CN113707281A - Dynamic disease group grouping realization method based on multi-dimensional disease characteristics - Google Patents
Dynamic disease group grouping realization method based on multi-dimensional disease characteristics Download PDFInfo
- Publication number
- CN113707281A CN113707281A CN202010442799.5A CN202010442799A CN113707281A CN 113707281 A CN113707281 A CN 113707281A CN 202010442799 A CN202010442799 A CN 202010442799A CN 113707281 A CN113707281 A CN 113707281A
- Authority
- CN
- China
- Prior art keywords
- disease
- diagnosis
- treatment
- dimensional
- library
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 201000010099 disease Diseases 0.000 title claims abstract description 123
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 title claims abstract description 123
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000003745 diagnosis Methods 0.000 claims abstract description 58
- 238000013145 classification model Methods 0.000 claims abstract description 21
- 230000011218 segmentation Effects 0.000 claims description 24
- 238000013016 damping Methods 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 8
- 230000000391 smoking effect Effects 0.000 claims description 8
- 230000001419 dependent effect Effects 0.000 claims description 5
- 238000013138 pruning Methods 0.000 claims description 4
- 208000024891 symptom Diseases 0.000 abstract description 5
- 241000894007 species Species 0.000 description 14
- 208000006545 Chronic Obstructive Pulmonary Disease Diseases 0.000 description 13
- 230000029058 respiratory gaseous exchange Effects 0.000 description 6
- 238000003066 decision tree Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 210000003128 head Anatomy 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 239000008280 blood Substances 0.000 description 2
- 210000004369 blood Anatomy 0.000 description 2
- 238000005399 mechanical ventilation Methods 0.000 description 2
- 238000002324 minimally invasive surgery Methods 0.000 description 2
- 238000001356 surgical procedure Methods 0.000 description 2
- 206010003445 Ascites Diseases 0.000 description 1
- 208000004429 Bacillary Dysentery Diseases 0.000 description 1
- 208000035473 Communicable disease Diseases 0.000 description 1
- 206010017915 Gastroenteritis shigella Diseases 0.000 description 1
- 206010022678 Intestinal infections Diseases 0.000 description 1
- 206010028980 Neoplasm Diseases 0.000 description 1
- 208000030852 Parasitic disease Diseases 0.000 description 1
- 241001354013 Salmonella enterica subsp. enterica serovar Enteritidis Species 0.000 description 1
- 208000037386 Typhoid Diseases 0.000 description 1
- 230000001580 bacterial effect Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 238000002586 coronary angiography Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000000502 dialysis Methods 0.000 description 1
- 208000016097 disease of metabolism Diseases 0.000 description 1
- 238000004945 emulsification Methods 0.000 description 1
- 230000002124 endocrine Effects 0.000 description 1
- 208000030172 endocrine system disease Diseases 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000003394 haemopoietic effect Effects 0.000 description 1
- 230000008076 immune mechanism Effects 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 208000015181 infectious disease Diseases 0.000 description 1
- 230000002458 infectious effect Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000002721 intensity-modulated radiation therapy Methods 0.000 description 1
- 230000000968 intestinal effect Effects 0.000 description 1
- 210000003734 kidney Anatomy 0.000 description 1
- 229940078979 liver therapy drug Drugs 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 238000000746 purification Methods 0.000 description 1
- 238000001959 radiotherapy Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 201000005113 shigellosis Diseases 0.000 description 1
- 238000011476 stem cell transplantation Methods 0.000 description 1
- 238000002560 therapeutic procedure Methods 0.000 description 1
- 238000002054 transplantation Methods 0.000 description 1
- 201000008297 typhoid fever Diseases 0.000 description 1
- 210000004127 vitreous body Anatomy 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
- G16H40/20—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- General Business, Economics & Management (AREA)
- Biomedical Technology (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Pathology (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention discloses a dynamic disease component grouping realization method based on multi-dimensional disease characteristics, which comprises the following steps: step 1: determining the disease category of each case in the cases to be classified, and acquiring one or more disease category sets; step 2: establishing a classification model, and classifying the cases in the disease category set according to treatment cost by adopting the classification model; the step 1 comprises the following steps: step 101: acquiring the disease name and the operation mode of the case to be classified; step 102: determining a diagnosis and treatment mode according to the operation mode; step 103: and determining the disease species of the case to be classified according to the disease name and the diagnosis and treatment mode. The invention can classify according to the clinical symptoms of patients and treatment cost by a classification model and a CART algorithm so as to realize reasonable medical insurance control cost and improve the utilization efficiency of medical resources of hospitals.
