A Smart Method for Monitoring Adverse Drug Reactions

By constructing a large sample database and using the Boyer-Moore algorithm to screen similar data and calculate the probability of adverse reactions, the problem of the inability to predict adverse drug reactions in existing technologies has been solved. This enables pre-medication analysis and judgment, reduces the probability of adverse reactions, and improves medication safety and personalized treatment.

CN123120367APending Publication Date: 2026-11-17CHONGQING UNIV CANCER HOSPITAL
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411888980.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-20
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing intelligent monitoring methods for adverse drug reactions cannot analyze and judge them before patients take medication, forcing medical staff to rely on experience to predict them, making it difficult to effectively reduce the probability of adverse reactions.

Method used

By constructing a large sample database, we acquire and store the personal data of the monitored subjects, including basic information, drug names, and adverse reaction data. We use the Boyer-Moore algorithm to match the personal basic information of patients to be given medication, filter similar data, calculate the probability of individual and overall adverse reactions, and assist medical staff in making medication decisions.

Benefits of technology

It enables the probability analysis of adverse reactions in patients before medication, reducing the probability of adverse reactions and improving the safety of medication and the effect of personalized treatment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123120367A_ABST
    Figure CN123120367A_ABST
Patent Text Reader

Abstract

This invention relates to the field of adverse drug reaction monitoring technology, specifically to an intelligent method for monitoring adverse drug reactions; comprising the following steps: acquiring personal data of the monitoring subjects and storing it in a large sample database, the personal data including basic personal information, drug name, drug dosage, and adverse reaction data; acquiring basic personal information of patients to be treated and the designated drug; identifying monitoring subjects who have used the designated drug based on the large sample database, and acquiring and storing the personal data corresponding to all monitoring subjects in a target database; extracting target item data based on the basic personal information of patients to be treated, and filtering similar personal data to the target item data from the target database; calculating the probability of a single adverse reaction based on the similar personal data; and calculating the overall adverse reaction probability by combining all the probabilities of single adverse reactions.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of adverse drug reaction monitoring technology, and in particular to an intelligent method for monitoring adverse drug reactions. Background Technology

[0002] Drug intervention is one of the important means of medicine to fight diseases, but some drugs may cause adverse reactions that are harmful to the body. Therefore, the monitoring of adverse drug reactions is particularly important. At present, although my country has initially established a drug adverse reaction monitoring and management organization system based on the national, provincial and municipal levels, it is still in the initial stage in terms of professional team building, regulations and resources.

[0003] By monitoring the effects and responses of drugs in patients' bodies, adverse reactions that may threaten patients' health can be detected and addressed in a timely manner, thereby reducing the harm suffered by patients due to drug use. At the same time, by collecting and analyzing data on adverse drug reactions, doctors can more accurately understand the efficacy and safety of drugs, thereby adjusting treatment plans and providing patients with more personalized and effective treatment.

[0004] Most existing intelligent monitoring methods for adverse drug reactions can only record the patient's actual reaction after medication. They cannot analyze and judge whether the patient will have adverse reactions before medication. As a result, when administering medication, they can only rely on the work experience of medical staff to make predictions, which cannot effectively reduce the probability of adverse reactions. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent monitoring method for adverse drug reactions, which can analyze and judge patients who need to take medication according to a set method, determine the probability of adverse reactions to such drugs, and thus assist medical staff in administering medication and reduce the probability of adverse reactions after medication.

[0006] To achieve the above objectives, the present invention employs an intelligent monitoring method for adverse drug reactions, comprising the following steps:

[0007] The personal data of the monitored subjects is acquired and stored in a large sample database. The personal data includes basic personal information, drug names, drug dosages, and adverse reaction data.

[0008] Obtain basic personal information and designated medications from patients awaiting medication.

