Medical institution management benchmark value analysis method based on medical insurance big data

Abstract

The invention relates to the field of medical institution management, and discloses a medical institution management benchmark value analysis method based on medical insurance big data, and the method comprises the steps: S1, extracting DRG data of each medical institution in a region from a medical insurance database; s2, calculating indexes of each medical institution on a DRG group level, wherein the indexes comprise a total weight, a time consumption index, a cost consumption index and a fifteen-day re-admission rate; s3, analyzing the index data of the medical institutions in the region, and calculating a region benchmark value of each index; s4, calculating to obtain a state coefficient of each medical institution according to each index of each medical institution, comparing the state coefficient with a state threshold interval, and dividing each medical institution into a high state, a middle state and a low state in combination with a result obtained by comparing and analyzing each index of each medical institution with the corresponding regional benchmark value; and S5, predicting a future change trend of the medical institution by using a machine learning algorithm, and putting forward improvement measures according to an analysis result.

Description

Technical Field

[0001] The present invention relates to the field of medical institution management, and in particular to a medical institution management benchmark value analysis method based on medical insurance big data. Background Art

[0002] At present, the payment method based on diagnosis-related groups (DRGs) is gradually being promoted nationwide. This payment method aims to group patients by disease severity and treatment complexity and determine the payment standard for each group, thereby promoting the rational allocation and use of medical resources. However, existing technologies still face many challenges in their specific application.

[0003] The main disadvantages of existing technologies include:

[0004] 1. Differences in service capabilities of medical institutions: Different medical institutions show significant differences when facing the same disease diagnosis group. Some institutions are able to provide efficient treatment while controlling costs and hospitalization time, while other institutions' treatment costs and time are significantly higher than the average, and the readmission rate is high. This difference affects the fair allocation of medical insurance resources and leads to uneven overall medical service quality.

[0005] 2. Low treatment efficiency: Many medical institutions still face the challenge of shortening hospital stay and ensuring medical quality when implementing the DRG payment model. Shortening hospital stay may affect the treatment effect of patients, while excessive focus on cost control may reduce the quality of medical services and increase the risk of recurrence of patients.

[0006] 3. High readmission rate: A high readmission rate reflects poor initial treatment results or patients’ failure to receive adequate rehabilitation and health management after discharge, which increases the expenditure of medical insurance funds and occupies medical resources.

[0007] 4. Insufficient data analysis capabilities: Although medical institutions have gradually realized the importance of data, most hospitals lack systematic analysis capabilities, resulting in the inability to fully utilize existing data for scientific decision-making when optimizing medical services and implementing DRG management.

[0008] These problems not only affect the efficiency and quality of medical services, but also increase the pressure on medical insurance funds. Summary of the invention

[0009] The purpose of the present invention is to provide a medical institution management benchmark value analysis method based on medical insurance big data to solve at least one of the above technical problems.

[0010] The purpose of the present invention can be achieved through the following technical solutions:

[0011] A medical institution management benchmark value analysis method based on medical insurance big data includes the following steps:

[0012] S1. Data collection and preprocessing: Extract the DRG data of each medical institution in the region from the medical insurance database, including hospitalization days, medical expenses, disease type, settlement ID information, and clean and standardize the collected data;

[0013] S2. Indicator calculation: Calculate the indicators of each medical institution at the DRG group level, including total weight, time consumption index, cost consumption index and 15-day readmission rate;

[0014] S3. Benchmark value calculation: by summarizing and analyzing the indicator data of medical institutions in the region, the regional benchmark value of each indicator is calculated;

[0015] S4. Multi-dimensional analysis: The status coefficient of each medical institution is calculated based on the various indicators of each medical institution, and the status coefficient is compared with the preset status threshold interval. Combined with the results of comparative analysis of the various indicators of each medical institution with the corresponding regional benchmark values, each medical institution is divided into three states: high, medium and low.

[0016] S5. Prediction and Improvement: Use machine learning algorithms to predict future trends of change in each medical institution and propose improvement measures based on the analysis results.

