KR102503931B1 - Apparatus and method for predicting medical expenditures through big data analytics - Google Patents

Abstract

An apparatus and method for predicting medical expenses through big data analysis are provided. The device includes a communication unit that receives medical-related big data, a memory having at least one program for predicting medical expenses, and executing the program to obtain statistical data for each variable included in the medical-related big data through the program. and a control unit that extracts first result data by using an analysis model based on the statistical data for each variable, and predicts medical expenses for medical services based on the first result data.

Description

Medical cost prediction device and method through big data analysis {APPARATUS AND METHOD FOR PREDICTING MEDICAL EXPENDITURES THROUGH BIG DATA ANALYTICS}

The present invention relates to an apparatus and method for predicting medical expenses through big data analysis.

Currently, Korea is advocating an inclusive welfare state and promoting community care policies. In order to meet the increasing needs of the elderly in an aging society for nursing care and care, 'visiting health care (including visiting nursing)' and 'home visit medical care' are provided for the purpose of health and medical security, while providing housing support and long-term care It is to provide integrated care within the community by linking in-home services such as food delivery and meal delivery. This is to ensure a comfortable old age and improve the life satisfaction of the elderly by providing community-centered medical and welfare services in an integrated way so that health care and daily life can be carried out efficiently at home.

However, as with the 'home nursing service' provided by professional nurses in medical institutions, the health needs of elderly patients with high severity and the provision of professional nursing have not yet been established. Therefore, it is necessary to conduct research for settlement in the local community, such as how to connect with community care in the future through an analysis of the current state of home nursing service.

Registered Patent Publication No. 10-0597289, 2006.06.28.

The problem to be solved by the present invention is to provide an apparatus and method for predicting medical expenses through big data analysis.

However, the problem to be solved by the present invention is not limited to the above problem, and other problems may exist.

An apparatus for predicting medical expenses through big data analysis according to an embodiment of the present invention for solving the above problems is a communication unit for receiving medical-related big data, a memory having at least one program for predicting medical expenses, and the program Execute, extract statistical data for each variable included in the medical big data through the program, extract first result data using an analysis model based on the statistical data for each variable, and extract the first result data Based on the control unit for predicting the medical expenses for home care.

In the present invention, the control unit extracts the first result data by applying current status data extracted from the statistical data for each variable to the analysis model, and the current status data may be extracted in consideration of at least one variable.

In the present invention, the control unit determines whether each variable is meaningful to the analysis model based on the first value for each variable included in the first result data, and predicts the medical cost excluding variables determined to be insignificant. can

In the present invention, the control unit may determine that the first value is not significant for a variable greater than a preset reference value.

In the present invention, the control unit may predict the treatment cost using the second value of the variable determined to be significant.

In the present invention, the first result data includes first result data related to the first group and first result data related to the second group, and the control unit, based on the first result data related to the first group, , It is possible to predict medical expenses for home care of the first group, and predict medical expenses for home care of the second group based on the first result data related to the second group.

In the present invention, the first group may include at least one private medical institution, and the second group may include at least one public medical institution.

In the present invention, the control unit may extract second result data using an analysis model based on statistical data for each variable, and predict the number of medical service uses based on the second result data.

In the present invention, the variable may include at least one of age, year, number of days of recuperation, operation status, region, gender, treatment course, main illness, minor illness, type of insurer, type of care institution, and group.

A method for predicting medical expenses through big data analysis according to an embodiment of the present invention for solving the above problems includes receiving medical-related big data, executing at least one program for predicting medical expenses, and executing the program. extracting statistical data for each variable included in the medical big data, extracting first result data using an analysis model based on the statistical data for each variable, and assuming based on the first result data. Estimating the cost of nursing care.

In the present invention, in the step of extracting the first result data, the first result data is extracted by applying the current status data extracted from the statistical data for each variable to the analysis model, and the current status data is obtained by considering at least one variable can be extracted.

In the present invention, the medical cost prediction step is the step of determining whether each variable is meaningful to the analysis model based on the first value for each variable included in the first result data, and excluding variables determined to be insignificant A step of estimating the treatment cost may be included.

In the present invention, in the significance determination step, it may be determined that the variable having the first value greater than a preset reference value is not significant.

In the present invention, the step of estimating the medical expense may include predicting the medical expense using a second value of a variable determined to be significant.

In the present invention, the first result data includes first result data related to the first group and first result data related to the second group, and the medical expense prediction step is based on the first result data related to the first group. Based on the first group, the treatment cost for home care may be predicted, and based on the first result data related to the second group, the treatment cost for home care for the second group may be predicted.

