JP5003004B2 - Variance root cause analysis support system - Google Patents

Description

  The present invention relates to hospital information system technology in the medical field, and more particularly to a system that supports root cause analysis of variance, which is the difference between a standard medical care plan and actual medical care.

  A clinical path (also called a critical path) that represents a standard treatment plan in order to improve the quality of medical care by realizing standardization of medical care in a social situation that promotes the improvement and efficiency of medical care. Is emphasized. In order to improve the quality of medical care using the clinical path, it is said that it is necessary to improve the clinical path based on the basis based on the accumulated medical data. In particular, by collecting and analyzing the variance that is the difference between the standard treatment plan described in the clinical path and the actual treatment, it is possible to create a clinical path with low occurrence frequency and high medical effect. Conceivable. Specifically, from the viewpoints of length of stay, treatment results, cost, etc., analysis was performed by extracting as a variance the treatments that were different from the standard treatment plan, and the outcomes that were clinically desirable outcomes and goals were not achieved. I do. As a method for collecting and analyzing variance, a discharge variance method, a gateway variance method, a sentinel method, and an all variance method have been proposed according to the granularity of variance ("Patent Document 1"). Among them, the all variance method is a method for analyzing differences from all medical treatments described in the clinical path in order to analyze the variance in detail, and the improvement effect of the clinical path is considered to be remarkable. Yes. With the spread of electronic medical records, it has become possible to collect a large amount of data, and expectations for realizing all variance analysis are increasing. In particular, there is an increasing expectation for the realization of the analysis of the root cause of variance, which is considered to have a high clinical path improvement effect. The reason is that, in the case of a method of manually performing a root cause analysis of variance that is widely performed at present, there are many cases where only a superficial cause of occurrence of variance can be found.

  Next, a conventional example related to a system that supports variance analysis and creation of a clinical path with low occurrence frequency and high medical effect will be described. In the method (conventional example 1) described in “Non-Patent Document 1”, first, a specific event such as surgery or movement from an ICU (Intensive Care Unit) to a general ward is set in advance. Next, the occurrence frequency of the variance that is sliced variously according to the occurrence factor of the variance for the set event is displayed in a graph. The system described in “Patent Document 1” (Conventional Example 2) is a clinically desirable outcome and target outcome for each disease name and postoperative day and criteria for determining whether the outcome has been achieved. This is an electronic medical record system equipped with a file describing a certain assessment as a database. The system described in “Patent Document 2” (conventional example 3) is a system that divides the entire medical process into several steps for each purpose, and sets conditions for moving to the next step in each step. This makes it possible to perform medical process management with a clear standard for process extension and suspension. The system described in “Patent Document 3” (conventional example 4) is based on clinical path standardization information, which is progress information composed of a plurality of pieces of medical practice information indicating a schedule related to a medical practice for a patient. Patient-specific clinical path information, which is information created for each patient, is acquired. Next, the execution result information implemented based on the instruction information created from the patient-specific clinical path information is acquired. Thereby, it is possible to flexibly cope with the occurrence of variance.

JP 2004-86506 A JP 2004-13219 A JP 2005-165513 A Variance Management System (Bing Net Press) pp.26-27, pp.41-42

  In the variance analysis in the prior art, emphasis was placed on the collection and management of variance. For this reason, the variance analysis method for the purpose of creating and improving a clinical path with low occurrence frequency and high medical effect has not been sufficiently considered. In particular, efficient analysis of the cause of the occurrence of variance was not considered. Further, in order to realize the all variance analysis, the number of objects of the variance analysis becomes enormous, so that even if all the variances are acquired, there is a problem that the analysis requires a great deal of labor.