[0017] As a further technical solution, the calculation formula for each indicator of each medical institution in S2 at the DRG group level is:

[0018] Through formula 1: Calculate the first The total weight of medical institutions ;in, For the DRG groups, For the The weight coefficient of each DRG group, For current medical institutions The number of cases under each DRG group;

[0019] Through formula 2: Calculate the first Time consumption coefficient of medical institutions ;in, For current medical institutions The average length of hospital stay in each DRG group was In the current area The average length of hospital stay in each DRG group;

[0020] Through formula three: Calculate the first Cost consumption coefficient of medical institutions ;in, For current medical institutions The average medical expenses under each DRG group, In the current area Average medical expenses under each DRG group;

[0021] Through formula 4: Calculate the first Fifteen-day readmission rate of medical institutions ;in, is the number of patients readmitted to the current medical institution within 15 days, It is the number of patients discharged from the current medical institution within fifteen days.

[0022] As a further technical solution, the method for calculating the regional benchmark value of each indicator in S3 is:

[0023] The regional benchmark value of each indicator is calculated by the formula: Calculate the first Regional benchmark values ​​for indicators ;

[0024] in, For the The medical institution The value of the indicator, For the The medical institution The weight factor corresponding to each indicator is: For the The medical institution The index coefficient of an indicator; is the total number of medical institutions in the current area, When the values ​​are 1, 2, 3, and 4, the corresponding indicators are total weight, time consumption coefficient, cost consumption coefficient, and 15-day readmission rate.

[0025] As a further technical solution, the exponential coefficient The acquisition process is:

[0026] Divide a monitoring cycle into equally spaced Time gradients, based on big data fitting, obtain the curves of each indicator changing over time under each time gradient ;

[0027] By formula: Calculate the exponential coefficient of the current indicator ;

[0028] in, For the time gradient, are the start and end time of the time gradient, is a reference curve based on big data. , are the maximum and minimum values ​​of the current indicator in each time gradient, , is the preset scale factor.

[0029] As a further technical solution, the method for obtaining the status coefficient of each medical institution is:

[0030] Substitute the calculated total weight, time consumption coefficient, cost consumption coefficient and 15-day readmission rate into the following formula:

[0031]

[0032] Calculate the status coefficient of each medical institution ;

[0033] in, is the preset weight coefficient, is the preset coefficient, is the total weight mean of the current region, They are efficiency factor, cost factor and nursing quality factor respectively.

[0034] As a further technical solution, the method of dividing each medical institution into three states: high, medium and low is as follows:

[0035] The status coefficient of each medical institution obtained will be calculated The preset status threshold interval Make a comparison;

[0036] like , then it is judged that the current medical institution is in a high state;

[0037] like , then the current medical institution is judged to be in a medium state;

[0038] like , it is judged that the current medical institution is in a low state.

[0039] As a further technical solution, the method for obtaining the efficiency factor, cost factor, and nursing quality factor is:

[0040] For the current medical institutions, the total number of random samples is Time consumption coefficient , Cost consumption coefficient and 15-day readmission rate ;

[0041] By formula:

[0042] ;

[0043] ;

[0044] ;

[0045] Calculate the efficiency factor of the current medical institution , cost factor , nursing quality factors ;

[0046] in, For the sampling points, They are the regional benchmark values ​​of the current regional time consumption coefficient, cost consumption coefficient and 15-day readmission rate, is the conversion coefficient.

[0047] As a further technical solution, the steps of S5 are as follows:

[0048] S51, inputting the historical data of each indicator of each medical institution at the DRG group level into a pre-trained recurrent neural network model, and outputting the predicted change trend of each indicator as good or bad;

[0049] S52. If the predicted change trends of all indicators are good, the overall change trend of the medical institution is judged to be good, and the current state is maintained unchanged;

[0050] If the predicted change trend of at least one indicator is bad, the medical institution is improved according to the corresponding indicator;

[0051] If the predicted change trends of all indicators are bad, the overall change trend of the medical institution is judged to be bad, and comprehensive rectification is carried out.