Description
Technical Field
The invention relates to the technical field of computer information management, in particular to a dynamic grouping and grouping implementation method based on multi-dimensional disease characteristics.
Background
The disease-related grouping is a grouping system for grouping cases into a plurality of groups according to the factors of age, sex, hospitalization days, clinical diagnosis, disease symptoms, operations, disease severity, complications, outcome and the like of patients. Cases grouped together have similar clinical symptoms and convergent resource expenditure. The grouping of disease diagnosis related is used as a tool for medical payment, and is one of the more advanced payment methods recognized in the world today. The medical resource utilization standardization is achieved through the establishment of the unified disease diagnosis classification quota payment standard. The method is beneficial to exciting the hospital to strengthen medical quality management, forcing the hospital to actively reduce cost, shorten the number of hospitalization days and reduce inductive medical expense payment for obtaining profits, and is beneficial to expense control.
The payment and calculation of the medical insurance cost are important links of a medical insurance system and are related to the benefits of all the medical insurance parties. At present, the medical expense is increased quickly, and medical insurance control cost is needed to prevent medical insurance from entering or not being laid out. The traditional medical insurance payment methods include methods of paying according to items and paying according to head. The existing pay-per-item is lack of a constraint mechanism, and excessive medical behaviors such as repeated inspection, large prescription and the like occur, so that the waste of medical resources is caused, and the medical expense is continuously and rapidly increased. Pay-per-head is a fixed amount that a medical insurance agency charges monthly or annually for the number of people and rules for hospital or doctor services, and prepays to the service provider a fixed amount. The head payment induces the hospital to selectively receive patients, such as patients with mild receiving symptoms and relatively short hospitalization time, and patients with perverting severe diseases; decomposing the hospitalization times of the patient to obtain more heads; hospitals lack competitive awareness, medical staff are not motivated to improve medical skills, and even a decline in medical quality may occur.
According to known case basic data including personal information, disease information, treatment information, hospitalization time, cost and the like of a case, various factors of a hospital are comprehensively considered by using a related algorithm, and the grouping of disease diagnosis of the case is realized, so that a foundation is laid for later application such as medical insurance payment pricing and the like, and the technical problem to be solved at present is solved.
Disclosure of Invention
The invention aims to provide a method for classifying cases based on a classification model and a CART algorithm.
In order to achieve the purpose, the invention is realized by adopting the following technical scheme:
a dynamic disease component grouping realization method based on multi-dimensional disease characteristics comprises the following steps:
step 1: determining the disease category of each case in the cases to be classified, and acquiring one or more disease category sets;
step 2: and establishing a classification model, and classifying the cases in the disease category set according to treatment cost by adopting the classification model.
Preferably, the step 1 comprises the following steps:
step 101: acquiring the disease name and the operation mode of the case to be classified;
step 102: determining a diagnosis and treatment mode according to the operation mode;
step 103: and determining the disease species of the case to be classified according to the disease name and the diagnosis and treatment mode.
Further, in step 102, a diagnosis and treatment method corresponding to the operation method is queried through a pre-established diagnosis and treatment code library.
Further, in step 103, the disease name and the disease type corresponding to the diagnosis and treatment manner are queried through a pre-established disease type code library.
Preferably, in step 2, the establishing of the classification model comprises the following steps:
step 201: all factors in a pre-established influencing factor library are used as independent variables, and the treatment cost of the cases in the disease category set is used as a dependent variable;
step 202: selecting an optimal segmentation point from all factors in the influence factor library by adopting a CART algorithm as a splitting node, and repeating the selection process of the optimal segmentation point under each splitting node until a preset stop condition of the CART algorithm is reached to generate a classification regression tree;
step 203: and pruning the classification regression tree to generate the classification model.