  Conventional Example 1 is limited to grasping the occurrence status of variance, and in order to formulate countermeasures for the generated variance, it is necessary to further examine individual cases. In addition, because the focus is on the frequency of occurrence in grasping the status of variance, in order to find a significant variance that is less frequent but leading to death, a comparison of the effect on treatment results and length of hospital stay is required. This approach was necessary. Conventional example 2 focuses on variance collection for quantitative variance analysis, but does not mention a method for analyzing the collected variance. Also, it does not take into account that the number of objects of variance analysis such as all variance analysis becomes enormous. Conventional example 3 and conventional example 4 prescribe a medical practice for a presumed variance, and it is necessary to create a rule in advance which medical practice is to be performed under what conditions. However, an effective variance analysis method for creating this rule was not considered.

  As described above, in order to effectively analyze a large amount of collected variance and analyze the root cause of variance having a high clinical path improvement effect, it is difficult to obtain a sufficient effect in the conventional example described above. It was.

  Therefore, an object of the present invention is to provide a variance root cause analysis support system that extracts candidates that can become root causes of variance from implementation information and history of root cause analysis performed in the past.

  In order to achieve the above object, the variance analysis system of the present invention including an input means for analyzing a variance that is a difference between a standard medical care plan and an actual medical care is a difference between the standard medical care plan and the actual medical care. An electronic medical record system including an electronic medical record database including an input means for analyzing a variance, and an electronic medical record input / output means, an electronic medical record control means, and implementation record data including performance data of variance An electronic medical record linking means that cooperates with the data, a root cause analysis database that accumulates the history of the cause of the occurrence of variance that has been captured through the input means, and a correlation rule that includes each variance in the conclusion section from the implementation record data by data mining Correlation generating means for different variances for generating the correlation rule, From the a variance conclusion part and analyte which is selected via the input means of association rules and having an input auxiliary knowledge search means for extracting at least partially coincident to correlate Rururu.

  Further, in the variance analysis system described above, the correlation generation unit according to variance creates a correlation rule including variance in a conclusion part by data mining from the history data of the execution record and the cause of occurrence of variance. And

  Furthermore, the variance analysis system described above further includes importance calculation means for calculating a variance importance representing the degree of severity of injury finally caused by the occurrence of variance from the execution record data.

  ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to provide the variance root cause analysis support system which extracts the candidate which can become the root cause of a variance from implementation information and the history of the root cause analysis conducted in the past.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a configuration diagram of a variance root cause analysis support system according to the present invention. The variance root cause analysis support system shown in FIG. 1 includes an electronic medical record input / output unit 100, an electronic medical record control unit 101, an electronic medical record database 102, an electronic medical record linkage unit 103, a correlation generation unit 104 according to variance, It comprises a cause analysis database 105, input auxiliary knowledge database 106, input auxiliary knowledge search means 107, screen configuration processing means 108, display means 109, and input means 110. The variance root cause analysis support system supports root cause analysis in which variance has occurred by utilizing past record data accumulated in the past when improving the clinical path.

  The hardware configuration of this configuration will be described. FIG. 2 shows a hardware configuration diagram for realizing the variance root cause analysis support system. The electronic medical record database 102 includes an HDD (Hard Disk Drive) device 2014 and the like. Similarly, the root cause analysis database 105 and the input auxiliary knowledge database 106 are configured by the HDD device 2024 and the like. The electronic medical chart control unit 101 can realize various processes by developing and starting a predetermined program in the central processing unit 2013, the memory 2012, and the like. Similarly, the electronic medical record linking means 103, the variance-specific correlation generating means 104, the input auxiliary knowledge searching means 107, and the screen configuration processing means 108 are expanded in a predetermined program in the central processing unit 2023, the memory 2022, or the like. Various processes can be realized by activation. The display means 109 can be realized by a monitor using a liquid crystal display 2021 or a CRT (Cathode-Ray Tube). The input unit 110 can be realized by a keyboard 2020, a mouse, or a pen tablet. Similarly, the electronic medical record input / output means 100 can be realized by a liquid crystal display 2011, a keyboard 2010, and the like.