In the present invention, the first group may include at least one private medical institution, and the second group may include at least one public medical institution.

In the present invention, the step of extracting second result data using an analysis model based on the statistical data for each variable, and predicting the number of cases of home care based on the second result data may further include. .

In the present invention, the variable may include at least one of age, year, number of days of recuperation, operation status, region, gender, treatment course, main illness, minor illness, type of insurer, type of care institution, and group.

In addition to this, another method for implementing the present invention, another device, another system, and a computer readable recording medium recording a computer program for executing the method may be further provided.

According to the present invention described above, it is possible to check the contribution to medical expenses for each variable through analysis of past medical-related big data.

In addition, it is easy to set the direction of linkage between home care service and community care by separately predicting the cost of medical service related to home care and the cost of medical service related to the community and comparing the two values.

Through this, it is possible to efficiently distribute medical resources or establish effective medical policies.

The effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description below.

1 is a block diagram of an apparatus for predicting medical expenses according to the present invention.
2 is a flowchart of a method for predicting medical expenses according to the present invention.
3 is an exemplary diagram for explaining statistical data for each variable.
4 is a diagram for explaining current status data related to the number of medical institutions providing home care by year.
5 and 6 are diagrams for explaining current status data related to the number of home care cases by year, by insurer type, and by region.
7 and 8 are diagrams for explaining current state data related to total home care medical expenses by year, by insurer type, and by region.
9 to 11 are diagrams for explaining current status data related to the number of cases of home care by year, age, insurer type, and gender.
12 and 13 are diagrams for explaining the current state data related to the number of home care use by year, by sick sickness and group, and by year, by injured soldier and by group.
14 and 15 are diagrams for explaining prediction of medical expenses using first result data.
16 is a diagram for explaining the number of uses using second result data.

Advantages and features of the present invention, and methods of achieving them, will become clear with reference to the detailed description of the following embodiments taken in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, only these embodiments are intended to complete the disclosure of the present invention, and are common in the art to which the present invention belongs. It is provided to fully inform the person skilled in the art of the scope of the invention, and the invention is only defined by the scope of the claims.

Terminology used herein is for describing the embodiments and is not intended to limit the present invention. In this specification, singular forms also include plural forms unless specifically stated otherwise in a phrase. As used herein, "comprises" and/or "comprising" does not exclude the presence or addition of one or more other elements other than the recited elements. Like reference numerals throughout the specification refer to like elements, and “and/or” includes each and every combination of one or more of the recited elements. Although "first", "second", etc. are used to describe various components, these components are not limited by these terms, of course. These terms are only used to distinguish one component from another. Accordingly, it goes without saying that the first element mentioned below may also be the second element within the technical spirit of the present invention.

Unless otherwise defined, all terms (including technical and scientific terms) used in this specification may be used with meanings commonly understood by those skilled in the art to which the present invention belongs. In addition, terms defined in commonly used dictionaries are not interpreted ideally or excessively unless explicitly specifically defined.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of an apparatus for predicting medical expenses according to the present invention.

2 is a flowchart of a method for predicting medical expenses according to the present invention.

3 is an exemplary diagram for explaining statistical data for each variable.

4 is a diagram for explaining current status data related to the number of medical institutions providing home care by year.

5 and 6 are diagrams for explaining current status data related to the number of home care cases by year, by insurer type, and by region.

7 and 8 are diagrams for explaining current state data related to total home care medical expenses by year, by insurer type, and by region.

9 to 11 are diagrams for explaining current status data related to the number of cases of home care by year, age, insurer type, and gender.

12 and 13 are diagrams for explaining the current state data related to the number of home care use by year, by sick sickness and group, and by year, by injured soldier and by group.

14 and 15 are diagrams for explaining prediction of medical expenses using first result data.

16 is a diagram for explaining the number of uses using second result data.

Referring to FIG. 1 , a medical expense prediction apparatus 100 (hereinafter referred to as apparatus) includes a communication unit 120, a memory 140, and a control unit 160.

However, the device 100 in the present invention may include fewer or more components than those shown in FIG. 1 .

In the present invention, the device 100 may mean a server device for predicting medical expenses. For example, it may be a server device of a medical institution or a server device of a government institution, but is not limited thereto.

The communication unit 120 may include one or more modules enabling wireless communication between the device 100 and a wireless communication system, between the device 100 and a user terminal, or between the device 100 and an external server. Also, the communication unit 12 0 may include one or more modules that connect the device 100 to one or more networks.

