JP2022167702A - Health check data analysis system and health check data analysis program - Google Patents

Abstract

To provide a health check data analysis system with which it is possible to perform analysis in group units and automatically analyze and create a report as to what measures can be taken for employers and constituents of health insurance associations to effectively reduce the disease risk of the whole of a group and cut back future medical expenses.SOLUTION: A health check data analysis system 110 comprises: a storage device 113 for storing a plurality of health check data and/or a plurality of health check data of a group having at least a plurality of inquiry results and/or at least a plurality of inquiry results; an analysis unit 112a for deriving one or more risk factors in a group from the plurality of health check data and/or the plurality of inquiry results, calculating a coefficient corresponding to each of the one or more risk factors in the group, and weighting the one or more risk factors in the group by the coefficient; and an output device 119 for outputting a report for the group.SELECTED DRAWING: Figure 2

Description

The present invention relates to a health checkup data analysis system.

Businesses and local governments conduct health examinations for their employees every year in accordance with laws and ordinances. Health checkups are the cornerstone of secondary prevention for early detection of various abnormalities and serious diseases. On the other hand, although primary prevention, which prevents diseases and abnormalities from occurring in the first place, is also important, there are few technologies that lead to primary prevention based on the results of health checkups.

As part of their data health plans, various health insurance associations accumulate and retain receipt data on medical expenses. The technology for estimating costs and analyzing measures to curb increases in medical costs is even scarcer.

For example, Patent Document 1 describes a disease onset factor analysis system for insurers, which generates a graphical model from healthcare data including medical information and health checkup information using techniques such as Bayesian networks. Generate a future distribution from the distribution, and based on the expected values calculated from these, improve future problem diseases, data items that will suppress the increase of problem diseases, test value items that will be risk factors, and lifestyle items Calculating the amount is described.

In addition, Patent Document 2 describes a health prediction device that uses a prediction model of machine learning. It states that lifestyle habits that have a large impact on health risks are ranked and selected based on the partial regression coefficient of the model and presented. In addition, it is described that multi-year prediction is possible from single-year data of health checkup data and interview result data. In addition, it is described that changes in the predicted value of health risk can be confirmed on a screen display or the like by changing lifestyle habits.

In addition, Patent Document 3 describes a health condition estimation device that targets individuals or groups, and estimates the future health condition of the target using a predictive model such as a regression model from biological information including health checkup results. and output the onset probability and the like. It also describes inputting biometric information different from the current one, estimating a hypothetical probability of the future using a prediction model, generating a difference, and outputting the result.

In addition, Patent Document 4 describes a health business planning and improvement system that formulates an implementation plan for a health guidance program in an organization such as a health institution based on the effect of suppressing medical expenses. Based on information, etc., it states that a medical cost prediction model will be created using a statistical model, etc., and for health checkup information, the model will be constructed by predicting the information for the following year onwards based on the information in the first year. . In addition, based on the medical prediction model, it is stated that the predicted medical expenses for each individual at the time of participation in the health guidance program will be calculated, the priority of participation in each insurance guidance program will be determined, and the implementation plan for the health guidance program will be drawn up. It is

WO2015/173917 WO2020/040253 JP 2020-166441 A WO2016/120986

However, the conventional system only shows the analysis results of existing health checkup data and receipt data as aggregate results, and what risk factors are primary prevention of diseases, estimation of future medical expenses, and control of medical expenses increase. It is not a system that automatically analyzes how it actually relates to the analysis of Furthermore, there is no system that performs group-by-group analysis and suggests what measures should be prioritized in the organization, and there is no system that automatically outputs the report.

One embodiment of the present invention solves the above-described problems, and analyzes are performed on a group-by-group basis. To provide a health checkup data analysis system capable of automatically analyzing whether or not it is possible to reduce the disease risk of the entire population and reduce medical costs in the future, and to create a report. Or, in one embodiment, for individuals who have undergone health checkups, what kind of measures can be taken to reduce the risk of future diseases and the risk of increased medical costs based on the results of group-based analysis? is automatically analyzed, and a health checkup data analysis system capable of creating a report is provided.

According to one embodiment of the present invention, using the plurality of health examination data and/or the at least plurality of interview results of a population having a plurality of health examination data and/or at least a plurality of interview results, a plurality of health examination data and/or / Or the processing device derives one or more risk factors in the population from a plurality of interview results, the processing device calculates coefficients corresponding to the one or more risk factors in the population, and one or more risk factors in the population A health examination data analysis system is provided wherein risk factors are weighted by factors and the processor outputs a report for the population.

To derive one or more risk factors in the population, a predictive model may be derived by regression analysis or machine learning using multiple health checkup data and/or multiple interview results of the population.

The plurality of health checkup data and/or the plurality of interview results of the population may be single year data or time series data.

Clinical data and/or medication data may further be used to derive one or more risk factors for the population.

Predictive models may be used to estimate population health care costs, population disease incidence, or population index averages.

One or more measures of the population corresponding to one or more risk factors of the population may be applied to a predictive model to estimate average medical costs for the population, incidence of disease in the population, or measures of the population. good.

One or more measures corresponding to one or more risk factors of the population may be applied to the predictive model to weight the multiple measures of the population.

One or more risk-bearing individuals who have one risk factor among the one or more risk factors of the population from which multiple risk factors have been derived for the purpose of weighting and presenting one or more countermeasures for the population The processing device calculates the proportion of risk-bearers for whom it is effective to implement one or more measures of the group, and for each of the one or more measures of the group, the improvement effect is calculated from the proportion and the coefficient. The expected value may be calculated by the processor.

