Disclosure of Invention
In order to facilitate improvement of medical insurance costs by hospitals, the application provides a medical insurance cost control recommendation method, a system and a storage medium based on regional big data.
In a first aspect, the application provides a medical insurance control fee recommendation method based on regional big data, which adopts the following technical scheme:
a medical insurance fee control recommendation method based on regional big data comprises the following steps:
acquiring medical service data of a plurality of hospitals;
splitting medical service data of each hospital into service sub-data corresponding to the groups one by one based on preset group classification;
determining medical insurance costs of corresponding disease groups according to the business sub-data;
judging whether the medical insurance costs of different disease groups of each hospital exceed preset marker post values in sequence, wherein a plurality of marker post values are preset, and the marker post values are in one-to-one correspondence with the disease groups;
if the medical insurance expense is lower than or equal to a preset marker post value, defining business sub-data corresponding to the medical insurance expense as first data;
if the medical insurance expense exceeds a preset marker post value, defining business sub-data corresponding to the medical insurance expense as second data;
determining medical expense standard information according to all first data under the same disease group;
determining hyperbranched data corresponding to each second data according to the medical expense standard information;
generating corresponding medical insurance hyperbranched behavior portraits according to the hyperbranched data, and transmitting the medical insurance hyperbranched behavior portraits to a corresponding hospital;
the business sub-data comprises a plurality of diagnosis and treatment projects and diagnosis and treatment fees corresponding to each diagnosis and treatment project, and the medical fee standard information is determined according to all first data under the same disease group, and the method comprises the following steps:
a corresponding blank set is created according to the disease group,
counting all diagnosis and treatment projects contained in the first data under the same disease group;
calculating the occurrence frequency of each diagnosis and treatment project in turn;
it is determined whether the frequency of occurrence is above a preset threshold,
if yes, calculating an average value of diagnosis and treatment costs of the corresponding diagnosis and treatment projects, and adding the diagnosis and treatment projects and the corresponding average value into a corresponding blank set of the corresponding disease group to form medical cost standard information;
the step of determining the hyperbranched data corresponding to each second data according to the medical expense standard information comprises the following steps:
screening diagnosis and treatment items different from medical expense standard information from the second data, and defining the diagnosis and treatment items as first hyperbranched items;
sequentially judging whether the diagnosis and treatment cost corresponding to the rest diagnosis and treatment projects in the second data exceeds the average value corresponding to the medical cost standard information,
if yes, defining the corresponding diagnosis and treatment item as a second hyperbranched item;
taking the first hyperbranched item, the second hyperbranched item and the corresponding diagnosis and treatment cost as hyperbranched data;
the method for generating the corresponding medical insurance hyperbranched behavior portraits according to the hyperbranched data comprises the following steps:
judging whether the first hyperbranched item exists in the hyperbranched data,
if yes, a preset first template is obtained, first description words are determined according to the number of the first hyperbranched items, and the first description words are added into the first template to form a first description text;
judging whether a second hyperbranched item exists in the hyperbranched data,
if yes, determining a second description word according to the diagnosis and treatment cost corresponding to the second hyperbranched item, adding the second hyperbranched item and the corresponding second description word into a preset second template to form a second description text,
and combining the first descriptive text and the second descriptive text to form the medical insurance hyperbranched behavioral portrayal.
Through the technical scheme, medical service data of hospitals are collected and classified according to different disease groups. And (3) aiming at the medical service data of each disease group, further screening the medical service data of different hospitals into first data which are not hyperbranched and second data which are hyperbranched according to whether the medical service data exceeds a preset marker post value. And determining medical insurance expense standard information through the first data, further comparing the second data with the medical insurance expense standard information to obtain hyperbranched data, and finally analyzing the hyperbranched data to obtain the medical insurance hyperbranched behavior portraits. The medical insurance expense standard information is generated by adopting the non-hyperbranched data, so that the medical insurance expense standard information can be ensured to have higher credibility, the hyperbranched data corresponding to the second data is determined by the medical insurance expense standard information, and the abnormal data can be screened out, so that the corresponding hospitals are guided for the abnormal data, the hospitals can manage the medical insurance expense more pertinently, and the occurrence probability of the medical insurance expense hyperbranched behavior is reduced.
