JP2006309585A - Disease name encoding method and disease name encoding program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a disease name encoding method and a disease name encoding program, for efficiently encoding a disease name recorded by a character string. <P>SOLUTION: A management computer 21 produces a dictionary (an electronic dictionary) by use of a modifier master of a modifier master data storage part 32 and a disease name master of a disease name master data storage part 31. A constituent character of an unencoded disease name is extracted one by one, a vocabulary candidate recording the extracted constituent character as a head character is extracted, and a candidate list associating an appearance position of the constituent character in the unencoded disease name with the vocabulary candidate is generated. Combination of the vocabulary candidates properly expressing the whole disease name is generated by use of the vocabulary candidate included in the candidate list on the basis of the appearance position of the constituent character in the unencoded disease name and the number of characters of the vocabulary candidate, and the unencoded disease name is encoded on the basis of the generated combination. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a wound and disease name coding method and a wound and disease name coding program for coding a wound and disease name described in a character string, and in particular, an electronic medical chart system in which a medical institution codes a wound and disease name and manages the medical record, Alternatively, the present invention relates to a wound disease name coding method and a wound disease name coding program used in a receipt computer processing system or the like in which an insurer or an examination / payment institution encodes a wound disease name described in a receipt (medical remuneration statement) and performs processing related to the receipt.

  When an individual receives a medical practice at a medical institution, the individual pays his / her own share to the medical institution at the same time, and the medical institution issues a receipt (medical remuneration statement) to the insurance group and pays the individual. Request payment for non-medical expenses. Then, the examination / payment institution examines whether the receipt is in error, and the insurance group accepts the receipt through the examination and pays the medical institution.

  In the past, this processing was done using paper-based receipts. However, for efficiency, an acceptance system for the “Receipt Computer Processing System” of the examination and payment institution, led by the Ministry of Health, Labor and Welfare, was established in April 1999, and each hospital It is now possible to submit a receipt electronically.

In an electronic receipt, it is fundamental to use a wound name / modifier code and a modifier code to describe the wound name, but it is also possible to express the wound name as a character string.
Thus, when handling the receipt data in which the name of the wound is described as a character string, a system for efficiently coding the name of the wound described in the character string has been developed (see, for example, Patent Document 1). ). In the system disclosed in Patent Document 1, before collating the character string corresponding to the wound name described in the receipt with the wound name master, various different notations that may occur on the received wound name string are mastered. Convert to a character string for the name of the wound. Then, the name of the master wound is searched for in the converted character string, and when a plurality of master wound names are searched, the acceptance / rejection is determined based on the appearance position relationship, and then the modifier is detected for the remaining character string. .
JP 2002-366648 A (page 3-17)

  In the above system, when a plurality of master disease names are detected, the master disease name is determined first without considering the modifier. In this case, even if there is a combination of a master disease name and a modifier that can code the disease name to be coded without excess or deficiency, it may not be selected. Further, the above system is not very efficient because it presents the extracted candidate names such as the names of the sickness and allows the user to select them, and requires the user's work for encoding.

  The present invention has been made to solve the above-mentioned problems, and its purpose is to provide a wound name and disease name coding method and a wound name and name coding for efficiently coding a wound name described in a character string. To provide a program.

In order to solve the above-mentioned problem, the invention described in claim 1 includes a vocabulary data storage means for recording vocabulary data including a wound name and a wound name code, and vocabulary data including a modifier and a modifier code, and a computer. And a disease name encoding method for encoding an uncoded disease name described in a character string, wherein the computer extracts a component character included in the uncoded disease name A vocabulary candidate extraction step of extracting from the vocabulary data storage means a vocabulary candidate recorded with the constituent character as the first character; and associating an appearance position of the constituent character in the uncoded wound name with the vocabulary candidate A candidate list generating step for generating a candidate list, and using the vocabulary candidates included in the candidate list based on the appearance position and the number of characters of the vocabulary candidates, A combination generating step of generating a whole without excess or deficiency represents a combination of lexical candidates name, and be required to perform the encoding step of encoding the uncoded sick name based on the combination that is generated.

  According to a second aspect of the present invention, in the wound name coding method according to the first aspect, when the computer generates a plurality of combinations in the combination generation step, the computer is based on the number of characters of the wound name in each combination. In summary, the method further includes the step of assigning a priority to the generated combination.

  The invention according to claim 3 is the wound disease name encoding method according to claim 1 or 2, further using a unified rule data storage means that records dictionary filter data for unifying notation, The gist of the present invention is to further execute the step of unifying the notation of the uncoded disease name using the dictionary filter data.

  The invention according to claim 4 is the wound disease name encoding method according to any one of claims 1 to 3, wherein when the computer detects a delimiter in an uncoded disease name, the delimiter The method further includes the step of dividing the uncoded disease name in order to code each divided character string.

  The invention according to claim 5 is the wound name coding method according to any one of claims 1 to 4, wherein the computer uses the wound name code and the modifier code included in the received receipt data as the code. The gist is to further execute the step of generating a non-coded disease name by converting into a vocabulary using the vocabulary data storage means, and encoding the generated uncoded disease name.

  The invention according to claim 6 is described in a character string using a computer and a vocabulary data storage means in which vocabulary data including a wound name and a wound name code and vocabulary data including a modifier and a modifier code are recorded. A coding program for wound name and disease name for encoding the encoded non-coded disease name, wherein the computer extracts component characters included in the uncoded disease name, and the component characters Vocabulary candidate extraction means for extracting the vocabulary candidate recorded as the first character from the vocabulary data storage means, and a candidate for generating a candidate list in which the appearance position of the constituent characters in the uncoded wound name is associated with the vocabulary candidate Based on the list generation means, the appearance position, and the number of characters of the vocabulary candidate, the entire uncoded disease name is overlooked using the vocabulary candidates included in the candidate list. Combination generating means for generating a combination of lexical candidates representing not be summarized in that to function as encoding means for encoding the uncoded sick name based on the combination that is generated.