The processing device receives designation of one or more attributes included in the plurality of health checkup data of the population and/or the at least the plurality of medical interview results of the population, and the processing device determines the plurality of populations based on the one or more attributes and/or the population that constitutes at least a plurality of medical examination results of the population is divided into two or more segments, and the processing device divides the segment into one or more segments for one segment that includes the specified attribute Deriving the risk factors, the processor may calculate coefficients respectively corresponding to the one or more risk factors of the segment.

The processing device comprises a columnar database storing a plurality of tips, the processing device extracting one or more tips from the columnar database based on expected values or based on the results of a cohort study or meta-analysis; A personalized report containing one or more recommendations may be output.

According to one embodiment of the present invention, analysis is performed on a group basis, and what kind of measures are taken for business operators and members of health insurance associations to effectively reduce the disease risk of the entire group, A health checkup data analysis system capable of automatically analyzing whether future medical expenses can be reduced and creating a report is provided. Or, in one embodiment, for individuals who have undergone health checkups, what kind of measures can be taken to reduce the risk of future diseases and the risk of increased medical costs based on the results of group-based analysis? is automatically analyzed, and a health checkup data analysis system capable of creating a report is provided.

1 is a schematic diagram showing a health checkup data analysis system 100 according to one embodiment of the present invention; FIG. 1 is a block configuration diagram showing an information processing device 110 according to an embodiment of the invention; FIG. FIG. 4 is a flow diagram illustrating processing for reading an analysis data set according to one embodiment of the present invention; FIG. 2 is a flow diagram presenting risk factor countermeasures for a population according to one embodiment of the present invention. FIG. 10 is a flow diagram for automatically creating a report by associating advice columns with risk factors and/or countermeasures for a population according to one embodiment of the present invention; FIG. 4 is a schematic diagram showing part of a report for a group according to one embodiment of the present invention; (a) and (b) are schematic diagrams showing an example of a report prompting improvement for an organization according to an embodiment of the present invention. FIG. 4 is a flow diagram illustrating processing for reading an analysis data set according to one embodiment of the present invention; FIG. 4 is a flow diagram showing a medical cost estimation process according to one embodiment of the present invention. FIG. 4 is a flow diagram for automatically creating a report by associating advice columns with risk factors and/or countermeasures according to an embodiment of the present invention; FIG. 4 is a flow diagram showing an outcome estimation process for individuals according to one embodiment of the present invention. FIG. 10 is a flow diagram for automatically creating a report by associating advice columns with factors and/or countermeasures according to an embodiment of the present invention; It is a schematic diagram which shows an example of the individual report concerning embodiment of this invention. FIG. 10 is a diagram showing improvement results of medical examination behavior by issuing an individualized report according to an embodiment of the present invention;

Hereinafter, a health checkup data analysis system according to the present invention will be described with reference to the drawings. The health checkup data analysis system according to the present invention should not be construed as being limited to the descriptions of the embodiments shown below. In the drawings referred to in this embodiment, the same parts or parts having similar functions are denoted by the same reference numerals, and repeated description thereof will be omitted.

FIG. 1 is a schematic diagram showing a health checkup data analysis system 100 according to one embodiment of the present invention. Health checkup data analysis system 100 includes at least an information processing device selected from server 10 or terminal (for example, personal computer) 30 . Server 10 and terminal 30 are connected to network 150 . The network 150 is a communication network that can be connected to the server 10 and the terminal 30 by means of a so-called directly connected communication cable, wired communication such as the Internet or an intranet, or wireless communication. In one embodiment, the health checkup data analysis system 100 may further include a mobile phone 50 including a smart phone or a tablet 70 .

FIG. 2 is a block configuration diagram showing the information processing device 110 according to one embodiment of the present invention. The information processing device 110 includes, for example, a control device 111 , a storage device 113 , a communication device 115 and a power supply device 117 . The control device 111 includes, for example, a central processing unit (CPU), an operating system (OS), an application program or module that controls the information processing device 110, and the like. The storage device 113 is, for example, an auxiliary storage device such as a hard disk or solid state disk (SSD). The communication device 115 is composed of, for example, a network adapter, an onboard communication chip, etc., and is capable of communicating with terminals via the network 150 . The power supply device 117 is a device that supplies power from the outside to each device of the information processing device 110, and is not particularly limited.

The information processing device 110 further includes an input device 118 and an output device 119 . The input device 118 is, for example, a keyboard, mouse, touch panel, microphone, or the like, and any known device can be arranged. The output device 119 is a display device or a printer, and any known device can be arranged. Furthermore, the information processing device 110 may also include various electronic devices that other known information processing devices have.

In one embodiment, when the server 10 constitutes the information processing device 110 , the above-described keyboard, mouse, or the like connected to the server 10 can be used as the input device 118 , but the input device 118 is connected to the server 10 via the network 150 . Terminal 30 may be used as input device 118 . As the output device 119 , the above-described display device, printer, or the like connected to the server 10 can be used.