Optionally, before sequentially judging whether the medical insurance costs of different disease groups of each hospital exceed the preset marker post values, the method comprises the following steps:
obtaining standard values and historical fees of different disease groups;
determining a correction value according to the difference between the historical expense and the standard value;
and adjusting the standard value according to the correction value to form a marker post value.
Optionally, before generating the corresponding medical insurance hyperbranched behavior portrait according to the hyperbranched data, the method comprises the following steps:
counting the number of hyperbranched data in each disease group respectively;
determining a comparison value of each disease group according to the number of the medical expense standard information corresponding to each disease group;
judging whether the number of the hyperbranched data in each disease group exceeds a comparison value,
if the number of the hyperbranched data in each disease group exceeds the comparison value, generating corresponding medical insurance hyperbranched behavior portraits according to the hyperbranched data;
if the number of the hyperbranched data in each disease group is lower than or equal to the comparison value, the corresponding medical insurance hyperbranched behavior portraits are not generated.
Optionally, after the corresponding medical insurance hyperbranched behavior portraits are not generated if the number of the hyperbranched data in each disease group is lower than or equal to the comparison value, the method comprises the following steps:
one first data is selected from several first data of the same disease group as reference data,
the reference data is transmitted to all hospitals corresponding to the group.
Optionally, the determining whether the occurrence frequency is higher than a preset threshold value includes the following steps:
the number of hospitals is determined according to the number of medical service data related to the disease group corresponding to the occurrence frequency,
determining a threshold according to the corresponding relation between the preset number of hospitals and the threshold;
and judging whether the occurrence frequency is higher than a preset threshold value.
In a second aspect, the application provides a medical insurance control fee recommendation system based on regional big data, which adopts the following technical scheme:
a medical insurance fee control recommendation system based on regional big data comprises a hospital end and a regional end,
the hospital end is used for acquiring medical service data; splitting medical service data of each hospital into service sub-data corresponding to the groups one by one based on preset group classification; determining medical insurance costs of corresponding disease groups according to the business sub-data; sequentially judging whether medical insurance costs of different disease groups exceed preset marker post values, wherein a plurality of marker post values are preset, and the marker post values are in one-to-one correspondence with the disease groups; if the medical insurance expense is lower than or equal to a preset marker post value, defining business sub-data corresponding to the medical insurance expense as first data; if the medical insurance expense exceeds a preset marker post value, defining business sub-data corresponding to the medical insurance expense as second data;
the regional end is used for acquiring the first data and the second data transmitted by the hospital end and determining medical expense standard information according to all the first data in the same disease group; determining hyperbranched data corresponding to each second data according to the medical expense standard information; and generating a corresponding medical insurance hyperbranched behavior portrait according to the hyperbranched data, and transmitting the medical insurance hyperbranched behavior portrait to a corresponding hospital terminal.
In a third aspect, the present application provides a readable storage medium storing a computer program capable of being loaded by a processor and executing a medical insurance premium recommendation method based on regional big data as described above.
In summary, the present application includes at least one of the following beneficial technical effects: medical service data of hospitals are collected and classified according to different disease groups. And (3) aiming at the medical service data of each disease group, further screening the medical service data of different hospitals into first data which are not hyperbranched and second data which are hyperbranched according to whether the medical service data exceeds a preset marker post value. And determining medical insurance expense standard information through the first data, further comparing the second data with the medical insurance expense standard information to obtain hyperbranched data, and finally analyzing the hyperbranched data to obtain the medical insurance hyperbranched behavior portraits. The medical insurance expense standard information is generated by adopting the non-hyperbranched data, so that the medical insurance expense standard information can be ensured to have higher credibility, the hyperbranched data corresponding to the second data is determined by the medical insurance expense standard information, and the abnormal data can be screened out, so that the corresponding hospitals are guided for the abnormal data, the hospitals can manage the medical insurance expense more pertinently, and the occurrence probability of the medical insurance expense hyperbranched behavior is reduced.
Detailed Description
The present application will be described in further detail with reference to the accompanying drawings.
The application discloses a medical insurance control fee recommendation method based on regional big data, which is shown in fig. 1 and comprises the following steps:
s100, acquiring medical service data of a plurality of hospitals.