[0052] Beneficial effects of the present invention:

[0053] (1) The present invention ensures the authority and accuracy of the data by directly extracting data from the medical insurance database. By cleaning and standardizing, redundant and erroneous data are removed, and the availability and reliability of the data are improved. High-quality data provides a solid foundation for subsequent indicator calculation, benchmark value analysis, and prediction improvement. By calculating indicators such as total weight, time consumption index, cost consumption index, and 15-day readmission rate, the performance of each medical institution at the DRG group level can be quantitatively evaluated. These indicators make it possible to compare different medical institutions, which helps to identify medical institutions with excellent performance and those that need improvement. These indicator data are the basis for calculating regional benchmark values ​​and provide an important basis for subsequent benchmark value analysis.

[0054] (2) The present invention provides a clear standard for medical institutions in the region through the calculation of benchmark values, which helps medical institutions understand their relative position in the region; by comparing the indicator data of each medical institution with the benchmark value, it can be found in which aspects the medical institutions have gaps, providing a direction for improvement; the publication of benchmark values ​​can promote competition and cooperation among medical institutions and promote the improvement of medical service quality in the entire region;

[0055] (3) The present invention can comprehensively evaluate each medical institution and understand its overall performance by calculating the status coefficient and combining it with the preset status threshold interval; the medical institutions are divided into three states: high, medium and low, which helps to classify and manage medical institutions in different states and improve management efficiency; combined with the comparative analysis results with the benchmark value, specific improvement measures can be proposed for the weak links of each medical institution. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The present invention will be further described below in conjunction with the accompanying drawings.

[0057] Figure 1 It is a diagram of the method steps of the present invention;

[0058] Figure 2 This is a step diagram for S5. DETAILED DESCRIPTION

[0059] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] See also Figure 1-Figure 2 As shown, the present invention is a medical institution management benchmark value analysis method based on medical insurance big data, comprising the following steps:

[0061] S1. Data collection and preprocessing: Extract the DRG data of each medical institution in the region from the medical insurance database, including hospitalization days, medical expenses, disease type, settlement ID information, and clean and standardize the collected data;

[0062] S2. Indicator calculation: Calculate the indicators of each medical institution at the DRG group level, including total weight, time consumption index, cost consumption index and 15-day readmission rate;

[0063] S3. Benchmark value calculation: by summarizing and analyzing the indicator data of medical institutions in the region, the regional benchmark value of each indicator is calculated;

[0064] S4. Multi-dimensional analysis: The status coefficient of each medical institution is calculated based on the various indicators of each medical institution, and the status coefficient is compared with the preset status threshold interval. Combined with the results of comparative analysis of the various indicators of each medical institution with the corresponding regional benchmark values, each medical institution is divided into three states: high, medium and low.

[0065] S5. Prediction and Improvement: Use machine learning algorithms to predict future trends of change in each medical institution and propose improvement measures based on the analysis results.

[0066] In this embodiment, in order to solve the technical problems raised in the background technology, the DRG data of each medical institution in the region is extracted from the medical insurance database, and then the total weight, time consumption index, cost consumption index and 15-day readmission rate index of each medical institution at the DRG group level are calculated. The above indicators can reflect the overall operational efficiency of the medical institution, the complexity and cost of cases, the time investment and the readmission situation. Then, the indicator data of the medical institutions in the region are summarized and analyzed to calculate the regional benchmark value of each indicator; the state coefficient of each medical institution is calculated according to the various indicators of each medical institution, and the state coefficient is compared with the preset state threshold interval. Combined with the results of comparative analysis of the various indicators of each medical institution with the corresponding regional benchmark values, each medical institution is divided into three states: high, medium and low; a machine learning algorithm is used to predict the future change trend of each medical institution, and improvement measures are proposed based on the analysis results;

[0067] Through the above technical solution, data is directly extracted from the medical insurance database to ensure the authority and accuracy of the data. Through cleaning and standardization, redundant and erroneous data are removed, and the availability and reliability of the data are improved. High-quality data provides a solid foundation for subsequent indicator calculation, benchmark value analysis and prediction improvement. By calculating indicators such as total weight, time consumption index, cost consumption index and 15-day readmission rate, the performance of each medical institution at the DRG group level can be quantitatively evaluated. These indicators make it possible to compare different medical institutions, which helps to identify medical institutions with excellent performance and those that need improvement. These indicator data are the basis for calculating regional benchmark values ​​and provide an important basis for subsequent benchmark value analysis. The calculation of benchmark values ​​provides a clear standard for medical institutions in the region, which helps medical institutions understand their relative position in the region. By comparing the indicator data and benchmark values ​​of each medical institution, it can be found in which aspects the medical institutions have gaps, providing a direction for improvement. The publication of benchmark values ​​can promote competition and cooperation among medical institutions and promote the improvement of medical service quality in the entire region.