In one embodiment, controller 111 includes analyzer 112 . The analysis unit 112 is composed of an application program, a module, or the like that aggregates and statistically analyzes data. The analysis unit 112 may be, for example, an application program or module stored in the storage device 113 and executed by the control device 111 . Considering the data processing amount and statistical analysis load in the analysis unit 112, when the server 10 constitutes the information processing device 110, the analysis unit 112 may be an application program or module executed by the control device 111 of the server 10. Preferably. On the other hand, application programs or modules stored in the storage device 113 may be provided from the server 10 to the terminal 30 and the analysis unit 112 may be configured by the control device 111 of the terminal 30 .

In one embodiment, storage device 113 includes, for example, health checkup result database 114a, item list 114b, and column database 114c. The health checkup result database 114a is a database in which the results of health checkups and the answers to questionnaires collected at the time of health checkups of each individual in a group to be analyzed are stored for a single year or two years or more. Here, "group (or population)" means a group of employees, insured persons, or dependents included in any organizational unit such as a company or health insurance association, or a group of residents included in a municipality. etc. is intended. In one embodiment, the "group (or population)" refers not only to all employees, insured persons, dependents, or residents, but also to employees and insured persons extracted under arbitrary conditions. , dependents or groups of residents. In one embodiment, the health checkup result database 114a is a database that stores the results of health checkups and answers to questionnaires provided by medical institutions. The results of the health checkup and the answers to the questionnaire may be provided to the server 10 via the network 150 from the terminal or server 90 of the medical institution, for example, and the health checkup provided in the form of a report 91 such as a paper medium. and the results of answers to the questionnaire may be input from the terminal 30 and converted into data.

The item list 114b contains items that may be medically risk factors for each of the health abnormalities and diseases to be prevented (hereinafter collectively referred to as outcomes) analyzed by the health checkup data analysis system. A database that stores a list of objective variables and explanatory variables.

The column database 114c stores articles (advice) such as sentences and drawings describing recommended measures based on medical grounds linked to possible output results, or medically known risk coefficients for various outcomes. database.

The health checkup result database 114a, the item list 114b, and the column database 114c are preferably stored in the storage device 113 of the server 10 from the viewpoint of data size and security. On the other hand, when the analysis unit 112 is built in the control device 111 of the terminal 30, these databases may be stored in the storage device 113 of the terminal 30 as long as the data size and security issues are ensured. . Alternatively, when the analysis unit 112 is built in the control device 111 of the terminal 30, even if necessary data is provided to the storage device 113 of the terminal 30 from these databases stored in the storage device 113 of the server 10 good.

[Presentation of a report on improvement measures for organizations for outcomes measured by health checkups]
Health outcomes measured in physical examinations include abnormal weight (obesity or leanness), abnormal blood pressure (hypertension or hypotension), lipid abnormalities, and glucose intolerance. In the following, a method for presenting effective measures for preventing obesity is shown as a specific example, and the health checkup data analysis system 100 according to one embodiment will be described.

<Using analysis data sets>
FIG. 3 is a flow diagram illustrating processing for reading an analysis dataset according to one embodiment of the present invention. A user such as a person in charge of human resources of a company or an employee of a health insurance association designates an outcome to be analyzed from the input device 118 (terminal 30) (S101). In one embodiment, items that can be selected as outcomes are stored in the item list 114b, and the stored outcomes are read from the item list 114b and displayed on the display device of the terminal 30. FIG. For example, the user can specify obesity as an analysis target from among the displayed outcomes.

In addition, the analysis unit 112 reads items that are factors that may be related to outcomes to be analyzed, stored in the item list 114b of the storage device 113 (S103). This item includes objective variables and explanatory variables for items that may be medically risk factors. The analysis unit 112 also reads the health checkup results (measurement values) corresponding to the acquired items for the group to be analyzed from the health checkup result database 114a (S105). In one embodiment, these two steps S103 and S105 obtain all the measurements stored in the physical examination result database 114a and only the measurements corresponding to the items stored in the item list 114b are selected. It may be configured as

If obesity is specified as an analysis target, the read measurement values include, as objective variables, the BMI value in the analysis target year, the difference in BMI values between the analysis target year and the time of the previous health checkup, and whether or not the person is obese. Information of 1 or 0 of this and information of 1 or 0 of whether the weight has increased may be included.

In addition, the read measurement values include sex, age, BMI, smoking habits, exercise habits of moderate intensity or more for 30 minutes or more, walking intensity or more for 1 hour a day at the time before the analysis target year as explanatory variables. The above exercise habits, walking speed, eating speed, eating habits within 2 hours before going to bed, snacking habits after dinner, skipping breakfast, drinking frequency, amount of drinking per drink, and ensuring rest during sleep may be included. .

It should be noted that the items included in these objective variable lists are merely examples of legal medical checkup items at the time of filing of the present application, and if data are available, body fat percentage, visceral fat mass, etc. may also be adopted. can.

In addition, the items included in these explanatory variable lists are only examples of the items listed in the standard health checkup questionnaire established by the Ministry of Health, Labor and Welfare. working hours, commuting time, means of commuting, time and regularity of each meal, snacking habits, eating habits of vegetables, seaweed, mushrooms, eating habits of fish, other proteins Intake habits of raw materials, carbohydrate intake habits, lipid intake habits, frequency of eating out, sleep schedules on weekdays, sleep schedules on holidays, presence/absence of sleep disorders, implementation/non-execution of specific health guidance, etc. can also be adopted.

