Return reason analysis device, notification device, and return reason analysis system

Abstract

To provide a return reason analysis apparatus, a notification apparatus, and a return reason analysis system capable of analyzing in detail the reason why a medical fee statement was returned.SOLUTION: A return reason analysis apparatus of the present disclosure includes an acquisition unit that acquires a return report record including a reason statement indicating a return reason, which is the reason for the return, generated at the medical facility when a medical fee statement is returned to the medical facility; and an analysis unit that performs predetermined statistical processing using the reason statement and analyzes the return reason.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present disclosure relates to a return reason analysis device, a notification device, and a return reason analysis system for analyzing reasons for which a medical fee statement has been returned.

Background Art

[0002] A medical facility creates a medical fee statement (also called a receipt) that describes the details of a medical procedure, etc., in order to bill an insurer for a portion of the fee for a medical procedure performed on an insured person, and requests a review from a review and payment agency. The review and payment agency reviews the content of the medical fee statement and, if there are no problems, causes the insurer to pay the fee. On the other hand, if there is a defect in the medical fee statement, the review and payment agency returns the medical fee statement to the medical facility. In this case, the medical facility can correct the defect in the medical fee statement and request a re-review from the review and payment agency. If the defect in the medical fee statement is improved, the medical facility can receive payment of the fee, but it is not operationally desirable for the medical facility to have a return occur. The reason for this is, for example, that it takes time and effort to improve the defect or request a re-review.

[0003] Patent Document 1 discloses a receipt analysis device that can calculate a doubtful amount, which is a medical fee amount that may not be paid due to a return of a past returned medical fee statement.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] In the technology disclosed in Patent Document 1, in order to reduce the number of returned medical fee statements, for past returns, the reasons for the return of the medical fee statements are analyzed for each doctor who created the medical fee statements. In Patent Document 1, it is stated that a person referring to the analysis results can easily formulate improvement measures for reducing deficiencies for each doctor.

[0006] However, there are various reasons for the return of medical fee statements. In the technology disclosed in Patent Document 1, only the analysis of the assumed reasons in advance is performed. Analyses capable of corresponding to various reasons regarding the return are required.

[0007] An object of the present disclosure is to provide a return reason analysis device, a notification device, and a return reason analysis system that can analyze in detail the reasons for the return of medical fee statements.

Means for Solving the Problems

[0008] The return reason analysis device of the present disclosure includes an acquisition unit that acquires a return report record including a reason sentence indicating a return reason, which is the reason for the return, generated at the medical facility when the medical fee statement is returned to the medical facility, and an analysis unit that performs predetermined statistical processing using the reason sentence and analyzes the return reason.

[0009] The notification device of the present disclosure includes a notification unit that performs notification regarding the return reason based on a category classified using the analysis result of the return reason by the above-described return reason analysis device.

[0010] The return reason analysis system of the present disclosure includes the above-described return reason analysis device and the above-described notification device.

Effects of the Invention

[0011] According to the present invention, the reasons for the return of medical fee statements can be analyzed in detail.

Brief Description of the Drawings

[0012]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Mode for Carrying Out the Invention

[0013] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. However, detailed descriptions that are more detailed than necessary, for example, detailed descriptions of well-known matters and duplicate descriptions of substantially the same configurations may be omitted in some cases.

[0014] FIG. 1 is a conceptual diagram illustrating a mode in which a return reason analysis system is used. FIG. 1 shows a plurality of dispensing pharmacies Ph and a headquarters HQ that oversees the plurality of dispensing pharmacies Ph. In the present embodiment, the plurality of dispensing pharmacies Ph correspond to each store belonging to a pharmacy chain, for example. The headquarters HQ is an organization that manages the operations of each store belonging to a pharmacy chain, for example. Note that, in the present embodiment, a dispensing pharmacy is exemplified as an example of a medical facility, but the present disclosure may be applied to a plurality of medical facilities, that is, hospitals, clinics, dental clinics, etc. that perform insurance medical treatment other than dispensing pharmacies, and a headquarters facility that oversees them.

