Medical quality safety management system

Abstract

The invention discloses a medical quality safety management system, and belongs to the field of medical management. The medical quality safety management system comprises an adverse event management module which is used for front-line medical personnel to input adverse events and a manager to read the beginning and the end of the adverse events in real time; the data summarization and analysis module is used for reading the indexes in the events recorded by the adverse event management module, respectively counting the indexes to form a form, and analyzing and calculating the incentive proportion of the adverse events by using a K-means clustering analysis algorithm; and the data statistics module is used for generating an adverse event analysis report by utilizing a CART model according to an analysis result of the data summarization analysis module. According to the technical scheme, the medical staff can input the adverse events by logging in the account number, the manager can read the beginning and the end of the adverse events in real time by logging in the account number, analysis is carried out through the clustering analysis algorithm and the CART model, the report is generated, and therefore the workload of the medical staff is relieved.

Description

Technical Field

[0001] The present invention belongs to the field of medical management, and in particular relates to a medical quality and safety management system. Background Art

[0002] With the continuous development of science and technology, adverse events often occur in hospitals. During the operation of the hospital, it is necessary to conduct statistical analysis of adverse events. The traditional statistical method of adverse events is that front-line medical staff write down the whole story of the adverse event by hand, convey it to the head nurse, the head nurse conveys it to the head nurse of the department, and finally conveys it to the nursing department; then manual analysis and summary are performed to generate analysis reports; such a process brings a workload to medical staff, and for this reason, a medical quality and safety management system is proposed. Summary of the invention

[0003] In view of the deficiencies of the prior art, the purpose of the present invention is to provide a medical quality and safety management system to solve the problems in the prior art.

[0004] The purpose of the present invention can be achieved through the following technical solutions:

[0005] Medical quality and safety management system, including:

[0006] Adverse event management module: used for frontline medical staff to input adverse events, and managers to read the whole process of adverse events in real time;

[0007] Data summary and analysis module: read the indicators of each event recorded in the adverse event management module and count them separately to form a table, and use the K-means clustering analysis algorithm to analyze and calculate the proportion of the causes of adverse events;

[0008] Data statistics module: Based on the analysis results of the data summary and analysis module, the CART model is used to generate an adverse event analysis report.

[0009] Medical staff can enter adverse events by logging into their accounts, and managers can read the entire process of adverse events in real time by logging into their accounts. They can also analyze and generate reports through cluster analysis algorithms and CART models, thereby reducing the workload of medical staff.

[0010] Furthermore, the types of adverse events include: accidental extubation, falls from bed, medication errors, skin problems, surgical safety, specimen collection, intravenous treatment and interception events.

[0011] Furthermore, the analysis process in the data summary and analysis module includes:

[0012] S1 extracts various event indicators recorded in the adverse event management module and counts the occurrence frequency;

[0013] S2 generates tables and charts to visualize data;

[0014] S3 first extracts features from event data, and then inputs the extracted features into the K-means clustering analysis algorithm for clustering; the algorithm groups similar events into the same group, helps identify the cause category, and calculates the cause ratio.

[0015] Furthermore, the specific process of S3 is as follows:

[0016] S31, first analyze and define the characteristics related to the causation of adverse events;

[0017] S32, performing data cleaning on the defined features;

[0018] S33, converting the categorical data features into binary values ​​through unique coding; extracting time features from the time data features, and calculating the time interval from the event processing time as a feature;

[0019] S34, standardizing all extracted features;

[0020] S35, input the standardized feature data into the K-means cluster analysis algorithm, the K-means cluster analysis algorithm divides the feature data into K clusters, each cluster represents an event cause category; and calculates the cause ratio.

[0021] Furthermore, the features defined in S31 include event type, departments involved, processing time, date and time of occurrence, and severity of the event.

[0022] Furthermore, the data cleaning includes:

[0023] 1) Handling missing values: filling or deleting missing data items;

[0024] 2) Standardized data format: ensure that all data is recorded in a unified format;

[0025] 3) Remove outliers: Identify and process abnormal data;

[0026] Furthermore, the specific steps of using the CART model to generate an analysis report include:

[0027] Step 1: Divide the feature data extracted in the data summary and analysis module into a training set and a test set in a ratio of 7:3;

[0028] Step 2: Use the training set to train the CART model and adjust the model parameters;

[0029] Step 3: Use the test set to evaluate the performance of the model and optimize the parameter settings;

[0030] Step 4: Use visualization tools to visualize the tree, display the tree structure and split nodes, and calculate the importance score of each feature to extract decision rules from the CART model;

[0031] Step 5: Summarize and count the occurrence frequency and distribution of various events in the training set; analyze the causes and their impacts based on the tree structure, feature importance scores, and cause ratios, and generate recommended measures based on the analysis results.