  According to a seventh aspect of the present invention, in the wound name encoding program according to the sixth aspect, when the computer further generates a plurality of combinations by the combination generation means, the number of characters of the wound name in each combination Based on the above, the gist is to function as a means for assigning a priority to the generated combination.

  The invention according to claim 8 is the sick and disease name coding program according to claim 6 or 7, further using a unified rule data storage means that records dictionary filter data for unifying notation, Further, the gist of the present invention is to allow the uncoded disease name to function as a means for unifying the notation using the dictionary filter data.

  The invention according to claim 9 is the wound name and disease name encoding program according to any one of claims 6 to 8, wherein the computer is further detected when a delimiter is detected in the uncoded wound name. The gist is to function as means for dividing the uncoded disease name by the delimiter and to code each divided character string.

  A tenth aspect of the present invention is the injury / illness name encoding program according to any one of the sixth to ninth aspects, wherein the computer is further processed with an injury / disorder name code and a modifier code included in the received receipt data. The gist of the present invention is to use the vocabulary data storage means to convert the vocabulary into a vocabulary and generate an uncoded disease name, and to code the generated uncoded disease name.

(Function)
According to the first or sixth aspect of the invention, the computer first extracts the constituent characters of the uncoded wound name. The vocabulary candidate recorded with the constituent characters as the first character is extracted from the vocabulary data storage means in which the vocabulary data including the wound name and the wound name code and the vocabulary data including the modifier and the modifier code are recorded, A candidate list is generated in which lexical candidates are associated with the appearance positions of constituent characters in the uncoded disease names. Then, based on the appearance position and the number of characters of the vocabulary candidates, a vocabulary candidate combination that expresses the whole uncoded disease name without excess or deficiency is generated using the vocabulary candidates included in the candidate list, and based on this combination Code the name of the uncoded wound. Thereby, based on the constituent characters of the uncoded disease name, a combination of the disease name and the modifier can be generated, and the uncoded disease name can be coded based on this combination. Therefore, it is possible to efficiently code the injury / illness name described in the character string.

  According to the invention described in claim 2 or 7, when a plurality of combinations are generated, a priority is given to the generated combinations based on the number of characters of the names of wounds in each combination. Thereby, when a plurality of combinations are generated, it is possible to specify the combination used for coding the wound name based on the number of characters of the wound name in each combination. In this case, for example, when preferentially specifying a combination with a large number of characters in the name of a disease, the computer may automatically specify a combination in which the name of the disease with the largest number of characters is used. The combinations may be output in the order of the largest number, and the user may specify the combination from among the combinations. As a result, it can be applied to batch processing and interactive processing. In batch processing, processing when multiple combinations are generated can be automated. In interactive processing, the names of the disease names are output in descending order. Therefore, the user can select a combination efficiently.

  According to the third or eighth aspect of the present invention, the unified rule data storage means in which dictionary filter data for standardizing notation is further used. Then, the computer unifies the notation of the uncoded disease name using the dictionary filter data. Thereby, the fluctuation | variation of the character string of the uncoded wound disease name in an uncoded wound disease name can be eliminated, and a description can be unified. Then, using this unified notation, it is possible to code the name of the wound.

  According to the invention of claim 4 or 9, when the computer detects a delimiter character in the uncoded disease name, the computer divides the uncoded disease name by the delimiter character, and the divided character string Code each one. Thereby, when a delimiter is included in the uncoded disease name, the disease name can be coded for each character string delimited by the delimiter.

According to the invention described in claim 5 or 10, the computer generates an uncoded wound name by converting the wound name and modifier code included in the received receipt data into a vocabulary. Code the name of the uncoded wound. As a result, it is possible to re-encode the wound name for the uncoded wound name that allows multiple combinations of wound names and modifiers regardless of the wound name and modifier code included in the receipt data. It becomes. Accordingly, a unified disease name code and modifier code can be used for the same uncoded disease name.

  According to the present invention, it is possible to efficiently code a wound name described in a character string.

  Hereinafter, an embodiment embodying the present invention will be described with reference to FIGS. In the present embodiment, a description will be given as a wound name coding method and a wound name coding program for coding an uncoded wound name recorded in an electronic receipt.

As shown in FIG. 1, the sick name coding server 20 is connected to a receipt computer processing system 40.
The receipt computer processing system 40 receives an electronic receipt (receipt data) and inputs it to the screen examination system. The wound and disease name encoding server 20 takes in the receipt data (data before coding) from the screen examination system, encodes the uncoded wound and disease names by batch processing, and encodes the uncoded wound and disease names ( The encoded data is transferred to the receipt computer processing system 40. The receipt computer processing system 40 performs processing related to the receipt using the encoded data. In this receipt computer processing system 40, the receipt is displayed on the screen based on the receipt data, and the examiner examines the receipt based on this screen display. And when the examination by the person in charge of examination in the receipt computer processing system 40 is completed, the receipt computer processing system 40 performs counting arrangement regarding the receipt. Based on the processing result by the receipt computer processing system 40, a bill payment process is performed.

  The wound name coding server 20 includes a management computer 21. The management computer 21 executes various processes for encoding an uncoded injury and disease name recorded in the receipt data fetched from the receipt computer processing system 40.

  The management computer 21 includes a CPU, a RAM, a ROM, and the like (not shown), and performs processing described later (processing including a constituent character extraction stage, a vocabulary candidate extraction stage, a candidate list generation stage, a combination generation stage, and a coding stage). . The management computer 21 functions as a constituent character extraction unit, a vocabulary candidate extraction unit, a candidate list generation unit, a combination generation unit, a coding unit, and the like by executing the injury / illness name encoding program for this purpose.

The constituent character extracting means extracts constituent characters included in the uncoded wound name.
The vocabulary candidate extracting means extracts the vocabulary candidates recorded with the constituent characters as the first character from a dictionary created based on the wound name master and the modifier master.

The candidate list generating means generates a candidate list in which appearance positions are associated with vocabulary candidates.
Based on the appearance position of the constituent characters in the uncoded wound name and the number of characters of the vocabulary candidate, the combination generation means uses a vocabulary candidate included in the candidate list to represent a combination that represents the entire uncoded wound name without excess or deficiency. Generate.