Further, in step 202, the selection process of the optimal segmentation point includes the following steps:
step 202 a: for each factor in the influence factor library, calculating a damping coefficient before segmentation and a damping coefficient after segmentation;
step 202 b: calculating the difference value between the damping coefficient before segmentation and the damping coefficient after segmentation;
step 202 c: and selecting the factor corresponding to the maximum value of the difference value as an optimal dividing point.
Further, in step 102, the diagnosis and treatment code library is built by:
step 102 a: classifying and coding the existing diagnosis and treatment modes;
step 102 b: and classifying the operation and operation code ICD-9-CM-3 and the name into the corresponding diagnosis and treatment mode, and forming the diagnosis and treatment code library by the operation and operation code ICD-9-CM-3, the name and the attributive diagnosis and treatment mode.
Further, in step 103, the disease seed code library is created by the following steps:
step 103 a: encoding ICD-10 according to international disease classification established by WHO, and intercepting the first three bits of the encoding as a preset disease major class;
step 103 b: subdividing the disease category, namely intercepting the first four digits of the ICD-10 code of the international disease classification code as a preset disease subclass;
step 103 c: and (3) corresponding the disease subclasses with the diagnosis and treatment modes to form a disease species coding library.
Further, the factors in the influencing factor library include the patient's age, sex, smoking history, physical fitness, rescue, complication status, and ventilator assistance.
Further, in step 202, the calculation formula of the kini coefficient gini (d) before slicing is:
gini coefficient after slicingA(D) The calculation formula is as follows:
wherein D is a clinical case and resource consumption dataset, m is the number of classifications made by classifying the clinical case and resource consumption dataset, piThe probability of any one of the classifications is selected for each case.
The invention has the following beneficial effects:
the invention classifies cases with approximately same clinical symptoms and treatment cost into one class, has higher classification speed and higher precision compared with the existing manual classification mode, and can realize scientific and reasonable medical insurance control cost by calculating the medical insurance cost according to the classification.
Drawings
FIG. 1 is a schematic representation of the classification of chronic obstructive pulmonary disease using the method of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings.
A dynamic disease component grouping realization method based on multi-dimensional disease characteristics comprises the following steps:
step 1: determining the disease category of each case in the cases to be classified, and acquiring one or more disease category sets.
Specifically, the step 1 includes the steps of:
step 101: acquiring the disease name and the operation mode of the case to be classified;
step 102: determining a diagnosis and treatment mode according to the operation mode;
step 103: and determining the disease species of the case to be classified according to the disease name and the diagnosis and treatment mode.
In step 102, the diagnosis and treatment method corresponding to the operation method is queried through a pre-established diagnosis and treatment code library. In step 103, the disease name and the disease type corresponding to the diagnosis and treatment method are searched through a pre-established disease type code library.
Specifically, in step 102, the diagnosis and treatment code library is built by the following steps:
step 102 a: classifying and coding the existing diagnosis and treatment modes;
step 102 b: and classifying the operation and operation code ICD-9-CM-3 and the name into the corresponding diagnosis and treatment mode, and forming the diagnosis and treatment code library by the operation and operation code ICD-9-CM-3, the name and the attributive diagnosis and treatment mode.
In step 103, the disease seed code library is created by the following steps:
step 103 a: encoding ICD-10 according to international disease classification established by WHO, and intercepting the first three bits of the encoding as a preset disease major class;
step 103 b: subdividing the disease category, namely intercepting the first four digits of the ICD-10 code of the international disease classification code as a preset disease subclass;
step 103 c: and (3) corresponding the disease subclasses with the diagnosis and treatment modes to form a disease species coding library.
