates to the technical field of medical security system research, and in particular to a method for researching factors affecting a medical security system. Background Art

[0002] Digital inclusive finance combines the advantages of inclusive finance and financial technology, and can effectively solve the limitations of traditional financial models in terms of coverage, cost-effectiveness and convenience. In particular, it has far-reaching significance for the optimization of the medical insurance system.

[0003] In this context, digital inclusive finance, with its low cost, wide coverage, high efficiency and convenience, can reduce residents' insurance costs, strengthen residents' awareness and acceptance of medical insurance, and significantly improve residents' willingness to participate in insurance, thus improving and expanding the depth and breadth of medical insurance coverage, especially playing an important role in medical insurance, a key level that benefits the general public. Therefore, in-depth research on the impact of digital inclusive finance on the multi-level medical insurance system, especially its effect on the bottom layer of security, is not only of great theoretical value, but also of great practical significance. To this end, we propose a research method for the influencing factors of the medical insurance system. Summary of the invention

[0004] The purpose of the present invention is to solve the problems mentioned in the above background technology. The present invention provides a method for studying the factors affecting the medical insurance system.

[0005] In order to achieve the above-mentioned purpose, the present invention specifically adopts the following technical solutions:

[0006] A research method for influencing factors of a medical security system comprises the following steps:

[0007] Step 1: Construct a multi-level medical security development index at the provincial level. Based on security, we construct evaluation index systems for different security levels. On this basis, we calculate the basic medical insurance development index, medical assistance development index, and commercial health insurance development index at the provincial level.

[0008] Step 2: Explore the impact of digital inclusive finance on the multi-level medical security system, build a fixed effect model, explore the structural effects of digital inclusive finance on the multi-level medical security system from the three dimensions of coverage breadth, usage depth, and digitalization degree, and analyze the promotion effect and potential mechanism of digital inclusive finance on the multi-level medical security system;

[0009] Step 3: Test and evaluate the reliability and universality of the impact mechanism, build a mediation effect model, test and evaluate the two possible impact mechanisms of digital inclusive finance on the multi-level medical security system, the income effect and the employment effect, analyze the impact path and degree of digital inclusive finance on the multi-level medical security system, and then use causal identification methods such as causal forest and dual machine learning to conduct heterogeneity analysis and robustness tests to verify the reliability and universality of the research results and eliminate possible interference factors and biases;

[0010] Step 4: Propose research conclusions and policy recommendations. Based on the in-depth discussion of the relationship between digital inclusive finance and medical insurance, provide suggestions and references for the optimization of the medical insurance system, and put forward policy recommendations and future research directions.

[0011] Furthermore, the construction of the multi-level medical security development index at the provincial level also includes: calculating the actual composition of the multi-level medical security system, the major disease insurance development index and constructing the multi-level medical security development index at the provincial level.

[0012] Furthermore, the calculation of the actual composition of the multi-level medical security system is generated by the Topsis algorithm based on the entropy weight method.

[0013] Furthermore, the reliability and universality of the testing and evaluation of the impact mechanism include: theoretical basis, impact mechanism, empirical testing and further exploration.

[0014] Furthermore, the theoretical basis includes: financial development theory, financial exclusion theory and social security theory.

[0015] Furthermore, the influencing mechanism includes: direct effect and mediating effect.

[0016] Furthermore, the empirical test includes: basic regression analysis and mediation transmission mechanism test.

[0017] Furthermore, basic regression analysis constructs a regression model:

[0018] In_score i，t =β 0 +β 1 In_dif i，t +β 2 control i，t +μ t +δ i +ε i，t #(1)

[0019] Among them, i represents a single province, t represents the year, and In_score i，t In_dif represents the level of the multi-level medical security system in province i in year t,i，t is the independent variable representing digital financial inclusion, control i，t Represents other factors that affect the multi-level medical security system, μ t represents the time fixed effect, δ i represents the fixed effect of each province, is the regression error term, and ε i，t is the regression error term.

[0020] Furthermore, the further exploration includes: dual machine model, robustness test, heterogeneity analysis and causal forest.

[0021] Furthermore, the dual machine model uses random forest as a regression method.

[0022] The beneficial effects of the present invention are as follows:

[0023] 1. This paper uses a fixed effect model to explore the overall impact of digital inclusive finance on the multi-level medical security system, and interprets social security from a new perspective of digital inclusive finance. Digital inclusive finance can also be subdivided into three aspects: coverage breadth, usage depth, and degree of digitization, further enriching the research scope of the impact of digital inclusive finance on social security.