[0032] Furthermore, the calculation indicators of the feature importance score P include information gain IG, Gini index Gini and mean square error MSE, and the feature importance score is calculated in a weighted manner:

[0033] P=x1×IG+x2×Gini+x3×MSE

[0034] Where x1, x2 and x3 are the weights of information gain, Gini index and mean square error respectively.

[0035] Furthermore, the information gain IG is calculated as follows:

[0036] IG(T,A)=Entropy(T)-Entropy(T,A)

[0037]

[0038] Where Entropy(T) is the entropy of data set T, Entropy(T, A) is the entropy of data set T under the condition of feature A, and pi is the probability of the i-th class in data set T;

[0039] The calculation formula of the Gini index Gini is:

[0040]

[0041] In the formula, p i is the probability of the i-th class in the data set T;

[0042] The calculation formula of the mean square error MSE is:

[0043]

[0044] In the formula, y i is the actual value, is the predicted value and n is the number of samples.

[0045] Beneficial effects of the present invention:

[0046] By utilizing the management system of the present invention, medical staff can enter adverse events by logging into their accounts, and managers can read the entire process of adverse events in real time by logging into their accounts, and analyze and generate reports through cluster analysis algorithms and CART models, thereby reducing the workload of medical staff. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0048] Figure 1 It is the overall architecture diagram of the management system of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] like Figure 1 As shown, the medical quality and safety management system includes: adverse event management module, data summary and analysis module and data statistics module.

[0051] Adverse Event Management Module

[0052] It is used for frontline medical staff to input adverse events, and managers can read the whole process of adverse events in real time.

[0053] In the adverse event management module, the types of adverse events entered by medical staff include: accidental extubation, fall from bed, medication errors, skin problems, surgical safety, specimen collection, intravenous therapy, interception events and other events;

[0054] Among them, interception events are also called Clinical Near Misses, which refer to events in which one or more links in the process of receiving medical care services have errors or abnormalities, which may cause harm to the patient, but are corrected due to intentional or unintentional real-time intervention and do not cause adverse effects on the patient. The error is discovered and corrected before implementation, and the patient ultimately does not receive the wrong medical care services.

[0055] Other events: In addition to the above 7 types of adverse nursing events, there are other nursing-related injuries or potential dangerous events caused by the subjective behavioral errors of nurses, such as the mistaken injection of nutrient solution into the vein, electric burns to patients during surgery, specimen errors including test tube errors (wrong barcode) and patient errors, blood transfusion errors, puncture of blood transfusion bags, etc.

[0056] In addition, in the adverse event management module, the content of adverse events includes:

[0057] 1. Basic information of the patient

[0058] Includes: time of occurrence, patient hospital number, gender, age, and patient diagnosis (including an introduction to the consequences of accidental extubation).

[0059] 2. Situation of relevant nurses

[0060] Including: name of related personnel, education background, years of full-time experience, years of work experience, relevance, etc.

[0061] 3. Department report

[0062] For example, in the process of handling unexpected extubation events, the following are included:

[0063] 3.1 Basic information of the incident: Severity classification: 1-4 (National Standard: The latest version of the classification standard for adverse medical quality and safety events: severity is divided into 4 categories; the degree of harm caused to patients is divided into 9 levels).

[0064] 3.2 Risk assessment: age, gender, date of admission, date of catheterization, consciousness, emotional state, etc.

[0065] 3.3 Occurrence: type of catheter removed, main reason for unplanned extubation, situation of occurrence, whether there is restraint during unplanned extubation, whether it is reset, time of discovery, and most recent inspection and observation (e.g., within half an hour, within 1 hour, etc.).

[0066] 3.4 Ratio of patients in charge: total number of patients, admissions, discharges, transfers, surgeries, deliveries, and bed relocations.

[0067] 4. The situation at that time

[0068] Enter treatment measures for patients

[0069] 5. Basic information of the parties

[0070] Date of admission to the department, actual working time in the department (days), the person’s work number, the person’s name, position, age, level, education level, etc.

[0071] 6. Event Description

[0072] After selecting the above indicators, enter the event description.