<Analysis of risk factors>
The analysis unit 112 selects an analysis method according to the read objective variable (S111). The analysis unit 112 performs analysis using the read explanatory variables and objective variables based on the selected analysis method. The method of statistical analysis, for example, multiple regression analysis, logistic regression analysis, can be selected from predictive models such as extended quantification type 1 analysis and machine learning, etc., but not limited thereto, known methods can be used. For example, when the objective variable is a continuous variable such as the latest BMI value or the difference between BMI values (S111), the analysis unit 112 performs multivariate multiple regression analysis (S113). Alternatively, if the objective variable is two or three categorical variables such as whether or not the subject is obese, the analysis unit 112 performs multivariate logistic regression analysis (S115). Alternatively, extended type quantification type 1 analysis or machine learning may be executed regardless of the type of objective variable. Coefficients indicating how much each item (explanatory variable) contributes to the objective variable, such as regression coefficient, standardization coefficient, risk ratio, or item score, etc. is output (S117). In addition, one or more explanatory variables are derived as risk factors, weighted by a standardization coefficient (β), and the risk factors in the group to be analyzed are output (S119). The weighted risk factors in the group to be analyzed may be displayed, for example, on the display device of the terminal 30, or may be output as a paper medium such as a report. Alternatively, as will be described later, a report for the analysis target group may be displayed on each individual's mobile phone 50 or tablet 70 .

In the above analysis, a group having a specific attribute included in the group may be specified as a segment, and analysis may be performed on the segment. The analysis unit 112 reads attributes contained in the health checkup result database 114a and displays them on the display device of the terminal 30 (S201). The user designates one or more of the displayed attributes and designates a group to be analyzed as a segment (S203). Thereby, a population comprising a plurality of medical checkup data and/or at least a plurality of interview results is divided into two or more segments based on one or more specified attributes. The analysis unit 112 can perform the above analysis by extracting from the health checkup result database 114a only the data of a group having attributes matching the specified attribute for one segment.

For example, when it is desired to know information for current countermeasures against obesity, a segment of only obese people can be designated as an analysis target from the population. Alternatively, if the target of analysis is to be limited to males, a male-only segment can be specified from the group, and if the target of analysis is to be limited to those aged 50 and over, a segment of age 50 and over can be specified from the group.

For example, for an organization with 5,000 samples, in a segment targeting "obese people" with a BMI of 25 or more, the range of increase in BMI for one year is specified as the objective variable, and the profile and lifestyle habits of one year ago Table 1 shows the results of performing multiple regression analysis as a method of statistical analysis using , as an explanatory variable.

For example, in this segment of the population, the explanatory variables "age", "previous year's BMI", "daily alcohol consumption" and "walking speed" are statistically significant (p < 0.05), It can be seen that this is a risk factor associated with BMI variation. That is, even if items other than these risk factors in the analyzed segment are improved, the effect on BMI cannot be statistically expected.

For example, in this segment of the population, a 60-year-old male with a current BMI of 27.0, who does not drink alcohol, walks fast, and does not engage in physical activity for more than 1 hour a day will have the following year's change in BMI: So, next year's BMI value can be estimated as follows.

Variation value = 3.868 - 0.013 * 60 - 0.038 - 0.1 * 27 - 0.08 * 0 - 0.168 = 0.182
Next year's BMI = 27 + 0.182 = 27.182

In addition, it can be seen that the absolute value of the standardization coefficient (β) is the largest for the "walking speed" among the lifestyle items. In this segment of the population, it can be seen that "walking speed" is the strongest simple influence within these items.

By referring to the room for improvement in the risk factors derived from the above analysis, it is possible to derive the items (countermeasures) that can be expected to have the greatest improvement effect by improving the entire population. FIG. 4 shows a flow diagram presenting risk factor countermeasures for a population. The analysis unit 112 reads items corresponding to risk factors for the population from the health checkup result database 114a or the item list 114b (S301). The analysis unit 112 extracts the number of persons corresponding to the item corresponding to the risk factor among the risk-bearers in this group from the health checkup result database 114a (S303). From the extracted number of people and the total number of groups, the analysis unit 112 calculates the ratio of the number of people corresponding to the item corresponding to the risk factor (S305).

The analysis unit 112 converts the calculated ratio of the number of people into the ratio of room for improvement of the measures in the group. The ratio of room for improvement is the ratio obtained by subtracting the ratio of the number of people from 100% for the entire population. The analysis unit 112 multiplies the room-for-improvement ratio by the regression coefficient (B) to calculate the expected improvement effect (expected value) in the analysis target group (S307). The analysis unit 112 weights each countermeasure with the calculated expected improvement effect and outputs it. At this time, the countermeasures may be arranged in descending order of expected improvement effect and output. The weighted countermeasures may be displayed on the display device of the terminal 30, or may be output as a paper medium such as a report (S309). Alternatively, as will be described later, a report for the analysis target group may be displayed on each individual's mobile phone 50 or tablet 70 .

For example, in the population shown in Table 1, for each lifestyle (risk factor) that affected the increase in BMI, the number of people with and without desirable habits is distributed as shown in Table 2. .

Here, for example, among the variables in Table 2, if all the people in this group who were "slow walking" were able to walk "fast", the BMI of the whole group in the next year would decrease by -0.087 on average. This is expected and will be the largest of the items.

