Multi-database data management method and system for IgA nephropathy treatment

Abstract

The invention discloses a multi-database data management method and system for IgA nephropathy treatment, and relates to the technical field of data management. According to the method, multi-source clinical data is collected, a structured data set is formed based on data quality scoring and conflict resolution, the disease course stage of a patient is dynamically recognized based on a biomarker and a glomerular filtration rate, and differentiated collection, storage, synchronization and access strategies based on the disease course stage of the patient are specifically generated; multi-database collaborative management is executed, and the management effect is iteratively optimized through the data collection integrity rate, the data synchronization timeliness rate, the storage resource utilization rate and the response time; the problem that the IgA nephropathy multi-source database management and synergy efficiency is low is effectively solved, and the accuracy and efficiency of IgA nephropathy multi-source data management are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of data management technology, specifically to a multi-database data management method and system for the treatment of IgA nephropathy. Background Technology

[0002] Currently, in the clinical treatment and research of IgA nephropathy, related data are usually stored in multiple heterogeneous medical databases, such as electronic medical record databases, laboratory test databases, and pathology information databases. Due to their different design intentions, functional positioning, and technical architectures, these databases have significant differences in data structure, storage protocols, and access interfaces, forming natural data silos.

[0003] This multi-source, heterogeneous data storage situation makes it difficult to achieve efficient collaborative integration and unified management among data elements; the integrity of data collection, the rationality of storage, and the timeliness of cross-database synchronization are difficult to guarantee, thus restricting the efficiency of precision diagnosis and treatment decision support based on a complete data chain. Therefore, under current technological conditions, the lack of a mechanism to effectively coordinate these heterogeneous databases and achieve intelligent collaborative data management has become a key bottleneck in improving the data-driven diagnosis and treatment of IgA nephropathy.

[0004] To address the aforementioned shortcomings, a technical solution is provided. Summary of the Invention

[0005] The purpose of this invention is to solve the problem of low efficiency in the management and collaboration of multi-source databases for IgA nephropathy, and to propose a multi-database data management method and system for the treatment of IgA nephropathy.

[0006] The objective of this invention can be achieved through the following technical solutions: A multi-database data management method for the treatment of IgA nephropathy includes: S1. Multi-source clinical data acquisition: Acquire basic diagnostic and treatment data, laboratory test data, and pathological data of IgA nephropathy patients from multiple clinical data sources through data acquisition equipment; S2. Dynamic identification of disease stage: Based on the collected clinical data, the disease stage of IgA nephropathy patients is dynamically identified and classified through the disease stage identification engine. S3. Generation of Differentiated Management Strategies: Based on the dynamic identification results of disease stages, generate corresponding differentiated data management strategies for different disease stages and clinical needs. S4. Multi-database collaborative management execution: Based on differentiated management strategies, perform collaborative management operations on IgA nephropathy diagnosis and treatment data distributed across multiple databases; S5. Management Effectiveness Evaluation and Iteration: Evaluate the effectiveness of multi-database collaborative management and dynamically iterate and optimize differentiated management strategies based on the evaluation results.

[0007] As a further improvement of the present invention, the specific implementation process of S1 is as follows: Based on the feature identifiers of commonly used clinical databases pre-stored in multi-source clinical data acquisition devices, the database type is identified, and connections are established based on multiple built-in data interaction protocols. At the same time, invalid data entries are filtered out. Based on the clinical characteristics of IgA nephropathy, a data quality scoring framework was constructed to assess data quality; Basic diagnosis and treatment, laboratory test data, pathology information data, medication management data, and patient follow-up data are collected from electronic medical record databases, laboratory test databases, pathology information databases, medication management databases, and patient follow-up databases, respectively. A combination of timed incremental data collection and real-time triggered data collection is adopted, with routine data collected on a timed basis and new medication and follow-up records collected in real time. Configure a cross-data source conflict detection engine to resolve data conflicts based on time priority, data source authority, and clinical rationality rules for IgA nephropathy; attach a unique collection identifier and data quality score label to each collected data, remove data entries without key information and low-quality data, and form a structured dataset.

[0008] As a further improvement of the present invention, the specific operation steps of S2 are as follows: We retrieved characteristic biomarker datasets and specific supplementary data from patients with IgA nephropathy, standardized the data format, and removed outliers. We divided the continuous time-series observation windows based on the date of the first diagnosis, extracted the three most recent test results within each window, and used a weighted average method to obtain the representative values ​​of each biomarker window. The glomerular filtration rate was obtained based on serum creatinine value. A glomerular filtration rate sequence was constructed and the annual rate of change of glomerular filtration rate was obtained to determine the renal function status. Based on the proteinuria load value of the time window, a cumulative proteinuria load curve is generated and the proteinuria trend is determined; The immune activity composite index is obtained by standardizing serum IgA, C3 and Gd-IgA1, and the immune activity status is determined. A decision matrix is ​​constructed that includes renal function status, proteinuria trend, and immune activity status. The disease stage is determined by combining the matrix with a fuzzy rule base. The determination result is output and all relevant data and intermediate calculation results are stored together.