[0015] A terminal device TE is arranged in each dispensing pharmacy Ph. The terminal device TE receives various input operations by a pharmacist or the like belonging to the dispensing pharmacy Ph.

[0016] The terminal device TE of each dispensing pharmacy Ph is connected to a return reason analysis system 100 arranged at the headquarters HQ via a network NW. The return reason analysis system 100 is a system that analyzes the reasons for returns based on past data regarding the medical treatment fee statements returned by the review and payment institution.

[0017] Note that the medical treatment fees include medical treatment fees related to medicine, dentistry, and dispensing. Since the dispensing pharmacy Ph is a dispensing pharmacy in the present embodiment, a medical treatment fee statement related to dispensing is created. When the medical facility is a hospital or a clinic, the medical facility creates a medical treatment fee statement related to medicine. When the medical facility is a dental clinic, the medical facility creates a medical treatment fee statement related to dentistry. The medical treatment fee statement is generally also called a receipt.

[0018] Note that the review and payment institution is, for example, the Social Insurance Medical Fee Payment Fund or the National Health Insurance Federation. The review and payment institution reviews the medical treatment fee statement created by the dispensing pharmacy Ph, and if there are no deficiencies in the content, causes the insurer of the medical insurance to pay the medical treatment fee to the dispensing pharmacy Ph, and if there are deficiencies in the content, returns the medical treatment fee statement to the dispensing pharmacy Ph.

[0019] The network NW is a public communication network such as the Internet, or an intranet or the like.

[0020] When the medical fee statement is returned, at each dispensing pharmacy Ph, an input of a return report reporting the return of the medical fee statement is performed via the terminal device TE. In the terminal device TE, an input of one return report is performed for one return. As a result, the terminal device TE generates one return report record for one input and transmits it to the return reason analysis system 100 arranged at the head office HQ via the network NW.

[0021] The return report record includes information (such as an identification number) for specifying the returned medical fee statement, information indicating the date of return, information regarding the sentence indicating the reason for return, and the like. In the following description, the sentence indicating the reason for return may be referred to as the reason sentence. The reason sentence has a body as a sentence, such as "I brought the family's insurance card by mistake.", "There is a misentry in the date of birth column.", etc., and also includes combinations of words or phrases such as "loss of qualification", "incorrect benefit rate". Further, the reason sentence may include two or more sentences, words or phrases.

[0022] The return report may be input at a predetermined timing or at an arbitrary timing at each dispensing pharmacy Ph. The predetermined timing is, for example, a specific day within a month. The arbitrary timing is, for example, the timing when the medical fee statement is returned, or the timing when the hands of an employee (such as a pharmacist) at the dispensing pharmacy Ph are free.

[0023] <Return reason analysis system 100> Hereinafter, the return reason analysis system 100 will be described in detail. FIG. 2 is a block diagram showing an example of the configuration of the return reason analysis system 100. As shown in FIG. 2, the return reason analysis system 100 includes a return reason analysis device 1, a notification device 2, and a storage unit 3.

[0024] The storage unit 3 stores various types of information used in the processes executed by the return reason analysis device 1 and the notification device 2. In particular, the storage unit 3 has a return report database (DB) that accumulates return report records received from the terminal device TE of each dispensing pharmacy Ph.

[0025] Based on the information stored in the storage unit 3, the return reason analysis device 1 analyzes the reasons for the return of the medical fee statement that occurred in the past (hereinafter referred to as the return reason), and classifies the return reasons into a predetermined number of categories based on the analysis results. The notification device 2 performs various notifications based on the information stored in the storage unit 3 and the results of the category classification (hereinafter referred to as the classification results) by the return reason analysis device 1.

[0026] Figure 3 is a flowchart for explaining an operation example of the return reason analysis system 100.

[0027] In step S1, the storage unit 3 accumulates the return report records received from the terminal device TE arranged at the dispensing pharmacy Ph via the network NW in the return report DB.