The medical service data comprises the data of the front page of the inpatient records and the detail data of inpatient cost,
the medical staff refines and gathers the relevant information of the patient in the hospital, and the data summary is obtained. The hospitalization expense detail data is the detailed summary of diagnosis and treatment information and expense information of the patient during the hospitalization period. The hospitalization case front page data and the hospitalization expense detail data are in one-to-one correspondence.
Specifically, the medical records first page data of the hospitalization include basic information of patients, information of hospitalization process, medical and treatment brief information and total cost information. The hospitalization cost detail data comprise diagnosis and treatment projects and diagnosis and treatment cost corresponding to the diagnosis and treatment projects one by one, and the diagnosis and treatment projects comprise medication information, material information and inspection and detection information.
S200, splitting medical service data of each hospital into service sub-data corresponding to the groups one by one based on preset group classification.
The preset disease group is DRG classification.
When medical service data are split, the medical service data are classified according to different disease groups, the split data still comprise the front page data of the inpatient records and the detail data of inpatient fees, and only the split data are related to the corresponding disease groups.
However, the medical items included in the hospital expense detail data are too complicated, and the direct use of such data may result in an excessive amount of data to be calculated later. Thus, hospitalization cost details are sorted by efficacy to reduce the number of hospitalization cost details. The medicine information is classified by pharmacological actions of medicines, such as pain relieving, antibiotics, antivirus, blood pressure lowering, blood vessel dilating, anticoagulation, immunosuppression, immunity enhancement, sedation, antiepileptic, nutrition and the like; the material information is gathered by the material efficacy, such as hemostatic materials, nursing materials, anastomotic suture materials and the like.
The hospitalization expense details corresponding to the medicines (materials, examination and detection) in the same class after the class aggregation are added together.
S300, determining medical insurance costs of corresponding disease groups according to the business sub-data.
The medical insurance cost of each patient group refers to the cost of medical insurance for the patient group. The number of the treatment persons and the total treatment cost of each disease group are required to be obtained from the business sub-data, wherein the number of the treatment persons can be directly obtained through the data of the first page of the inpatient records. The total treatment expense can be obtained by only adding corresponding data in the front page data of the hospitalization medical records, so that the purpose of convenient operation is achieved. If the total cost of medical insurance is required to be verified to be true and reliable, the sum of diagnosis and treatment costs in the hospitalization cost detail data can be additionally calculated, and if the two calculated total costs of medical insurance are similar (the difference value of the two total costs of medical insurance is smaller than or equal to a preset parameter), the data are considered to be true; if the two calculated total medical insurance costs differ greatly (the difference between the two total medical insurance costs is larger than a preset parameter), the total medical insurance cost with larger value is selected.
S400, judging whether the medical insurance costs of different disease groups of each hospital exceed preset marker post values in sequence.
The number of the preset marker post values is multiple, and the marker post values are in one-to-one correspondence with the disease groups.
The benchmarking value is the upper limit value that people in the corresponding group are allowed to spend on medical insurance.
When the medical insurance expense of a certain disease group exceeds the marker post value, the condition that the medical insurance expense is hyperbranched for the disease group is indicated; otherwise, there is no cost.
The target value is not fixed, but is affected by historical medical insurance costs. For example, if the medical insurance expense of a certain disease group in a month area is increased as a whole, the marker post value corresponding to the disease group in the current month is also increased appropriately.
In one embodiment, before determining whether the medical insurance costs of different groups of patients in each hospital exceeds a preset target value in turn, see fig. 2, the method comprises the following steps:
s410, obtaining standard values and historical fees of different disease groups.
The standard value is a fixed value and is uniformly formulated by the region. The historical expense is the medical insurance expense counted in the past in the area, and the historical expense is of various types, such as the medical insurance expense in the past year, the medical insurance expense in the last month, the medical insurance expense in the same period and the like, and the specific selection of which type of the historical expense is preset by staff.
S420, determining a correction value according to the difference between the historical expense and the standard value.
The historical cost is subtracted from the standard value to obtain the difference. When the historical cost is greater than the standard value, the difference is an integer; when the historical cost is less than the standard value, the difference is negative.
The correction values are a plurality of, and the correction values correspond to different preset ranges, and all the preset ranges are continuous in value. For example, correction values of-1000, 0 and 1000, the three preset ranges corresponding to each are (- ≡, -1000], (-1000, 1000) and [1000, ≡).