The memory 140 may store data supporting various functions of the device 100 . The memory 140 may store a plurality of application programs (application programs or applications) running in the device 100 , data for operation of the device 100 , and commands. At least some of these application programs may exist for basic functions of the device 100 . Meanwhile, the application program may be stored in the memory 140, installed on the device 100, and driven by the control unit 160 to perform an operation (or function) of the device 100.

The controller 160 may control general operations of the device 100 in addition to operations related to the application program. The control unit 160 may provide or process appropriate information or functions to the user by processing signals, data, information, etc. input or output through the components described above or by running an application program stored in the memory 140.

In addition, the controller 160 may control at least some of the components discussed in conjunction with FIG. 1 in order to drive an application program stored in the memory 140 . Furthermore, the controller 160 may combine and operate at least two or more of the components included in the device 100 to drive the application program.

Although not shown in FIG. 1 , the apparatus 100 may further include a display unit to display a predicted result on the display unit. At this time, the display unit may be implemented with various display devices such as a liquid crystal display (LCD), a light emitting diode (LED), a plasma display panel (PDP), and an organic light emitting diode (OLED).

Hereinafter, a specific method of predicting medical expenses through big data analysis using the device 100 including the communication unit 120, the memory 140, and the control unit 160 will be described.

The communication unit 120 receives medical-related big data (S310).

Specifically, the communication unit 120 may receive medical-related big data from a server of an external institution, for example, a server of the Health Insurance Review and Assessment Service.

Here, the medical related big data may include general specifications, medical treatment details, and injury/disease details.

The general details of the statement are the statement join key, insurer type classification code, examinee's personal identification replacement key, gender group, examinee's age, examinee's statistical age, nursing institution identification replacement key, medical institution type classification code, area code, format classification code, address Disease code, injured disease code, treatment subject code, treatment start date, treatment end date, hospitalization days, treatment days, review decision co-payment, review decision insurer co-payment, whether or not surgery, treatment result classification code, hospitalization arrival route classification code, medical It may include information related to a salary type code, a marked subject code, etc., but is not limited thereto, and the included information may differ depending on the user's (or manager's) selection.

Treatment details are specification join key, item code, classification code division, classification code, daily dose or number of times of administration, total number of days of administration or number of times of administration, total amount of use, unit price, amount, additional amount applied, common name code, treatment exception classification It may include information related to code, 1_2 division code, etc., but is not limited thereto, and the included information may be different depending on the user's (or manager's) selection.

Injury details may include, but are not limited to, information related to specification join key, serial number, private disease code, disease code, disease classification, treatment department code, internal medicine subspecialty, etc., and the user (or manager) Depending on your selection, the information included may vary.

Next, the controller 160 executes at least one program for predicting medical expenses (S320).

Next, the control unit 160 extracts statistical data for each variable included in medical big data through the program (S330) .

Here, the variable may be determined based on various pieces of information included in the medical big data.

Specifically, the variable may include at least one of age, year, number of days of medical treatment, whether or not surgery was performed, region, gender, treatment category, primary disease, minor disease, type of insurer, type of medical institution, and group.

3 is a diagram showing statistical data for each variable generated by classifying according to variables using a total of 32,774,706 pieces of collected medical big data.

3A shows statistical data for age.

For example, the controller 160 sets the age to 9 or less, 10 or more to 19 or less, 20 or more to 29 or less, 30 or more to 39 or less, 40 or more to 59 or less, 50 or more to 59 or less, 60 or more to 69 or less, 70 or more to 79 or less, 80 Statistical data representing the frequency of home nursing care can be extracted by classifying them into 89 or more, 90 or more, 99 or less, and 100 or more.

Referring to Figure 3a, it can be seen that the frequency of use is high in the 70's and 80's.

Figure 3b shows statistical data for each type of health care institution.

For example, the control unit 160 classifies the types of nursing institutions into tertiary general hospitals, general hospitals, hospitals, nursing hospitals, clinics, public health centers, health branch offices, health clinics, health care centers, and oriental medicine hospitals, and uses home nursing for each medical institution. Statistical data representing frequencies can be extracted.

Referring to Figure 3b, it can be seen that the frequency of use of the clinic is high.

3C shows statistical data for groups.

For example, the control unit 160 may classify groups into private medical institutions and public medical institutions, and extract statistical data indicating frequency of home care use for each group.

At this time, private medical institutions may include tertiary general hospitals, general hospitals, hospitals, nursing hospitals, and clinics, which are expressed as the first group in this specification.