[0028] In step S2, the return reason analysis device 1 analyzes the return reason and classifies it into categories based on the accumulated return report records.

[0029] In step S3, the notification device 2 displays the classification result of the return reason.

[0030] In step S4, the notification device 2 performs a notification based on the classification result.

[0031] In step S5, the notification device 2 presents a method for eliminating the return reason based on the classification result.

[0032] <Return Reason Analysis Device 1> Figure 4 is a block diagram showing an example of the configuration of the return reason analysis device 1. As shown in Figure 4, the return reason analysis device 1 includes an acquisition unit 11 and an analysis unit 12.

[0033] The acquisition unit 11 acquires return report records received from the terminal devices TE of each dispensing pharmacy Ph via the network NW (see FIG. 1).

[0034] Based on the reason sentences included in the return report records stored in the return report DB, the analysis unit 12 performs predetermined statistical processing and analyzes the reasons for the return of the medical fee statements generated at the dispensing pharmacy Ph. The analysis unit 12 may perform the analysis using all of the return report records stored in the return report DB, or may perform the analysis using a part of the return report records stored in the return report DB, for example, the return report records generated during a predetermined period.

[0035] FIG. 5 is a flowchart for explaining the analysis process of the return reason in the return reason analysis device 1. The flowchart shown in FIG. 5 corresponds to step S2 in FIG. 3.

[0036] In step S11, the analysis unit 12 extracts the reason sentences from each of the return report records stored in the storage unit 3.

[0037] In step S12, the analysis unit 12 divides all the reason sentences into words. Note that the analysis unit 12 may divide the reason sentences into clauses instead of words. In the present disclosure, the analysis unit 12 dividing the reason sentences into words or clauses may sometimes simply be described as the analysis unit 12 dividing the reason sentences into words.

[0038] Also, when dividing the reason sentences into words, the analysis unit 12 may refer to a dictionary including medical-related terms. Thereby, even when general terms, such as industry terms and abbreviations, are included in the reason sentences, they can be accurately divided into words. Specifically, examples of the dictionary include words such as "rececomputer", which is an abbreviation of the receipt computer used for creating the medical fee statement (receipt).

[0039] Then, in step S13, the analysis unit 12 counts the words included in all the reason sentences. FIG. 6 is a diagram showing an example of the word count result. In FIG. 6, the words are shown in the left column, and the number of occurrences of each word is shown in the right column, respectively.

[0040] In step S13, the analysis unit 12 may classify the separated words by part of speech and then count them. FIG. 7 is a diagram showing an example of the count result when the words are classified by part of speech.

[0041] Next, in step S14, the analysis unit 12 extracts the words whose count results are equal to or greater than the threshold number of times. The threshold number of times is a value preset by an administrator of the return reason analysis system 100 or the like. The threshold number of times may be appropriately set based on, for example, the number of reason sentences acquired by the analysis unit 12 in step S11.

[0042] Next, in step S15, the analysis unit 12 performs predetermined statistical processing based on the extracted word group and the count results of each word. Examples of the predetermined statistical processing include correlation analysis, cluster analysis, and multidimensional scaling method.

[0043] FIG. 8 is a diagram showing an example of the analysis result when the multidimensional scaling method is used as the statistical processing. FIG. 8 shows the result of clustering for a plurality of words based on the distances and similarities between the plurality of words. In FIG. 8, the regions corresponding to each word are shown based on the coefficient results for each word, and the regions corresponding to words with close distances and similarities to each other are arranged close to each other. Also, the size of the region corresponding to each word is based on the total count result of the number of occurrences of each word, and the size of the region corresponding to a word with a large number of occurrences is large.

[0044] As a result, clusters are formed by groups of words where the frequency of occurrence of each word is relatively high and the distance or similarity between the words is close to each other. In FIG. 8, words belonging to the same cluster are shown with the same shading. The group of words included in each cluster is likely to be a group of words included in one reason sentence. In FIG. 8, eight clusters are formed, but the number of clusters may be appropriately set by, for example, the administrator of the return reason analysis system 100.