[0009] Based on the large sample database, the drug use monitoring subjects who have used the specified drugs are identified, and the personal data corresponding to all the drug use monitoring subjects are stored in the target database;

[0010] Based on the basic personal information of the patients who need medication, target item data is extracted, and similar personal data that are similar to the target item data are filtered out according to the target database;

[0011] The probability of a single adverse reaction is calculated based on the aforementioned similar individual data.

[0012] The overall adverse reaction probability is calculated by combining all the individual adverse reaction probabilities.

[0013] The specific methods for obtaining the personal data of the monitored subjects are as follows:

[0014] The personal data of the monitored subjects is obtained by constructing a medical big data platform;

[0015] The acquired personal data is categorized and organized.

[0016] The personal data, once processed, is stored.

[0017] The specific method for extracting the target project data is as follows:

[0018] The basic personal information of the patients awaiting medication is categorized by item;

[0019] Extract data corresponding to a specified item based on the category.

[0020] The specific methods for filtering similar personal data are as follows:

[0021] The personal basic information of the medication monitoring subjects is classified according to the classification method of the personal basic information items of the patients to be prescribed medication;

[0022] Determine the types of items that require calculation;

[0023] The similar personal data are filtered based on the specified project type.

[0024] The specific method for calculating the probability of a single adverse reaction is as follows:

[0025] Obtain the similar personal data corresponding to the specified project type;

[0026] Filter out adverse similar personal data that contain the adverse reaction data from the acquired similar personal data;

[0027] The probability of a single adverse reaction is calculated based on the total number of similar personal data extracted and the number of adverse similar personal data.

[0028] The specific method for calculating the overall adverse reaction probability is as follows:

[0029] Determine the total number of project types;

[0030] Obtain the probability of each individual adverse reaction for all project types;

[0031] The calculated individual adverse reaction probabilities are then filtered.

[0032] The overall adverse reaction probability is the average of all individual adverse reaction probabilities calculated based on the total number of project types.

[0033] The specific method for classifying the basic personal information of the patients awaiting medication is as follows:

[0034] They are categorized based on the types of information in the individual's basic information, including gender, weight, height, body fat percentage, and blood type.

[0035] The specific method for filtering similar personal data based on specified project types is as follows:

[0036] The similarity data for a specified project type is set, and the similarity data includes similarity range values ​​and similarity values;

[0037] Filtering is performed based on the specified similarity data.

[0038] This invention discloses an intelligent drug adverse reaction monitoring method. First, it acquires and stores the personal data of the monitoring subjects in a large-sample database. The personal data includes basic personal information, drug name, drug dosage, and adverse reaction data. It then acquires the basic personal information of patients awaiting medication and the designated drug. Based on the large-sample database, it identifies monitoring subjects who have used the designated drug and stores the personal data corresponding to all monitoring subjects in a target database. Based on the basic personal information of the patients awaiting medication, it extracts target item data and filters similar personal data from the target database. It calculates the probability of a single adverse reaction based on the similar personal data and calculates the overall adverse reaction probability by combining all the individual adverse reaction probabilities. This method enables the analysis and judgment of patients requiring medication according to a pre-defined method, determining the probability of adverse reactions to such drugs, thereby assisting medical personnel in medication administration and reducing the probability of adverse reactions after medication. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a flowchart of an intelligent monitoring method for adverse drug reactions according to the present invention.

[0041] Figure 2 This is a flowchart of the present invention for obtaining personal data of the monitored object.

[0042] Figure 3 This is a flowchart of the present invention for extracting target project data.

[0043] Figure 4 This is a flowchart of the present invention for filtering similar personal data.

[0044] Figure 5 This is a flowchart of the present invention for filtering similar personal data.

[0045] Figure 6 This is a flowchart of the present invention for calculating the probability of a single adverse reaction.

[0046] Figure 7 This is a flowchart for calculating the overall adverse reaction probability according to the present invention. Detailed Implementation

[0047] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0048] In the description of this invention, it should be understood that "a plurality of" means two or more, unless otherwise explicitly specified.