In addition, public medical institutions may include public health centers, health branch offices, health clinics, health care centers, and oriental medicine hospitals, which are represented as a second group in this specification.

Referring to FIG. 3C , it can be seen that the frequency of use of private medical institutions (ie, the first group) is much higher.

3D shows statistical data for each insurer type.

For example, the controller 160 may classify the type of insurer into health insurance and medical benefit, and extract statistical data representing the frequency of using home care for each type.

Referring to FIG. 3D , it can be seen that the frequency of use of insurers applying health insurance is high.

3E shows statistical data for regions.

For example, the controller 160 classifies the region into Seoul, Busan, Incheon, Daegu, Gwangju, Daejeon, Ulsan, Gyeonggi, Gangwon, Chungbuk, Chungnam, Jeonbuk, Jeonnam, Gyeongbuk, Gyeongnam, Jeju, and Sejong, Statistical data representing the frequency of home care use can be extracted.

Referring to Figure 3e, it can be seen that the frequency of use in Seoul is high.

3F shows statistical data for gender.

For example, the controller 160 may classify gender into male and female, and extract statistical data representing the frequency of using home care for each gender.

Referring to Figure 3f, it can be seen that the frequency of use of women is high.

Figure 3g shows statistical data for year.

For example, the controller 160 classifies the year into 2014, 2015, 2016, 2017, 2018, and 2019, and extracts statistical data representing the frequency of home care use for each year.

Referring to Figure 3g, it can be seen that the frequency of use in 2014 is the highest.

Fig. 3h shows statistical data for treatment subjects.

For example, the controller 160 may classify the treatment subjects into internal medicine, surgery, other technical medicine, dentistry, and oriental medicine, and extract statistical data representing the frequency of home care use for each treatment subject.

Referring to Figure 3h, it can be seen that the frequency of use of the internal medicine system is the highest.

Figure 3i shows statistical data for main illness.

For example, the controller 160 classifies the main illness into hypertension, ischemic heart disease, heart failure, diabetes, stroke, dementia, cancer, dyslipidemia, and renal failure, and extracts statistical data representing the frequency of home care use for each main illness. can do.

Referring to FIG. 3i , it can be seen that the frequency of use of dementia, hypertension, and cancer is sequentially high.

3J shows statistical data for injured soldiers.

For example, the controller 160 classifies injured diseases into hypertension, ischemic heart disease, heart failure, diabetes, stroke, dementia, cancer, dyslipidemia, and renal failure, and extracts statistical data representing the frequency of home care use for each injured disease. there is.

Referring to FIG. 3j , it can be seen that the frequency of use of diabetes and hypertension is sequentially high.

Next, the controller 160 extracts first result data using an analysis model based on the statistical data for each variable (S340).

Specifically, the controller 160 extracts first result data by applying current status data extracted from statistical data for each variable to an analysis model.

At this time, the current status data may be extracted in consideration of at least one variable.

Hereinafter, with reference to FIGS. 4 to 12 , current status data extracted in consideration of at least one variable will be described in detail.

Figure 4 shows the current data extracted by considering the year and the type of health care institution among the variables.

Specifically, FIG. 4a shows the frequency of use by year of private medical institutions (first group) among the health care institutions included in the type of health care institution, and FIG. Indicates the frequency of use by year.

Referring to Figure 4a, it can be seen that among private medical institutions (first group), the institution with the highest total frequency of use during the total period (2014-2019) is a clinic, and the year with the highest frequency of use among each year is 2014 can

Referring to Figure 4b, it can be seen that among public medical institutions (group 2), the public health center has the highest total frequency of use during the total period (2014-2019), and the year with the highest frequency of use among each year is 2014 can

5 and 6 show current data extracted by considering year, insurer type, and region among variables together.

5A shows the frequency of use by year and by type of insurer for Seoul among regions. Referring to Figure 5a, it can be seen that the frequency of use of insurers applying health insurance is higher than that of medical benefits during the total period (2014-2019), and it can be seen that 2014 is the year with the highest frequency of use among each year.

5B shows the frequency of use by year and by insurer type for Busan among regions. Referring to FIG. 5B, it can be seen that the frequency of use of insurers applying health insurance is higher than that of medical benefits during the total period (2014-2019), and it can be seen that 2014 is the year with the highest frequency of use among each year.