[0045] Note that the distance between words means the distance between the words included in the reason sentence. That is, in the reason sentence "There is an error in the date of birth column.", the distance between the word "date of birth" and the word "column" is 1 (the particle "no" is not counted). Similarly, the distance between the word "date of birth" and the word "error" is 2.

[0046] Similarity is the degree indicating the semantic proximity between two or more different words. Similarity is represented by, for example, the correlation coefficient between two or more different words calculated by correlation analysis. Alternatively, similarity may be represented by a value obtained by learning, such as by machine learning, how often one word is used together with the other word in the reason sentence.

[0047] Note that at the time of cluster analysis, information indicating the dispensing pharmacy Ph that created the return report record including the reason sentence in which each word was extracted, information indicating the area where the dispensing pharmacy Ph that created the return report record exists, or information indicating the time (year, month, day, etc.) when the return report record was created may be used for clustering. As a result, for example, it becomes possible to extract return reasons highly relevant to a specific medical facility, region, or time.

[0048] FIG. 9 is a diagram showing an example of another analysis result based on the counting result when the words are classified by part of speech as shown in FIG. 7. In FIG. 9, the result of clustering is shown such that the distance to a sahen noun or an adjectival noun having a high relevance to a certain noun becomes close. In FIG. 9, using the same group of words as in FIG. 8, an analysis result different from that in FIG. 8 is obtained.

[0049] The result of performing statistical processing such as cluster analysis in this way is the analysis result by the analysis unit 12. In addition to the multidimensional scaling method, an appropriate method can be adopted for the predetermined statistical processing performed by the analysis unit 12 to obtain the analysis result.

[0050] Next, in step S16, the analysis unit 12 classifies the return reasons based on the analysis result of step S15 and the category data indicating the existing return reasons. The category data indicating the existing return reasons is data showing the categories of return reasons classified by analyzing past return reasons and is stored in the storage unit 3 in advance. The existing category data may be data generated by the return reason analysis device 1 in the past, or may be data introduced from outside the return reason analysis system 100, for example. The existing category data may be generated by a person belonging to the head office HQ, for example.

[0051] FIG. 10 is a diagram showing an example of the category classification of return reasons. The category of return reasons shown in FIG. 10 includes three elements: classification name, reason, and cause. The classification name is the name indicating the category. The reason shows the outline of the return reason in each category. The cause shows the cause (the hospital where the prescription was created, the pharmacy that sold the medicine based on the prescription, the patient, etc.) for which the return occurred in each category. Although 10 categories are shown in FIG. 10, this is an example, and the number of categories can be arbitrarily set by the administrator of the return reason analysis system 100, for example.

[0052] Based on the analysis results and the existing category data, the analysis unit 12 newly classifies the return reasons indicated by the reason sentences included in the return report records stored in the storage unit 3 into a plurality of new categories. Specifically, the analysis unit 12 sets a word group that satisfies specific conditions among the word groups with close word distances and similarities as one category of return reasons. Satisfying specific conditions means, for example, belonging to one cluster as a result of clustering, or even a word group belonging to different clusters, as long as it corresponds to the category of return reasons classified in the existing category data.

[0053] In FIGS. 8 and 9, the solid ellipses show examples of the categories of return reasons classified by the analysis unit 12. As shown in FIGS. 8 and 9, the analysis unit 12 may classify the return reasons corresponding to the word groups included in one cluster into one category, or may classify the return reasons corresponding to the word groups included in the same cluster into different categories.