[0073] 7. Analysis of the cause of the incident;

[0074] Analyze the reasons from the perspectives of people, equipment, consumables, systems or site.

[0075] 8. Improvement measures

[0076] Support uploading improvement plan files or entering improvement plans.

[0077] Data summary and analysis module

[0078] Read the indicators of each event recorded in the adverse event management module and count them separately to form a form, and use cluster analysis to help managers analyze the proportion of the causes of adverse events;

[0079] At the same time, the data summary and analysis module can also monitor and count various key quality indicators in real time, such as: the incidence of falls in hospitalized patients, the proportion of falls and injuries in hospitalized patients, the incidence of in-hospital pressure injuries of stage 2 and above in hospitalized patients, the unplanned extubation rate of catheterized patients, the unplanned extubation rate of endotracheal tubes, the unplanned extubation rate of nasogastric tubes, the unplanned extubation rate of urinary catheters, the unplanned extubation rate of central venous catheters (CVCs), and the unplanned extubation rate of peripherally inserted central venous catheters (PICCs). If the statistical data exceeds the threshold of each indicator, a red alarm reminder will be issued.

[0080] The analysis process in the data summary analysis module includes:

[0081] S1: Extract various event indicators recorded in the adverse event management module and count the occurrence frequency;

[0082] S2: Generate forms and charts. The chart types are bar charts or pie charts to achieve data visualization;

[0083] S3: First, extract features from the event data, and then input the extracted features into the K-means clustering analysis algorithm for clustering; the algorithm groups similar events into the same group, helps identify the main cause categories, and calculates the cause ratio;

[0084] The specific process of S3 is:

[0085] S31, first analyze and define the characteristics related to the causation of adverse events;

[0086] For example, the type of incident, the departments involved, the processing time, the date and time of occurrence, the severity of the incident, etc.;

[0087] S32, performing data cleaning on the defined features;

[0088] Data cleaning includes:

[0089] 1) Handling missing values: filling or deleting missing data items;

[0090] 2) Standardized data format: Ensure that all data is recorded in a unified format, such as date format, time unit, etc.;

[0091] 3) Remove outliers: Identify and process abnormal data, such as time values ​​that are outside a reasonable range;

[0092] S33, converting categorical data features (such as event type, involved departments) into binary values ​​through unique coding; extracting time features from time data features (such as time occurrence date, time, etc.), and calculating the time interval from the event processing time as a feature;

[0093] S34, all extracted features are standardized to ensure consistency in subsequent analysis and modeling.

[0094] S35, input the standardized feature data into the K-means cluster analysis algorithm, the K-means cluster analysis algorithm divides the feature data into K clusters, each cluster represents an event cause category; and calculates the cause ratio.

[0095] Data statistics module

[0096] Based on the analysis results of the data summary and analysis module, an adverse event analysis report can be generated with one click; the analysis report includes: event statistics, cause analysis and recommended measures.

[0097] Specifically, the CART (Classification and Regression Tree) model is used to generate analysis reports, automatically generate decision support information, and clearly present the results, further reducing the workload of medical staff;

[0098] The specific steps to generate an analysis report using the CART model include:

[0099] Step 1: Divide the feature data extracted in the data summary and analysis module into a training set and a test set in a ratio of 7:3;

[0100] Step 2: Use the training set to train the CART model and adjust the model parameters;

[0101] During the training process, the CART model will split according to the feature value and build a tree structure to maximize information gain or reduce the Gini index. The tuned model parameters include: maximum tree depth, minimum sample split number, etc., to prevent overfitting and improve the generalization ability of the model.

[0102] Step 3: Use the test set to evaluate the performance of the model and optimize the parameter settings.

[0103] Step 4: Use visualization tools (such as Graphviz or the plot_tree function in Scikit-learn) to visualize the tree, show the tree structure and split nodes, and calculate the importance score of each feature to extract decision rules from the CART model;

[0104] The calculation indicators of the feature importance score P include information gain IG, Gini index Gini and mean square error MSE, and the importance score of the feature is calculated in a weighted manner:

[0105] P=x1×IG+x2×Gini+x3×MSE

[0106] Where x1, x2 and x3 are the weights of information gain, Gini index and mean square error respectively.

[0107] Among them, the calculation formula of information gain IG is:

[0108] IG(T,A)=Entropy(T)-Entropy(T,A)

[0109]

[0110] Where Entropy(T) is the entropy of data set T, Entropy(T, A) is the entropy of data set T under the condition of feature A, and pi is the probability of the i-th class in data set T;

[0111] The calculation formula of Gini index is:

[0112]

[0113] Where pi is the probability of the i-th class in the dataset T.