In this way, when the frequency of use is calculated for the remaining regions (Incheon, Daegu, Gwangju, Daejeon, Ulsan, Gyeonggi, Gangwon, Chungbuk, Chungnam, Jeonbuk, Jeonnam, Gyeongbuk, Gyeongnam, Jeju, and Sejong), the current status data as shown in FIG. is extracted

Referring to FIG. 6 , it can be seen that 2014 is the year in which the frequency of use of health insurance and medical benefits is the highest in each year.

7 and 8 show current data extracted by considering year, insurer type, and region among variables together.

7a to 7f show tables calculated by year of the average total medical expenses when using home nursing once in Seoul among regions.

In FIGS. 7A to 7F , the number of observations represents the total number of cases of home nursing care. That is, FIGS. 7a to 7f show the average of the total medical expenses, which are the sum of all medical expenses paid each time patients in the Seoul area used home nursing care in 2014, 2015, 2016, 2017, 2018, and 2019, respectively. is a table representing In the following, the average total medical expenses for one home care use is indicated as the total medical expenses.

Referring to FIG. 7A , it can be seen that in 2014, the total medical cost applied by health insurance in Seoul is calculated as 63191.5, and the total medical cost applied by medical benefit is calculated as 63162.8.

Referring to FIG. 7B, it can be seen that in 2015, the total medical cost applied by health insurance in Seoul is calculated as 70522.3, and the total medical cost applied by medical benefit is calculated as 71608.3.

Referring to FIG. 7c , it can be seen that in 2016, the total medical cost applied by health insurance in Seoul is calculated as 79914.6, and the total medical cost applied by medical benefit is calculated as 75997.5.

Referring to FIG. 7D , it can be seen that in 2017, the total medical cost applied by health insurance in Seoul is calculated as 90238.5, and the total medical cost applied by medical benefit is calculated as 74756.1.

Referring to FIG. 7E, it can be seen that in 2018, the total medical cost applied by health insurance in Seoul is calculated as 98240.5, and the total medical cost applied by medical benefit is calculated as 78714.3.

Referring to FIG. 7f , it can be seen that in 2019, the total medical cost applied by health insurance in Seoul is calculated as 100071, and the total medical cost applied by medical benefit is calculated as 88102.8.

In this way, if the total medical expenses by year are calculated for the remaining regions (Busan, Incheon, Daegu, Gwangju, Daejeon, Ulsan, Gyeonggi, Gangwon, Chungbuk, Chungnam, Jeonbuk, Jeonnam, Gyeongbuk, Gyeongnam, Jeju, Sejong), Figure 8 and The same status data is extracted.

Referring to FIG. 8 , it can be seen that 2019 is the year in which the total medical expenses combined with health insurance and medical benefits is the highest in each year.

9 to 11 show current data extracted by considering year, age, insurer type, and gender among variables.

Figure 9a shows the frequency of use of health insurance and medical benefit by year for males aged 9 years or younger, and Figure 9b shows the frequency of using health insurance and medical benefit applied for females aged 9 or younger by year. indicate

Referring to Figure 9a, it can be seen that in the case of males aged 9 years or younger, the frequency of use of applying health insurance is much higher than that of medical benefits during the total period (2014-2019), and although the frequency of use is similar for each year, Among them, it can be seen that the frequency of use was the highest in 2015.

Referring to Figure 9b, in the case of women aged 9 years or younger, it can be seen that the frequency of use of applying health insurance is much higher than that of medical benefit during the total period (2014-2019), and although the frequency of use is similar for each year, Among them, it can be seen that the frequency of use was the highest in 2016.

10 shows a table in which the average age of people (recipients) who received treatment during 2014 was calculated.

Referring to FIG. 10 , it can be seen that the average age of all examinees who received treatment in 2014 was 72 years old.

As shown in Figure 9, the remaining ages (10 or more 19 or less, 20 or more 29 or less, 30 or more 39 or less, 40 or more 59 or less, 50 or more 59 or less, 60 or more 69 or less, 70 or more 79 or less, 80 or more 89 or less, 90 or more 99 Hereinafter, 100 or more), the frequency of use of health insurance and medical benefits is calculated by gender, and the average age of examinees is calculated for the remaining years (2015-2019) as shown in FIG. 10, the same status as in FIG. data is extracted.

That is, FIG. 11 is current status data showing the frequency of use of health insurance and the frequency of use of medical benefit by year according to the calculated age and gender, and the average age of examinees by year.

Referring to FIG. 11, it can be seen that the average age of examinees in each year (2014-2019) is almost similar to 70, and it can be seen that the age groups showing the highest frequency of use in all years are in their 70s and 80s. And, as a whole, the frequency of use of health insurance is higher than that of medical insurance, and when the age is young, the frequency of use of men is high, but as the age increases, the frequency of use of women increases.