Hereinafter, each of these components will be described in more detail.
The electronic medical record linkage unit 103 acquires master data and execution record data from the electronic medical record system. FIG. 3 shows two tables relating to master data stored in the electronic medical record database 102. In the action master table, the medical practice to be performed, the medical practice code, and the elapsed days to be implemented are recorded. In this example, three medical treatments of orientation, vital check and regular medicine check are performed on the first day of hospitalization. In the variance content master table, the content of variance that may occur in each medical practice and its variance code are recorded. FIG. 4 shows an implementation record table representing implementation record data stored in the electronic medical record database 102. The implementation record data is a history of the practiced medical practice and a progress record regarding the patient condition. In the implementation record table, data on the date when the medical practice was performed for each patient and the variances that occurred were recorded.

  First, the correlation generation unit 104 according to variance creates knowledge about a specific medical practice, patient condition, variance, etc. at the time of occurrence of variance. The knowledge is, for example, an association rule having a variance as a conclusion part. FIG. 5 is a sequence diagram relating to the correlation generation unit 104 according to variance. In order to generate a correlation rule in the correlation generation unit 104 according to variance, an acquisition request for execution record data and the like to the electronic medical record linkage unit 103 and a root cause acquisition request to the root cause analysis database 105 are made. Here, the acquisition request of the root cause data to the root cause analysis database 105 and the acquisition of the root cause from the root cause analysis database 105 are not essential. An association rule having a variance as a conclusion part is generated by data mining from the execution record data and the root cause data acquired in this way. The generated correlation rules are accumulated in the correlation table of the input auxiliary knowledge database 106.

  FIG. 6 shows a correlation table. In this embodiment, an event having a high correlation with the occurrence of a variance of “bleeding (variance code: V5)” is recorded as a correlation rule with respect to the medical treatment of “puncture (medical practice code: T4)”. The correlation rule is shown. The first is that the certainty of the association rule that “bleeding” occurs when a patient who is “65 years of age or older” and “Balun insertion” is performed is 80%, and is generated from an implementation record. Second, the confidence of the correlation rule that “bleeding” occurs in the case of “insufficiently fixed sheath” and “fluctuating body position” is 90%, which is generated from the root cause data described later. The difference between the correlation rule generated from the implementation record and the correlation rule generated from the root cause data will be described in detail in the explanation of the root cause analysis database 105.

In the input auxiliary knowledge search means 107, correlation rules relating to the medical practice specific to the occurrence of the variance of the analysis target selected through the input means 110, the patient state, and other variances are obtained from the correlation table of the input auxiliary knowledge database 106. Explore. In order to select the analysis target variance, it is carried out with reference to a screen displayed on the display means 109 described later. When searching from the correlation table shown in FIG. 6 , a record holding the correlation code of the correlation table corresponding to the variance selected with reference to the screen displayed on the display unit 109 is acquired. The acquired record is passed to the screen configuration processing means 108.

  The root cause analysis database 105 stores the result of the user analyzing the cause of the occurrence of variance. Specifically, the cause of the occurrence of variance that has been captured via the input unit 110 described later is stored in the root cause table of the root cause analysis database 105. FIG. 7 shows a root cause table. In this embodiment, the result of analyzing the causal relationship in which the variance of “bleeding (variance code: V5)” with respect to the medical practice of “puncture (medical practice code: T4)” is expressed in a tree structure format Is shown. For example, the causes of the occurrence of the variance “bleeding (node ID: 1)” are “natural sheath removal (node ID: 1-1)” and “catheter insertion failure (node ID: 1-2)”. Show. In addition, the cause of the “natural removal of the sheath” is “insufficient fixing of the sheath (node ID: 1-1-1)” and “many posture change (node ID: 1-1-2)”. The situation is also shown. Such a causal relationship is described in the “node ID” field. Note that the correlation generation unit 104 according to variance described above may generate a correlation rule having a variance as a conclusion part from the root cause table shown in FIG. 7 by data mining. As a result, the amount of root cause analysis data increases as the user performs root cause analysis. Therefore, it becomes possible to create a more fulfilling association rule, and it becomes possible to perform root cause analysis more effectively.