The encoding means encodes the uncoded disease name based on the generated combination.
Further, the management computer 21 includes a wound name master data storage unit 31, a modifier master data storage unit 32, a notation unified master data storage unit 33 as a unified rule data storage unit, a pre-encoding data storage unit 34, and a post-encoding data storage unit 34. The data storage unit 35 is connected.

  The wound name master data storage unit 31 records a wound name master that constitutes a correspondence table between wound names and wound name codes. Each data (injured disease name master data) of the injured disease name master is constituted by data relating to the injured disease name and the injured disease name code corresponding to the injured disease name. The wound name master is acquired and recorded from the wound name master managed by the public institution, and when the wound name master managed by the public institution is updated, it is reacquired and updated.

  The modifier master data storage unit 32 stores a modifier master that constitutes a correspondence table between modifiers and modifier codes. Each data of the modifier master (modifier master data) is composed of data relating to a modifier and a modifier code corresponding to the modifier. The modifier master is updated as appropriate by newly registering a modifier.

  In the notation unified master data storage unit 33, a notation unified master that constitutes a replacement table for unifying the fluctuation of the notation is recorded. Each data of the notation unified master (notation unified master data as dictionary filter data) is composed of data relating to the notation before replacement and the notation after replacement. FIG. 10 shows a replacement table 140 for explaining the notation unified master. In the replacement table 140, the content of replacement is shown as “Remarks” together with the data before replacement and the data after replacement of the unified notation master. The notation unified master is appropriately updated by newly registering data before and after replacement.

In the pre-encoding data storage unit 34, receipt data (pre-encoding data) taken from the receipt computer processing system 40 is recorded. Of the wound name records included in this receipt data, those with the wound name code “0000999” are to be coded later. The injury and illness name record has a specification according to “Reception Computer Processing System Magnetic Receipt Creation Guide-Medical Department” (Creation of Social Insurance Medical Fee Payment Fund).

The wound name record includes data relating to record identification information, a wound name code, a modifier code, and a wound name.
In the record identification information data area, data relating to identification information representing a disease name record is recorded.

The wound name code corresponding to the corresponding wound name is recorded in the wound name code data area. In the case of recording a coded injury / illness name, data relating to the injury / illness name code corresponding to the corresponding injury / illness name is recorded. For example, in the case of “appendicitis”, the wound name code “5419001”
Is recorded. When recording a non-coded injury name (uncoded injury name), “0000999” is recorded.

  In the modifier code data area, a modifier code corresponding to the relevant modifier is recorded. The wound name codes are recorded in the order in which they are displayed on the receipt, and up to 20 can be recorded. Specifically, when a modifier is used and prefixed, for example, “4012” is recorded in the case of “acute appendicitis”. When a modifier is used and suffixed, for example, “8002” is recorded in the case of “suspected appendicitis”. When both are attached when using a modifier, for example, “40128002” is recorded in the case of “suspected acute appendicitis”. In the case where no modifier is used or in the case of an uncoded disease name, no modifier code is recorded.

In the wound name data area, data relating to a wound name (uncoded wound name) represented by a character string is recorded when an uncoded wound name code “0000999” is used as the wound name code. It is not recorded if the wound name code is other than the uncoded wound name code “0000999”.

  Thus, in the wound name record, the wound name is represented by a combination of one wound name code and zero or more modifier codes. As in the case of the disease name record, in the process described later, the disease name is encoded by a combination of one disease name code and zero or more modifier codes.

  In the post-coding data storage unit 35, receipt data (post-coding data) coded with respect to the disease name record to be coded is recorded. In this post-coding data, the encoded injury / sickness name record is appended with the injury / sickness name code and modifier code coded based on the injury / sickness name (uncoded injury / sickness name) represented by a character string. Has been.

(Summary of sickness name coding)
Next, an overview of the coding of injury and illness names realized in the present embodiment will be described.
The coding of injury and illness names realized in the present embodiment can be classified into the following three types. That is, there are three types: (1) perfect match with injury / disease name, (2) complete match with injury / disease name / modifier, and (3) partial match with injury / disease name. (1) In the case of an exact match with the wound name, search for the wound name that completely matches the wound name before coding from the wound name master and replace it with the corresponding code. (2) In the case of a perfect match with the wound name / modifier, the wound name before coding is decomposed as a sequence of one wound name and a plurality of modifiers and replaced with the corresponding code. As will be described later, for example, “acute gastritis” has a method of encoding the injury name “acute gastritis” and a method of encoding the combination of the modifier “acute” and the injury name “gastritis” After the combination is tried, the former that matches the longest injury name is adopted. (3) In the case of partial coincidence of wound and disease names, a wound and disease name that partially matches the wound and disease names before coding is searched from the wound and disease name master and replaced with the corresponding code.

  In addition, any dictionary can be used as long as there is a dictionary of wound names and modifiers that can be associated with wound name codes and modifier codes used in the receipt computer processing system. For example, “Standard disease name master for ICD10 compatible electronic medical record” distributed by MEDIS-DC (Medical Information System Development Center) via the Internet can be used. The name and modifier of the wound and disease listed in the master can be associated with the name and modifier code of the receipt computer system through the exchange code.

  FIG. 5 shows an example 100 of the coding method when the wound name and disease name are completely matched, the wound and disease name / modifier is completely matched, the conversion by the uncoded wound and disease name conversion master, etc., and the partially matched wound name and disease name. In this case 100, for each case, the wound name before coding (uncoded wound name), the wound name after coding, the wound name, the modifier code, and the modifier are shown.

Furthermore, in order to improve the coding accuracy, the following functions are realized.
Parenthesis modifiers: Some of the uncoded wound names have modifiers written in parentheses. Since these cannot be converted correctly by the method described above, the conversion is attempted without the parentheses. FIG. 6 shows a case 110 in this case.