And matching a preset range in which the difference value falls, and then determining a corresponding correction value according to the corresponding relation between the preset range and the correction value.
S430, adjusting the standard value according to the correction value to form a marker post value.
The standard value is obtained by adding the correction value to the standard value.
S500, if the medical insurance expense is lower than or equal to a preset marker post value, defining business sub-data corresponding to the medical insurance expense as first data.
And S600, if the medical insurance expense exceeds a preset marker post value, defining the business sub-data corresponding to the medical insurance expense as second data.
When the service sub-data is defined as first data, a tag of the first data is assigned to the corresponding service sub-data. Similarly, when the service sub-data is defined as the second data, a tag of the second data is assigned to the corresponding service sub-data. And the service sub data can be quickly identified as the first data or the second data by searching the label.
And S700, determining medical expense standard information according to all the first data under the same disease group.
Because the first data are business sub-data with medical insurance expense not exceeding the standard pole value, the method is more suitable for popularization in hospitals in the area than the second data. It is desirable to determine medical insurance expense medical expense standard information from the first data.
Of course, there is necessarily a certain difference between the first data generated by different hospitals, and a plurality of first data needs to be filtered and summarized to form one medical expense standard information.
In one embodiment, the determining medical fee standard information according to all the first data under the same disease group, see fig. 3, includes the following steps:
s710, creating a corresponding blank set according to the disease group.
The blank set is one data storage unit.
S720, counting all diagnosis and treatment projects contained in the first data under the same disease group.
Since the first data may come from different hospitals and the treatment items in different hospitals are often different, the treatment items cannot be determined directly from only one first data, but the treatment items of all the first data need to be counted.
S730, sequentially calculating the occurrence frequency of each diagnosis and treatment project.
The frequency of occurrence of a diagnosis and treatment item is obtained by dividing the frequency of occurrence of the same diagnosis and treatment item by the total number of diagnosis and treatment items in the same disease group.
S740, judging whether the occurrence frequency is higher than a preset threshold value.
The threshold is set by the staff.
And S750, if so, calculating the average value of the diagnosis and treatment cost of the corresponding diagnosis and treatment project, and adding the diagnosis and treatment project and the corresponding average value into the corresponding blank set of the corresponding disease group to form medical cost standard information.
If not, S760 does not use the diagnosis cost of the diagnosis item as the medical cost standard information.
When the frequency of occurrence is higher than the threshold value, the frequency of occurrence of the corresponding diagnosis and treatment item is considered to be higher, so that the corresponding diagnosis and treatment item is considered to be used in most hospitals, and the diagnosis and treatment item and the corresponding average value are added into the blank set as medical expense standard information relatively more rule.
When the frequency of occurrence is lower than or equal to the threshold value, the corresponding diagnosis and treatment project is considered to be unusual, and can be set by individual hospitals or applied to the disease group by mistake, so that the diagnosis and treatment project is not used as medical expense standard information.
In one embodiment, determining whether the frequency of occurrence is above a preset threshold comprises the steps of:
s741, determining the number of hospitals according to the number of medical service data related to the disease group corresponding to the occurrence frequency.
S742, determining a threshold according to the corresponding relation between the preset number of hospitals and the threshold.
S743, judging whether the occurrence frequency is higher than a preset threshold value.
The corresponding relation between the preset number of hospitals and the threshold value is set manually. Theoretically, the smaller the number of hospitals, the higher the corresponding threshold. This is because if the number of hospitals is smaller, the calculated frequency of occurrence will be higher once the unusual medical service occurs, and if a threshold value with a lower value is adopted, it may be judged that the medical service is routine, and obvious misjudgment occurs. Conversely, if the number of hospitals is large, it is likely that some hospitals do not perform the treatment of the group according to the conventional medical service, and therefore the frequency of occurrence may not always be kept high, so that the corresponding threshold value is appropriately lowered. For example, when the number of hospitals is 5 or less, the corresponding threshold value is 100%; when the number of hospitals is 5 to 10, the corresponding threshold value is 80%.
S800, determining hyperbranched data corresponding to each second data according to the medical expense standard information.
The hyperbranched data refers to diagnosis and treatment items which are additionally added or abnormal in the second data and corresponding diagnosis and treatment fees compared with the medical fee standard information.