  The screen configuration processing unit 108 configures a screen layout for analyzing the root cause of variance, transfers data to the display unit, and acquires information captured by the input unit. Three examples of the screen layout presented on the display means 109 will be described.

  FIG. 8 is a first example of a screen layout. This example includes a variance basic information presentation unit 801, a variance root cause analysis construction unit 802, and a comment description unit 803. The variance basic information presentation unit 801 further includes an analysis target selection unit 8011, a statistical data presentation unit 8012, a rule presentation unit 8013, and a data storage button unit 8014. The analysis target selection unit 8011 selects a medical practice, a patient, and a variance to be analyzed. The statistical data presentation unit 8012 presents basic statistical data regarding the selected patient. In this example, patient B's length of stay and balance are presented. The rule presentation unit 8013 presents the correlation rule for variance obtained from the correlation table via the input auxiliary knowledge search means 107. Two processes are performed to present. First, a record holding a correlation rule corresponding to the variance selected by the analysis target selection unit 8011 is acquired from the correlation table via the input auxiliary knowledge search means 807. Next, the value of the prerequisite knowledge condition of the acquired record is displayed. The correlation rules to be presented may be presented in descending order of certainty, or a predetermined number of presentations may be extracted and presented from correlation rules with high certainty. The data save button unit 8014 stores the analysis result constructed by the user in the variance root cause analysis construction unit 802 via the root cause analysis database 105. The variance root cause analysis construction unit 802 describes the result of the user analyzing the cause of occurrence of variance. This example shows the result of analysis using a characteristic factor diagram widely used as a method for analyzing the cause of the problem. The user may construct the causal event by a text-based free sentence description, or may construct semiautomatically by an association rule selection method using the rule presentation unit 8013. In constructing the analysis result, it is possible to support the analysis of the cause of occurrence by adopting a method of selecting the correlation rule presented by the rule presentation unit 8013. A comment description unit 803 describes comments for other departments in a text format. When constructing the analysis result in this way, by utilizing the correlation rule presented by the rule presentation unit 8013, it is possible to search not only the superficial cause of the occurrence of the variance but also the root cause, and analyze the cause of the occurrence. Can help.

  FIG. 9 is a second example of the screen layout. A difference from FIG. 8 is a variance root cause analysis construction unit 802. In this example, an actor is shown on the horizontal axis and time is shown on the vertical axis, and the causal relationship until the analysis target variance occurs is shown. As in FIG. 8, the rule presentation unit 8013 is used to describe the result of the user analyzing the cause of occurrence of variance.

  10 and 11 show a third example of the screen layout. FIG. 10 is a screen for selecting the analysis target variance and constructing the variance root cause analysis, and includes a variance analysis target unit 1001 and a variance root cause analysis construction unit 1002. In FIG. 11, in order to support the construction of the variance root cause analysis in FIG. 10, the correlation rule of the variance obtained from the correlation table via the input auxiliary knowledge search unit 107 is presented and operates in cooperation with FIG. 10. This example includes a rule presentation unit 1101. Here, an example of screen operation will be described. The variance analysis target unit 1001 selects a medical practice, patient, and variance to be analyzed, and builds a root cause analysis result in the variance root cause analysis construction unit 1002. In FIG. 10, the same content described in the characteristic factor diagram is displayed in the left part of the diagram in the form of a tree structure. At the time of construction, the user may construct the causal phenomenon by a text-based free sentence description, or may construct semiautomatically by a correlation rule selection method utilizing the screen shown in FIG. In order to analyze using FIG. 11, a method of transitioning to FIG. 11 by pressing the right mouse button in the variance root cause analysis construction unit 1002 can be considered. In FIG. 11, the correlation rule of variance obtained from the correlation table via the input auxiliary knowledge search means 107 is presented to the rule presentation unit 1101. Two processes are executed before presentation. First, a record holding a correlation rule corresponding to the variance selected by the variance analysis target unit 1001 is acquired from the correlation table via the input auxiliary knowledge search means 107. Next, the value of the premise part knowledge condition of the acquired record is displayed on the rule presentation unit 1101. In the rule presentation unit 1101, the horizontal axis indicates the number of days elapsed since hospitalization, and the vertical axis indicates the work classification of the medical practice, and a rectangle is arranged according to the number of elapsed days when the medical practice was performed. Further, the color of the rectangle is determined according to the certainty factor. By visually displaying the association rule in this way, it is easy to grasp a medical practice that can be a cause of occurrence, and support for root cause analysis becomes possible.