[0024] 2. By constructing a multi-level medical security development index, the present invention breaks the barrier that makes it difficult to make regional comparisons of comprehensive security effects under multiple security systems, and achieves regional comparability at the overall level of the multi-level medical security system. From the perspective of the multi-level medical security system, the present invention systematically examines the impact of digital inclusive finance on different security levels such as basic medical insurance, major disease insurance, medical assistance, commercial health insurance, and the structural and heterogeneous effects of its impact, filling the gap in the research field and enriching the research content and methods.

[0025] 3. The present invention uses the causal forest algorithm to perform heterogeneity analysis, verify the reliability and universality of the research results, eliminate possible interference factors and deviations, and can effectively process large-scale, high-dimensional, nonlinear, and non-stationary data, explore the potential laws and characteristics of the data, discover the heterogeneous effects between digital inclusive finance and medical insurance, examine the impact of different factors on the research results, test the robustness of the research hypothesis, and improve the universality and adaptability of the research results. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a work flow chart of the present invention;

[0027] Figure 2 It is a construction diagram of the causal decision tree in the present invention. DETAILED DESCRIPTION

[0028] To make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0029] See also Figure 1 - Figure 2 The present invention provides a method for studying factors affecting the medical security system, comprising the following steps:

[0030] Step 1: Construct a multi-level medical security development index at the provincial level. Based on the security, an evaluation index system for different security levels is constructed respectively, and on this basis, the basic medical insurance development index, medical assistance development index and commercial health insurance development index at the provincial level are calculated; based on the complex dimensions of the construction of a multi-level medical security system, with the help of the calculation results of the above two steps, by constructing a multi-level medical security development index at the provincial level, the comprehensive security level and geographical distribution characteristics of the multi-level medical security system are revealed.

[0031] Step 2: Explore the impact of digital inclusive finance on the multi-level medical security system, construct a fixed effect model, and explore the structural effects of digital inclusive finance on the multi-level medical security system from the three dimensions of coverage breadth, usage depth, and degree of digitization. Analyze the promoting role and potential mechanism of digital inclusive finance on the multi-level medical security system; use the fixed effect model to explore the overall impact of digital inclusive finance on the multi-level medical security system, and interpret social security from the new perspective of digital inclusive finance. Digital inclusive finance can also be subdivided into three aspects: coverage breadth, usage depth, and degree of digitization, which further enriches the research scope of the impact of digital inclusive finance on social security.

[0032] Step 3: Test and evaluate the reliability and universality of the impact mechanism, build a mediation effect model, test and evaluate the two possible impact mechanisms of digital inclusive finance on the multi-level medical security system, the income effect and the employment effect, analyze the impact path and degree of digital inclusive finance on the multi-level medical security system, and then use causal identification methods such as causal forest and dual machine learning to conduct heterogeneity analysis and robustness test to verify the reliability and universality of the research results and eliminate possible interference factors and biases; by constructing a multi-level medical security development index, it breaks the barrier that makes it difficult to compare the comprehensive security effects in different regions under multiple security systems, realizes the regional comparability of the multi-level medical security system at the overall level, and From the perspective of the barrier system, this paper systematically examines the impact of digital inclusive finance on different levels of protection, such as basic medical insurance, major disease insurance, medical assistance, commercial health insurance, as well as its structural effects and heterogeneous effects, filling the gap in the research field and enriching the research content and methods. It uses the causal forest algorithm to conduct heterogeneity analysis, verifies the reliability and universality of the research results, and eliminates possible interference factors and deviations. It can effectively process large-scale, high-dimensional, nonlinear, and non-stationary data, explore the potential laws and characteristics of the data, discover the heterogeneous effects between digital inclusive finance and medical insurance, examine the impact of different factors on the research results, test the robustness of the research hypothesis, and improve the universality and adaptability of the research results.

[0033] Step 4: Propose research conclusions and policy recommendations. Based on the in-depth discussion of the relationship between digital inclusive finance and medical insurance, provide suggestions and references for the optimization of the medical insurance system, and put forward policy recommendations and future research directions.

[0034] In this embodiment, preferably, constructing a multi-level medical security development index at the provincial level also includes: calculating the actual composition of the multi-level medical security system, the serious illness insurance development index and constructing a multi-level medical security development index at the provincial level; extending from the single dimension of basic medical insurance to a composite dimension including basic medical insurance, serious illness insurance, medical assistance and commercial health insurance, filling the gap in the research field, enriching the research content and methods, and being able to enhance the breadth of coverage and depth of use, thereby enhancing comprehensiveness.