In one embodiment, determining the hyperbranched data corresponding to each second data according to the medical expense standard information includes the following steps:
s810, screening diagnosis and treatment items different from medical expense standard information from the second data, and defining the diagnosis and treatment items as first hyperbranched items.
S820, judging whether the diagnosis and treatment cost corresponding to the rest diagnosis and treatment items in the second data exceeds the average value corresponding to the medical cost standard information or not.
And S830, if yes, defining the corresponding diagnosis and treatment item as a second hyperbranched item. If not, the corresponding diagnosis and treatment item is not defined as the second hyperbranched item.
And S840, taking the first hyperbranched item, the second hyperbranched item and the corresponding diagnosis and treatment cost as hyperbranched data.
The first hyperbranched item is a diagnosis and treatment item and corresponding diagnosis and treatment cost which are additionally added in the second data compared with the medical cost standard information.
The second hyperbranched item is diagnosis and treatment expense with the exception of the second data compared with the medical expense standard information and the corresponding diagnosis and treatment item.
When the second hyperbranched item is screened, besides using the average value of diagnosis and treatment costs in the medical cost standard information as a screening standard, a preset permission error can be added on the basis of the average value to form a standard cost range, and only when the diagnosis and treatment costs in the second data exceed the standard cost range, the second hyperbranched item can be defined, so that the probability of misjudging normal second data as abnormal data is reduced.
S900, generating corresponding medical insurance hyperbranched behavior portraits according to hyperbranched data, and transmitting the medical insurance hyperbranched behavior portraits to the corresponding hospitals.
The medical insurance hyperbranched behavior representation is text information which is formed based on hyperbranched data and has summary properties on diagnosis and treatment projects related to expense hyperbranched.
In one embodiment, generating a corresponding medical insurance hyperbranched behavioral portrayal from the hyperbranched data comprises the steps of:
s910, judging whether a first hyperbranched item exists in the hyperbranched data.
S920, if yes, acquiring a preset first template, determining first description words according to the number of the first hyperbranched items, and adding the first description words into the first template to form a first description text.
All first hyperbranched items correspond to one first template, which is "unrelated drug (material, inspection) xx". Wherein the content of xx is a first descriptor.
The first descriptor is what the number of first hyperbranched items is expressed. In this embodiment, the first descriptor includes too many and too many. When the number of the first hyperbranched items is 1, the matched first descriptors are more; when the number of the first hyperbranched items is 2 or more, the matched first descriptors are too many.
If the medical expense standard information includes hemostatic material, anastomotic suture material, nutrient solution, antibiotics, anesthetics and the like for the disease group GB29 (large operation of small intestine and large intestine), if the business data provided by a certain hospital for the disease group GB29 includes hemostatic material, anastomotic suture material, nutrient solution, antibiotics, anesthetics, antihypertensive drugs and anti-hepatitis drugs, the comparison with the medical expense standard information shows that the antihypertensive drugs and the anti-hepatitis drugs belong to the first hyperbranched items, the first template is matched, and the number of the first hyperbranched items is two, the matched first descriptor is too many, so that the finally formed first description text is "irrelevant drugs too many".
S930, judging whether a second hyperbranched item exists in the hyperbranched data.
And S940, if so, determining a second description word according to the diagnosis and treatment cost corresponding to the second hyperbranched item, and adding the second hyperbranched item and the corresponding second description word into a preset second template to form a second description text.
The second descriptor expresses how much diagnosis and treatment expense the second hyperbranched item corresponds to.
Each second hyperbranched item is provided with a second template correspondingly.
The preset second template is AA and BB, wherein AA is a specific diagnosis and treatment item in the second hyperbranched item, and BB is a second descriptor.
The second descriptor includes high, and too high. The second descriptor is matched according to the hyperbranched proportion of the diagnosis and treatment cost in the second hyperbranched item in the medical cost standard information, for example, the hyperbranched proportion is 0 to 10 percent, and the matched second descriptor is higher; the hyperbranched proportion is 10% to 30%, and the matched second descriptor is high; the proportion of the hyperbranched is more than 30%, and the matched second descriptor is too high.
The hyperbranched proportion is obtained by subtracting the diagnosis and treatment cost corresponding to the medical cost standard information from the diagnosis and treatment cost corresponding to the second hyperbranched item and dividing the diagnosis and treatment cost by the diagnosis and treatment cost corresponding to the medical cost standard information.