[0054] Through such processing, the analysis unit 12 can not only classify the return reasons indicated by the reason sentences included in the return report records stored in the storage unit 3 into the same categories as the existing category data, but also generate new categories. The new category is, for example, a category that does not match the existing category, and may include return reasons specific to a particular medical facility, region, or time period. Specifically, for example, if there are deficiencies in the business process of a certain pharmacy, many returns may occur at that pharmacy. In this case, the analysis unit 12 can set the fact that the pharmacy has generated a medical fee statement as a new category of return reasons. Or, when an infectious disease is locally prevalent in a certain region, the returns related to the infectious disease may increase at the pharmacies existing in that region. In this case, the analysis unit 12 can set the fact that the pharmacies in that region have generated a medical fee statement as a new category of return reasons. Furthermore, for example, when a change occurs in the calculation rule of medical fees due to the impact of the revision of medical fees, it is also possible to classify the return reasons that increase with the change into a new category different from the existing categories during a predetermined period from the medical fee revision.

[0055] In addition, the analysis unit 12 may determine whether the category classification in the existing category data is accurate based on the clustering result and the existing category data. For example, when the clustering result and the existing category data are far apart, it can be determined that either the clustering result or the existing category classification is not accurate. When it is determined that either the clustering result or the existing category classification is not accurate, the analysis unit 12 may include that fact in the classification result.

[0056] In addition, when the existing category data is generated by another return reason analysis system, the analysis unit 12 may learn using data related to the word groups included in the return reasons used at the time of generating the existing category data, and perform category classification based on the results of the learning. Alternatively, when the return reason analysis system 100 has performed category classification in the past using past return report records, the analysis unit 12 may learn using data related to the word groups included in the return reasons used at the time of the past category classification, and perform category classification based on the results of the learning.

[0057] Return to the description of FIG. 5. In step S17, the analysis unit 12 outputs data related to the result of the category classification to the notification device 2 and the storage unit 3. The storage unit 3 stores data related to the new category classification. Thereby, the analysis process of the return reason in the analysis unit 12 is completed.

[0058] As described above, the return reason analysis device 1 can classify the return reasons into a predetermined number of categories based on the return report records stored in the storage unit 3. The category classification by the return reason analysis device 1 is performed as a result of clustering the words included in the reason sentences of the return report records while referring to the existing category classification of the return reasons. In this way, the return reason analysis device 1 can classify the return reasons indicated by the reason sentences included in the return report records into various categories. Also, it is possible to determine whether the existing category classification is accurate.

[0059] In addition, since the return reason analysis device 1 performs category classification based on a large number of return report records input by a plurality of dispensing pharmacies Ph, it can perform a more accurate category classification compared to the case of performing category classification based on return report records input only by a specific medical facility.

[0060] In addition, when the reason for return in the return report record stored in the storage unit 3 has a high correlation with a specific medical facility, region, or time period, the return reason analysis device 1 can set a new category corresponding to the specific medical facility, region, or time period. Therefore, for example, when operations are stagnant due to a shortage of staff in a specific medical facility, when an infectious disease is prevalent in a specific region, or when the number of returns is increasing due to the impact of the revision of medical fees, the category classification of the reason for return can be performed immediately and appropriately. As a result, the return reason analysis device 1 can analyze not only the reasons for return classified into existing categories but also various reasons for return.

[0061] <Notification device 2> Next, the notification device 2 will be described. FIG. 11 is a block diagram showing an example of the configuration of the notification device 2. As shown in FIG. 11, the notification device 2 includes a receiving unit 21, a counting unit 22, and a notification unit 23.

[0062] The receiving unit 21 receives information regarding the category classification of the reason for return output from the return reason analysis device 1.

[0063] The counting unit 22 counts, for each category, the return report records included in the return report DB of the storage unit 3 that contain a reason statement indicating the reason for return classified into the category classified by the return reason analysis device 1.

[0064] The notification unit 23 has, for example, a display unit such as a liquid crystal display or a speaker, and performs various notifications. The notification by the notification unit 23 includes the display of characters or symbols on the display unit or the output of sound from the speaker. The display on the display unit and the sound output from the speaker may be used in combination.

[0065] The notification unit 23 displays, for example, the result of the category classification by the return reason analysis device 1 (corresponding to step S3 in FIG. 3). As a result, an administrator of the return reason analysis system 100 or the like can confirm what category classification has been performed by the return reason analysis device 1.