[0068] At the same time, by calculating the status coefficient and combining it with the preset status threshold range, each medical institution can be comprehensively evaluated to understand its overall performance; dividing medical institutions into three states: high, medium, and low will help to classify and manage medical institutions in different states and improve management efficiency; combined with the comparative analysis results with the benchmark values, specific improvement measures can be proposed for the weak links of each medical institution.

[0069] The calculation formula for each indicator of each medical institution at the DRG group level in S2 is:

[0070] Through formula 1: Calculate the first The total weight of medical institutions ;in, For the DRG groups, For the The weight coefficient of each DRG group, For current medical institutions The number of cases under each DRG group;

[0071] Through formula 2: Calculate the first Time consumption coefficient of medical institutions ;in, For current medical institutions The average length of hospital stay in each DRG group was In the current area The average length of hospital stay in each DRG group;

[0072] Through formula three: Calculate the first Cost consumption coefficient of medical institutions ;in, For current medical institutions The average medical expenses under each DRG group, In the current area Average medical expenses under each DRG group;

[0073] Through formula 4: Calculate the first Fifteen-day readmission rate of medical institutions ;in, is the number of patients readmitted to the current medical institution within 15 days, It is the number of patients discharged from the current medical institution within fifteen days.

[0074] In this embodiment, calculation formulas for the above four indicators are provided, through which the above indicators can be specifically quantified, which is convenient for the subsequent calculation and analysis of regional benchmark values.

[0075] The method for calculating the regional benchmark value of each indicator in S3 is:

[0076] The regional benchmark value of each indicator is calculated by the formula: Calculate the first Regional benchmark values ​​for indicators ;

[0077] in, For the The medical institution The value of the indicator, For the The medical institution The weight factor corresponding to each indicator is determined based on historical data analysis. For the The medical institution The index coefficient of an indicator; is the total number of medical institutions in the current area, When the values ​​are 1, 2, 3, and 4, the corresponding indicators are total weight, time consumption coefficient, cost consumption coefficient, and 15-day readmission rate.

[0078] The exponential coefficient The acquisition process is:

[0079] Divide a monitoring cycle into equally spaced Time gradients, based on big data fitting, obtain the curves of each indicator changing over time under each time gradient ;

[0080] By formula: Calculate the exponential coefficient of the current indicator ;

[0081] in, For the time gradient, are the start and end time of the time gradient, is a reference curve based on big data. , are the maximum and minimum values ​​of the current indicator in each time gradient, , is the preset scale factor, It is a reference coefficient, which is determined comprehensively based on historical data and experimental data.

[0082] In this embodiment, a method for obtaining a regional benchmark value is provided. Specifically, by substituting each indicator into the formula Calculate the first Regional benchmark values ​​for indicators , where the exponential coefficient The expression is:

[0083] , since the regional benchmark values ​​of the four indicators all use the same calculation formula, for the convenience of expression, we let When 1, 2, 3, and 4, the corresponding indicators are total weight, time consumption coefficient, cost consumption coefficient, and 15-day readmission rate. It can be seen from the above formula that The larger the value, the greater the difference between the extreme values ​​of the current indicator, the larger the numerical amplitude of the indicator, and the exponential coefficient The larger it is, the greater the impact on the regional benchmark value. The larger the value is, the greater the deviation between the actual change curve of the current indicator and the reference curve. Therefore, the greater the cumulative change of the indicator at each time gradient, and the greater the impact on the regional benchmark value. By combining the range and the cumulative change, the exponential coefficient is iterated to accurately reflect the changing status of the corresponding indicator in the historical period, thereby improving the accuracy of the regional benchmark value of each indicator.