The disease species are the disease species with more definite diagnosis and treatment scheme and hospital entrance and exit standard, more mature diagnosis and treatment technology, stable clinical path and little difference of comprehensive service cost. In specific implementation, the diagnosis and treatment coding library is established by adopting the following method:
firstly, the commonly used diagnosis and treatment methods are classified and coded, and the diagnosis and treatment methods mainly comprise 20 diagnosis and treatment methods, which are shown in the following table:
code | Diagnosis and treatment method | Code | Diagnosis and treatment method |
0 | Conservative treatment | 10 | Conformal radiotherapy |
1 | Traditional surgery | 11 | Conformal intensity modulated radiation therapy |
2 | Minimally invasive surgery | 12 | Ascites reinfusion |
3 | Interventional therapy | 13 | Repair surgery |
4 | Artificial liver therapy | 14 | Extracorporeal lithotripsy |
5 | Dialysis treatment | 15 | Volume intensity modulation |
6 | Stem cell transplantation | 16 | Blood purification |
7 | Mechanical ventilation | 17 | Mechanical continuous cooling |
8 | Ultrasonic emulsification | 18 | Kidney transplantation |
9 | Cutting vitreous body | 19 | Coronary angiography |
Then, the common operation and manipulation codes ICD-9-CM-3 and names are classified into corresponding diagnosis and treatment modes, so that the common operation and manipulation codes ICD-9-CM-3 and names and diagnosis and treatment modes belonging to the common operation and manipulation codes form a diagnosis and treatment code library, and the total number of the common operation and manipulation codes ICD-9-CM-3 is 5398, wherein the following table shows 5 examples:
in the specific implementation process, the disease species coding library is established in the following way: firstly, according to international disease classification code ICD-10 established by WHO, the first three bits of the code are intercepted as the preset disease major classes, and the total number of the disease major classes is 20, including: some infectious and parasitic diseases A00-B99, tumors C00-D48, blood and hematopoietic organ diseases and some diseases involving immune mechanisms D50-D89, endocrine, nutritional and metabolic diseases E00-E90, and the like; subdividing the disease species major class, and then cutting the first four digits encoded by ICD-10 as preset disease subclasses, wherein the total number of 187 disease subclasses comprises: typhoid A01.0, unspecified paratyphoid A01.4, Salmonella enteritidis A02.0, unspecified bacillary dysentery A03.9, other intestinal colibacillus infection A04.4, etc. Then, the 187 disease subclasses described above and the main treatment methods already identified in this example were matched to form a "disease species coding library" of 4654 disease species in total. For example, if the disease subclass is "bacterial intestinal infection as other specific indications of A04.8" and the diagnosis and treatment mode is "0 conservative treatment", the case is determined to be the disease species "BZ 1", and the following table exemplifies 5 disease species in the disease species coding library:
when the case is under the same ICD-10 sub-vision and several different diagnosis and treatment modes exist at the same time, the priority is judged by priority numerical values, and the higher the numerical value is, the higher the priority is judged. For example, in a case with the sub-order ICD-10 of "a 16.2", if there are three diagnosis and treatment modes, namely "conservative treatment", "minimally invasive surgery" and "mechanical ventilation", the case is judged to belong to the "BZ 5" with the highest priority by priority. The "priority" is an evaluation made by a clinical specialist according to the resource consumption, technical difficulty, and risk level of diagnosis and treatment modes of various diseases.
After the diagnosis and treatment code library and the disease species code library are established, the operation of a case and the operation code ICD-9-CM-3 can be matched with a pre-established diagnosis and treatment code library, and the diagnosis and treatment mode corresponding to the operation and the operation code can be obtained; then, according to the diagnosis and treatment mode of the case, the pre-constructed disease category coding library is matched, and the specific disease category corresponding to the case can be obtained.
Step 2: and establishing a classification model, and classifying the cases in the disease category set according to treatment cost by adopting the classification model.
Specifically, in step 2, establishing a classification model includes the following steps:
step 201: all factors in a pre-established influencing factor library are used as independent variables, and the treatment cost of the cases in the disease category set is used as a dependent variable;
step 202: selecting an optimal segmentation point from all factors in the influence factor library by adopting a CART algorithm as a splitting node, and repeating the selection process of the optimal segmentation point under each splitting node until a preset stop condition of the CART algorithm is reached to generate a classification regression tree;
step 203: and pruning the classification regression tree to generate the classification model.
In a specific implementation, the establishing of the classification model includes: all factors in a pre-established influence factor library are used as independent variables; the factors in the influencing factor library include: the patient's age, sex, smoking history, physical fitness, rescue, complication status, ventilator assistance; using as a dependent variable a treatment cost of the cases in the race set; selecting an optimal segmentation point from all factors in the influence factor library as a splitting node by adopting a CART algorithm; repeating the selection process of the optimal segmentation point under each splitting node until a preset stop condition of the CART algorithm is reached, and generating a classification regression tree; and pruning the classification regression tree to generate the classification model. Factors in the influencing factor library include the patient's age, sex, smoking history, fitness, whether to rescue, complication status, and whether to ventilator aid.