[0066] When the notification unit 23 displays the result of the category classification, it may be displayed in the form of a list as shown in FIG. 10, for example, or may be displayed in the form of a diagram showing the relationship between the word group and the category as shown in FIG. 8 or FIG. 9.

[0067] In addition, the notification unit 23 performs various notifications when, for example, the counting result of the counting unit 22 satisfies a predetermined condition (corresponding to step S4 in FIG. 3). The predetermined condition is, for example, a condition set by an administrator of the return reason analysis system 100 or the like. As an example of one of the predetermined conditions, for example, whether or not the number of return report records including a reason sentence corresponding to the return reason of a specific category among the return report records for a predetermined period is greater than a predetermined number. As another example of the predetermined condition, for example, whether or not the number of return report records input by a specific dispensing pharmacy Ph among the return report records for a predetermined period is greater than a predetermined number.

[0068] A plurality of predetermined conditions for determining whether or not the notification unit 23 performs notification are set in advance, and the notification unit 23 may determine whether or not to perform notification based on any one of the conditions, or may determine whether or not to perform notification based on a combination of a plurality of conditions. For example, when there is a category in which the number of return report records including the reason sentence of the return reason is equal to or more than a predetermined number at a specific dispensing pharmacy Ph, the notification unit 23 may perform notification regarding the return reason classified into the category.

[0069] Specific examples will be given to explain the content of the notification by the notification unit 23. For example, as a predetermined condition, it is assumed that it is set that within a predetermined period, at a specific pharmacy, the counting result is a category greater than a predetermined number. In this case, the counting unit 22 counts for each category of the return report records generated at a specific pharmacy A within a predetermined period, and extracts a category in which the counting result is greater than a predetermined number. In this case, the notification unit 23 performs notification regarding a category in which the counting result is greater than a predetermined number, such as "At pharmacy A, returns in the information error category occur frequently."

[0070] Note that the predetermined conditions may be set in more detail. For example, the conditions may be set using the elements (classification name, reason, cause) of each category shown in FIG. 10. For example, as a predetermined condition, among the return input records for the past one month generated at a specific pharmacy, the number of records where the return reason in the reason sentence included in each return input record is "typing error" and the cause is "pharmacy" is set to be a predetermined number or more (see FIG. 10). In this case, the counting unit 22 counts the records where the reason is "typing error" and the cause is "pharmacy" among the return input records for the past one month input at a specific pharmacy A, and the notification unit 23 performs notification when the counting result is more than the predetermined number. The content of the notification is, for example, "At pharmacy A, there are many returns due to typing errors." The content of the notification may be set, for example, by the administrator of the return reason analysis system 100 together with the predetermined conditions. Note that the notification by the notification unit 23 may be performed not only by a message like the above example, but also, for example, by displaying an icon indicating that there are many returns for a specific reason at a specific pharmacy, or by playing a specific alarm sound indicating that there are many returns for a specific reason at a specific pharmacy.

[0071] Also, as a predetermined condition, it is assumed that the number of records where the classification name of the reason sentence included in each return input record is "revision of medical fee" among the return input records in the one month after the revision of medical fees is set to be a predetermined number or more. In this case, the counting unit 22 counts the records where the classification name of the reason sentence included in each return input record is "revision of medical fee" among the return input records in the one month after the revision of medical fees input at all dispensing pharmacies Ph, and the notification unit 23 performs notification when the counting result is more than the predetermined number. The content of the notification is, for example, "Due to the revision of medical fees, the number of cases where the calculation rules are not observed is increasing. Please call the attention of each pharmacy."

[0072] Note that in the above example, it was assumed that the content of the notification was set by the administrator of the return reason analysis system 100 or the like together with predetermined conditions. However, for example, a learning model may be generated using as teacher data a predetermined condition set in the past and the content of the notification made based on the condition, and when a new notification based on a predetermined condition is to be made, the appropriate content of the notification may be generated based on the learning model. In this case, the labor of setting the content of the notification by the administrator of the return reason analysis system 100 or the like can be reduced.