[0114] The calculation formula for mean square error MSE is:

[0115]

[0116] In the formula, y i is the actual value, is the predicted value and n is the number of samples.

[0117] The extracted decision rules are, for example: if the event type is a medication error and the department involved is internal medicine, it may lead to a serious event.

[0118] Step 5: Summarize and count the occurrence frequency and distribution of various events in the training set; analyze the main causes and their impacts based on the tree structure, feature importance scores, and cause ratios, and generate recommended measures based on the analysis results.

[0119] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0120] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. Medical quality and safety management system, characterized by: include: Adverse event management module: used for frontline medical staff to input adverse events, and managers to read the whole process of adverse events in real time; Data summary and analysis module: read the indicators of each event recorded in the adverse event management module and count them separately to form a table, and use the K-means clustering analysis algorithm to analyze and calculate the proportion of the causes of adverse events; Data statistics module: Based on the analysis results of the data summary and analysis module, the CART model is used to generate an adverse event analysis report.

2. The medical quality and safety management system according to claim 1, characterized in that: The types of adverse events described included: unplanned extubation, falls in bed, medication errors, skin problems, procedure safety, specimen collection, intravenous therapy, and interception events.

3. The medical quality and safety management system according to claim 1, characterized in that: The analysis process in the data summary and analysis module includes: S1 extracts various event indicators recorded in the adverse event management module and counts the occurrence frequency; S2 generates tables and charts to visualize data; S3 first extracts features from event data, and then inputs the extracted features into the K-means clustering analysis algorithm for clustering; the algorithm groups similar events into the same group, helps identify the cause category, and calculates the cause ratio.

4. The medical quality and safety management system according to claim 3, characterized in that: The specific process of S3 is as follows: S31, first analyze and define the characteristics related to the causation of adverse events; S32, performing data cleaning on the defined features; S33, converting the categorical data features into binary values ​​through unique coding; extracting time features from the time data features, and calculating the time interval from the event processing time as a feature; S34, standardizing all extracted features; S35, input the standardized feature data into the K-means cluster analysis algorithm, the K-means cluster analysis algorithm divides the feature data into K clusters, each cluster represents an event cause category; and calculates the cause ratio.

5. The medical quality and safety management system according to claim 4, characterized in that: The characteristics defined in S31 include event type, departments involved, processing time, date and time of occurrence, and severity of the event.

6. The medical quality and safety management system according to claim 4, characterized in that: The data cleaning includes: 1) Handling missing values: filling or deleting missing data items; 2) Standardized data format: ensure that all data is recorded in a unified format; 3) Remove outliers: Identify and process abnormal data.

7. The medical quality and safety management system according to claim 1, characterized in that: The specific steps of using the CART model to generate an analysis report include: Step 1: Divide the feature data extracted in the data summary and analysis module into a training set and a test set in a ratio of 7:3; Step 2: Use the training set to train the CART model and adjust the model parameters; Step 3: Use the test set to evaluate the performance of the model and optimize the parameter settings; Step 4: Use visualization tools to visualize the tree, display the tree structure and split nodes, and calculate the importance score of each feature to extract decision rules from the CART model; Step 5: Summarize and count the occurrence frequency and distribution of various events in the training set; analyze the causes and their impacts based on the tree structure, feature importance scores, and cause ratios, and generate recommended measures based on the analysis results.

8. The medical quality and safety management system according to claim 7, characterized in that: The calculation indicators of the feature importance score P include information gain IG, Gini index Gini and mean square error MSE, and the importance score of the feature is calculated in a weighted manner: P=x1×IG+x2×Gini+x3×MSE Where x1, x2 and x3 are the weights of information gain, Gini index and mean square error respectively.

9. The medical quality and safety management system according to claim 8, characterized in that: The calculation formula of the information gain IG is: IG(T,A)=Entropy(T)-Entropy(T,A) Where Entropy(T) is the entropy of data set T, Entropy(T, A) is the entropy of data set T under the condition of feature A, and pi is the probability of the i-th class in data set T; The calculation formula of the Gini index Gini is: Where pi is the probability of the i-th class in the data set T; The calculation formula of the mean square error MSE is: In the formula, y i is the actual value, is the predicted value and n is the number of samples.

Images