In this way, by weighting the countermeasures against risk factors, it is possible to clarify the items that are considered to be significant in countermeasures and the items that are not considered to be significant. In addition, by comparing the order of the absolute value of the standardization coefficient (β) and/or the numerical value obtained by multiplying the room for improvement by the coefficient, it is possible to prioritize the measures that can be expected to be effective through improvement in the group, and the analysis results. It is possible to weight the output of

In addition, since the coefficient of each risk factor for the increase or decrease of the outcome is clarified, it is possible to estimate how the outcome will be distributed among the members at any point in the future. For example, in this embodiment, it is possible to estimate the percentage of obese people in the future.

If there is only a single time-series result in the health checkup result database 114a, the above analysis is performed using the explanatory variables of the current year instead of the previous data to calculate a provisional value and refer to it. You can also print the report as However, in that case, the analysis is a cross-sectional analysis, and the ability to predict the future is lower than the longitudinal analysis in which data from two or more time points are analyzed.

<Use of column database>
In one embodiment, the column database 114c contains articles (advice columns) of recommended actions based on medical evidence corresponding to a list of explanatory variables and whether intervention is possible or recommended for that item. A flag indicating whether or not is stored.

FIG. 5 illustrates a flow diagram for automatically generating a report with associated advice columns for risk factors and/or countermeasures according to one embodiment. The analysis unit 112 reads measures derived for the group to be analyzed (S401). The analysis unit 112 reads an advice column corresponding to the read countermeasure from the column database 114c (S403). The analysis unit 112 integrates the derived risk factors and/or countermeasures and the advice columns, and outputs a report for the group from the output device 119 (S405). In reports to populations, risk factors and/or countermeasures are weighted to clearly indicate the importance of risk factors and the effectiveness of countermeasures.

For example, in the above analysis, the column database 114c contains information on how important brisk walking is for maintaining good health, burning fat and maintaining muscle strength, and the importance of continuous exercise. Advice column is invoked.

On the other hand, “age”, “gender” and “BMI in the previous year” are items that cannot be intervened, so there are no columns corresponding to these items, and recommendations for improvement are made based on the results of the analysis. articles are not output. In addition, "amount of alcohol consumed per day" has a significant (p<0.05) and negative coefficient in the analysis, indicating that heavy drinking reduces body weight, but heavy drinking is not medically recommended. Therefore, there is no column corresponding to the recommendation to increase the "drinking amount per day", and an article corresponding to the "drinking amount per day" is not output.

The output device 119 integrates the results of the above analysis and the articles called up in the column database 114c, and produces a report that encourages improvement for the group in the organization as an electromagnetic document or an automatically printed document. can be issued. As an example, part of the report for the group is shown in FIG. FIG. 6 is an example of a report visually showing changes in the proportion of obese people with a BMI of 25 next year when the derived measures are implemented and not implemented.

7(a) and 7(b) show an example of a report for promoting improvement for organizations such as companies and health insurance associations. FIG. 7(a) is an example report for obesity, and FIG. 7(b) is an example report for hypertension. Each report presents a risk factor 1 and a countermeasure 3 with age segments. 6, 7(a), and 7(b) are examples of reports, and reports for groups output by the health checkup data analysis system 100 according to the present invention are not limited to these, Based on the analysis results, risk factors for the population, their countermeasures, and advice columns can be displayed in any format.

[Presentation of a report on improvement measures for organizations to reduce medical expenses or the risk of developing a specific disease]
In one embodiment, the health checkup data analysis system 100 uses the medical insurance claim data owned by the health insurance association to analyze the future medical expenses of the population to be analyzed, the risk of developing a specific disease, and the average of the fluctuations or indicators of the number of patients. can be predicted. In one embodiment, the storage device 113 of the server 10 includes a receipt database 114d. Alternatively, the health checkup data analysis system 100 may be set to refer to the medical insurance claim database 114d of another server that stores medical insurance claim data. The receipt database 114d is a database that stores the disease morbidity and medical expense status of each individual in a group to be analyzed for a single year or two years or more.

The receipt database 114d includes information on the name of a disease indicating whether or not a disease has developed, and information (diagnosis data) indicating whether drug administration data and treatment used for the development of a disease have been used. For example, if the disease name "acute myocardial infarction" is newly added, it means that the subject has suffered an acute myocardial infarction, and if the procedure name "percutaneous coronary angioplasty" newly exists, It means that the subject underwent catheterization for acute myocardial infarction or angina pectoris.

The receipt database 114d also includes information on medical expenses. In addition, information on medical expenses is segmented by multiple organizations, including the Ministry of Health, Labor and Welfare, and segmented into categories such as "malignant neoplasm," "infectious disease," "cerebrovascular disease," and "lifestyle-related disease." Aggregation in any area is also possible.

The health checkup data analysis system 100 according to the present embodiment uses the past lifestyle habits of the analysis target group and the basic disease risk such as the presence of hypertension as explanatory variables to predict the future disease onset and medical treatment of the analysis target group. By conducting an analysis using the occurrence of costs as an objective variable, it is possible to present weighted measures for reducing the future disease risk of the analyzed population or reducing medical costs.

<Using analysis data sets>
FIG. 8 is a flow diagram illustrating processing for reading an analysis data set according to one embodiment of the present invention. A user such as a person in charge of human resources of a company or an employee of a health insurance association designates an outcome to be analyzed from the input device 118 (terminal 30) (S501). In one embodiment, disease items that can be selected as outcomes are stored in the item list 114b, and the stored outcomes are read from the item list 114b and displayed on the display device of the terminal 30. FIG. For example, the user can specify lifestyle-related disease-related medical expenses as an analysis target from among the displayed outcomes.