[0073] The timing at which the notification unit 23 makes a notification may be, for example, a predetermined timing or a timing when a predetermined condition is satisfied. Examples of the predetermined timing include, for example, the beginning of the month (the 1st of each month), the end of the month (the last day of each month), a specific day (the 10th of each month, etc.).

[0074] Furthermore, the notification unit 23 may make a notification based on a predetermined condition and present a resolution method for eliminating the return reason (corresponding to step S5 in FIG. 3). The presentation of the resolution method is made, for example, by displaying a message such as "Please improve the type mistakes at pharmacy A" when there are many returns due to type mistakes at a specific pharmacy A. The resolution method may be set in advance for each return reason by the administrator of the return reason analysis system 100 or the like, or may be generated based on a learning model generated in advance using as teacher data the return reason and the resolution method presented in the past, at the timing when a notification based on a predetermined condition is made.

[0075] As described above, the notification device 2 displays the result of the category classification of the return reason performed by the return reason analysis device 1. Thereby, when the return reason of the return report record stored in the storage unit 3 has a high correlation with a specific medical facility, region, or time, the administrator of the return reason analysis device 1 or the like can recognize that a new category corresponding to the specific medical facility, region, or time has been set. For this reason, the administrator of the return reason analysis device 1 or the like can easily recognize problems such as, for example, business stagnation due to a shortage of staff in a specific medical facility, an epidemic of an infectious disease in a specific region, or an increase in returns due to the influence of the revision after the revision of the medical fee schedule.

[0076] Further, when the count result of the return report records stored in the storage unit 3 satisfies a predetermined condition, the notification device 2 performs notification corresponding to the predetermined condition. For this reason, the administrator of the return reason analysis device 1 or the like can grasp in advance the points to be noted in order to reduce the return of the medical fee statement by setting the conditions.

[0077] Furthermore, since the notification device 2 presents a resolution method for resolving the return reason together with the notification based on the predetermined condition, the administrator of the return reason analysis device 1 or the like can easily grasp what measures should be taken to reduce the return of the medical fee statement.

[0078] <Modification Example> In the above description, one embodiment of the present disclosure has been described. The present disclosure is not limited to the above-described embodiment, and various modifications as described below may be made.

[0079] In the above-described embodiment, the return reason analysis device 1 has been described as classifying return reasons into a predetermined number of categories based on the return report records stored in the storage unit 3. However, the return reason analysis device 1 may, for example, in the terminal device TE of the dispensing pharmacy Ph, while the reason sentence of the return report record is being input, obtain the reason sentence in the middle of the input in real time via the network NW, and determine which category the return reason indicated by the obtained reason sentence corresponds to. At this time, if the reason sentence input in the dispensing pharmacy Ph is not suitable for the determined category, the return reason analysis device 1 may transmit, via the network NW, a message or the like indicating that the reason sentence should be corrected to the terminal device TE of the dispensing pharmacy Ph for display. Further, the return reason analysis device 1 may present information necessary for creating a return report record according to the category to which the reason sentence input in the dispensing pharmacy Ph belongs. Specifically, the return reason analysis device 1 may display a message such as "The reason sentence being input now is considered to correspond to the return reason of the ○○ category. The return report of the ○○ category requires ×× data." on the terminal device TE.

[0080] Also, based on the category of the return reason classified by the return reason analysis device 1, when inputting a return report record in the terminal device TE of the dispensing pharmacy Ph, instead of allowing the reason sentence to be input in plain text, it may be made selectable from among the preset categories of return reasons. In this case, the return reason analysis device 1 may, for example, every time a predetermined period elapses, reflect the category of the return reason classified based on the return report records obtained within the predetermined period in the input field used for inputting the return report record in the terminal device TE of each dispensing pharmacy Ph in the next predetermined period. Thereby, in the terminal device TE of each dispensing pharmacy Ph, an immediate and appropriate return reason can always be easily selected without the labor of inputting plain text.