[0035] In this embodiment, preferably, the actual composition of the multi-level medical security system is calculated based on the entropy weight method and generated by the Topsis algorithm; the provincial multi-level medical security index is formulated using the entropy weight method and the TOPSIS method to evaluate the performance of each region in establishing a medical security system. The entropy weight method assigns weights according to the degree of variability of each indicator, thereby reducing the subjectivity in the weight allocation process. Using these weights, the TOPSIS method calculates the distance between each province and the positive and negative ideal solutions, and then obtains a comprehensive score. This method can objectively and quantitatively evaluate the relative performance of the medical systems in each province, thereby ensuring the accuracy of the system.

[0036] In this embodiment, preferably, testing and evaluating the reliability and universality of the impact mechanism includes: theoretical basis, impact mechanism, empirical testing and further exploration.

[0037] In this embodiment, preferably, the theoretical basis includes: financial development theory, financial exclusion theory and social security theory, which can ensure the comprehensiveness of the basic theory.

[0038] In this embodiment, preferably, the impact mechanism includes: direct effect and mediating effect, and the mediating effect model is used to analyze the impact channels of digital inclusive finance on social security.

[0039] In this embodiment, preferably, the empirical test includes: basic regression analysis and mediating transmission mechanism test; through basic regression analysis, it is possible to find out which influencing factors and the law of their influence on the results, and with the cooperation of basic regression analysis and mediating transmission mechanism test, the accuracy of the research can be guaranteed.

[0040] In this embodiment, preferably, the basic regression analysis constructs a regression model:

[0041] In_score i，t =β 0 +β 1 In_dif i，t +β 2 control i，t +μ t +δ i +ε i，t #(1)

[0042] Among them, i represents a single province, t represents the year, and In_score i，t In_dif represents the level of the multi-level medical security system in province i in year t, i，t is the independent variable representing digital financial inclusion, control i，t Represents other factors that affect the multi-level medical security system, μ t represents the time fixed effect, δ irepresents the fixed effect of each province, is the regression error term, and ε i，t is the regression error term.

[0043] In this embodiment, preferably, further exploration includes: dual machine model, robustness test, heterogeneity analysis and causal forest; further exploration can improve the accuracy of the research and thus ensure the effect of the research.

[0044] Constructing linear models in dual machine models,

[0045] In_socre it+1 =θ 0 In_dif it +g(X it )+U it #(2)

[0046] E(U it |In_dif it , X it )=0#(3)

[0047] Where i represents the province, t represents the year, In_socre is the logarithm of the multi-layered medical security system score, and In_dif represents the logarithm of digital inclusive finance. 0 is the coefficient of digital financial inclusion, X it represents the set of control variables, and the specific function form is g(X it ) is estimated using a machine learning algorithm, g(X it ) represents the error term,

[0048] If equations (2) and (3) are estimated directly, the results may be biased and θ may be biased. In high-dimensional or complex model settings, machine learning models usually introduce regularization terms to reduce the dimension. Although regularization helps prevent the variance of the estimated value from becoming too large, it also introduces regularization bias, making it difficult to converge to . In order to solve this problem, the following auxiliary regression model is constructed:

[0049] In_dif it =m(X it )+V it #(4)

[0050] E(V it |X it )=0#(5)

[0051] Among them, m(X it ) represents the regression function of the treatment variable on the high-dimensional control variable, and its specific form is It is necessary to use a machine learning algorithm to estimate. The specific process is as follows: First, use the machine learning model to estimate m(Xit ) to get the estimated value Then, the residual is calculated as Next step, As In_dif it Finally, the machine learning algorithm is used again to estimate

[0052]

[0053]

[0054] Constructing a continuous causal tree includes the following steps:

[0055] Step 1: The algorithm iterates over each variable in the list of control variables and each unique value of the corresponding feature in the training set and calculates the metrics explicitly described in step 2;

[0056] θθ

[0057] Step 2: For each feature and value, the treatment effect is expressed as and the node splitting score is expressed as as the splitting variable and threshold, respectively. The treatment effect is defined as:

[0058]

[0059] Among them, Y is the dependent variable, representing the multi-level medical insurance index; W is the independent variable or treatment variable, representing the digital inclusive financial index, Cov(Y, W) is the covariance, which indicates the strength and direction of the linear relationship between the dependent variable and the treatment variable, and Var(W) is the variance, which measures the variability of the treatment variable.