In addition, the analysis unit 112 reads items that are factors that may be related to outcomes to be analyzed, stored in the item list 114b of the storage device 113 (S503). This item includes objective variables and explanatory variables for items that may be medically risk factors. Also, the analysis unit 112 reads the health checkup result (measurement value) corresponding to the acquired item from the health checkup result database 114a, and compares the total medical expenses of the group with the total medical expenses of the year to be analyzed and the time before that. The difference, the presence or absence of the disease related to the disease, the subtotal of medical expenses, etc. are read from the receipt database 114d (S505).

When reading the above data, a group having a specific attribute included in the group can be designated as a segment, and analysis can be performed on the segment. The analysis unit 112 reads attributes contained in the health checkup result database 114a and/or the medical insurance claim database 114d, and displays them on the display device of the terminal 30 (S601). The user designates one or more displayed attributes to designate a group to be analyzed as a segment (S603). As a result, a plurality of health checkup data and/or at least a plurality of medical interview results and/or a group that constitutes the claim database 114d is divided into two or more segments based on one or more designated attributes. The analysis unit 112 extracts only data of a group having attributes matching the specified attribute from the health checkup result database 114a and the receipt database 114d, and performs the following analysis. Further, when a segment to be analyzed is specified in the terminal 30, the analysis unit 112 can extract only data of attributes matching the segment to be analyzed and perform the following analysis.

For example, if you want to know information for measures against lifestyle-related disease-related medical expenses, you can specify a segment only for lifestyle-related diseases as an analysis target. Alternatively, if the target of analysis is to be limited to males, the male only segment can be specified, and if the target of analysis is to be limited to those aged 50 and over, the segment of age 50 and over can be specified.

If lifestyle-related disease-related medical expenses are specified as the target of analysis, the objective variable will be the total lifestyle-related disease-related medical expenses in the analysis target year and the lifestyle habits in the analysis target year and before. The difference from the disease-related medical expenses total, the occurrence of cardiovascular diseases included in lifestyle-related diseases or subtotal medical expenses, more specifically the occurrence of acute myocardial infarction or the total medical expenses, or other specific Major diseases include cerebral infarction, diabetic nephropathy, non-alcoholic hepatitis, chronic obstructive pulmonary disease and related medical costs. In addition, the presence or absence of onset of malignant neoplasms and the total amount of related medical expenses, the presence or absence of onset of infectious diseases and the total amount of related medical expenses, etc. can be used.

In addition, among the explanatory variables that correspond to the risk of underlying diseases, the presence or absence of hypertension or blood pressure, the presence or absence of dyslipidemia or LDL cholesterol, HDL cholesterol, triglycerides, total cholesterol, LDL/HDL ratio , total cholesterol/HDL ratio, presence or absence of diabetes or fasting blood sugar, HbA1c level, presence or absence of renal dysfunction or creatinine level, eGFR, presence or absence of liver dysfunction or AST value, ALT value, γ-GT value, red blood cell count, hematocrit Values, MCV, MCH, MCHC, gender, age, BMI, etc. may be included.

Of the explanatory variables, those that correspond to the risk of underlying diseases are only examples from the items of the health checkup results included in the standard health checkup specified by the Ministry of Health, Labor and Welfare. fraction, serum iron and ferritin levels, CRP, serum zinc, fatty acid fraction, thyroid function, albumin level, serum total protein level, cholinesterase level, serum electrolytes, platelet count, Fib-4 index, or implementation status of cancer screening As a result, the implementation status of various vaccinations can be included in the analysis.

In addition, as explanatory variables that correspond to lifestyle habits, smoking habits, exercise habits of moderate intensity or more for 30 minutes or more, exercise habits of walking intensity or more for 1 hour or more a day, Walking speed, eating speed, habit of eating dinner within 2 hours before going to bed, habit of snacking after dinner, skipping breakfast, frequency of drinking, amount of alcohol per drink, ensuring rest during sleep, and the like may be included.

Of the explanatory variables, the items included in the list of items that correspond to lifestyle habits are only examples from the items listed in the standard health checkup questionnaire established by the Ministry of Health, Labor and Welfare. Situation, whether or not you live alone, whether or not you have a job, working hours per week, commuting time, means of commuting, time and regularity of each meal, eating habits between meals, eating habits of vegetables, seaweed, and mushrooms, eating fish Habits, other protein intake habits, carbohydrate intake habits, fat intake habits, frequency of eating out, sleep schedules on weekdays, sleep schedules on holidays, sleep disorders, implementation of specific health guidance, etc. can also

<Analysis of risk factors>
The analysis unit 112 selects an analysis method according to the read objective variable (S511). The analysis unit 112 performs analysis using the read explanatory variables and objective variables based on the selected analysis method. Since the method of statistical analysis has been described above, detailed description thereof will be omitted. For example, when the objective variable is a continuous variable (S511), the analysis unit 112 performs multivariate multiple regression analysis (S513). Alternatively, when the objective variable is a categorical variable, the analysis unit 112 performs multivariate logistic regression analysis (S515). Alternatively, extended type quantification type 1 analysis or machine learning may be executed regardless of the type of objective variable. Coefficients indicating how much each item (explanatory variable) contributes to the objective variable, such as regression coefficient, standardization coefficient, risk ratio, or item score, etc. is output (S517). In addition, one or more explanatory variables are derived as risk factors of the group to be analyzed, weighted by a standardization coefficient (β), and output as risk factors (S519). The weighted risk factors of the group to be analyzed may be displayed, for example, on the display device of the terminal 30, or may be output as a paper medium such as a report. Alternatively, as will be described later, a group report may be displayed on each individual's mobile phone 50 or tablet 70 .