[0081] The notification device 2 counts the return report records for each dispensing pharmacy Ph and performs notification. Further, based on the return report records in a predetermined period, for example, for each dispensing pharmacy Ph, the amount that will not be paid for returns (hereinafter referred to as the expected non-payment amount) may be estimated and notified. Generally, the remuneration based on the medical treatment remuneration statement is paid to the dispensing pharmacy Ph after a certain period (for example, two months) from the submission of the medical treatment remuneration statement for the review by the review and payment institution. Therefore, the notification device 2 can easily estimate the expected non-payment amount based on the return report records in a predetermined period.

[0082] At this time, the notification device 2 may calculate and notify the expected non-payment amount for each category based on the counting results for each dispensing pharmacy Ph and for each category of return reasons. Further, a learning model for estimating the expected non-payment amount in a predetermined period may be generated based on the number of past return report records for each dispensing pharmacy Ph and the number of return report records for each dispensing pharmacy Ph and for each category of return reasons, and the expected non-payment amount in a future predetermined period may be notified based on the learning model.

Industrial Applicability

[0083] The present disclosure is useful for a return reason analysis system that analyzes the return reasons of medical treatment remuneration statements.

Explanation of Signs

[0084] 100 Return Reason Analysis System 1 Return Reason Analysis Device 2 Notification Device 3 Storage Unit 11 Acquisition Unit 12 Analysis Unit 21 Reception Unit 22 Counting Unit 23 Notification Unit

Claims

1. An acquisition unit that, when a medical fee statement is returned to a medical facility, acquires a return report record including a reason sentence indicating the reason for the return, which is generated at the medical facility; An analysis unit that performs predetermined statistical processing using the reason sentence and analyzes the reason for the return; A return reason analysis device comprising:

2. The analysis unit classifies the return reason into a plurality of categories based on the analysis result of the return reason. The return reason analysis device according to Claim 1.

3. The analysis unit classifies the return reason into the plurality of categories based on the analysis result and category data indicating existing return reasons. The return reason analysis device according to Claim 2.

4. In the statistical processing, the analysis unit divides the reason sentence into words, counts the number of occurrences of each word that appears in the reason sentences included in a plurality of the return report records, and extracts words whose number of occurrences is greater than a threshold number of occurrences. The return reason analysis device according to Claim 2 or 3.

5. In the statistical processing, the analysis unit performs clustering on the extracted words and uses the result of the clustering as the analysis result. The return reason analysis device according to Claim 4.

6. While a reason sentence for generating the return report record is being input, the analysis unit determines into which of the plurality of categories the return reason indicated by the input reason sentence is classified. The return reason analysis device according to any one of Claims 2 to 5.

7. A notification device comprising a notification unit that performs notification based on a classification result by the return reason analysis device according to any one of Claims 2 to 6. Notification device.

8. The notification unit notifies the classification result. The notification device according to Claim 7.

9. Further comprising a counting unit that counts the return report records including the reason sentences of the return reasons classified into each category, The notification unit performs the notification based on the counting result of the counting unit. The notification device according to Claim 7 or 8.

10. When there is a category in which the number of return report records including the reason sentences of the classified return reasons is equal to or greater than a predetermined number, the notification unit performs notification regarding the return reasons classified into the category. The notification device according to Claim 9.

11. The counting unit counts the return report records generated by each medical facility for each medical facility. When the count result of the return report records generated by one medical facility is greater than a predetermined number, the notification unit performs notification regarding the one medical facility. The notification device according to claim 9. **Claim 12** The counting unit counts the return report records generated by each medical facility. When, in one medical facility, the count result of the return report records including a reason sentence corresponding to a return reason of one category is greater than a predetermined number, the notification unit performs notification by associating the one medical facility with the one category. The notification device according to claim 9. **Claim 13** A return reason analysis device according to any one of claims 1 to 6, A notification device according to any one of claims 7 to 12, A return reason analysis system comprising the same.

Images