[0049] Please see Figures 1 to 7 This invention provides an intelligent monitoring method for adverse drug reactions, comprising the following steps:

[0050] S1: Acquire the personal data of the monitored subjects and store it in a large sample database. The personal data includes basic personal information, drug names, drug dosages, and adverse reaction data.

[0051] S101: The specific methods for obtaining the personal data of the monitored subjects are as follows:

[0052] S1011: Obtain the personal data of the monitored object by constructing a medical big data platform;

[0053] S1012: Classify and organize the acquired personal data;

[0054] S1013: Store the personal data after it has been processed;

[0055] Specifically, by collecting data from the entire population nationwide, basic information about the entire population can be determined. Then, by establishing a nationwide universal medical database, a comprehensive and systematic collection of medical information and data on the healthcare needs of the entire population can be achieved, facilitating a better understanding of the healthcare situation of the population. Data collection can be conducted through a corresponding medical service data storage website. The entire medical service data storage website mainly involves aspects such as website structure design, data collection layer settings, data storage layer configuration, and system monitoring.

[0056] At the same time, the data collected from different regions can also intuitively reflect the development of medical standards and medical disparities in different places, which can better help managers to develop and improve the overall medical system.

[0057] When collecting personal data, sorting and organizing it allows for the filtering of items, removing incorrect or missing data and retaining only valid information. Specifically, when classifying and organizing personal data, it's necessary to first clarify the classification criteria, i.e., the nature of the data. Data can be classified according to its nature, such as basic information (name, age, gender, etc.), contact information (phone, email, address, etc.), occupational information (position, employer, etc.), and health information (medical history, medical examination reports, etc.). It can also be classified according to its source, such as internal system data, third-party platform data, and user-provided data, making subsequent data filtering and processing easier. Before classification and organization, the data needs to be cleaned to remove duplicates, errors, or invalid data. Secondly, the data should be formatted into a uniform format, such as date or number format, to facilitate subsequent data processing and analysis. An index should be created for the classified and organized data for quick retrieval. The index can be created based on data category tags, keywords, or attributes. Finally, personal data may change over time; therefore, it needs to be updated regularly to ensure accuracy and timeliness.

[0058] When establishing the entire medical data network, existing medical data platforms can be utilized, such as the National Medical Insurance Service Platform. This platform provides nationwide medical insurance public services, including functions such as medical insurance personal account inquiries, medical insurance payment record inquiries, and nationwide designated medical institution inquiries. It is an official channel with accurate and reliable data, serving as an important window for understanding medical insurance policies and services. The Chinese Center for Disease Control and Prevention (China CDC), as a public institution under the supervision of the National Health Commission, not only implements national-level disease prevention and control and public health technology management and services, but also provides a wealth of authoritative medical science knowledge and information through its official website. This content includes, but is not limited to, information on disease prevention, health promotion, and vaccination, which is of guiding significance to public health. Finally, the National Health Information Platform, established in accordance with the requirements of the "Healthy China 2030" Plan Outline, aims to achieve information sharing and business collaboration between medical and public health regulatory agencies. This platform, based on a comprehensive population information database, resident electronic health record database, electronic medical record database, and public health database, provides all-round health information services. Through this platform, users can learn more about disease prevention, treatment, and rehabilitation, while also contributing to the development of "Internet + medical services" in the region. The National Medical Products Administration (NMPA) is the highest regulatory body in China's food and drug sector. Its official website provides review information for all food and drugs sold domestically. Users can use this website to query detailed information on food, drugs, medical devices, cosmetics, and other products to ensure medication safety and product quality. The National Hospital Database is an information platform integrating domestic hospital data, covering multiple provinces and cities nationwide. It contains a large amount of basic information on hospitals. Users can use this database to query detailed information on hospitals in various provinces and cities, such as their nature, type, level, establishment year, outpatient volume, number of medical staff, and number of beds. This is of great significance for patients choosing medical locations, for medical practitioners conducting market research, and for governments formulating medical policies.