  As a result, it is possible to efficiently search for the cause of the occurrence of the variance, so that the root cause analysis can be supported. By grasping the root cause, it is considered that improvement to a clinical path with low occurrence of variance and high medical effect can be realized. In addition, by generating association rules not only from the electronic medical record system but also from root cause data constructed by the user, the possible cause candidates are gradually enriched, enabling root cause analysis to be carried out more effectively. Become.

  FIG. 12 is a configuration diagram of a variance root cause analysis support system according to the present invention. In particular, in order to effectively narrow down the variance to be analyzed from the large amount of variance, the importance calculation means 111 is added to the configuration diagram shown in FIG. The importance level calculation unit 111 can realize various processes by developing and starting a predetermined program in the central processing unit 2023 and the memory 2022 shown in FIG. The importance calculation unit 111 can determine the priority of the variance by calculating the variance importance by quantifying the influence on the clinical path improvement for each variance. FIG. 13 shows a flowchart for calculating the variance importance. First, execution record data including variance information is acquired via the electronic medical record linkage unit 103 (S1301). Next, the frequency is calculated for each variance code from the implementation record data to calculate the variance frequency (S1302). Next, the hospital days influence degree, which is the degree of influence on the hospital days for each variance code, is calculated from the implementation record data (S1303). As an example of the hospital days influence degree, an average value obtained by normalizing the hospital days can be considered. Finally, the variance importance is calculated from the variance frequency and the hospital stay influence (S1304). As an example of calculating the variance importance, a value obtained by multiplying the variance frequency and the hospital stay influence degree may be considered. In the calculation method shown in FIG. 13, the variance importance is calculated from the variance frequency and the hospital stay influence. However, the variance importance is calculated in consideration of factors necessary for improving the clinical path such as income and treatment results. It may be calculated.

  Next, an example of a fundamental analysis of variance using the importance of variance is shown. FIG. 14 is a diagram showing the importance of variance. First of all, in the variance selection means display unit 1401 on the left side in the figure, a medical practice or outcome (target to be achieved) of interest is selected. When paying attention to all medical treatments and outcomes, the all selection button 14011 is pressed. On the other hand, in order to cancel the selection in all the medical treatments and outcomes, the all cancel button 14012 is pressed. An index related to the importance of the variance generated in the medical practice or outcome selected in this way is presented to the right-side variance importance presentation section 1402 in the figure. In FIG. 14, six indicators such as hospital stay, income, cost, treatment result, frequency, and difficulty in detection are displayed in a radar chart format. Since the area of the hexagon formed by each line segment corresponds to the magnitude of the importance of variance, it is possible to identify a medical practice or outcome in which a variance that requires countermeasures occurs immediately. In FIG. 14, a medical practice or outcome “(1) Ready” and “(5) Pain / fever can be controlled in an outpatient” is selected on the variance selection means display unit 1401. As a result, each index is presented to the variance importance degree presentation unit 1402. The user can determine that countermeasures should be taken for “(5) pain and fever can be controlled in the outpatient” with a large hexagonal area and a high value of variance importance, and pressing the fundamental analysis button 14013 Thus, it becomes possible to shift to the variance root cause analysis shown in FIG. This makes it possible to efficiently select a variance having a high clinical path improvement effect from a large amount of variances and easily shift to effective and detailed fundamental analysis.