12 shows current data extracted by considering year, disease, and group among variables together.

Figure 12a shows the change in the frequency of use of private medical institutions (ie, the first group) by disease according to the year. Referring to FIG. 12A, it can be seen that private medical institutions (ie, the first group) had the highest frequency of use for cancer during the entire period (2014-2019), and that the year with the highest frequency of use among each year was 2017.

Figure 12b shows the change in the frequency of use for each main disease according to the year of the public medical institution (ie, the second group). Referring to FIG. 12B, it can be seen that public medical institutions (ie, the second group) had the highest frequency of use for hypertension during the entire period (2014-2019), and that the year with the highest frequency of use among each year was 2014.

13 shows current data extracted by considering the year, injured soldier, and group among the variables together.

Figure 13a shows the change in the frequency of use for each injured soldier according to the year of the private medical institution (ie, the first group). Referring to Figure 13a, private medical institutions (i.e., the first group) had the highest frequency of use for hypertension during the entire period (2014-2019), and among each year, 2016 and 2017 had the highest frequency of use at a similar number sound can be heard

Figure 13b shows the change in the frequency of use for each injured soldier according to the year of the public medical institution (ie, the second group). Referring to FIG. 13B, it can be seen that public medical institutions (ie, the second group) had the highest frequency of use for hypertension during the entire period (2014-2019), and the year with the highest frequency of use among each year was 2017.

As described above, the control unit 160 extracts various current status data by complexly analyzing various variables, and extracts first result data by applying it to an analysis model.

Here, the analysis model may use regression analysis, but is not limited thereto, and for example, descriptive statistical analysis and ANOVA may be used.

Here, the first result data includes first result data related to the first group and first result data related to the second group.

That is, the control unit 160 separates and extracts result data for private medical institutions and result data for public medical institutions. 14 is a table showing first result data for predicting medical expenses for private medical institutions, and FIG. 15 is a table showing first result data for predicting medical expenses for public medical institutions.

Next, the controller 160 predicts the medical expenses for home care based on the first result data (S350).

The controller 160 predicts the medical expenses for the home care of the first group based on the first result data related to the first group, and based on the first result data related to the second group, the second group of home care predict the cost of treatment for

As described above, the first group includes at least one private medical institution. For example, private medical institutions in the first group may include, but are not limited to, tertiary general hospitals, general hospitals, hospitals, nursing hospitals, and clinics.

Also, the second group includes at least one public medical institution. For example, public health institutions may include, but are not limited to, public health centers, health branch offices, health clinics, health care centers, oriental hospitals.

The controller 160 determines whether each variable is meaningful to the analysis model based on the first value for each variable included in the first result data.

Here, the first value is a p-value (Pr) value determined according to the absolute value of tValue of the first result data shown in FIGS. 14 and 15 (ie, the probt value shown in FIGS. 14 and 15) it means.

The control unit 160 may determine that the first value is not significant for a variable greater than a preset reference value.

Here, the preset reference value may be 0.05, but is not limited thereto.

And, according to the determination result, the control unit 160 predicts the medical expenses excluding variables determined to be insignificant.

That is, the medical expenses are predicted using only the variables determined to be significant, and at this time, the controller 160 may predict the medical expenses using the second value of the variables determined to be significant.

Here, the second value means an estimate (contribution) value of the first result data shown in FIGS. 14 and 15 .

Referring to FIG. 14 , when the reference value is set to 0.05, variables having a first value greater than 0.05 are Jeju variables among regions and oriental, dental, surgical, and internal medicine variables among treatment subjects, so the control unit 160 excludes them. And the second value of the remaining variables, that is, the contribution, can be used to predict the medical expenses for private medical institutions.

Referring to FIG. 15 , when the reference value is set to 0.05, variables having a first value greater than 0.05 are either oriental and internal medicine variables among treatment subjects, and renal failure and heart failure variables among main diseases, so the control unit 160 excludes them. The treatment cost for the public medical institution may be predicted using the second value of the remaining variables, that is, the contribution.

In the above, it has been described that a univariate regression analysis method was used to predict medical expenses, but it is not limited thereto, and a multivariate regression analysis method may also be used.

Meanwhile, although not shown in FIG. 2 , the present invention further includes extracting, by the control unit 160, second result data using an analysis model based on the statistical data for each variable.