[0059] Adverse reaction data can be recorded according to the actual situation. For example, if fever, high blood pressure, vomiting, or diarrhea occur, the above adverse reactions can all be recorded as adverse reaction data. However, when the system makes a judgment in the later stage, the above reactions are all marked as adverse reactions, so that the data processing pressure is reduced during the subsequent screening and analysis. The detailed recording is for the purpose of subsequent analysis by medical staff on the type and specific situation of the adverse reactions.

[0060] S2: Obtain basic personal information of patients awaiting medication and the designated medication;

[0061] S3: Based on the large sample database, identify the drug use monitoring subjects who have used the specified drugs, and obtain the personal data corresponding to all the drug use monitoring subjects and store it in the target database;

[0062] S4: Extract target item data based on the personal basic information of the patient to be treated, and filter out similar personal data that are similar to the target item data according to the target database;

[0063] S401: The specific method for extracting the target project data is as follows:

[0064] S4011: Classify the basic personal information of the patients awaiting medication;

[0065] S4012: Extract data corresponding to a specified item based on the category;

[0066] S402: The specific method for filtering similar personal data is as follows:

[0067] S4021: Classify the basic personal information of the medication monitoring subjects according to the classification method of the basic personal information items of the patients to be treated;

[0068] S4022: Determine the type of project that needs to be calculated;

[0069] S4023: Filter out the similar personal data according to the specified project type;

[0070] S40111: The specific method for classifying the basic personal information of the patients awaiting medication is as follows:

[0071] Classified according to the types of information in the basic personal information, including gender, weight, height, body fat percentage, and blood type.

[0072] S40231: The specific method for filtering similar personal data based on the specified project type is as follows:

[0073] S402311: Set similar data for a specified project type, wherein the similar data includes similar range values ​​and similar values;

[0074] S402312: Filter based on the specified similar data;

[0075] Specifically, medical staff first need to input the patient's basic personal information and the specified medication. Then, based on the Boyer-Moore algorithm, a search and matching process is performed on the large sample database. The Boyer-Moore algorithm, also known as the Boyer-Moore string search algorithm (BM algorithm for short), is an efficient string matching algorithm. It is often used to find the occurrence of a pattern string in a target string (usually called the "main string"). Its core idea is to preprocess the pattern string and use the mismatch information of character comparisons to skip as many target characters as possible, thereby quickly locating possible matching positions and reducing the number of comparisons. The basic idea of ​​the BM algorithm is to start matching from the end of the pattern string and then skip some unnecessary comparisons based on the position of the mismatched characters in the pattern string. It mainly includes two key heuristic strategies: the Bad Character Rule and the Good Suffix Rule. During the matching process, when a character in the main string does not match the corresponding character in the pattern string, the character is called a "bad character". The number of positions to slide the pattern string to the right (i.e., badShift) is determined based on the last occurrence position of the bad character in the pattern string. If the bad character does not exist in the pattern string, the entire pattern string is usually moved to the right, and the character is skipped directly. When a suffix of the pattern string matches a substring of the main string, the suffix is ​​called a "good suffix". The number of positions to slide the pattern string to the right (i.e., goodShift) is determined based on the matching length and position of the good suffix in the pattern string. The good suffix rule also considers whether there are other substrings in the pattern string that match the good suffix, and whether the suffix substring of the good suffix can match the prefix substring of the pattern string.

[0076] The Boyer-Moore algorithm is implemented by constructing a bad character rule table, which records the last occurrence position of each character in the pattern string. For ASCII encoding, the size of the bad character rule table is typically 256.

[0077] Construct a suffix rule table: record the matching length of each suffix substring in the pattern string, as well as the corresponding number of sliding positions. This usually involves preprocessing the suffix and prefix arrays of the pattern string.