Specifically, in step 202, the selection process of the optimal segmentation point includes the following steps:
step 202 a: for each factor in the influence factor library, calculating a damping coefficient before segmentation and a damping coefficient after segmentation;
step 202 b: calculating the difference value between the damping coefficient before segmentation and the damping coefficient after segmentation;
step 202 c: and selecting the factor corresponding to the maximum value of the difference value as an optimal dividing point.
Further, in step 202, the calculation formula of the kini coefficient gini (d) before slicing is:
gini coefficient after slicingA(D) The calculation formula is as follows:
wherein D is a clinical case and resource consumption dataset, m is the number of classifications made by classifying the clinical case and resource consumption dataset, piThe probability of any one of the classifications is selected for each case. If the kini coefficients of the continuous values are calculated, the middle points of the sorted adjacent values are required to be used as splitting points, and the calculation is also carried out by using the above formula.
The following parameters are also determined when the classification model is established:
1. complexity of the decision tree: control decision tree growth (i.e., control number of classes);
2. the minimum number of cases each category should contain, as a condition for stopping the algorithm.
In addition, when the CART algorithm selects the complications as segmentation characteristic variables, determining the complications with the segmented node cost average value higher than a preset threshold value as important complications, and determining the complications with the segmented node cost average value lower than the preset threshold value as general complications.
The following describes the classification process by taking the classification of cases of chronic obstructive pulmonary disease as an example. Firstly, the disease category of a case is determined to be chronic obstructive pulmonary disease through a pre-constructed diagnosis and treatment coding library and a disease category coding library. Aiming at chronic obstructive pulmonary disease, case treatment cost is set as a dependent variable, independent variables are all multi-dimensional characteristic information such as age, sex, smoking history, physical quality, rescue, complication and ventilator assistance, a CART algorithm is used for selecting proper influencing factors, and the most effective grouping factors and grouping conditions are quickly and adaptively searched by the CART algorithm. When a classification model is established, the complexity of a decision tree is set to be 0.01 so as to control the growth of the decision tree (control the number of classifications); the minimum number of cases that each category should contain is set to 50 as a condition for stopping the algorithm.
According to the CART algorithm, firstly, complications are selected as segmentation variables and divided into non-complications and complications, cases with the complications are further subdivided, the rules are that the complications with the segmented node cost average value higher than a preset threshold value are determined as important complications, and the complications with the segmented node cost average value lower than the preset threshold value are determined as general complications; for the uncomplicated cases, the breathing machine auxiliary time is further divided into non-breathing machine support and breathing machine support, the breathing machine support is further divided into breathing machine support accompanied by more than 96 hours and breathing machine support accompanied by less than 96 hours; ventilator support with 96 hours or less is further subdivided by age, into 66 years and up to 65 years; for patients over 66 years old, the number of hospitalization days in one year is further subdivided into no hospitalization record in the last year and a hospitalization record in the last year, and the hospitalization record in the last year is further subdivided into the number of hospitalization days in the last year of 1-30 days and the number of hospitalization days in the last year of 30 days; for patients who have no hospitalization record in the last year, a final classification model is formed according to whether the patients have a smoking history or not.
As shown in fig. 1, the case data set is divided into 9 groups according to the CART algorithm, which are: 1) chronic obstructive pulmonary disease with major complications (average cost 13555.9 yuan); 2) chronic obstructive pulmonary disease with general complications (average cost 10039.7 yuan); 3) chronic obstructive pulmonary disease is uncomplicated and without ventilator support (average cost 6110.3 yuan); 4) chronic obstructive pulmonary disease is without complications and with ventilator support for more than 96 hours (average cost 11636.3 yuan); 5) chronic obstructive pulmonary disease without complications and with ventilator support for 96 hours or less, less than or equal to 65 years old (average cost 8684.9 yuan); 6) chronic obstructive pulmonary disease without complications and with ventilator support for 96 hours or less, age 66 or more, last year hospital stay 30 days or more (average cost 12436.3 yuan); 7) chronic obstructive pulmonary disease has no complications and is supported by a ventilator for 96 hours or less, the patient is over 66 years old, and the patient has a hospitalization day for the last year for 1-30 days (the average cost is 11636.3 yuan); 8) chronic obstructive pulmonary disease has no complications and ventilator support for 96 hours or less, is over 66 years old, has no hospitalization record in the last year, and has a history of smoking (average cost is 9345.7 yuan); 9) chronic obstructive pulmonary disease is without complications and with ventilator support for 96 hours or less, over the age of 66 years, no hospitalization for the last year, no history of smoking (average cost 8543.7 yuan). The average value of the group costs can be used as a reference value for predicting medical service costs.