For example, for an organization with a sample of 10,000 people, specify the increase in medical expenses for lifestyle-related diseases per person per year as the objective variable, and explain the profile and lifestyle of a year ago. Table 3 shows the results of performing multiple regression analysis as a method of statistical analysis using variables.

For example, in this group, the explanatory variables "participation in specific medical examinations", "gender", "snacking habits", "eating speed", "drinking frequency" and "sleep" significantly affect the growth of medical expenses. I know you do. In particular, when we look at the standardization coefficient (β), we can see that the effects on the growth of medical expenses are in the order of “gender,” “snacking habits,” “sleep,” “participation in specific health checkups,” “frequency of alcohol consumption,” and “eating speed.” I know it's big.

In addition, since the coefficient of each factor for the increase or decrease of the outcome is clarified, it is possible to estimate how the outcome will be distributed among the members of the analyzed group at any point in the future. For example, in the present embodiment, it is possible to estimate the medical expenses for lifestyle-related diseases per person in the future.

FIG. 9 is a flowchart showing a medical cost estimation process according to one embodiment of the present invention. The analysis unit 112 reads the items (explanatory variables) derived from the analysis of the risk factors described above, the variation width coefficients corresponding to the items, and the average values (ratios) from the storage device 113 (S701). The analysis unit 112 uses the loaded parameters to estimate the lifestyle-related disease-related disease medical expenses per person (S703). The analysis unit 112 outputs the estimated lifestyle-related disease-related disease medical expenses per person from the output device 119 (S705). The lifestyle-related disease-related disease medical expenses per person may be displayed, for example, on the display device of the terminal 30, or may be output as a paper medium such as a report. Alternatively, the report may be displayed on the mobile phone 50 or tablet 70 of each individual.

Lifestyle-related disease-related disease medical expenses per person in total can be calculated by the following formula.

Σ (lifestyle-related disease-related medical expenses for each person: item A × coefficient A + item B × coefficient B + item C × coefficient C …) / n = average value (ratio) of item A × coefficient A + average value of item B (ratio) x factor B + average value of item C (ratio) x factor C...

For example, as shown in Table 4, the lifestyle-related disease-related medical expenses per person in this group in the following year are expected to decrease by 799 yen.

If there is only a single time-series result in the health checkup result database 114a, the above analysis is performed using the explanatory variables of the current year instead of the previous data to calculate a provisional value and refer to it. You can also print the report as However, in that case, the analysis is a cross-sectional analysis, and the ability to predict the future is lower than the longitudinal analysis in which data from two or more time points are analyzed.

<Use of column database>
FIG. 10 illustrates a flow diagram for automatically generating a report with associated advice columns for risk factors and/or countermeasures according to one embodiment. The analysis unit 112 reads the measures derived for the group (S501). The analysis unit 112 reads an advice column corresponding to the read countermeasure from the column database 114c (S503). The analysis unit 112 integrates the derived risk factors and/or countermeasures with the advice columns, and outputs a report for the group from the output device 119 (S505). In the report for the population, the risk factors and/or countermeasures in the analyzed population are weighted, and the importance of the risk factors and the effect of the countermeasures are clearly indicated.

For example, in the above analysis, the correct snacking method, sleep advice, importance of specific medical examination, recommendation of appropriate drinking, and importance of slowly chewing and eating well are included in the column database 114c. A medical advice column is called.

On the other hand, since it is impossible to intervene for "sex", there is no column for improvement associated with this item, and articles recommending improvement based on the results of analysis are not output.

The output device 119 integrates the results of the above analysis and the items called up in the column database 114c, and produces an electromagnetic document or an automatically printed document showing the increase in medical expenses at the group level for the organization concerned. issue a report to encourage measures to curb

In addition, if a similar analysis is performed in the same way as the above analysis, for example, with the onset of cardiovascular disease as the target variable, it will be important to identify lifestyle habits such as exercise habits related to the disease and basic risk factors such as hypertension. It can be derived while weighting with degrees. Based on this, a report can be issued prompting measures to reduce disease risk in the population.

[Improvement of health checkup outcomes, reduction of the risk of developing specific diseases, presentation of improvement recommendation reports for individuals to reduce future medical costs]
The health checkup data analysis system 100 calculates coefficients by which various risk factors contribute to health checkout outcomes, the risk of developing specific diseases, or increases in future medical costs through the above-described processing.

Depending on the presence or absence of those risk factors of individuals included in the analyzed group and the situation, by multiplying the calculated coefficients and adding or adding them and statistically processing, it is possible It is possible to estimate whether the patient has such health checkup outcomes, the degree of risk of developing each disease, and the degree of future medical expenses.

FIG. 11 is a flow diagram showing an outcome estimation process for individuals according to one embodiment. The analysis unit 112 reads the risk factors and coefficients derived by the above process (S901). For example, the risk factors and coefficients derived by the above processing may be directly used, or the risk factors and coefficients stored in the storage device 113 may be read.