[0078] Start comparing the pattern string with the main string from the end of the pattern string;

[0079] When a mismatched character is encountered, the number of positions that the pattern string needs to slide to the right is calculated based on the bad character rule table and the good suffix rule table (taking the maximum value of the two).

[0080] Repeat the above process until a matching position is found or the pattern string slides out of the main string range;

[0081] When performing a search and matching process, the search for the specified drug is straightforward because the information of the specified drug is fixed. However, the personal basic information of the patient awaiting medication contains multiple items, so it is necessary to first classify the personal basic information of the patient awaiting medication and then match it according to the corresponding items in the classification. When classifying the personal basic information of the patient awaiting medication, it is necessary to first set the corresponding similar data. The similar data can be divided into similar range values ​​and similar values, depending on the actual situation. For example, when matching the height, weight, and body fat percentage of the personal basic information of the patient awaiting medication, since the data collected for height, weight, and body fat percentage are dynamic values, it is necessary to set an appropriate range when setting the similar data. For example, if the height of the patient awaiting medication is 160 cm, the similar data can be set between 158 cm and 162 cm. In this way, when performing a search, the data of the monitoring subjects who use the same drug and whose height is between 158 cm and 162 cm will be searched and matched. The specific setting range can be adjusted according to the actual situation, so that the similar personal data of a single item can be filtered out.

[0082] Since the initial collection of basic personal information was not perfect, when recording for a long period of time, the basic personal information items can be appropriately added. For example, historical medication data can be added to the search items of basic personal information to calculate the probability that the set drug will cause adverse effects on the drug to be used, as well as the probability that adverse effects will occur due to other basic personal information of the patient to be used.

[0083] S5: Calculate the probability of a single adverse reaction based on the aforementioned similar individual data;

[0084] S501: The specific method for calculating the probability of the aforementioned individual adverse reaction is as follows:

[0085] S5011: Obtain the similar personal data corresponding to the specified project type;

[0086] S5012: Filter out adverse similar personal data that contains the adverse reaction data from the acquired similar personal data;

[0087] S5013: Calculate the corresponding probability of the single adverse reaction based on the total number of similar personal data extracted and the number of adverse similar personal data;

[0088] S6: Calculate the overall adverse reaction probability by combining all the individual adverse reaction probabilities;

[0089] S601: The specific method for calculating the overall adverse reaction probability is as follows:

[0090] S6011: Determine the total number of project types;

[0091] S6012: Obtain the probability of each individual adverse reaction for all item types;

[0092] S6013: Filter the calculated individual adverse reaction probabilities;

[0093] S6014: Calculate the average of all individual adverse reaction probabilities based on the total number of project types, which is the overall adverse reaction probability.

[0094] Specifically, after screening and collecting similar personal data for the specified items, the similar personal data will contain both those who experienced adverse reactions and those who did not. By performing a second screening on all the similar personal data, the data of those who experienced adverse reactions can be extracted. Then, by combining the overall similar personal data, the probability of the corresponding single adverse reaction can be calculated. For example, if the blood type of the patient to be treated is A, and there are 3,000 patients with the same blood type who are using the same drug, and 300 patients with adverse reactions occur among the 3,000 patients, then the probability of the single adverse reaction is 10%. Similarly, the probability of the single adverse reaction corresponding to all the items to be monitored can be calculated. Then, the overall adverse reaction probability and the average probability of all the single adverse reaction probabilities are calculated based on the monitored items.

[0095] When calculating the probability of a single adverse reaction and the overall probability of an adverse reaction, the user can perform rapid calculations by setting the corresponding basic algorithm in conjunction with the extracted data. Specifically, first initialize the sum and counter: set a variable (usually called sum or total) to store the sum of the values, with an initial value of 0; set a counter (usually called count or n) to record the number of values, also with an initial value of 0; then iterate through the list of values: for each input value, perform the following operations:

[0096] Add the value to the total variable;

[0097] The counter increments by 1;

[0098] Finally, calculate the average:

[0099] After iterating through all the values, the sum is divided by the counter value to obtain the average.