[0084] The method to obtain the status coefficient of each medical institution is:

[0085] Substitute the calculated total weight, time consumption coefficient, cost consumption coefficient and 15-day readmission rate into the following formula:

[0086]

[0087] Calculate the status coefficient of each medical institution ;

[0088] in, is the preset weight coefficient. It is determined based on historical data analysis. is the preset coefficient, is the total weight mean of the current region, They are efficiency factor, cost factor and nursing quality factor respectively.

[0089] The method for obtaining the efficiency factor, cost factor and nursing quality factor is:

[0090] For the current medical institutions, the total number of random samples is Time consumption coefficient , Cost consumption coefficient and 15-day readmission rate ;

[0091] By formula:

[0092] ;

[0093] ;

[0094] ;

[0095] Calculate the efficiency factor of the current medical institution , cost factor , nursing quality factors ;

[0096] in, For the sampling points, They are the regional benchmark values ​​of the current regional time consumption coefficient, cost consumption coefficient and 15-day readmission rate, is the conversion coefficient, which is selected and determined based on historical experience data.

[0097] In this embodiment, the four calculated indicators are substituted into Calculate the status coefficient of each medical institution ,in Used to adjust the impact and sensitivity of the corresponding value, the initial setting is Through the calculation process of the above four indicators, the status of each medical institution is quantitatively calculated, and the status coefficient reflecting the actual status of each medical institution is obtained. Obviously, the larger the status coefficient, the better the operation status of the medical institution. At the same time, in order to take into account the comparison between each indicator of each medical institution and the regional benchmark value, the formula is used. , , Calculate the efficiency factor of the current medical institution , cost factor , nursing quality factors ; Obviously the efficiency factor , cost factor , nursing quality factors The larger it is, the better the medical institution's performance in efficiency, cost and nursing quality is, the greater the impact on the status coefficient of the medical institution is, and the more accurately it reflects the status of the medical institution.

[0098] The method of classifying each medical institution into three statuses: high, medium, and low is as follows:

[0099] The status coefficient of each medical institution obtained will be calculated The preset status threshold interval Make a comparison;

[0100] like , then it is judged that the current medical institution is in a high state;

[0101] like , then the current medical institution is judged to be in a medium state;

[0102] like , it is judged that the current medical institution is in a low state.

[0103] The steps of S5 are as follows:

[0104] S51, inputting the historical data of each indicator of each medical institution at the DRG group level into a pre-trained recurrent neural network model, and outputting the predicted change trend of each indicator as good or bad;

[0105] S52. If the predicted change trends of all indicators are good, the overall change trend of the medical institution is judged to be good, and the current state is maintained unchanged;

[0106] If the predicted change trend of at least one indicator is bad, the medical institution is improved according to the corresponding indicator;

[0107] If the predicted change trends of all indicators are bad, the overall change trend of the medical institution is judged to be bad, and comprehensive rectification is carried out.

[0108] It should be noted that the calculation formulas and various parameters involved in the calculation in the present invention have been dimensionally processed in advance, and the process of dimensionless processing is well known in the industry and will not be described here.

[0109] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A medical institution management benchmark value analysis method based on medical insurance big data, characterized in that: The following steps are involved: S1. Data collection and preprocessing: Extract the DRG data of each medical institution in the region from the medical insurance database, including hospitalization days, medical expenses, disease type, settlement ID information, and clean and standardize the collected data; S2. Indicator calculation: Calculate the indicators of each medical institution at the DRG group level, including total weight, time consumption index, cost consumption index and 15-day readmission rate; S3. Benchmark value calculation: by summarizing and analyzing the indicator data of medical institutions in the region, the regional benchmark value of each indicator is calculated; S4. Multi-dimensional analysis: The status coefficient of each medical institution is calculated based on the various indicators of each medical institution, and the status coefficient is compared with the preset status threshold interval. Combined with the results of comparative analysis of the various indicators of each medical institution with the corresponding regional benchmark values, each medical institution is divided into three states: high, medium and low. S5. Prediction and Improvement: Use machine learning algorithms to predict future trends of change in each medical institution and propose improvement measures based on the analysis results.