The analysis unit 112 performs statistical processing using risk factors and coefficients to estimate outcomes (S903). For example, when estimating the risk of developing myocardial infarction, each item such as age, gender, blood pressure, lipid test results, smoking status, glucose intolerance status, etc. is assigned by the above process. Multiply the risk of each item by the coefficient of risk, or calculate the final morbidity probability after summing in the logit model, etc., and calculate the probability of developing myocardial infarction for each person in an arbitrary period in the future as a percentage can be visualized.

At this time, if the group does not have a sufficient amount of data and the number of items that reach the significance level is insufficient, analyze an expanded dataset that combines data from other analysis groups, or , cohort studies, their meta-analyses, and the like, the coefficients of risk factors medically clarified can be called from the column database 114c and estimated.

The estimated outcome may be displayed on the display device of the terminal 30, for example. Further, the report may be output as a paper medium such as a report for personal use, or the report may be displayed on the mobile phone 50 or tablet 70 of the individual.

FIG. 12 illustrates a flow diagram for automatically generating a report with associated advice columns for risk factors and/or countermeasures according to one embodiment. The analysis unit 112 reads the derived measures (S1001). The analysis unit 112 reads an advice column corresponding to the read countermeasure from the column database 114c (S1003). The analysis unit 112 integrates the estimation of health checkup outcomes, the probability of contracting a disease, the prediction of future medical expenses, the derived risk factors and/or countermeasures, and the advice columns clarified from the results of the analysis. Then, the report is output from the output device 119 as an electromagnetic document or an automatically printed document (S1005). In the report, the risk factors and/or countermeasures are weighted, the importance of the risk factors and the effect of the countermeasures are clearly shown, and improvement can be encouraged for each individual.

FIG. 13 shows an example of an individualized report. The individual reports show risk factor 1 and countermeasures and advice column 5. FIG. 13 is an example of an individualized report, and the individualized report output by the health checkup data analysis system 100 according to the present invention is not limited to these. • Columns can be displayed in any format.

Visualizing the risk of developing a serious disease and communicating it to individuals will help to effectively recommend secondary examinations and consultations following the results of health checkups. For example, before and after using the individual report according to the present embodiment in a group of companies with about 1000 employees, the percentage of those who were determined to be "requiring medical examination" in the medical examination actually took the medical examination action , as shown in Fig. 14, the number increased three to four times by issuing individual reports.

As described above, by predicting future risks to an organization or individual and visualizing and weighting the items that have a large improvement effect, we can improve health conditions, prevent diseases, and provide future medical care. The present invention makes it possible to keep costs down.

1 risk factor, 3 countermeasure, 5 advice column, 10 server, 30 terminal, 50 mobile phone, 70 tablet, 90 server, 91 report, 100 health checkup data analysis system, 110 information processing device, 111 control device, 112 analysis Unit, 113 Storage device, 114a Health checkup result database, 114b Item list, 114c Column database, 114d Receipt database, 115 Communication device, 117 Power supply device, 118 Input device, 119 Output device, 150 Network

Claims (10)

Using the plurality of health checkup data and/or the at least a plurality of interview results of a population having a plurality of health checkup data and/or at least a plurality of interview results,
a processor deriving one or more risk factors in the population from the plurality of health examination data and/or plurality of interview results;
the processor calculates coefficients each corresponding to one or more risk factors in the population;
weighting one or more risk factors in the population by the factor;
A health checkup data analysis system, wherein the processor outputs a report for the population. 2. Deriving a predictive model by regression analysis or machine learning using multiple health checkup data and/or multiple interview results of the population to derive one or more risk factors in the population. health checkup data analysis system. 3. The health checkup data analysis system according to claim 1, wherein the plurality of health checkup data and/or the plurality of interview results of the group are single-year data or time-series data. 4. The medical examination data analysis system of claim 3, further using examination data and/or medication data to derive one or more risk factors for the population. 3. The health checkup data analysis system according to claim 2, wherein said predictive model is used to estimate the medical expenses of said group, the number of patients with a disease of said group, or the average index of said group. applying one or more measures of the population corresponding to one or more risk factors of the population to the predictive model to reduce medical costs of the population, disease incidence of the population, or indicators of the population; 6. The health checkup data analysis system according to claim 5, which estimates an average. applying one or more countermeasures corresponding to one or more risk factors of the population to the predictive model;
6. The health checkup data analysis system of claim 5, wherein the population measures are weighted. one or more of the population from which the plurality of risk factors have been derived have one or more risk factors among the one or more risk factors of the population for weighting and presenting one or more countermeasures of the population the processing device calculates the proportion of risk-holders for whom it is effective to implement any one or more measures of the population, and
8. The health checkup data analysis system according to claim 7, wherein for each of the one or more measures of the group, the processing device calculates an expected value of improvement effect from the ratio and the coefficient. The processing device accepts designation of one or more attributes included in the plurality of health checkup data of the group and/or the results of at least a plurality of interviews of the group,
The processing device divides a group comprising a plurality of medical examination data of the group and/or at least a plurality of interview results of the group into two or more segments based on the one or more attributes;
The processing unit derives one or more risk factors of the segment for a segment containing the specified attribute;
2. The medical examination data analysis system of claim 1, wherein the processor calculates coefficients respectively corresponding to one or more risk factors of the segments. the processing unit comprises a columnar database storing a plurality of tips;
The processor extracts one or more recommendations from the columnar database based on the expected value or based on results of a cohort study or meta-analysis;
9. The health checkup data analysis system of claim 8, outputting a personalized report including the one or more recommendations.

Images