  Notation fluctuation: Due to the notation fluctuation (notation is different but synonymous), it may not match the wound name in the wound name master. Specifically, there are the following. In other words, the difference between the sound-drawing “-” and the hyphen “-”, or the presence or absence thereof, the roar “nya” and the direct sound “ya”, the correct character “neck” and the different character “neck”, the Chinese character notation “欝” and the hiragana notation “depression” , Corresponds to the fluctuation of the notation. These are dealt with by using the unified master recorded in the unified master data storage unit 33 as an additional dictionary before matching with the disease name master. FIG. 7 shows a case (conversion example) 120 in this case.

Further, there may be a case where a plurality of wound names are described in one wound name column by separating them with characters such as “•”, “+”, “,”. About this, it divides | segments with a delimiter and performs the above-mentioned encoding about each. FIG. 8 shows a case 130 in this case.

(Processing procedure)
Next, a processing procedure in the case of encoding an uncoded wound name using the system as described above will be described with reference to FIGS. The management computer 21 of the wound name coding server 20 takes in the receipt data (pre-coding data) from the receipt computer processing system 40 and records it in the pre-coding data storage unit 34. Then, the management computer 21 performs the following processing for the disease name code of “0000999” in the disease name record of the receipt data.

  As shown in FIG. 2, first, the management computer 21 of the wound / patient name coding server 20 reads the wound / patient name master and the modifier master from the wound / patient name master data storage unit 31 and the modifier master data storage unit 32, respectively. A dictionary (electronic dictionary) is created (step S1-1). The created dictionary is recorded in a predetermined storage unit (not shown) of the management computer 21. By recording this dictionary, this storage unit functions as vocabulary data storage means. This dictionary is expressed by an index with the first character and a link to the name of the disease. This dictionary will be described with reference to FIG.

  As shown in FIG. 9, the management computer 21 includes a dictionary including an index part 60 composed of the first character of the name of the sick or sickness or a modifier and a word part 70. Here, the index part 60 is composed of an array of the first character of the wound name and modifier. The word part 70 comprises a list of wound names and modifiers. Each record of the word part 70 is composed of a classification 71 for identifying the name of the wound / disease name / qualifier, a code 72 of the name of the wound / disease or modifier, and a name of the wound / disease name or modifier (dictionary registration name) 73. . The category 71 is used to identify the name of the wound / disease name / modifier. Specifically, “b” represents the name of the wound / disease and “z” represents the modifier. The code 72 represents a disease name code or a modifier code. The name 73 represents a wound name or a modifier. In this way, the management computer 21 registers the wound name and modifier in a mixed manner in this dictionary. Further, an index table or the like is registered in the same dictionary as necessary.

  When registering the name of the sickness and the modifier in the dictionary, the management computer 21 uses the unified notation master recorded in the unified notation master data storage unit 33 to replace the character string having the notation. Register from. An example of the registration data of the notation unified master is shown as the replacement table 140 in FIG. 10 as described above. For example, if “Harler-Scheier Syndrome” is before replacement, the sound-drawing is deleted to convert it to “Hara-Shayer Syndrome” after replacement.

In the following processing, processing is performed using the dictionary created in this way. In addition, when there are multiple disease name records to be encoded (those whose disease name code is “0000999”), use the dictionary created and recorded for all the disease name records to be encoded. Process. That is, for the second and subsequent encoding target disease name records, the processing from step S1-2 is executed.

  Next, the management computer 21 unifies the notation (uncoded wound name) of the wound name in the wound name record to be coded (step S1-2). More specifically, the management computer 21 uses the notation unified master recorded in the unified master data storage unit 33 to replace the uncoded wound disease name in the same manner as in the case of creating the dictionary. With this processing, the management computer 21 performs coding by replacing character strings using the notation unified master, as in each case of the case 120 shown in FIG.

Next, the management computer 21 performs candidate extraction processing (step S1-3). In this candidate extraction process, all the wound names and modifiers in the dictionary that appear in the uncoded wound names (or the uncoded wound names after replacement if replacement is necessary) are extracted. The procedure of this candidate extraction process will be described with reference to FIG.

  As shown in FIG. 3, first, the management computer 21 scans the wound name (uncoded wound name) of the wound name record to be encoded one character at a time from the beginning, and specifies one character of the constituent characters (step S2). -1). The case where this treatment is performed for “acute ulcerative colitis” will be described below. That is, scanning is performed character by character in the uncoded disease name character string 80 of FIG. Here, as shown in FIG. 11, the management computer 21 searches for “abrupt” as the first character of “acute ulcerative colitis”, which is the uncoded wound name string 80.

  Then, the management computer 21 extracts a dictionary registration name candidate having the identified constituent character as the first character (step S2-2). Specifically, a list of candidate dictionary registration names is acquired using an index corresponding to one specified constituent character. Here, as shown in FIG. 11, the management computer 21 acquires a dictionary registration name whose first character is “sudden”.

  Then, the management computer 21 sequentially specifies the acquired list, and checks whether or not all of the dictionary registration names match the portion of the uncoded injury / illness name starting from the specified constituent character (step S2). -3). For example, the dictionary registration name “rapidly progressive meniscus-forming glomerulonephritis” does not match the portion of the uncoded wound disease name that starts with the specified constituent character. If the entire dictionary registration name does not match the portion of the uncoded injury / defect name starting from the specified constituent character (NO in step S2-3), the management computer 21 selects from the extracted candidates. The next candidate is specified (step S2-6). And the process of step S2-3-step S2-5 is performed about the specified next candidate.

  When a matching candidate is acquired (YES in step S2-3), the management computer 21 records the dictionary registration name, code, and appearance position in the candidate list (step S2-4). Specifically, the management computer uses a classification, code (sickness name code or modifier code), dictionary registration name (sickness name or modifier), and position of appearance as one record of the candidate list for identifying the name of the sickness name / qualifier. 21 is recorded in a storage unit (not shown). For example, in the case of the above-mentioned “acute ulcerative colitis”, by repeating the above-described processing sequentially, as indicated by a bold line in FIG. , “Acute ulcerative colitis”, “acute ulcerative”, and “acute” are obtained.

  When not all the candidates extracted for the constituent characters have been processed (NO in step S2-5), the management computer 21 identifies the next candidate from the extracted candidates. (Step S2-6). And the process of step S2-3-step S2-5 is performed about the specified next candidate. When all the candidates extracted for the identified constituent characters have been processed (in the case of YES in step S2-5), the management computer 21 includes the next constituent in the uncoded wound name to be coded. Whether there is a character is checked (step S2-7). If there is the next constituent character (YES in step S2-7), the management computer 21 returns to the process of step S2-1. Then, the uncoded disease name is scanned character by character from the beginning, and the next constituent character in the uncoded disease name is identified (step S2-1). In the case of “acute ulcerative colitis”, the management computer 21 specifies “sex” as the second character, and sequentially specifies “sex”, “ulcer”, “ulcer”.

In this way, the uncoded wound name is scanned one by one from the beginning, and if there is no next constituent character in the uncoded wound name (NO in step S2-7), the management computer 21 This candidate extraction process is terminated. As a result of searching for all the constituent characters of “acute ulcerative colitis”, a candidate list shown as candidate list explanation display 150 in FIG. 12 is obtained. In the candidate list explanation display 150, for each candidate obtained by sorting the search results into injury and disease names and modifiers, each candidate is classified (type of injury and disease name / modifier), code (injury and disease name code or modifier code), The dictionary registration name (injury name or modifier) and the application position are shown. Specifically, “Acute ulcerative colitis” (injury name), “Acute ulcerative” (modifier), and “Acute” (modifier) are assumed to have “sudden” at the appearance position “0” as the first character. ) Is obtained. “Ulcerative colitis” (injured disease name) and “ulcerative” (modifier) are obtained with “collapse” at the appearance position “2” as the first character. “Colitis” (injured disease name) is obtained with “Large” at the appearance position “5” as the first character.

  Next, the management computer 21 performs a combination listing process (step S1-4). In this combination enumeration process, all combinations that are composed of one wound name and zero or more modifiers and that represent the entire wound name are selected from the candidates extracted in the candidate extraction process. And enumerate.

  Specifically, the combination enumeration process is performed by executing the following “combination enumeration process (N)” with the start position “0” as an argument. This process finds all combinations by calling itself recursively. This combination enumeration process will be described with reference to FIG. Here, the above-described case will be described with reference to FIG. In FIG. 13, each candidate in the candidate list is displayed in association with each other according to the combination for the purpose of explaining the combination. Here, for each candidate in the candidate list, a classification 91 for identifying the name of the wound / symptom / modifier, a dictionary registered name 92 and an appearance position 93 of the candidate wound / patient name or modifier are shown.

  As shown in FIG. 4, first, the management computer 21 checks whether N is equal to or longer than the length (number of characters) of the name of the disease to be coded (uncoded disease name) (step S3-1). At the start of processing, N is the start position “0”. When N is not equal to or longer than the length of the encoded disease name (uncoded injury disease name) (NO in step S3-1), the management computer 21 extracts the first candidate from the candidate list (step S3- 2). In the above case, as shown in FIG. 13, the extracted first candidate is “acute ulcerative colitis” (category “b (sickness name)”, appearance position “0”).

  If N matches the candidate appearance position (NO in step S3-3), the management computer 21 records this candidate as a combination target candidate (step S3-4). More specifically, the classification, code (wound / sickness name code or modifier code), dictionary registration name (wound / sickness name or modifier), and position of appearance are recorded. The appearance position of the above “acute ulcerative colitis” is “0”, which matches N (= 0). Therefore, this “acute ulcerative colitis” is recorded as a candidate for combination. Specifically, the classification “b (name of injury)”, code “0091015”, dictionary registration name “acute ulcerative colitis”, and appearance position “0” are recorded.

  Next, the management computer 21 sets X to N plus the number of characters in the candidate dictionary registration name (step S3-5). In the above case, the number of characters of “acute ulcerative colitis” is “8”, so “X = 8”. Then, the combination enumeration process (X) is called (step S3-6).

  In step S3-1 of the combination enumeration process (X (= 8)), X (= 8) is equal to the number of characters of the name of the disease to be encoded, that is, equal to or greater than the number of characters of the name of the disease to be encoded (step S3). Since this is YES at −1), the combination enumeration process (X (= 8)) is terminated.

  The management computer 21 checks whether there is a next candidate in the combination enumeration process (N) (step S3-7). When there is a next candidate (YES in step S3-7), the management computer 21 takes out the next candidate (step S3-8). And it returns to the process of step S3-3 and performs the process of step S3-3-step S3-6 about the taken-out candidate. If there is no next candidate (NO in step S3-7), the combination enumeration process (N) is terminated.

Hereinafter, the above example will be described in detail.
In the above case, in the combination enumeration process (N (= 0)), candidates are sequentially extracted from the candidate list by the process of step S3-8. Here, for “ulcerative colitis” (category “b (sickness name)”, appearance position “2”), N (= 0) does not match the candidate appearance position “2” (step S 3-3). 3 is YES), the processing of step S3-4 to step S3-6 is skipped and the next candidate is examined (step S3-7). In addition, regarding “colitis” (category “b (sickness name)”, appearance position “5”), N (= 0) does not match the candidate appearance position “5” (YES in step S3-3). Therefore, the processing from step S3-4 to step S3-6 is skipped and the next candidate is examined (step S3-7).

  For “acute” (section “z (modifier)”, appearance position “0”), N (= 0) matches the candidate appearance position “0” (because it is NO in step S3-3). Then, it is recorded as a combination target candidate (step S3-4). In this case, X (= 2) is obtained by adding N (= 0) to the number of characters (2) of the dictionary registration name (step S3-5). Then, the combination enumeration process (X (= 2)) is called (step S3-6). In this combination listing process (X (= 2)), for “ulcerative colitis” (category “b (sickness name)”, appearance position “2”), X (= 2) is the candidate appearance position “2”. Therefore, “ulcerative colitis” is recorded as a combination target candidate in association with “acute” as shown in FIG. 13 (step S3-4). In this case, X (= 2) plus the number of characters of the dictionary registration name (6) is set as X (= 8) (step S3-5), and the combination enumeration process (X (= 8)) is called (step S3). -6). In step S3-1 of the combination enumeration process (X (= 8)), X (= 8) is equal to or greater than the number of characters of the name of the disease to be encoded (because it is YES in step S3-1). The process (X (= 8)) is terminated.

  Then, the management computer 21 continues the combination enumeration process (X (= 2)). In this combination enumeration process (X (= 2)), when “ulcerative” (category “z (modifier)”, appearance position “2”) is taken out as the next candidate, , X (= 2) matches the candidate appearance position “2” (NO in step S3-3). Therefore, “ulcerative” is recorded as a combination target candidate in association with “acute” as shown in FIG. 13 (step S3-4). In this case, X (= 2) plus the number of characters (3) of the dictionary registration name is set as X (= 5) (step S3-5), and the combination enumeration process (X (= 5)) is called (step S3). -6). In this combination listing process (X (= 5)), X (= 5) matches the candidate appearance position “5” for “colitis” (category “b (sickness name)”, appearance position “5”). Therefore, “colitis” is recorded in association with “acute” and “ulcerative” as shown in FIG. 13 (step S3-4). In this case, X (= 5) plus the number of characters (3) of the dictionary registration name is set as X (= 8) (step S3-5), and the combination enumeration process (X (= 8)) is called (step S3). -6). In step S3-1 of the combination enumeration process (X (= 8)), X (= 8) is equal to or greater than the number of characters of the name of the disease to be encoded (because it is YES in step S3-1). The process (X (= 8)) is terminated.

Then, the management computer 21 continues the combination enumeration process (X (= 5)), and if there is no next candidate (NO in step S3-7), the management computer 21 determines that this combination enumeration process (X ( = 5)) is terminated. Further, the management computer 21 continues the combination enumeration process (X (= 2)), and if there is no next candidate (NO in step S3-7), the management computer 21 determines that this combination enumeration process (X ( = 2)) is finished.

  Then, the management computer 21 continues the combination listing process (N (= 0)). In this combination enumeration process (N (= 0)), when “acute ulcer” (category “z (modifier)”, appearance position “0”) is extracted as the next candidate, this “acute ulcer” Since N (= 0) matches the candidate appearance position “0” (NO in step S3-3), it is recorded as a candidate for combination (step S3-4). In this case, the combination enumeration process (X (= 5)) is called (step S3-5), where X (= 5) is obtained by adding N (= 0) to the number of characters (5) of the dictionary registration name (step S3-5). -6). In this combination listing process (X (= 5)), X (= 5) matches the candidate appearance position “5” for “colitis” (category “b (sickness name)”, appearance position “5”). Therefore, “colitis” is recorded as a candidate for combination in association with “acute ulcer” as shown in FIG. 13 (step S3-4). In this case, X (= 5) plus the number of characters (3) of the dictionary registration name is set as X (= 8) (step S3-5), and the combination enumeration process (X (= 8)) is called (step S3). -6). In step S3-1 of the combination enumeration process (X (= 8)), X (= 8) is equal to or greater than the number of characters of the name of the disease to be encoded (because it is YES in step S3-1). The process (X (= 8)) is terminated.

  Then, the management computer 21 continues the combination listing process (X (= 5)). If there is no next candidate (NO in step S3-7), the management computer 21 ends this combination listing process (X (= 5)).

  Then, the management computer 21 continues the combination listing process (N (= 0)). If there is no next candidate (NO in step S3-7), the management computer 21 ends this combination listing process (N (= 0)).

As a result, in the case described above, four combinations are obtained as in the combination result 160 shown in FIG.
Next, the management computer 21 selects candidates from the listed combination results (step S1-5). Specifically, the selection of candidates is performed as follows. (1) Perfect match with injury / sickness name: If a single injury / sickness name is found that matches the whole, select it. (2) Perfect match with injury / illness name / qualifier: When there is no match for injury / illness name / exact match and a single injury / disaster name and one or more modifiers are found Choose it. If there are multiple applicable items, select the one with the longest name. In the case of the above case, the second, third, and fourth lines of the combination result 160 in FIG. 14 correspond to this, and the second line with the longest wound name is selected (however, it corresponds to “exact match with wound name”). , Not actually selected). (3) Partial match of injury and illness name: If there is no match for “exact match with injury or illness name” and “exact match with injury or illness name / qualifier”, Choose it. If there are multiple applicable items, select the one with the longest name. Of the listed combinations, those that include a plurality of injury and illness names (category “b”) are discarded.

As a result of selecting candidates in this way, in the case described above, only the wound name “acute ulcerative colitis” (wound name code “0091015”) corresponding to the first line of the combination result 160 in FIG. A combination consisting of

If no result is obtained in this candidate selection, parentheses and delimiters included in the uncoded disease name are excluded, and the above steps S1-2 to S1-5 are executed again. To do. For example, if the name of the uncoded injury is “keloid (shoulder, back)”, it is converted to “keloid shoulder back” by excluding parentheses and separators. And it tries to code about this "keloid shoulder back". Examples of parentheses and delimiters to be excluded include “” (blank character), “•”, “−”, “−”, “(”, “)”, and the like. Note that each case of the case 110 shown in FIG. 6 described above shows the result of the encoding attempt excluding the delimiter character.

  Furthermore, when a result is not obtained by the procedure so far, an uncoded disease name is divided by a delimiter that appears in the uncoded disease name, and step S1-2 to step S1- Step 5 is executed again. In addition, in the division by this delimiter, “” (blank character), “•”, “,”, “.”, “,”, Etc. are treated as delimiters. However, if a delimiter appears in a character string enclosed in parentheses, it is not regarded as a delimiter. For example, an uncoded wound name “ulcer, keloid (shoulder, back)” is divided into “ulcer” and “keloid (shoulder, back)”. Note that the case 130 shown in FIG. 8 described above shows the result of trying to code each of the results by dividing the uncoded disease name by the delimiters that appear in the uncoded disease name as described above. Yes. In the case 130, three combinations of wound names and modifiers encoded for one wound name record are created. In addition, when there is a result for which the candidate for encoding could not be specified among the results divided in this way, parentheses and delimiters are excluded as described above for those for which the candidate could not be specified. Then, the above-described processing of step S1-2 to step S1-5 is executed again.

  Next, the management computer 21 encodes the selected candidate, adds the selected combination of the disease name code and the modifier code to the injury name record of the pre-encoded data, and stores the encoded data storage unit. 35 (step S1-6). If there are a plurality of modifiers in the selected combination, a plurality of modifier codes are recorded, and if there are no modifiers, no modifier code is recorded.

In this way, the encoded data storage unit 35 uses the combination selected for the wound name record for the wound name record in which the uncoded wound name is used (the wound name code is “0000999”). An injury / sickness name code and a modifier code are added and recorded. Note that, for the wound name and disease name records that were originally encoded (those with names other than “0000999”), the wound name records are recorded in the encoded data storage unit 35 as they are. As for the data other than the injury / patient name record of the receipt data, the same data as the pre-encoding data is recorded in the post-encoding data storage unit 35.

  Then, the wound name coding server 20 delivers the receipt data (data after coding) in which the uncoded wound name is coded to the receipt computer processing system 40. The receipt computer processing system 40 uses this receipt data (encoded data) to perform a process for reviewing the receipt.

As described above, according to the present embodiment, the following effects can be obtained.
In the above embodiment, the management computer 21 creates a dictionary (electronic dictionary) using the modifier master in the modifier master data storage unit 32 and the wound name master in the wound name master data storage unit 31. Then, the constituent characters of the uncoded disease name are extracted one by one, the vocabulary candidate recorded with the extracted component character as the first character is extracted, and the appearance of the constituent characters in the uncoded disease name is extracted from this vocabulary candidate A candidate list in which positions are associated is generated. Then, based on the appearance position of the constituent characters in the uncoded disease name and the number of characters in the vocabulary candidate, the vocabulary candidate included in the candidate list is used to generate a combination of vocabulary candidates that represents the entire disease name without excess or deficiency. An uncoded wound name is encoded based on the combination. As a result, the combination of the wound name and the modifier can be specified based on the constituent characters of the uncoded wound name, and the uncoded wound name can be coded based on this combination. Therefore, it is possible to efficiently code the injury / illness name described in the character string.

  In the above embodiment, when a plurality of combinations are generated, the management computer 21 specifies the combination having the largest number of characters of the wound name in each combination, and the wound and disease corresponding to the wound name and modifier included in this combination A name code and a modifier code are assigned. As a result, when a plurality of combinations of wound names and modifiers can be generated, it is possible to use the one having the largest number of characters in the wound name.

  In the above embodiment, the management computer 21 uses the notation unified master recorded in the notation unified master data storage unit 33 to unify the notation of the uncoded wound name. Thereby, the fluctuation of the character string of the uncoded injury / illness name can be eliminated, and the notation can be unified. Then, using this unified notation, it is possible to code the name of the wound.

  In the above embodiment, when the management computer 21 detects a delimiter in an uncoded disease name, the management computer 21 divides the uncoded disease name by the detected delimiter, and encodes each divided character string. I do. Thereby, when a delimiter is included in the uncoded disease name, the disease name can be coded for each character string delimited by the delimiter.

  In the above embodiment, when the management computer 21 detects two or more names of sicknesses from the generated combination, this combination is discarded. As a result, it is possible to specify the combination of the wound name and the modifier and exclude the combination including two or more wound names so as to encode the wound name.

  In the above embodiment, the management computer 21 lists all combinations of wound names and modifiers that express uncoded wound names without excess or deficiency, and selects the one having the largest number of characters in the wound name. As a result, when a combination of a wound name and a modifier that expresses an uncoded wound name without excess or deficiency using a wound name master and a modifier master is possible, an uncoded wound name must be used without excess or deficiency. The combinations to be expressed can be specified.

  In the above-described embodiment, the management computer 21 unifies the notation of the injury and illness name master and the modifier master using the notation unified master recorded in the notation unified master data storage unit 33. As a result, even when there is a fluctuation of the notation (character string) in each registered data of the sickness name master and the modifier master, the fluctuation of the character string can be eliminated and the notation can be unified. Then, using this unified notation, it is possible to code the name of the wound.

In addition, you may change the said embodiment into the following aspects.
In the above embodiment, an uncoded wound name is coded when the wound name is not coded in the wound name record of the receipt data. Instead, the management computer 21 converts the wound name and modifier code contained in the wound name record of the receipt data into a vocabulary using the dictionary (electronic dictionary) of the above-described embodiment, thereby converting the uncoded wound name. May be generated, and the name of the wound and disease may be encoded again for the uncoded wound and disease name. As a result, for uncoded wound names that allow multiple combinations of wound names and modifiers, re-encoding the wound names regardless of the wound name and modifier code included in the wound name record in the receipt data. Can be performed. Accordingly, a unified disease name code and modifier code can be used for the same uncoded disease name.

In the above embodiment, a dictionary is created when coding a disease name. Instead of this, a dictionary may be created when the disease name master data storage unit 31, the modifier word master data storage unit 32, or the unified master data storage unit 33 is updated. Then, the created dictionary data may be recorded, and this dictionary data may be used when coding a disease name. Thereby, the process at the time of coding a sick-and-sickness name can be made efficient.

  In the above embodiment, in the candidate extraction process (step S1-3), the constituent characters of the uncoded injury / illness name are specified one by one, and the candidates having the identified constituent characters as the first character are extracted. Instead of this, a candidate having the first character as a constituent character of an uncoded disease name may be extracted in accordance with the candidate extraction status. For example, in the case described above, only “acute ulcerative colitis” (8 characters), “acute” (2 characters), and “acute ulcerative” (5 characters) are searched at the appearance position “0”. Since one character is not searched, a search for a candidate having “sex” (appearance position “1”) as the first character is not performed, and “crush” (appearance position “2”) is set as the first character. You may search for the candidate to do. Thereby, processing can be made efficient.

  In the above embodiment, when a plurality of combinations are generated, the combination having the largest number of characters in the name of the disease in each combination is specified. Instead of this, a combination having the smallest number of characters in the name of the injury or illness may be specified. Thus, when selecting the combination with the least number of characters in the wound name, the wound name excluding the modifier can be obtained as a code, which is convenient when used for statistical purposes, for example.

  ○ In the above embodiment, when the wound name coding process is performed in a batch process and a plurality of combinations are generated, the combination having the largest number of characters of the wound name in each combination is identified, and the wound corresponding to this combination Name code and modifier code were set automatically. Instead, when the present invention is applied to the case where a disease name is encoded by interactive processing, and a plurality of combinations are generated, the management computer 21 assigns a priority to the generated combinations, and A combination may be output according to the priority order, and the user may specify the combination from among the combinations. For example, when preferentially selecting a character with a large number of characters in the name of the disease, the management computer 21 outputs the combinations in the order of the number of characters in the name of the disease, and the user specifies the combination. In this case, since the number of characters in the name of the injury / illness is output in the order in which the number of characters is increased, the user can confirm the combination of the injury / disaster name and the modifier in the output order and select the combination efficiently. Can do.

  In the above embodiment, the present invention is embodied when the wound name (uncoded wound name) in the wound name record included in the receipt data is coded, but the present invention is not limited to this. The present invention can be used when coding a disease name recorded in a character string, for example, when coding a disease name recorded in an electronic medical record or the like.

1 is a schematic diagram of a system according to an embodiment of the present invention. Explanatory drawing of the process sequence of one Embodiment of this invention. Explanatory drawing of the process sequence of one Embodiment of this invention. Explanatory drawing of the process sequence of one Embodiment of this invention. Explanatory drawing which shows the example of the encoding method of the wound name of this invention. Explanatory drawing which shows the example of the encoding method of the wound name of this invention. Explanatory drawing which shows the example of the encoding method of the wound name of this invention. Explanatory drawing which shows the example of the encoding method of the wound name of this invention. Explanatory drawing of dictionary data. Explanatory drawing of a notation unified master. Explanatory drawing about a candidate extraction process. Explanatory drawing about a candidate extraction process. Explanatory drawing about a combination enumeration process. Explanatory drawing about a combination enumeration process.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 21 ... Management computer, 31 ... Disease name master data storage part, 32 ... Modifier master data storage part, 33 ... Unified master data storage part as a unified rule data storage means.

Claims (10)

Using a computer and a vocabulary data storage means in which vocabulary data including the wound name and wound name code and vocabulary data including the modifier and the modifier code are recorded, an uncoded wound disease name written in a character string is coded. A method of encoding a disease name
The computer is
A component character extraction step of extracting a component character included in the uncoded wound name;
A vocabulary candidate extraction step of extracting vocabulary candidates recorded with the constituent characters as the first character from the vocabulary data storage means;
A candidate list generation step of generating a candidate list in which the lexical candidates are associated with the appearance positions of the constituent characters in the uncoded disease names;
Based on the appearance position and the number of characters of the vocabulary candidates, a combination generation step of generating a combination of vocabulary candidates that represents the entire uncoded disease name using the vocabulary candidates included in the candidate list,
And a coding step for coding an uncoded wound name based on the generated combination. The computer is
2. The method according to claim 1, further comprising: when a plurality of combinations are generated in the combination generation step, a step of assigning a priority to the generated combinations based on the number of characters of the name of the disease in each combination. The wound name coding method described in 1. Further using a unified rule data storage means that records dictionary filter data for unifying the notation,
The computer is
The method according to claim 1 or 2, further comprising the step of unifying the notation of the uncoded disease name using the dictionary filter data. The computer is
If a delimiter is detected in the uncoded wound name, further performing a step of dividing the uncoded wound name by the delimiter,
The wound name encoding method according to any one of claims 1 to 3, wherein encoding is performed for each of the divided character strings. The computer is
Further executing the step of converting the wound name and modifier code included in the received receipt data into a vocabulary using the vocabulary data storage means to generate an uncoded wound name;
5. The wound name encoding method according to any one of claims 1 to 4, wherein the generated uncoded wound name is encoded. Using a computer and a vocabulary data storage means in which vocabulary data including the wound name and wound name code and vocabulary data including the modifier and the modifier code are recorded, an uncoded wound disease name written in a character string is coded. A sick and disease name coding program for
The computer,
A component character extracting means for extracting a component character included in the uncoded wound name;
Vocabulary candidate extraction means for extracting from the vocabulary data storage means vocabulary candidates recorded with the constituent characters as leading characters;
Candidate list generation means for generating a candidate list that associates the appearance positions of the constituent characters in the uncoded disease names with the vocabulary candidates;
Based on the appearance position and the number of characters of the vocabulary candidates, combination generation means for generating a combination of vocabulary candidates that represents the entire uncoded wound name using the vocabulary candidates included in the candidate list;
A wound name coding program for functioning as a coding means for coding an uncoded wound name based on the generated combination. Said computer further
7. The apparatus according to claim 6, wherein when a plurality of combinations are generated by the combination generation unit, the combination generation unit functions as a unit that gives priority to the generated combinations based on the number of characters of the name of the disease in each combination. Described disease name encoding program. Further using a unified rule data storage means that records dictionary filter data for unifying the notation,
Said computer further
The wound / morbid name coding program according to claim 6 or 7, wherein the uncoded wound / morbid name is made to function as means for unifying the notation using the dictionary filter data. Said computer further
When a delimiter is detected in an uncoded wound name, function as a means for dividing the uncoded wound name by the delimiter,
The wound name / encoding program according to claim 6, wherein encoding is performed for each of the divided character strings. Said computer further
The wound name and modifier code included in the received receipt data is converted into a vocabulary using the vocabulary data storage means to function as a means for generating an uncoded wound name,
10. The wound name encoding program according to claim 6, wherein the generated uncoded wound name is encoded.

Images