It is a 1st block diagram of the variance root cause analysis assistance system in this invention. It is a hardware block diagram of the variance root cause analysis support system in this invention. It is a figure showing the master in an electronic medical chart system in this invention. It is a figure showing the implementation record table in an electronic medical chart system in this invention. It is a sequence diagram which produces | generates correlation rule data in this invention. It is a figure showing a correlation table in this invention. It is a figure showing a root cause table in the present invention. It is a 1st example which shows the screen in the variance root cause analysis assistance system in this invention. It is a 2nd example which shows the screen in the variance root cause analysis support system in this invention. It is a 3rd example which shows the screen in the variance root cause analysis support system in this invention. It is a 3rd example which shows the screen in the variance root cause analysis support system in this invention. It is a 2nd block diagram of the variance root cause analysis assistance system in this invention. It is a flowchart regarding the importance calculation means in this invention. It is a figure which shows the screen which presents the data relevant to the variance importance in this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Electronic medical record input / output means, 101 ... Electronic medical record control means, 102 ... Electronic medical record database, 103 ... Electronic medical record cooperation means, 104 ... Correlation generating means classified by variance, 105 ... Root cause analysis database, 106 ... Input auxiliary knowledge database , 107 ... input assist knowledge search means, 108 ... screen configuration processing means, 109 ... display means, 110 ... input means, 111 ... importance calculation means, 2010 ... keyboard, 2011 ... display, 2012 ... memory, 2013 ... central processing unit , 2014 ... HDD device, 2020 ... keyboard, 2021 ... display, 2022 ... memory, 2023 ... central processing unit, 2024 ... HDD device, 801 ... variance basic information presentation unit, 8011 ... analysis target selection unit, 8012 ... statistical data presentation unit , 8013 ... rule presentation unit, 8 14 ... data storage button section, 802 ... Variance Variance root cause analysis construction unit, 803 ... comments describing unit, 1001 ... analysis of variance objective portion, 1002 ... variance root cause analysis construction unit, 1101 ... rule presentation unit,
1401 ... Variance selection means display section, 14011 ... All selection button, 14012 ... All release button, 14013 ... Root analysis button, 1402 ... Variance importance degree presentation section.

Claims (2)

A variance analysis system including a variance analysis means for analyzing a variance that is a difference between a standard medical care plan and an actual medical care, an input means for receiving an input, and a display means for displaying a result,
The variance analysis means includes
Electronic medical record input / output means, electronic medical record control means, and electronic medical record linking means that cooperates with an electronic medical record system having an electronic medical record database in which execution record data including at least information on medical practice and variance is stored as performance data of variance ,
A root cause analysis database that accumulates the variance captured through the input means and the cause of occurrence of the variance as a set;
Data mining outputs a first rule with the implementation record data as input and a second rule with the data accumulated in the root cause analysis database as input, and the first rule is extracted from the implementation record data A rule is generated in which a medical practice is a premise and a variance extracted from the execution record data is a conclusion. In the second rule, the cause of the variance that is extracted from the root cause analysis database is a premise and the variance is concluded. A correlation-specific correlation generating means for generating a rule to be a part,
The rule generated by the correlation generation unit according to variance and the variance of the analysis target selected via the input unit are input, and the first rule and the second rule are input via the input unit. A variance root cause analysis support system, comprising: input auxiliary knowledge search means for extracting a rule in which the selected analysis target variance is included in a conclusion section and outputting the extracted rule to the display means. The variance analysis system according to claim 1,
From the implementation record data, there is an importance degree calculation means for calculating the importance degree of variance that represents the degree of severity of injury finally caused by occurrence of variance,
In the importance calculation means, a variance influence degree representing the degree of serious injury caused by the occurrence of the variance from the execution record data is used, and an indicator of the length of stay in the subject case is used as the presence or absence of variance. 2. Variance root cause analysis support characterized in that it is calculated by comparison between two groups, and then the variance influence degree is calculated from the variance frequency for calculating the frequency for each variance from the execution record data and the variance influence degree. system.

Images