[0100] Note that if the counter is 0 (i.e., no value is entered), a special value (such as NaN, null, or an exception) should usually be returned because the average cannot be calculated.

[0101] It is important to note that when selecting the search item type, it is necessary to filter according to different medications. Medical staff should select items with significant differences or practical significance for calculation. The type of search item needs to be analyzed and judged based on existing pathological cases and statistical medical data. The large sample database used in this solution can greatly assist medical staff in analyzing drug usage data, thereby determining the type of items to be analyzed, so as to more accurately calculate the probability of the patient taking the specified drug.

[0102] The above description discloses only one preferred embodiment of the present invention, and should not be construed as limiting the scope of the present invention. Those skilled in the art will understand that all or part of the processes of the above embodiments can be implemented, and equivalent changes made in accordance with the claims of the present invention are still within the scope of the invention.

Claims

1. A method for intelligent monitoring of adverse drug reactions, characterized in that, Includes the following steps: The personal data of the monitored subjects is acquired and stored in a large sample database. The personal data includes basic personal information, drug names, drug dosages, and adverse reaction data. Obtain basic personal information and designated medications from patients awaiting medication. Based on the large sample database, the drug use monitoring subjects who have used the specified drugs are identified, and the personal data corresponding to all the drug use monitoring subjects are stored in the target database; Based on the basic personal information of the patients who need medication, target item data is extracted, and similar personal data that are similar to the target item data are filtered out according to the target database; The probability of a single adverse reaction is calculated based on the aforementioned similar individual data. The overall adverse reaction probability is calculated by combining all the individual adverse reaction probabilities.

2. The intelligent monitoring method for adverse drug reactions as described in claim 1, characterized in that, The specific methods for obtaining the personal data of the monitored subjects are as follows: The personal data of the monitored subjects is obtained by constructing a medical big data platform; The acquired personal data is categorized and organized. The processed personal data is then stored.

3. The intelligent monitoring method for adverse drug reactions as described in claim 2, characterized in that, The specific method for extracting the target project data is as follows: The basic personal information of the patients awaiting medication is categorized by item; Extract data corresponding to a specified item based on the category.

4. The intelligent monitoring method for adverse drug reactions as described in claim 3, characterized in that, The specific method for filtering similar personal data is as follows: The personal basic information of the medication monitoring subjects is classified according to the classification method of the personal basic information items of the patients to be prescribed medication; Determine the types of items that require calculation; The similar personal data are filtered based on the specified project type.

5. The intelligent monitoring method for adverse drug reactions as described in claim 4, characterized in that, The specific method for calculating the probability of the individual adverse reaction is as follows: Obtain the similar personal data corresponding to the specified project type; Filter out adverse similar personal data that contain the adverse reaction data from the acquired similar personal data; The probability of a single adverse reaction is calculated based on the total number of similar personal data extracted and the number of adverse similar personal data.

6. The intelligent monitoring method for adverse drug reactions as described in claim 5, characterized in that, The specific method for calculating the overall adverse reaction probability is as follows: Determine the total number of project types; Obtain the probability of each individual adverse reaction for all project types; The calculated individual adverse reaction probabilities are then filtered. The overall adverse reaction probability is the average of all individual adverse reaction probabilities calculated based on the total number of project types.

7. The intelligent monitoring method for adverse drug reactions as described in claim 6, characterized in that, The specific method for classifying the basic personal information of the patients awaiting medication is as follows: They are categorized based on the types of information in the individual's basic information, including gender, weight, height, body fat percentage, and blood type.

8. The intelligent monitoring method for adverse drug reactions as described in claim 7, characterized in that, The specific method for filtering similar personal data based on a specified project type is as follows: The similarity data for a specified project type is set, and the similarity data includes similarity range values ​​and similarity values; Filtering is performed based on the specified similarity data.