At this time, second result data is extracted by applying the statistical data for each variable extracted in S350 to the analysis model. 16A is a table showing second result data for predicting the number of uses per person in one year, and FIG. 16B is a table showing second result data for predicting the number of uses per person for six years.

Similar to the medical cost prediction, the controller 160 determines whether each variable is meaningful to the analysis model based on the first value for each variable included in the second result data.

Here, the first value means a p-value (Pr) value determined according to the magnitude of the absolute value of tValue of the second result data shown in FIG. 16 .

The control unit 160 may determine that the first value is not significant for a variable greater than a preset reference value.

Here, the preset reference value may be 0.05, but is not limited thereto.

And, according to the determination result, the control unit 160 predicts the number of uses (frequency of use) excluding variables determined to be insignificant.

That is, the number of uses is predicted using only the variables determined to be significant. At this time, the controller 160 may predict the number of uses using the second value of the variable determined to be significant.

Here, the second value means an estimate (contribution) value of the second result data shown in FIG. 16 .

Referring to FIG. 16A , when the reference value is set to 0.05, since a variable having a first value greater than 0.05 is a Seoul variable among regions, the controller 160 excludes this variable and uses the second value of the remaining variables, that is, the contribution, to calculate 1 It is possible to predict the number of uses per person per year.

Referring to FIG. 16B, when the reference value is set to 0.05, since a variable having a first value greater than 0.05 is a Seoul variable among regions, the control unit 160 excludes this variable and uses the second value of the remaining variables, that is, the contribution, to calculate 1 It is possible to predict the number of uses per person for 6 years.

By comparing the predicted treatment costs for private medical institutions and public medical institutions, variables that have a large impact in each case are identified, and the reasons for the non-preference for public medical institutions and the preference for private medical institutions are determined by which variables. can figure out whether In addition, by comparing the number of cases of use for 6 years per person with the number of cases of use for 6 years per person in private medical institutions, it is possible to predict the number of cases for home care in the short term (1 year) and long term (6 years). Through this, appropriate policies can be established and medical resources can be efficiently allocated, so that patients can use home nursing conveniently and with high satisfaction in private medical institutions as well as public medical institutions.

2 describes that steps S310 to S350 are sequentially executed, but this is merely an example of the technical idea of this embodiment, and those skilled in the art to which this embodiment belongs will Since it will be possible to change and execute the order described in FIG. 3 without departing from the essential characteristics or to perform various modifications and variations by executing one or more steps of steps S310 to S350 in parallel, FIG. 2 is shown in a time-series order. It is not limited.

The method in one embodiment of the present invention described above may be implemented as a program (or application) to be executed in combination with a computer, which is hardware, and stored in a medium.

The above-mentioned program is C, C++, JAVA, Ruby, C, C++, JAVA, Ruby, which the processor (CPU) of the computer can read through the device interface of the computer so that the computer reads the program and executes the methods implemented as a program. It may include a code coded in a computer language such as machine language. These codes may include functional codes related to functions defining necessary functions for executing the methods, and include control codes related to execution procedures necessary for the processor of the computer to execute the functions according to a predetermined procedure. can do. In addition, these codes may further include memory reference related codes for which location (address address) of the computer's internal or external memory should be referenced for additional information or media required for the computer's processor to execute the functions. there is. In addition, when the processor of the computer needs to communicate with any other remote computer or server in order to execute the functions, the code uses the computer's communication module to determine how to communicate with any other remote computer or server. It may further include communication-related codes for whether to communicate, what kind of information or media to transmit/receive during communication, and the like.

The storage medium is not a medium that stores data for a short moment, such as a register, cache, or memory, but a medium that stores data semi-permanently and is readable by a device. Specifically, examples of the storage medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc., but are not limited thereto. That is, the program may be stored in various recording media on various servers accessible by the computer or various recording media on the user's computer. In addition, the medium may be distributed to computer systems connected through a network, and computer readable codes may be stored in a distributed manner.

The above description of the present invention is for illustrative purposes, and those skilled in the art can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as distributed may be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and equivalent concepts should be construed as being included in the scope of the present invention. do.