The present invention is capable of other embodiments, and various changes and modifications may be made by one skilled in the art without departing from the spirit and scope of the invention.
Claims (10)
1. A dynamic disease component grouping realization method based on multi-dimensional disease characteristics is characterized by comprising the following steps:
step 1: determining the disease category of each case in the cases to be classified, and acquiring one or more disease category sets;
step 2: and establishing a classification model, and classifying the cases in the disease category set according to treatment cost by adopting the classification model.
2. The method for realizing dynamic disease component grouping based on multi-dimensional disease features as claimed in claim 1, wherein the step 1 comprises the following steps:
step 101: acquiring the disease name and the operation mode of the case to be classified;
step 102: determining a diagnosis and treatment mode according to the operation mode;
step 103: and determining the disease species of the case to be classified according to the disease name and the diagnosis and treatment mode.
3. The method for realizing dynamic disease grouping based on multi-dimensional disease features as claimed in claim 2, wherein: in step 102, the diagnosis and treatment method corresponding to the operation method is queried through a pre-established diagnosis and treatment code library.
4. The method for realizing dynamic disease grouping based on multi-dimensional disease features as claimed in claim 2, wherein: in step 103, the disease name and the disease type corresponding to the diagnosis and treatment method are searched through a pre-established disease type code library.
5. The method for realizing dynamic disease grouping based on multi-dimensional disease features according to claim 1, wherein: in step 2, establishing a classification model comprises the following steps:
step 201: all factors in a pre-established influencing factor library are used as independent variables, and the treatment cost of the cases in the disease category set is used as a dependent variable;
step 202: selecting an optimal segmentation point from all factors in the influence factor library by adopting a CART algorithm as a splitting node, and repeating the selection process of the optimal segmentation point under each splitting node until a preset stop condition of the CART algorithm is reached to generate a classification regression tree;
step 203: and pruning the classification regression tree to generate the classification model.
6. The method for realizing dynamic disease grouping based on multi-dimensional disease features as claimed in claim 5, wherein: in step 202, the selection process of the optimal cut point includes the following steps:
step 202 a: for each factor in the influence factor library, calculating a damping coefficient before segmentation and a damping coefficient after segmentation;
step 202 b: calculating the difference value between the damping coefficient before segmentation and the damping coefficient after segmentation;
step 202 c: and selecting the factor corresponding to the maximum value of the difference value as an optimal dividing point.
7. The method for realizing dynamic disease grouping based on multi-dimensional disease features according to claim 3, wherein: in step 102, the diagnosis and treatment code library is built by the following steps:
step 102 a: classifying and coding the existing diagnosis and treatment modes;
step 102 b: and classifying the operation and operation code ICD-9-CM-3 and the name into the corresponding diagnosis and treatment mode, and forming the diagnosis and treatment code library by the operation and operation code ICD-9-CM-3, the name and the attributive diagnosis and treatment mode.
8. The method for realizing dynamic disease grouping based on multi-dimensional disease features as claimed in claim 4, wherein: in step 103, the disease seed code library is created by the following steps:
step 103 a: encoding ICD-10 according to international disease classification established by WHO, and intercepting the first three bits of the encoding as a preset disease major class;
step 103 b: subdividing the disease category, namely intercepting the first four digits of the ICD-10 code of the international disease classification code as a preset disease subclass;
step 103 c: and (3) corresponding the disease subclasses with the diagnosis and treatment modes to form a disease species coding library.
9. The method for realizing dynamic disease component grouping based on multi-dimensional disease features according to claim 5 or 6, wherein: factors in the influencing factor library include the patient's age, sex, smoking history, fitness, whether to rescue, complication status, and whether to ventilator aid.