[0060] The node split score is defined as:

[0061]

[0062] Where P and C represent the parent node and child node respectively, n represents the sample size, and θ represents the treatment effect. The optimal split node is found from the covariates to maximize the difference in treatment effects between child nodes. Within each node, it is assumed that all samples are homogeneous, so the treatment effect can be calculated using formula (1);

[0063] Step 3: Then split the sample into left and right subtrees based on the feature and threshold with the highest node splitting score in the double loop;

[0064] Step 4: This recursive partitioning algorithm is applied to both the left and right subtrees, gradually partitioning all samples into different nodes in a top-down manner until a pre-set stopping criterion is reached, such as the minimum sample size in a node, the imbalance of treatment variables, or the minimum threshold of information gain, thereby completing the construction of the causal decision tree.

[0065] In this embodiment, preferably, the dual machine model uses random forest as the regression method; random forest can produce unbiased estimates of the generalized errors internally, can handle a large number of input variables, and for unbalanced classification data sets, it can balance the errors, that is, it can ensure the accuracy of the research.

[0066] The working principle and use process of the present invention are as follows: Step 1, constructing a multi-level medical security development index at the provincial level, based on the security, constructing an evaluation index system for different security levels, and on this basis calculating the basic medical insurance development index, medical assistance development index and commercial health insurance development index at the provincial level; based on the composite dimension of the construction of a multi-level medical security system, with the help of the calculation results of the above two steps, by constructing a multi-level medical security development index at the provincial level, the comprehensive security degree and geographical distribution characteristics of the multi-level medical security system are revealed. Constructing a multi-level medical security development index at the provincial level also includes: calculating the actual composition of the multi-level medical security system, the major disease insurance development index and constructing a multi-level medical security development index at the provincial level; extending from the single dimension of basic medical insurance to the basic medical security system. The composite dimension including medical insurance, major disease insurance, medical assistance and commercial health insurance fills the gap in the research field, enriches the research content and methods, can improve the coverage breadth and application depth, thereby improving the comprehensiveness, and calculating the actual composition of the multi-level medical security system on the basis of the entropy weight method through the Topsis algorithm; the provincial multi-level medical security index is formulated using the entropy weight method and the TOPSIS method to evaluate the performance of various regions in establishing a medical security system. The entropy weight method assigns weights according to the variability of each indicator, thereby reducing the subjectivity in the weight allocation process. Using these weights, the TOPSIS method calculates the distance between each province and the positive and negative ideal solutions, and then obtains a comprehensive score. This method can objectively and quantitatively evaluate the relative performance of the medical systems in each province, thereby ensuring the accuracy of the system.

[0067] Step 2: Explore the impact of digital inclusive finance on the multi-level medical security system, construct a fixed effect model, and explore the structural effects of digital inclusive finance on the multi-level medical security system from the three dimensions of coverage breadth, usage depth, and degree of digitization. Analyze the promoting role and potential mechanism of digital inclusive finance on the multi-level medical security system; use the fixed effect model to explore the overall impact of digital inclusive finance on the multi-level medical security system, and interpret social security from the new perspective of digital inclusive finance. Digital inclusive finance can also be subdivided into three aspects: coverage breadth, usage depth, and degree of digitization, which further enriches the research scope of the impact of digital inclusive finance on social security.

[0068] Step 3: Test and evaluate the reliability and universality of the impact mechanism, build a mediation effect model, test and evaluate the two possible impact mechanisms of digital inclusive finance on the multi-level medical security system, the income effect and the employment effect, analyze the impact path and degree of digital inclusive finance on the multi-level medical security system, and then use causal identification methods such as causal forest and dual machine learning to conduct heterogeneity analysis and robustness tests to verify the reliability and universality of the research results and eliminate possible interference factors and biases; by constructing a multi-level medical security development index, it breaks the barrier that the comprehensive security effect is difficult to compare across regions under multiple security systems , achieving the regional comparability of the multi-level medical security system at the overall level. From the perspective of the multi-level medical security system, it systematically examines the impact of digital inclusive finance on different security levels such as basic medical insurance, major disease insurance, medical assistance, and commercial health insurance, as well as its structural effects and heterogeneous effects. It fills the gap in the research field, enriches the research content and methods, uses the causal forest algorithm to conduct heterogeneity analysis, verifies the reliability and universality of the research results, eliminates possible interference factors and deviations, and can effectively process large-scale, high-dimensional, nonlinear, and non-stationary data, mine the potential laws and characteristics of the data, and discover the data. The heterogeneous effects between digital inclusive finance and medical insurance are investigated, the impact of different factors on the research results is examined, the robustness of the research hypothesis is tested, the universality and adaptability of the research results are improved, and the reliability and universality of the impact mechanism are tested and evaluated, including: theoretical basis, impact mechanism, empirical test and further exploration. The theoretical basis includes: financial development theory, financial exclusion theory and social security theory; it can ensure the comprehensiveness of the basic theory. The impact mechanism includes: direct effect and mediating effect. The mediating effect model is used to analyze the impact channel of digital inclusive finance on social security. The empirical test includes: basic regression analysis and mediating transmission mechanism test; through Basic regression analysis can find out which influencing factors have an impact on the results. The cooperation of basic regression analysis and mediating transmission mechanism test can ensure the accuracy of the research. Further exploration includes: dual machine model, robustness test, heterogeneity analysis and causal forest. Through further exploration, the accuracy of the research can be improved, and the effect of the research can be guaranteed. The dual machine model uses random forest as the regression method. Random forest can produce unbiased estimates of the generalized errors internally, and can handle a large number of input variables. For unbalanced classification data sets, it can balance the errors, that is, it can ensure the accuracy of the research.