S950, combining the first description text and the second description text to form the medical insurance hyperbranched behavior portrait.
In one embodiment, before generating the corresponding medical insurance hyperbranched behavioral portraits from the hyperbranched data, the method comprises the following steps:
and A910, respectively counting the number of the hyperbranched data in each disease group.
And A920, determining the comparison value of each disease group according to the number of the medical expense standard information corresponding to each disease group.
And A930, judging whether the number of the hyperbranched data in each disease group exceeds a comparison value.
And A940, if the number of the hyperbranched data in each disease group exceeds the comparison value, generating a corresponding medical insurance hyperbranched behavior portrait according to the hyperbranched data.
And A950, if the number of the hyperbranched data in each disease group is lower than or equal to the comparison value, not generating a corresponding medical insurance hyperbranched behavior portrait.
The purpose of generating the medical insurance hyperbranched behavior portraits is to enable a hospital with medical insurance hyperbranched to clearly know which aspects are hyperbranched, so that the hospital can improve the hyperbranched behavior more pertinently.
However, since the number of groups is large, a large number of operations are certainly required to generate a corresponding hyperbranched behavioral portraits for each group. The comparison value is added to screen the hyperbranched data. When the number of the hyperbranched data is lower than the comparison value, the hyperbranched data indicates that the disease group has the hyperbranched condition, but the hyperbranched data is not serious, and the disease group can be improved by a hospital without special guidance.
The comparison value is determined according to the number of diagnosis and treatment items in the medical expense standard information corresponding to the disease group. In this embodiment, the comparison value is calculated by dividing the medical item in the medical cost standard information by 10 and rounding up.
If the number of diagnosis and treatment items in the medical expense standard information corresponding to the disease group GB29 is 9, the corresponding comparison value is 1.
In one embodiment, if the number of hyperbranched data in each disease group is lower than or equal to the comparison value, after the corresponding medical insurance hyperbranched behavior portraits are not generated, the method comprises the following steps:
and B910, selecting one first data from a plurality of first data of the same disease group as reference data.
And B920, transmitting the reference data to all hospitals corresponding to the disease group.
In order to facilitate the control of the medical insurance expense in the hospital, the data of the hospital with proper medical insurance expense control can be used as a reference when the medical insurance hyperbranched behavior representation is not specially given for guidance, so as to improve the management of the medical insurance expense.
The embodiment of the application also discloses a medical insurance control fee recommendation system based on the regional big data, which comprises a hospital end and a regional end. An equal number of hospital ends are set in an area according to the number of hospitals, namely, each hospital corresponds to one hospital end; and only one area end needs to be arranged. The hospital terminal comprises an intra-hospital terminal and a plurality of intra-hospital terminals, and the intra-hospital terminals correspond to each department in the hospital. Medical staff in the department uploads medical service data of the department through intra-hospital terminals.
The intra-hospital terminal is used for acquiring medical service data of the corresponding department. The hospital terminal is used for receiving the medical service data transmitted by the hospital terminal so as to collect the medical service data of each department in the hospital, and splitting the medical service data of the hospital into service sub-data corresponding to the groups one by one based on preset group classification; determining medical insurance costs of corresponding disease groups according to the business sub-data; sequentially judging whether medical insurance costs of different disease groups exceed preset marker post values, wherein a plurality of marker post values are preset, and the marker post values are in one-to-one correspondence with the disease groups; if the medical insurance expense is lower than or equal to a preset marker post value, defining business sub-data corresponding to the medical insurance expense as first data; if the medical insurance expense exceeds a preset marker post value, defining business sub-data corresponding to the medical insurance expense as second data;
the regional end is used for acquiring the first data and the second data transmitted by the hospital end and determining medical expense standard information according to all the first data in the same disease group; determining hyperbranched data corresponding to each second data according to the medical expense standard information; and generating a corresponding medical insurance hyperbranched behavior portrait according to the hyperbranched data, and transmitting the medical insurance hyperbranched behavior portrait to a corresponding hospital terminal.
The embodiment of the application also discloses a readable storage medium which stores a computer program capable of being loaded by a processor and executing the medical insurance control fee recommendation method based on the regional big data.
The above embodiments are not intended to limit the scope of the present application, so: all equivalent changes in structure, shape and principle of the application should be covered in the scope of protection of the application.