2. The method for analyzing the benchmark value of medical institution management based on medical insurance big data according to claim 1 is characterized in that: The calculation formula for each indicator of each medical institution at the DRG group level in S2 is: Through formula 1: Calculate the first The total weight of medical institutions ;in, For the DRG groups, For the The weight coefficient of each DRG group, For current medical institutions The number of cases under each DRG group; Through formula 2: Calculate the first Time consumption coefficient of medical institutions ;in, For current medical institutions The average length of hospital stay in each DRG group was In the current area The average length of hospital stay in each DRG group; Through formula three: Calculate the first Cost consumption coefficient of medical institutions ;in, For current medical institutions The average medical expenses under each DRG group, In the current area Average medical expenses under each DRG group; Through formula 4: Calculate the first Fifteen-day readmission rate of medical institutions ;in, is the number of patients readmitted to the current medical institution within 15 days, It is the number of patients discharged from the current medical institution within fifteen days.

3. The method for analyzing the benchmark value of medical institution management based on medical insurance big data according to claim 2 is characterized in that: The method for calculating the regional benchmark value of each indicator in S3 is: The regional benchmark value of each indicator is calculated by the formula: Calculate the first Regional benchmark values ​​for indicators ; in, For the The medical institution The value of the indicator, For the The medical institution The weight factor corresponding to each indicator is: For the The medical institution The index coefficient of an indicator; is the total number of medical institutions in the current area, When the values ​​are 1, 2, 3, and 4, the corresponding indicators are total weight, time consumption coefficient, cost consumption coefficient, and 15-day readmission rate.

4. The method for analyzing the benchmark value of medical institution management based on medical insurance big data according to claim 3 is characterized in that: The exponential coefficient The acquisition process is: Divide a monitoring cycle into equally spaced Time gradients, based on big data fitting, obtain the curves of each indicator changing over time under each time gradient ; By formula: Calculate the exponential coefficient of the current indicator ; in, For the time gradient, are the start and end time of the time gradient, is a reference curve based on big data. , are the maximum and minimum values ​​of the current indicator in each time gradient, , is the preset scale factor, is the reference coefficient.

5. The method for analyzing the benchmark value of medical institution management based on medical insurance big data according to claim 4 is characterized in that: The method to obtain the status coefficient of each medical institution is: Substitute the calculated total weight, time consumption coefficient, cost consumption coefficient and 15-day readmission rate into the following formula: Calculate the status coefficient of each medical institution ; in, is the preset weight coefficient, is the preset coefficient, is the total weight mean of the current region, They are efficiency factor, cost factor and nursing quality factor respectively.

6. The method for analyzing the benchmark value of medical institution management based on medical insurance big data according to claim 1 or 5, characterized in that: The method of classifying each medical institution into three statuses: high, medium, and low is as follows: The status coefficient of each medical institution obtained will be calculated The preset status threshold interval Make a comparison; like , then it is judged that the current medical institution is in a high state; like , then the current medical institution is judged to be in a medium state; like , it is judged that the current medical institution is in a low state.

7. The method for analyzing the benchmark value of medical institution management based on medical insurance big data according to claim 5 is characterized in that: The method for obtaining the efficiency factor, cost factor and nursing quality factor is: For the current medical institutions, the total number of random samples is Time consumption coefficient , Cost consumption coefficient and 15-day readmission rate ; By formula: ； ； ； Calculate the efficiency factor of the current medical institution , cost factor , nursing quality factors ; in, For the sampling points, They are the regional benchmark values ​​of the current regional time consumption coefficient, cost consumption coefficient and 15-day readmission rate, is the conversion coefficient.

8. The method for analyzing the benchmark value of medical institution management based on medical insurance big data according to claim 1 is characterized in that: The steps of S5 are as follows: S51, inputting the historical data of each indicator of each medical institution at the DRG group level into a pre-trained recurrent neural network model, and outputting the predicted change trend of each indicator as good or bad; S52. If the predicted change trends of all indicators are good, the overall change trend of the medical institution is judged to be good, and the current state is maintained unchanged; If the predicted change trend of at least one indicator is bad, the medical institution is improved according to the corresponding indicator; If the predicted change trends of all indicators are bad, the overall change trend of the medical institution is judged to be bad, and comprehensive rectification is carried out.

Images