10. The method for realizing dynamic disease grouping based on multi-dimensional disease features as claimed in claim 6, wherein: in step 202, the calculation formula of the kini coefficient gini (d) before the slicing is:
gini coefficient after slicingA(D) The calculation formula is as follows:
wherein D is a clinical case and resource consumption dataset, m is the number of classifications made by classifying the clinical case and resource consumption dataset, piThe probability of any one of the classifications is selected for each case.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010442799.5A CN113707281A (en) | 2020-05-22 | 2020-05-22 | Dynamic disease group grouping realization method based on multi-dimensional disease characteristics |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010442799.5A CN113707281A (en) | 2020-05-22 | 2020-05-22 | Dynamic disease group grouping realization method based on multi-dimensional disease characteristics |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113707281A true CN113707281A (en) | 2021-11-26 |
Family
ID=78646348
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010442799.5A Pending CN113707281A (en) | 2020-05-22 | 2020-05-22 | Dynamic disease group grouping realization method based on multi-dimensional disease characteristics |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113707281A (en) |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106845098A (en) * | 2017-01-17 | 2017-06-13 | 天津艾登科技有限公司 | A kind of implementation method of the medical diagnosis on disease packet based on decision Tree algorithms |
-
2020
- 2020-05-22 CN CN202010442799.5A patent/CN113707281A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106845098A (en) * | 2017-01-17 | 2017-06-13 | 天津艾登科技有限公司 | A kind of implementation method of the medical diagnosis on disease packet based on decision Tree algorithms |
Non-Patent Citations (1)
Title |
---|
陈维雄 等: "病案首页应用于按病种分值付费的影响因素分析", 中国医疗保险, no. 1, pages 51 - 52 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111696660B (en) | Patient grouping method, device, equipment and storage medium based on artificial intelligence | |
US20220254493A1 (en) | Chronic disease prediction system based on multi-task learning model | |
DE102007007190A1 (en) | Patient classification predicting method for medical treatment, involves predicting classifications of medical treatment for patient based on patient treatment target data and set of predicted modules | |
US20100100394A1 (en) | Displaying clinical predicted length of stay of patients for workload balancing in a healthcare environment | |
CN111048190A (en) | DRG grouping method based on artificial intelligence | |
CN112086195A (en) | Admission risk prediction method based on self-adaptive ensemble learning model | |
CN112381598A (en) | Product service information pushing method and device | |
CN114266729A (en) | Chest tumor radiotherapy-based radiation pneumonitis prediction method and system based on machine learning | |
CN110895580B (en) | ICD operation and operation code automatic matching method based on deep learning | |
CN116910172A (en) | Follow-up table generation method and system based on artificial intelligence | |
CN117035368B (en) | Doctor dispatching method based on Internet | |
Basu et al. | Gene Therapy Versus Common Care for Eligible Individuals With Sickle Cell Disease in the United States: A Cost-Effectiveness Analysis | |
CN109740669A (en) | A kind of breast cancer pathology image classification method based on depth characteristic polymerization | |
CN109558461A (en) | A kind of medical data classification storage method and apparatus | |
CN113707281A (en) | Dynamic disease group grouping realization method based on multi-dimensional disease characteristics | |
CN117497149A (en) | Hospital management method and system based on artificial intelligence | |
CN111897857A (en) | ICU (intensive care unit) duration prediction method after aortic dissection cardiac surgery | |
TWM598504U (en) | System capable of generating a treatment planning sheet and integrating medical records/information of a patient thereof | |
Goodman et al. | Emergency department frequent user subgroups: Development of an empirical, theory-grounded definition using population health data and machine learning. | |
US20230068453A1 (en) | Methods and systems for determining and displaying dynamic patient readmission risk and intervention recommendation | |
CN115631823A (en) | Similar case recommendation method and system | |
CN115691820A (en) | Automatic construction system for multi-typing standard diagnosis and treatment scheme of disease types | |
CN116344005A (en) | Multi-objective optimization method for nurse allocation stage in HHCSRP | |
CN112364924A (en) | Deep learning-based oral medical image identification method | |
TW202119429A (en) | System and method capable of generating a treatment planning sheet and integrating medical records/information of a patient thereof |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |