JP2008234003A - Medicine information management program, medicine information management device and medicine information management method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a medicine information management device generating a retrieval expression for obtaining a retrieval result fit for a situation of a questioner based on side effect representation that is a representation example by a general person related to a side effect of medicine. <P>SOLUTION: This medicine information management device management information related to the side effect of the medicine has: a retrieval expression generation part 5 acquiring identification information of the medicine, acquiring side effect representation and a side effect kind corresponding to the acquired identification information from a side effect representation DB 1 wherein the identification information of the medicine, the side effect type that is a medical term showing the type of the side effect of the medicine, and the side effect representation that is representation of the side effect of the medicine are associated and stored, and generating the retrieval expression based on the acquired side effect representation and the acquired identification information; and a side effect representation extraction part extracting the side effect representation from a case of an answer to a question and the question related to the side effect of the medicine to generate the side effect representation DB 11. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a drug information management program, a drug information management apparatus, and a drug information management method for managing information on side effects of drugs.

  It is recommended to consult with the doctor or pharmacist who prescribed the drug if you suspect that the drug will cause side effects. Before consulting with the doctor or pharmacist, ask for advice on an electronic bulletin board on the Internet, Many patients search for related information. Here, when a patient writes questions on an electronic bulletin board or the like, there is a situation where it is difficult to accurately express his or her symptoms due to lack of specialized knowledge.

  In conventional electronic bulletin boards, respondents with a relatively high level of expertise are able to guess and respond to the patient's situation, or confirm the situation to the patient who wrote the questionnaire and answer. It has been broken. In this method, if the respondent has sufficient knowledge and skills, the respondent can present an answer that satisfies the patient who wrote the questionnaire.

As a technology related to the preparation of such answers, for example, a drug database storing information provided by a pharmaceutical package insert issued by a pharmaceutical company, and initial symptoms (drug prescriptions) collected during the process of taking medication Based on an initial symptom database storing symptom experienced by a patient who has been treated, a technique for assisting a pharmacist so that appropriate medication instruction including initial symptoms for side effect information can be performed is disclosed (for example, Patent Document 1). . In addition, side effect information and contraindication information described in the package insert (such as information on interaction between drugs that may cause side effects due to contraindications or administration of multiple drugs) Is disclosed in which an expert having expertise in pharmaceutical information creates a database and distributes it to end users (patients) via the Internet (for example, Patent Document 2). Electronic bulletin boards applying these technologies can support the process of creating answers for respondents with specialized knowledge.
JP 2004-290377 A JP 2002-132939 A

  However, the electronic bulletin board as described above is considered to be able to relatively reduce the fluctuation of the expression of the symptom to be described, but it will be described in an expression that is difficult for a general patient who lacks expertise. As a result, expressions that are difficult to come up with are stored as data on the electronic bulletin board. Therefore, when a general patient searches for past question and answer examples stored in the electronic bulletin board, the search keyword cannot be conceived, and as a result, the patient obtains a search result with low accuracy. In addition, for a document in which a patient with insufficient expertise describes a symptom, the expression describing the symptom greatly fluctuates, so even if one patient searches for a document written by another patient, the appropriate search keyword There is a problem that it is difficult to specify.

  Similarly, when searching for related information on the Internet, it is difficult for the patient himself to set an appropriate search condition, and thus there is a problem that a search result with sufficient quality cannot be obtained. For example, if you search for an appropriate keyword that identifies the name of the drug and the patient's symptoms, even if the appropriate information can be found, the patient cannot come up with the keyword that identifies the symptoms, so the appropriate information cannot be obtained. It is possible.

  The present invention has been made in order to solve the above-described problems. A drug information management program, a drug information management apparatus, and a drug information management for creating a search expression based on information accumulated systematically with examples of expressions related to side effects. It aims to provide a method.

  In order to solve the above-described problems, the present invention provides a pharmaceutical information management program for causing a computer to manage information related to side effects of a pharmaceutical, an identification information acquisition step of acquiring pharmaceutical identification information, and the identification information of a pharmaceutical The side effect type corresponding to the identification information acquired in the identification information acquisition step is acquired from first storage information that is stored in association with a side effect type that is a medical term indicating the type of side effect of the drug. The identification information from the second accumulation information accumulated in association with the side effect type acquisition step, at least one of the identification information of the medicine and the side effect classification of the medicine, and the side effect expression that is an expression of the side effect of the medicine The identification information acquired in the acquisition step and the small number of side effect types acquired in the side effect type acquisition step. The side effect expression acquisition step for acquiring the side effect expression corresponding to at least one, and the identification information acquired in the identification information acquisition step and the side effect expression for each side effect type acquired in the side effect type acquisition step A search expression generation step for generating a search expression based on the side effect expression acquired in the acquisition step is executed by a computer.

  Furthermore, the present invention provides an extraction step of extracting the side effect expression from a question answer example which is an example of a question about a drug side effect and an answer to the question in the drug information management program described above to generate the second accumulated information. It is what is executed by a computer.

  Further, in order to solve the above-described problem, the present invention is a drug information management device that manages information related to side effects of a drug, an identification information acquisition unit that acquires drug identification information, and the identification information of a drug, The side effect type corresponding to the identification information acquired by the identification information acquisition unit is acquired from first storage information that is stored in association with a side effect type that is a medical term indicating the type of side effect of the drug. The identification information acquisition from the second accumulation information in which the side effect type acquisition unit, at least one of the identification information of the medicine and the side effect classification of the medicine, and the side effect expression that is an expression of the side effect of the medicine are associated and stored The side effect expression corresponding to at least one of the identification information acquired by the unit and the side effect type acquired by the side effect type acquiring unit is acquired. For each side effect type acquired by the action expression acquisition unit and the side effect type acquisition unit, a search is performed based on the identification information acquired by the identification information acquisition unit and the side effect expression acquired by the side effect expression acquisition unit. And a search expression generation unit that generates an expression.

  Furthermore, the present invention provides an extraction unit for extracting the side effect expression from the question answer example which is an example of a question about a drug side effect and an answer to the question in the drug information management device described above to generate the second accumulated information. It is characterized by comprising.

  Further, in order to solve the above-described problem, the present invention provides a drug information management method for managing information on side effects of a drug, an identification information acquisition step for acquiring drug identification information, and the drug identification information The side effect type corresponding to the identification information acquired in the identification information acquisition step is acquired from first storage information that is stored in association with a side effect type that is a medical term indicating the type of side effect of the drug. The identification information from the second accumulation information accumulated in association with the side effect type acquisition step, at least one of the identification information of the medicine and the side effect classification of the medicine, and the side effect expression that is an expression of the side effect of the medicine At least one of the identification information acquired in the acquisition step and the side effect type acquired in the side effect type acquisition step For each side effect type acquired in the side effect expression acquisition step for acquiring the corresponding side effect expression and in the side effect type acquisition step, the identification information acquired in the identification information acquisition step and acquired in the side effect expression acquisition step And a search expression generation step of generating a search expression based on the side effect expression.

Furthermore, the present invention provides the aforementioned pharmaceutical information management method,
The extraction step of extracting the side effect expression from the question answer example which is an example of the question about the side effect of the medicine and the answer to the question is executed to generate the second accumulated information.

  According to the present invention, it is possible to create a search expression based on a side effect expression that is an example of an expression related to a side effect of a medicine.

  Embodiments of the present invention will be described below with reference to the drawings.

  The functional block of the pharmaceutical information management device 1 according to the present embodiment is shown in FIG. The pharmaceutical information management apparatus 1 includes a question answer process control unit 2, a side effect expression extraction unit 3, a related case search unit 4, a search formula creation unit 5, a question answer example DB 10, and a side effect expression DB 11.

  The question answer example DB 10 includes the question items from the questioner (patient), the answers by the respondents (hereinafter, a combination of the question items and the answer is a question answer example), and the side effect type related to the question items. The database to store. The side effect type is a medical term indicating the type of side effect of the drug as a symptom name, and typically, it is not commonly used, such as rhabdomyolysis, liver disorder, etc. It is a term that is widely known and used by experts with expertise in medicine.

  The side effect expression DB 11 is a side effect expression (for example, “muscle hurts”, “limb and numbness”, etc.) that is a drug name (identification information) and a linguistic expression related to the side effect (expression expressing a side effect symptom by a general person) Is stored in association with each drug name and each side effect type. The side effect expression DB 11 stores drug names and side effect types as initial settings in advance.

  The question answer process control unit 2 controls a process in which the answerer answers the question items from the questioner, and stores the question answer example in the question answer example DB 10.

  The side effect expression extraction unit 3 extracts a side effect expression from the question answer examples stored in the question answer example DB 10 and stores the side effect expression in the side effect expression DB 11.

  The related case search unit 4 searches the question answer case DB 10 for a question answer case regarding the medicine name designated by the questioner.

  The search formula creation unit 5 creates a full-text search formula used by the related case search unit 4 for a search for a question answer case based on the side effect expression stored in the side effect expression DB 11.

  The pharmaceutical information management apparatus 1 in the present embodiment is a computer, and the above functions are controlled and calculated by using a CPU and a memory of the computer. Various data stored in the question answer example DB 10 and the side effect expression DB 11 are held in a storage device such as an HDD of a computer.

  Next, the process of the pharmaceutical information management device 1 in the present embodiment will be described with reference to the drawings. The process of the pharmaceutical information management apparatus 1 is divided into three processes: a question answer case accumulation process, a side effect expression expansion process, and a related case search process. Hereinafter, these three processes will be described.

  First, question answer case accumulation processing will be described. The question answer process control unit 2 executes the question answer case accumulation process by accepting a question about the side effect of the medicine from the questioner. FIG. 2 shows a schematic diagram of the question answer case accumulation process. The question / answer case accumulation process is processed by the question / answer process control unit 2 inputting / outputting various data to / from the question / answer case DB 10 and the side effect expression DB 11 and acquiring question items and answers from the questioner and the answerer. The Further, in this overview diagram, in order to deepen understanding, main processing is described in association with a processing flow described later.

  Further, information (side effect document) described in a document attached to a commonly distributed medicine is accumulated in the side effect expression DB 11 as initial setting data.

  As shown in FIG. 3, the side effect document includes various information such as component name, product name, alias, efficacy, and serious side effect. The component name and product name representing the drug name in the side effect document are accumulated in the side effect expression DB 11 as the value of the drug name table shown in FIG. The drug name table is formed with items of drug ID, drug name, and name type indicating the type of drug name, which are drug management codes, and is held as initial setting data. In addition, when there are a plurality of drug names representing the same drug (a drug whose representation is different with a representative name, component name, alias, etc. as shown in FIG. 4A), the same drug ID is given.

  In addition, the side effect expression DB 11 includes a side effect expression table (FIG. 4B) having medicine ID, side effect type, state expression, related part, and number of cases as items. The drug name and the side effect type are associated with each other through the drug ID, which is a common item in the side effect expression table and the above drug name table. In FIG. 4B, values of state expression, related part, and number of cases are set, but in the initial state, only drug ID and side effect type values are set as initial setting data. The values of the expression, the related part, and the number of cases are accumulated by processing of the side effect expression extraction unit 3 described later. The description of the state expression, related parts, and number of cases will be described later.

  Next, the process flow of the question answer case accumulation process will be described. As shown in FIG. 5, the question answering process control unit 2 acquires the question title and the question items from the questioner (Step S101 in FIGS. 2 and 5), and the acquired title and question items are stored in the question answering example DB 10. By writing (step S102 in FIG. 5), the title and the question items are accumulated in the question answer example DB 10.

  The question answer process control unit 2 shifts the process to the process flow shown in FIG. 6 by accepting an answer request from an answerer. The question answer process control unit 2 obtains an answer request from the respondent (step S110), and a question item that has not been answered from the question answer example DB 10 (a question item in a question answer content table described later in the question answer example DB 10) Is obtained but the answer has no value) (step S111).

  The question answering process control unit 2 further performs morphological analysis on each sentence in the acquired question items, and extracts nouns and compound nouns as drug name candidates (step S112). An example of the drug name candidate extraction process is shown in FIG. The question answering process control unit 2 performs morpheme analysis to decompose the extraction target sentence into morpheme units, and extracts those parts of speech with nouns or compound nouns as drug name candidates. For example, when the extraction target sentence is “Is it a side effect of mebatiron?”, It is decomposed into parts of speech as shown in the morphological analysis result of FIG. 7 (“/” in the figure is a morpheme break, “//” is a phrase Separator.) Among the parts of speech that have been decomposed in this way, the nouns “mevalotin” and “side effects” are drug name candidates.

  The question answering process control unit 2 searches the side effect expression DB 11 with the drug name candidates extracted as described above, and the relationship including the representative drug name (the drug name whose name type is the representative name in the drug name table) and the side effect type Data is acquired (step S113). The question answering process control unit 2 extracts the representative drug name and the side effect type corresponding to the drug only when a matching drug name is found as a result of the search (therefore, in the example of the drug name candidate in FIG. Only “mevalotin” is taken out, “side effects” are not taken out). Next, the question answering process control unit 2 groups the side effect types for each representative drug name and creates a side effect description form (step S114), and presents the representative drug name and the side effect type as candidates to the respondent ( Step S115).

  An example of the side effect description form is shown in FIG. The side effect description form shows the title and question from the questioner in the “question” section. In the first line of the “Related side effects” part of the side effect description form, the part indicated as “Mevalotin” is the representative drug name presented as a candidate drug name related to the question by the Question Answer Process Control Unit 2. The previous check box is a field for entering the result of the respondent confirming whether the candidate is correct. FIG. 8 shows a state after the respondent confirms that it is appropriate (marking a check box).

  The table below the location where the representative drug name (indicated as “mevalotin” in the related side effects in FIG. 8) is shown is a side effect type corresponding to the drug “mevalotin” and a side effect type related to the questionnaire. And a field for the respondent to enter the severity of a symptom caused by a side effect (the “related” item in FIG. 8). In FIG. 8, the respondent determines that rhabdomyolysis is related as a side effect type of the questionnaire, and the severity determined from the description content of the questionnaire is 2 (the higher the value, the more serious An example of an evaluation result assuming an evaluation) is shown by the respondent. In addition, the “answer column” in FIG. 8 is a column for the respondent to write an answer to the question item.

  When the respondent completes the input for each item in the side effect description form and the “Confirm” button in the side effect description form is pressed, the question answer process control unit 2 is entered in the answer column from the respondent The answer, the medicine ID associated with the representative drug name selected by the respondent, the side effect type and severity selected by the respondent are written in the question answer example DB 10 together with the title and the question (step S116 in FIG. 6). .

  Here, an example of data stored in the question answer case DB 10 is shown in FIG. The question answer example DB 10 includes a question answer content table (FIG. 9 (a)) in which a case ID, a title, a question item, and an answer item, which are identifiers of question answer cases, are recorded as one record, a case ID, a medicine ID, and a side effect. Each data is managed in the related side effect information table (FIG. 9B) in which the type and severity items are one record. The case ID is automatically given at the timing (Step S101 in FIG. 5) when the question item is acquired from the questioner. The case ID is a common item in each table, and the question answer content table and the related side effect information table are associated by this case ID.

  In response to the browsing request from the questioner, the question answering process control unit 2 acquires the question items from the questioner and the answers from the answerer from the question answer example DB 10 and presents them (step S117 in FIG. 6).

  The pharmaceutical information management device 1 expands the question answer example DB 10 by storing the question items from the questioner and the answers from the answerer by the question answer example accumulation process as described above.

  Next, the side effect expression expansion process will be described. The side effect expression extraction unit 3 executes the side effect expression expansion process when a certain number of question answer examples are accumulated in the question answer example DB 10. FIG. 10 shows a schematic diagram of the side effect expression expansion process. The side effect expression expansion process is a process in which the side effect expression extracting unit 3 expands the side effect expression by extracting the side effect expression from the question answer examples accumulated in the question answer example DB 10 and writing it in the side effect expression DB 11 as described above. It is. Also in this overview diagram, in order to deepen the understanding, the main processing is described in association with the processing flow described later.

  The side effect expression expansion process may be performed sequentially when the question answer example accumulation process is completed, or may be performed at a regular time.

  Next, details of the processing flow of the side effect expression expansion processing will be described with reference to the drawings. FIG. 11 shows a flow of side effect expression extraction processing by the side effect expression extracting unit 3.

  The side effect expression extraction unit 3 acquires a question answer case, a medicine ID corresponding to the question answer case, and a side effect type (hereinafter collectively referred to as case related data) from the question answer case DB 10 (step S120). The side effect expression extraction unit 3 performs morpheme analysis on the acquired question related to the case-related data to decompose each sentence of the question into morphemes (parts of speech) (step S121). The side effect expression extraction unit 3 extracts a side effect expression based on the decomposed morpheme (step S122).

  In the present embodiment, the side effect expression refers to the body part (“head”, “hand”, etc.) in the sentence of the questionnaire, and physiological substances (“sweat”, “urine”, “ Nouns that represent names such as “tears” (hereinafter referred to as related parts), state expressions that express the state of related parts (adjective “Kayu”, state verbs “Zukizuki”, and their derivatives As a combination with the noun “pain”, etc.). In the case where only the state expression or only the related part is shown in the question item depending on the content of the sentence, the side effect expression is processed as only the related part or only the state expression.

  Here, an example of a method for extracting a side effect expression by the side effect expression extracting unit 3 is shown in FIG. 12 shows two question items accumulated in the question answer example DB 10 (FIG. 12A), a morphological analysis result for the question item (FIG. 12B), and a side effect expression extraction result (FIG. 12C )). The side effect expression extraction unit 3 processes the result into the following format based on the morpheme break, part of speech, and utilization information output by a general morpheme processing technique, and outputs a morpheme analysis result.

Each part separated by a slash (/) is a morpheme. For verbs, adjectives, and auxiliary verbs, the vocabulary includes one morpheme. The clauses are delimited by (//).
-For each morpheme, output in the form of “part of speech: supplementary information” within the notation <>.

  Here, supplementary information on the above morphological analysis results will be described with reference to FIG. Part-of-speech such as verbs and adjectives have different notations due to differences in usage (for example, in the case of verb “yellowing” (final form), “yellowing” (blank form), “yellowing” (continuous form), Etc.), a usage form to be expressed in units of parts of speech is set in advance, and a notation converted into the set usage form is shown as supplementary information.

  When the part of speech is a noun, the side effect expression extraction unit 3 does not output supplementary information in the morphological analysis result. When the part of speech is a verb, the side effect expression extraction unit 3 outputs, as supplementary information, a notation obtained by converting the target verb utilization form into a continuous form. Similarly, the side effect expression extraction unit 3 outputs supplementary information as an adjunctive form in the case of an adjective and as a word expression as it is in the case of a particle (however, the accusative form represents the original particle). Further, the side effect expression extraction unit 3 outputs auxiliary information in the case of an auxiliary verb or auxiliary word by converting it to a final form, and does not output additional information in the case of parts of speech other than those described above.

  The side effect expression extracting unit 3 extracts the state expression and the related part as shown in FIG. 12C based on the result of the morphological analysis as described above. The extraction method by the side effect expression extraction unit 3 is as follows. Regarding the extraction of the related part, whether the target noun is included in the related part dictionary data prepared in advance determines whether the relevant part is a related part. Regarding the extraction of the state expression, when the target verb is a state verb, it is determined whether or not the state expression is made using state verb dictionary data prepared in advance. Furthermore, in the case of state expression, it is extracted as the notation in the supplementary information described above. Note that the side effect expression extraction unit 3 does not treat the state expression as an adjective and does not treat the state expression related to the side effect expression in a sentence in which no related part appears.

  Returning to the processing flow of FIG. 11, the side effect expression extraction unit 3 then tabulates the appearance status of the side effect expression for each medicine ID and side effect type (step S123). The number of cases in FIG. 12C is a total result of the appearance status of the side effect expression described above. For example, in the case where the drug ID is “mev1” (representative drug name is “mevalotin”), the side effect type is “rhabdomyolysis”, the state expression is “numbness”, and the related part is “calf”. When there are two cases, the number of cases is “2”.

  Next, the side effect expression extraction unit 3 associates the extracted side effect expression with the drug ID and the side effect type, and updates the side effect expression table of the side effect expression DB 11 (step S124). The side effect expression extraction unit 3 also stores the number of cases described above in the side effect expression table of the side effect expression DB 11.

  Here, the correspondence of various data in the side effect expression DB 11 is shown in FIG. The correspondence relationship between the drug ID, the side effect type, the side effect expression (state expression and related part), and the number of cases in the side effect expression DB 11 is as illustrated. That is, one or a plurality of side effect types correspond to the medicine ID, and one or a plurality of side effect expressions (state expressions and related parts) and the number of cases correspond to the side effect types.

  A record as shown in FIG. 14B is added to the side effect expression table of the side effect expression DB 11. When a record having the same drug ID, side effect type, state expression, and related part already exists in the side effect expression table of the side effect expression DB 11, only the number of cases in the existing record is updated (added). Become.

  This side effect expression expansion process enables the collection of side effect expressions by questions from the general public in units of drug name and side effect type without placing a heavy burden on respondents, greatly reducing the work efficiency of respondents. Can do.

  Next, related case search processing will be described. By receiving a request for information on drug side effects from the questioner, the related case search unit 4 executes related case search processing. FIG. 15 shows a schematic diagram of the related case search process. In the related case search process, the related case search unit 4 and the search expression creation unit 5 use each data of the question answer case DB 10 and the side effect expression DB 11 accumulated as described above in response to the drug name input from the questioner. This is a process of providing the questioner with an outline of the question answer example as information on the side effect of the drug name. Also in this overview diagram, in order to deepen the understanding, the main processing is described in association with the processing flow described later.

  Next, the details of the processing flow of the related case search process will be described with reference to the drawings.

  First, the processing of the related case search unit 4 is shown in FIG. The related case search unit 4 receives the input of the drug name from the questioner, and acquires the drug name (representative name or ingredient name or alias name of the drug) (step S130 in FIGS. 15 and 16). The related case search unit 4 outputs the acquired drug name to the search formula creation unit 5 and creates a search formula (created based on the drug name and the side effect expression for each side effect type. Details of the process will be described later) (step S131). ).

  Based on the search formula created by the search formula creation unit 5, the related case search unit 4 further asks a question answer that matches the search condition of the search formula from the question answer content table (FIG. 9A) in the question answer example DB 10. Cases are searched (step S132), and summary information including the appearance location of the side effect expression is created from the question items of each question answer case data and the sentence of the answer (step S133). Details of the method for creating summary information will be described later.

  After that, the related case search unit 4 groups the summary information by the side effect type (obtained by the processing of the search formula creation unit 5 described later), together with the representative drug name (acquired by the processing of the search formula creation unit 5 described later). A list of summary information is displayed to the user (step S134).

  Next, the processing flow of the search expression creation unit 5 will be described with reference to FIG. The search formula creation unit 5 acquires a drug name (a representative name or a component name or an alias name of a drug) from the related case search unit 4 (step S140). Next, the search formula creation unit 5 acquires the drug ID corresponding to the drug name acquired from the related case search unit 4 from the side effect expression DB 11, and all drug names (representative name, component name) corresponding to the acquired drug ID. , All drug names of aliases), side effect type and side effect expression (state expression, related part) are acquired (step S141). Finally, the search formula creation unit 5 creates a search formula that combines all drug names and side effect expressions for each acquired side effect type, and outputs the created search formula, representative drug name, and side effect type to the related case search unit 4 (Step S142).

  Here, the search formula created by the search formula creation unit 5 will be described. The search expression consists of “search conditions related to drug names” specified by OR combination (logical sum) of one or more drug names (representative names, ingredient names, etc.) and one or more side effect expressions (the side effect expression is a state expression). And the related part, the AND of the search condition related to the side effect specified by the OR combination of the logical AND of the state expression and the related part is used. ).

  An example of creating a search formula when the questioner inputs “Mebatorte” as the drug name will be described with reference to FIG. FIG. 18A shows an example of a record of the side effect expression table of the side effect expression DB 11 which is the original data for creating the search expression, and FIG. 18B shows an example of the search expression creation result. In addition, it is assumed that the original medicine name is acquired from the medicine name table shown in FIG.

  The search formula creation unit 5 creates a search formula for each side effect type of the target drug in the following procedure. First, the search formula creation unit 5 creates “search conditions related to drug names” by ORing all drug names shown in FIG. Next, the search expression creation unit 5 creates “search conditions related to side effects” according to the processing flow of FIG. 19 using each side effect expression record of FIG. 18A as original data.

  Referring to FIG. 19, the search expression creation unit 5 first OR-expands the state expression based on a “word form development pattern” (described later) (step S150). If there is no state expression, step S150 is skipped. Next, when there is a related part for the state expression (step S151, Yes), the search expression creating unit 5 ANDs the search condition of the state expression obtained as described above and the related part to obtain “for each side effect expression. A “search condition” is created (step S152). When there is no related part (step S151, No), the search formula creation unit 5 ends the process as it is. The search expression creation unit 5 performs the processing from step S150 to step S152 on the target side effect expression (combination of the state expression and the related part). Thereafter, the search expression creation unit 5 creates a “search condition related to side effects” by ORing the created “search conditions for each side effect expression” (step S153).

  FIG. 20 shows a “word form development pattern” of the state expression. When the state expression is a noun or an undefined part of speech (for example, “muscle pain” or the like), the search expression creating unit 5 processes the state expression as it is without performing OR expansion. When the state expression is a verb and is used in the upper one stage or the lower one stage, the search expression creation unit 5 processes the state expression as a continuous form without performing OR expansion. When the state expression is a verb having a utilization form other than those described above, the search expression creation unit 5 expands as a “formal form OR combined use / derived noun form OR combined use form OR assumed form OR termination form”. If the state expression is an adjective, the search expression creation unit 5 expands as “unformed form OR continuous use OR end form OR stem +“ sa ”OR stem +“ mi ””.

  The search expression creation unit 5 creates a search expression by AND-connecting the “search condition related to drug name” and the “search condition related to side effect” obtained as described above for each side effect type (see FIG. 18B). ).

  Next, the related case search unit 4 displays search results for the user (steps S133 and S134), and the process of creating a list of search results will be described with reference to the flowchart of FIG. The purpose of this process is that a search expression set Q consisting of N types of search expressions corresponding to a certain side effect type and a search result set D consisting of M documents obtained by the search by Q are used as D summary information D Is a step of extracting a representative phrase (summary information p_i) including a side effect expression from each element d_i of D (S161 to S163). And summarizing the summary information p_i to create a search result list.

  The related case search unit 4 acquires a search result set D corresponding to the side effect type (a set of question answer cases searched in step S132) and a set Q of search formulas created by the search formula creation unit 5. (Step S160). The search result set D and the search expression set Q are expressed as follows.

D = {d_i: i∈ [1. . M]}
Q = {q_j: j∈ [1. . N]}

  Next, the related case search unit 4 compares i and M. If i is equal to or less than M (Yes in step S161), the document (question answer case) d_i is checked against the search expression set Q, and the summary information p_i Is created (step S162). Thereafter, the related case search unit 4 increments i by 1, and returns to the determination process of step S161 (step S163).

  The related case search unit 4 performs the above-described processing (step S161 to step S163) until i becomes larger than M. When i becomes larger than M (step S161, No), the related case search unit 4 creates a search result list in which the documents d_i are aggregated with the value of the summary information p_i (step S164).

  Next, FIG. 22 shows details of the process of creating the summary information p_i (step S162) in FIG. This process creates summary information p_i that extracts the appearance of side effect expressions for a given document d_i. The summary information p_i is one of the search expressions q_j (j∈ [1... N]). If q_j includes both a search condition related to a related part and a search condition related to a state expression, it is composed of two phrases (a phrase p1 corresponding to the related part and p2 corresponding to the state expression). , Q_j includes only a search condition related to the state expression, it is composed of one clause (the clause p2 corresponding to the state expression).

  The related case retrieval unit 4 acquires the document d_i and the retrieval formula set Q as input values for processing (step S170), and converts the document d_i into a sentence (a character string ending with a punctuation mark) s_k (k∈ [1... K]). Divide (step S171).

  Next, the related case search unit 4 substitutes the initial value 1 for k (step S172), and clears the summary information second candidate (q2) (step S173). The related case search unit 4 compares k and K. If k is equal to or less than K (step S174, Yes), the sentence s_k is morphologically analyzed and divided into phrase units (step S175).

  Next, the related case search unit 4 substitutes 1 for j (step S176), and compares j with N. When j is N or less (step S177, Yes), the related case search unit 4 selects the phrase p1 that matches the related part of q_j (the related part included in the above search formula) (step S178). Furthermore, the related case search unit 4 selects a clause p2 that matches any state expression of q_j (the state expression included in the above search expression) (step S179).

  The related case search unit 4 determines the above phrase selection result (results of step S178 and step S179) (step S180), and when it is determined that both p1 and p2 are valid, the summary information p_i is determined from p1 and p2. Is created (step S181), and the summary information creation process is terminated.

  In addition, when the related case search unit 4 determines that only p2 is valid and q2 is not valid in the determination of the phrase selection result (step S180), the related case search unit 4 creates the second candidate q2 of the summary information from p2 (step S182), j is incremented by 1 (step S183), and the process returns to the determination of step S177.

  Further, in the determination of the phrase selection result (step S180), the related case search unit 4 increments j by 1 (step S183), and processes the determination in step S177 when it is determined that it is other than the above-described phrase selection result. return.

  If the related case search unit 4 determines that j is larger than N in step S177 (No in step S177), k is incremented by 1 (step S184), and the process returns to the determination in step S174.

  If the related case search unit 4 determines that k is larger than K in step S174 (No in step S174), q2 is substituted for p_i (step S185), and the outline information creation process is terminated.

  For creating summary information (step S181, step S185), for example, all the sentences including p1 and p2 (or q2) are extracted and used as summary information or up to 50 characters before and after the clauses of p1 and p2 (or q2). However, it is created by setting a standard such as summary information.

  Next, FIG. 23 shows an example of a search screen that the related case search unit 4 presents to the questioner. When the questioner inputs the name of the drug for which he / she wants to know the side effect information in the drug name input field at the top, and the “Search” button is pressed, the related case search unit 4 executes the above-described processing, and the representative drug name A list of summary information is displayed for each side effect type together with (“pharmaceutical name” shown in the second row in FIG. 23).

  Two rows under the row of “rhabdomyolysis” as a search result are a list of summary information of question answer examples corresponding to the side effect type of rhabdomyolysis. For example, the line “… like calf numb ...” indicates that the parts of “calf” and “numbness” highlighted with large letters match the related part and the state expression. The number of cases displayed at the end of the line (such as “2 cases”) indicates the number of related cases corresponding to the summary information. When the questioner specifies the number of cases using a pointing device, the related case search unit 4 , Call a screen that displays the full text of all corresponding question answer cases. In addition, the related case search part 4 may call the screen which displays the corresponding seriousness degree with the whole sentence of a question answer example.

  In the related case search process, the question answer case DB 10 is a search target. However, since the creation of the search formula and the related document search can be performed independently, the database of other systems and the like are set as the search target in the same procedure. Is possible. For example, it is possible to search a full text search DB that collects web pages on the Internet.

  By the related case search process, the questioner can search the question answer examples related to the medicine comprehensively and accurately only by specifying the name of the medicine. Further, since the summary information is displayed from the set of search results, the questioner can easily select information suitable for his / her symptoms using easy-to-understand side effect expressions.

  According to the present embodiment, it is possible to automatically extract side effect expressions or the like included by the general public included in the questionnaire, organize them for different types of drugs and side effects, and register them in the side effect expression DB 11. It is possible to systematically accumulate examples of side effects expressed by ordinary people without greatly increasing.

  Further, according to the present embodiment, when searching for a question answer example, a symptom name (side effect type) that is difficult for a questioner (patient) to specify can be supplemented with a side effect expression that is familiar to the general public. The questioner can obtain a search result with less noise narrowed down by the supplemented side effect expression only by giving a keyword that is easy to specify, which is a drug name.

  Also, according to the present embodiment, a search expression corresponding to the side effect type can be created, so that the searched question answer cases can be classified for each side effect type and presented to the user. Furthermore, it is possible to create summary information by extracting the neighborhood of the appearance of the side effect expression included in the search expression corresponding to the side effect type for the retrieved question answer example, and show it to the user. Makes it easy to understand and select the characteristics of search results (such as what symptoms are related).

  The identification information acquisition unit, the side effect type acquisition unit, the side effect expression acquisition unit, and the search expression generation unit correspond to the search expression generation unit in the embodiment. The search unit and the search result display unit correspond to the related case search unit in the embodiment. The question answer example storage unit corresponds to the question answer process control unit in the embodiment. The extraction unit corresponds to the side effect expression extraction unit in the embodiment.

  Further, the first accumulation information and the second accumulation information are stored in the side effect expression DB 11, and the third accumulation information is stored in the question answer example DB 10. In the first accumulated information, the drug name (identification information) and the side effect type are associated with each other via the drug ID. Specifically, the drug name and the side effect type are associated with each other from the correspondence between the drug ID and the drug name in FIG. 4A and the correspondence between the drug ID and the side effect type in FIG. Furthermore, regarding the second accumulated information, as shown in FIG. 4B, the side effect type and the state expression are associated with each other, and the side effect type and the drug name are associated with each other through the drug ID as described above. Therefore, it can be said that the state expression is associated with the drug name.

  Furthermore, the program which performs each step mentioned above in the computer which comprises a pharmaceutical information management apparatus can be provided as a pharmaceutical information management program. By storing the above-described program in a computer-readable recording medium, the computer constituting the pharmaceutical information management apparatus can be executed. Here, examples of the recording medium readable by the computer include an internal storage device such as a ROM and a RAM, a portable storage such as a CD-ROM, a flexible disk, a DVD disk, a magneto-optical disk, and an IC card. It includes a medium, a database holding a computer program, another computer and its database, and a transmission medium on a line.

(Supplementary note 1) A pharmaceutical information management program for causing a computer to manage information on side effects of pharmaceuticals,
An identification information acquisition step of acquiring identification information of the medicine;
Corresponds to the identification information acquired in the identification information acquisition step from the first accumulation information accumulated by associating the identification information of the drug with the side effect type that is a medical term indicating the type of side effect of the drug A side effect type acquisition step of acquiring the side effect type;
Acquired in the identification information acquisition step from the second accumulated information accumulated by associating at least one of the identification information of the drug and the side effect type of the drug with the side effect expression that is an expression of the side effect of the drug A side effect expression acquiring step for acquiring the side effect expression corresponding to at least one of identification information and the side effect type acquired in the side effect type acquiring step;
A search expression that generates a search expression for each side effect type acquired in the side effect type acquisition step based on the identification information acquired in the identification information acquisition step and the side effect expression acquired in the side effect expression acquisition step Generation step;
Drug information management program that allows computers to execute.
(Appendix 2) In the drug information management program described in Appendix 1,
A search step for searching for a question answer example that matches the search formula generated in the search formula generation step from a third storage information in which a question about drug side effects and a question answer case that is an example of an answer to the question are stored When,
A predetermined range including the side effect expression is extracted from the question answer examples searched in the search step, and at least one question answer example in which the contents of the predetermined range are the same is attached with the contents of the predetermined range. A search result display step to display,
A pharmaceutical information management program characterized by causing a computer to execute.
(Appendix 3) In the drug information management program described in Appendix 2,
Further, a question about the side effect of the drug is obtained, and based on the question, at least one of identification information candidate and side effect type candidate is extracted and displayed from the first accumulated information, an answer to the question is obtained, and the question is obtained. A pharmaceutical information management program that causes a computer to execute a question answer case accumulation step of adding an answer to the question as the question answer example to the third accumulation information.
(Appendix 4) In the drug information management program described in Appendix 1,
A pharmaceutical information management program which causes a computer to execute an extraction step of extracting the side effect expression from a question answer example which is an example of a question about a drug side effect and an answer to the question, and generating the second accumulated information.
(Appendix 5) In the pharmaceutical information management program described in Appendix 1,
The side effect expression is at least one of a related part that is either a body part or a physiological substance excreted outside the body, and a state expression that is an expression of the state of the related part or the body state. A featured drug information management program.
(Appendix 6) In the drug information management program described in Appendix 4,
The extraction step extracts the side effect expression by performing morphological analysis on the question answer case.
(Appendix 7) In the drug information management program described in Appendix 5,
In the search expression generation step, when the state expression is a predetermined part of speech to be utilized, the state expression is converted into at least one utilization form, and the search expression is calculated based on the side effect expression including the converted state expression. A pharmaceutical information management program characterized by generating.
(Appendix 8) In the drug information management program described in Appendix 7,
The pharmaceutical information management program characterized in that the retrieval formula generation step generates a retrieval formula by expanding the state expression after the conversion by logical sum.
(Appendix 9) In the pharmaceutical information management program described in Appendix 5,
The pharmaceutical information management program characterized in that the retrieval formula generation step generates a retrieval formula by combining the identification information and the side effect expression by logical product.
(Supplementary note 10) A drug information management apparatus for managing information on side effects of drugs,
An identification information acquisition unit for acquiring identification information of a medicine;
Corresponds to the identification information acquired by the identification information acquisition unit from the first accumulation information accumulated by associating the identification information of the drug with the side effect type that is a medical term indicating the type of side effect of the drug A side effect type acquisition unit for acquiring the side effect type;
Acquired by the identification information acquisition unit from the second accumulation information accumulated by associating at least one of the identification information of the drug and the side effect type of the drug with a side effect expression that is an expression of the side effect of the drug A side effect expression acquiring unit that acquires the side effect expression corresponding to at least one of identification information and the side effect type acquired by the side effect type acquiring unit;
For each side effect type acquired by the side effect type acquisition unit, a search formula that generates a search formula based on the identification information acquired by the identification information acquisition unit and the side effect expression acquired by the side effect expression acquisition unit A generator,
Drug information management device that makes a computer execute.
(Supplementary note 11) In the pharmaceutical information management apparatus according to supplementary note 10,
A search unit for searching for a question answer example that matches the search formula generated by the search formula generation unit from a third storage information in which a question about drug side effects and a question answer example that is an example of an answer to the question are stored When,
A predetermined range including the side effect expression is extracted from the question answer examples searched by the search unit, and at least one question answer example having the same contents of the predetermined range is attached with the contents of the predetermined range. A search result display section to be displayed,
A pharmaceutical information management apparatus comprising:
(Supplementary note 12) In the pharmaceutical information management apparatus according to supplementary note 11,
Further, a question about the side effect of the drug is obtained, and based on the question, at least one of identification information candidate and side effect type candidate is extracted and displayed from the first accumulated information, an answer to the question is obtained, and the question is obtained. And a drug answer management unit for adding an answer to the question as the question answer example to the third stored information.
(Supplementary note 13) In the pharmaceutical information management apparatus according to supplementary note 10,
A pharmaceutical information management apparatus comprising: an extraction unit that extracts the side effect expression from a question about a side effect of a medicine and a question answer example that is an example of an answer to the question, and generates the second accumulated information.
(Supplementary note 14) In the pharmaceutical information management apparatus according to supplementary note 10,
The side effect expression is at least one of a related part that is either a body part or a physiological substance excreted outside the body, and a state expression that is an expression of the state of the related part or the body state. A pharmaceutical information management device.
(Supplementary note 15) In the pharmaceutical information management apparatus according to supplementary note 13,
The said extraction part extracts the said side effect expression by performing a morphological analysis with respect to the said question answer example, The pharmaceutical information management apparatus characterized by the above-mentioned.
(Supplementary Note 16) A drug information management method for managing information on side effects of drugs,
An identification information acquisition step of acquiring identification information of the medicine;
Corresponds to the identification information acquired in the identification information acquisition step from the first accumulation information accumulated by associating the identification information of the drug with the side effect type that is a medical term indicating the type of side effect of the drug A side effect type acquisition step of acquiring the side effect type;
Acquired in the identification information acquisition step from the second accumulated information accumulated by associating at least one of the identification information of the drug and the side effect type of the drug with the side effect expression that is an expression of the side effect of the drug A side effect expression acquiring step for acquiring the side effect expression corresponding to at least one of identification information and the side effect type acquired in the side effect type acquiring step;
A search expression that generates a search expression for each side effect type acquired in the side effect type acquisition step based on the identification information acquired in the identification information acquisition step and the side effect expression acquired in the side effect expression acquisition step Generation step;
Pharmaceutical information management method to execute.
(Supplementary note 17) In the pharmaceutical information management method according to supplementary note 16,
A search step for searching for a question answer example that matches the search formula generated in the search formula generation step from a third storage information in which a question about drug side effects and a question answer case that is an example of an answer to the question are stored When,
A predetermined range including the side effect expression is extracted from the question answer examples searched in the search step, and at least one question answer example in which the contents of the predetermined range are the same is attached with the contents of the predetermined range. A search result display step to display,
Pharmaceutical information management method characterized by performing
(Supplementary note 18) In the pharmaceutical information management method according to supplementary note 17,
Further, a question about the side effect of the drug is obtained, and based on the question, at least one of identification information candidate and side effect type candidate is extracted and displayed from the first accumulated information, an answer to the question is obtained, and the question is obtained. And a drug answer management method for executing a question answer case accumulation step of adding an answer to the question as the question answer example to the third accumulated information.
(Supplementary note 19) In the pharmaceutical information management method according to supplementary note 16,
A pharmaceutical information management method, comprising: performing an extraction step of extracting the side effect expression from a question about a side effect of a pharmaceutical and a question answer example which is an example of an answer to the question to generate the second accumulated information.
(Supplementary note 20) In the pharmaceutical information management method according to supplementary note 16,
The side effect expression is at least one of a related part that is either a body part or a physiological substance excreted outside the body, and a state expression that is an expression of the state of the related part or the body state. Pharmaceutical information management method characterized.

It is a block diagram which shows an example of a structure of the pharmaceutical information management apparatus which concerns on embodiment of this invention. It is an outline figure of "question answer example accumulation processing" concerning an embodiment of the invention. It is a figure which shows an example of the side effect document which concerns on embodiment of this invention. It is a figure which shows the information accumulate | stored in the side effect expression DB which concerns on embodiment of this invention. It is a flowchart which shows the processing flow (question matter acquisition) of the "question answer example accumulation | storage process" which concerns on embodiment of this invention. It is a flowchart which shows the processing flow (accumulation of a question matter and an answer) of the "question answer example accumulation | storage process" which concerns on embodiment of this invention. It is a figure which shows an example of the extraction process of the pharmaceutical name candidate which concerns on embodiment of this invention. It is a figure which shows an example of the side effect description form which concerns on embodiment of this invention. It is a figure which shows an example of the data of question answer example DB which concerns on embodiment of this invention. It is a flowchart of "side effect expression expansion processing" concerning an embodiment of the invention. It is a flowchart which shows the extraction processing flow of the side effect expression of the "side effect expression expansion process" which concerns on embodiment of this invention. It is a figure which shows the extraction method of the side effect expression of the "side effect expression expansion process" concerning embodiment of this invention. It is a figure which shows the supplementary information of the morphological analysis result which concerns on embodiment of this invention. It is a figure which shows the correspondence of various data in the side effect expression DB which concerns on embodiment of this invention, and update data. It is a flowchart of the related case search process which concerns on embodiment of this invention. It is a flowchart which shows the processing flow of the related case search part which concerns on embodiment of this invention. It is a flowchart which shows the processing flow of the search formula preparation part which concerns on embodiment of this invention. It is a figure which shows an example of preparation of the search expression which concerns on embodiment of this invention. It is a flowchart which shows the processing flow of search condition preparation which concerns on embodiment of this invention. It is a figure which shows an example of the "word form expansion pattern" of the state expression which concerns on embodiment of this invention. It is a flowchart which shows the creation process flow of the search result list | wrist of the related case search part which concerns on embodiment of this invention. It is a flowchart which shows the preparation process flow of the summary information of the related case search part which concerns on embodiment of this invention. It is a figure which shows an example of the search screen shown to the questioner which concerns on embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Pharmaceutical information management apparatus, 2 Question answer process control part, 3 Side effect expression extraction part, 4 Related case search part, 5 Search formula preparation part, 10 Question answer example DB, 11 Side effect expression DB.

Claims (5)

A drug information management program for causing a computer to manage information on side effects of drugs,
An identification information acquisition step of acquiring identification information of the medicine;
Corresponds to the identification information acquired in the identification information acquisition step from the first accumulation information accumulated by associating the identification information of the drug with the side effect type that is a medical term indicating the type of side effect of the drug A side effect type acquisition step of acquiring the side effect type;
Acquired in the identification information acquisition step from the second accumulated information accumulated by associating at least one of the identification information of the drug and the side effect type of the drug with the side effect expression that is an expression of the side effect of the drug A side effect expression acquiring step for acquiring the side effect expression corresponding to at least one of identification information and the side effect type acquired in the side effect type acquiring step;
A search expression that generates a search expression for each side effect type acquired in the side effect type acquisition step based on the identification information acquired in the identification information acquisition step and the side effect expression acquired in the side effect expression acquisition step Generation step;
Drug information management program that allows computers to execute. In the pharmaceutical information management program according to claim 1,
A search step for searching for a question answer example that matches the search formula generated in the search formula generation step from a third storage information in which a question about drug side effects and a question answer case that is an example of an answer to the question are stored When,
A predetermined range including the side effect expression is extracted from the question answer examples searched in the search step, and at least one question answer example in which the contents of the predetermined range are the same is attached with the contents of the predetermined range. A search result display step to display,
A pharmaceutical information management program characterized by causing a computer to execute. In the pharmaceutical information management program according to claim 2,
Further, a question about the side effect of the drug is obtained, and based on the question, at least one of identification information candidate and side effect type candidate is extracted and displayed from the first accumulated information, an answer to the question is obtained, and the question is obtained. A pharmaceutical information management program that causes a computer to execute a question answer case accumulation step of adding an answer to the question as the question answer example to the third accumulation information. A drug information management device for managing information on side effects of drugs,
An identification information acquisition unit for acquiring identification information of a medicine;
Corresponds to the identification information acquired by the identification information acquisition unit from the first accumulation information accumulated by associating the identification information of the drug with the side effect type that is a medical term indicating the type of side effect of the drug A side effect type acquisition unit for acquiring the side effect type;
Acquired by the identification information acquisition unit from the second accumulation information accumulated by associating at least one of the identification information of the drug and the side effect type of the drug with a side effect expression that is an expression of the side effect of the drug A side effect expression acquiring unit that acquires the side effect expression corresponding to at least one of identification information and the side effect type acquired by the side effect type acquiring unit;
For each side effect type acquired by the side effect type acquisition unit, a search formula that generates a search formula based on the identification information acquired by the identification information acquisition unit and the side effect expression acquired by the side effect expression acquisition unit A generator,
A pharmaceutical information management apparatus comprising: A drug information management method for managing information on side effects of drugs,
An identification information acquisition step of acquiring identification information of the medicine;
Corresponds to the identification information acquired in the identification information acquisition step from the first accumulation information accumulated by associating the identification information of the drug with the side effect type that is a medical term indicating the type of side effect of the drug A side effect type acquisition step of acquiring the side effect type;
Acquired in the identification information acquisition step from the second accumulated information accumulated by associating at least one of the identification information of the drug and the side effect type of the drug with the side effect expression that is an expression of the side effect of the drug A side effect expression acquiring step for acquiring the side effect expression corresponding to at least one of identification information and the side effect type acquired in the side effect type acquiring step;
A search expression that generates a search expression for each side effect type acquired in the side effect type acquisition step based on the identification information acquired in the identification information acquisition step and the side effect expression acquired in the side effect expression acquisition step Generation step;
Pharmaceutical information management method characterized by performing

Images