100: medical cost prediction device
120: communication department
140: memory
160: control unit

Claims (19)

A communication unit receiving medical big data;
a memory having at least one program for predicting medical expenses; and
run the program,
Extract statistical data for each variable included in the medical big data through the program;
Extracting first result data using an analysis model based on the statistical data for each variable;
A control unit for predicting a medical expense for home care based on the first result data;
The control unit,
Extracting the first result data by applying the current data extracted from the statistical data for each variable to the analysis model,
The current status data is extracted by considering two or more variables of age, year, number of days of medical treatment, surgery status, region, gender, treatment category, main disease, injured disease, type of insurer, type of medical institution, and group,
The control unit,
When the year and the type of health care institution are considered as variables, the medical institution with the highest frequency of use among a plurality of medical institutions during a specific period and the year with the highest total frequency of use by summing the frequencies of use of each of the plurality of medical institutions during the specific period Extract the current status data,
When the year, the type of insurer, and the region are considered as variables, the insurance type with the highest frequency of use among a plurality of insurance types during the specific period for each region, and among the specific period for each region Extracting current status data indicating the year in which the total frequency of use is the highest by summing the frequency of use of each of the plurality of types of insurance,
Medical cost prediction device through big data analysis. delete According to claim 1,
The control unit,
Determine whether each variable is significant to the analysis model based on a first value for each variable included in the first result data;
Excluding variables judged to be insignificant and predicting the medical expenses,
Medical cost prediction device through big data analysis. According to claim 3,
The control unit,
Determining that the first value is not significant for a variable greater than a preset reference value,
Medical cost prediction device through big data analysis. According to claim 3,
The control unit,
Predicting the medical expense using the second value of the variable determined to be significant,
Medical cost prediction device through big data analysis. According to claim 1,
The first result data,
Includes first result data related to the first group and first result data related to the second group;
The control unit,
Based on the first result data related to the first group, predicting the medical expenses for home care in the first group;
Based on the first result data related to the second group, predicting the treatment cost for home care in the second group,
Medical cost prediction device through big data analysis. According to claim 6,
The first group includes at least one private medical institution, and the second group includes at least one public medical institution.
Medical cost prediction device through big data analysis. According to claim 1,
The control unit,
Extracting second result data using an analysis model based on the statistical data for each variable;
Predicting the number of uses for home care based on the second result data,
Medical cost prediction device through big data analysis. delete In the method of predicting medical expenses through big data analysis performed by a medical expense predicting device,
Receiving medical big data;
Executing at least one program for predicting medical expenses;
extracting statistical data for each variable included in the medical big data through the program;
extracting first result data using an analysis model based on the statistical data for each variable; and
Predicting a medical expense for home care based on the first result data; Including,
The first result data extraction step,
Extracting the first result data by applying the current data extracted from the statistical data for each variable to the analysis model,
The current status data is extracted by considering two or more variables of age, year, number of days of medical treatment, surgery status, region, gender, treatment category, main disease, injured disease, type of insurer, type of medical institution, and group,
When the year and the type of health care institution are considered as variables, the current status data is the total frequency of use, which is the sum of the medical institution with the highest frequency of use among a plurality of medical institutions during a specific period and the frequency of use of each of the plurality of medical institutions during the specific period. is extracted to indicate the year with the highest,
When the year, the type of insurer, and the region are considered as variables, the current status data is the insurance type with the highest frequency of use among a plurality of insurance types during the specific period for each region, and the specific period for each region. Extracted to indicate the year in which the total frequency of use of each of the plurality of insurance types is the highest,
A method for predicting medical expenses through big data analysis. delete According to claim 10,
In the step of predicting the medical cost,
determining whether each variable is meaningful to the analysis model based on a first value for each variable included in the first result data; and
Predicting the medical cost excluding variables determined to be insignificant; including,
A method for predicting medical expenses through big data analysis. According to claim 12,
In the significance judgment step,
Determining that the first value is not significant for a variable greater than a preset reference value,
A method for predicting medical expenses through big data analysis. According to claim 12,
In the step of predicting the medical cost,
Predicting the medical expense using a second value of a variable determined to be significant; including,
A method for predicting medical expenses through big data analysis. According to claim 10,
The first result data,
Includes first result data related to the first group and first result data related to the second group;
In the step of predicting the medical cost,
Based on the first result data related to the first group, the medical expenses for home care of the first group are predicted, and based on the first result data related to the second group, the medical expenses for home care of the second group are predicted. doing,
A method for predicting medical expenses through big data analysis. According to claim 15,
The first group includes at least one private medical institution, and the second group includes at least one public medical institution.
A method for predicting medical expenses through big data analysis. According to claim 10,
extracting second result data using an analysis model based on the statistical data for each variable; and
Further comprising: predicting the number of uses for home care based on the second result data;
A method for predicting medical expenses through big data analysis. delete A computer program stored in a computer readable recording medium to perform the method of predicting medical expenses through big data analysis of any one of claims 10, 12 to 17 in combination with a computer.

Images