[0069] Step 4: Propose research conclusions and policy recommendations. Based on the in-depth discussion of the relationship between digital inclusive finance and medical insurance, provide suggestions and references for the optimization of the medical insurance system, and put forward policy recommendations and future research directions.

[0070] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A research method for factors affecting the medical security system, characterized in that: The steps include: Step 1: Construct a multi-level medical security development index at the provincial level. Based on security, we construct evaluation index systems for different security levels. On this basis, we calculate the basic medical insurance development index, medical assistance development index, and commercial health insurance development index at the provincial level. Step 2: Explore the impact of digital inclusive finance on the multi-level medical security system, build a fixed effect model, explore the structural effects of digital inclusive finance on the multi-level medical security system from the three dimensions of coverage breadth, usage depth, and digitalization degree, and analyze the promotion effect and potential mechanism of digital inclusive finance on the multi-level medical security system; Step 3: Test and evaluate the reliability and universality of the impact mechanism, build a mediation effect model, test and evaluate the two possible impact mechanisms of digital inclusive finance on the multi-level medical security system, the income effect and the employment effect, analyze the impact path and degree of digital inclusive finance on the multi-level medical security system, and then use causal identification methods such as causal forest and dual machine learning to conduct heterogeneity analysis and robustness tests to verify the reliability and universality of the research results and eliminate possible interference factors and biases; Step 4: Propose research conclusions and policy recommendations. Based on the in-depth discussion of the relationship between digital inclusive finance and medical insurance, provide suggestions and references for the optimization of the medical insurance system, and put forward policy recommendations and future research directions.

2. A method for studying the influencing factors of the medical security system according to claim 1, characterized in that: The construction of the multi-level medical security development index at the provincial level also includes: calculating the actual composition of the multi-level medical security system, the major disease insurance development index and constructing the multi-level medical security development index at the provincial level.

3. The method for studying the influencing factors of the medical security system according to claim 2, characterized in that: The calculation of the actual composition of the multi-level medical security system is generated by the Topsis algorithm based on the entropy weight method.

4. The method for studying the influencing factors of the medical security system according to claim 1, characterized in that: The reliability and universality of testing and evaluating the impact mechanism include: theoretical basis, impact mechanism, empirical testing and further exploration.

5. A method for studying the influencing factors of the medical security system according to claim 4, characterized in that: The theoretical basis includes: financial development theory, financial exclusion theory and social security theory.

6. A method for studying the influencing factors of the medical security system according to claim 4, characterized in that: The influencing mechanisms include: direct effect and mediating effect.

7. The method for studying the influencing factors of the medical security system according to claim 4, characterized in that: The empirical tests include: basic regression analysis and mediating transmission mechanism test.

8. A method for studying the influencing factors of the medical security system according to claim 7, characterized in that: The basic regression analysis constructs a regression model: In_score i，t =β0+β1In_dif i，t +β2control i，t +m t +d i +e i，t #(1) Among them, i represents a single province, t represents the year, and In_score i，t In_dif represents the level of the multi-level medical security system in province i in year t, i，t is the independent variable representing digital financial inclusion, control i，t Represents other factors that affect the multi-level medical security system, μ t represents the time fixed effect, δ i represents the fixed effect of each province, is the regression error term, and ε i，t is the regression error term.

9. The method for studying the influencing factors of the medical security system according to claim 4, characterized in that: Further exploration includes: dual machine model, robustness test, heterogeneity analysis and causal forest.

10. A method for studying factors affecting the medical security system according to claim 9, characterized in that: The dual machine model uses random forest as the regression method.

Images