[0009] As a further improvement of the present invention, the specific operation steps of S3 include: Based on multi-source clinical data and disease stage identification results, combined with clinical guidelines and multi-center clinical research data, a three-dimensional feature and demand mapping matrix is ​​constructed, including data characteristics, clinical needs, and individual characteristics. To address the different data characteristics and clinical needs of the stable phase, slow progression phase, active progression phase, and renal failure phase, differentiated acquisition, storage, synchronization, and access strategies are developed. This includes generating acquisition cycles, data item ranges, acquisition triggering conditions, storage media types, storage levels and data retention rules, synchronization modes, synchronization frequencies and latency tolerance windows, access permissions, and operation log retention requirements for each phase. The differentiated management strategy parameters for each stage of the disease are encapsulated into a structured strategy configuration package that includes a unique version number and a disease stage identifier. After being associated with the corresponding disease stage identifier, the package is stored in the strategy management database.

[0010] As a further improvement of the present invention, the specific operation steps of S3 also include: For the stable period, a combination of scheduled automatic data collection and self-service supplementary data collection is adopted. The core indicator set is collected on a quarterly basis. The collection cycle is adjusted based on the coefficient of variation of the core indicators. The data is stored in layers and asynchronously delayed for synchronization. Access is restricted to attending physicians and department administrators. For the slow progression period, the data collection cycle is monthly, collecting core indicator sets and extended monitoring sets. If the trend deviation or any parameter deviates from the corresponding benchmark value greater than the preset maximum allowable deviation, enhanced data collection is triggered. The data is stored in a hierarchical aggregated manner and synchronized in batches on a regular basis. Access is authorized to attending physicians, department directors and clinical pharmacists. During the progress phase of the activity, a weekly data collection cycle is used to collect all indicators. When patients receive immunosuppressive therapy, the collection cycle is automatically adjusted in relation to the treatment intensity. If the full indicator data deviates from the expected efficacy curve and triggers an alert, high-speed SSD storage and real-time synchronization are used, supporting encrypted sharing by multiple roles. For patients with renal failure, dialysis and vital signs data are collected continuously and at high frequency. If any parameter exceeds the threshold, an alert is issued in real time. Redundant hot storage is used and the emergency team's permissions are bound to the system. After the emergency treatment is completed, the permissions of non-core members are automatically revoked.

[0011] As a further improvement of the present invention, the specific operation steps of S4 are as follows: Extract the collection, storage, synchronization and access parameters from the structured strategy configuration package, and after verifying the format compliance, numerical rationality and logical consistency, load them into the distributed data collaboration engine and generate a management task list associated with patient and disease course identifiers; The terminal acquisition cycle is adjusted based on the acquisition strategy. When the cycle is shortened, supplementary acquisition is triggered. The acquisition process verifies the data in real time and records the supplementary acquisition log. Data tiered migration is performed based on storage strategies. Cold storage data is generated into a summary index after lossless compression. The capacity of each tier is monitored in real time, and data degradation and archiving are performed. Based on the synchronization strategy, real-time, timed batch, asynchronous delay and strong consistency synchronization modes are configured, and transmission bandwidth is allocated based on priority, with high-priority data occupying an independent channel. All strategy loading, data collection and adjustment, storage migration and synchronous execution operations generate standardized logs including operation timestamps, types, association identifiers and results, which are stored in an independent log database.

[0012] As a further improvement of the present invention, the specific operation steps of S5 are as follows: Extract data collection completeness rate, data synchronization timeliness rate, storage resource utilization rate and query response time for each stage of the disease course within the assessment period, statistically analyze slow query information and generate a comprehensive assessment score; If the comprehensive evaluation score meets the standard, the current management strategy will be maintained; if it does not meet the standard, iterative optimization will be carried out in stages: in the collection stage, the retry mechanism and adaptation rules will be optimized for interface anomalies or format incompatibility; in the synchronization stage, data sharding or reduction of single batch processing volume will be used to solve timeout problems; in the storage stage, encrypted archiving or degradation will be performed on non-core data that has not been accessed for a long time; and in the query stage, composite indexes will be built and cache preloading rules will be configured. Based on the optimized strategy parameters, a temporary test configuration package is constructed, and typical patient data from each stage of the disease are selected for simulation. One evaluation cycle is used as the test cycle. If the comprehensive evaluation score is still less than the preset threshold after the test, the parameters of the optimization measures are back-analyzed, and the parameters that have not met the standards are readjusted until the management effect meets the standards. Once the test is passed, the structured strategy configuration package version for the corresponding disease stage is updated and the optimization content is recorded. In the next management cycle, the distributed data collaboration engine automatically loads the updated configuration package.

[0013] A second aspect of the present invention provides a multi-database data management system for the treatment of IgA nephropathy, comprising: Multi-source clinical data acquisition module: Connects to multiple databases through multi-source clinical data acquisition devices, collects multi-source data, performs data quality scoring and conflict resolution, and forms a structured dataset; Dynamic disease stage identification module: Based on structured datasets, it divides time-series observation windows and analyzes them, constructs a decision matrix and combines it with a fuzzy rule base to dynamically determine the disease stage; Differentiated management strategy generation module: Based on the three-dimensional feature and demand mapping matrix, differentiated management strategies are formulated for each stage of the disease process, and a structured strategy configuration package is generated; Multi-database collaborative management execution module: loads and verifies strategy parameters, adjusts the collection cycle, performs data hierarchical storage migration, configures multi-mode synchronization channels, records all collaborative operation logs and stores them in an independent log database; Management effectiveness evaluation and iteration module: Based on multi-dimensional evaluation indicators, it generates a comprehensive score, judges the management effectiveness status, and performs iterative optimization.

[0014] Compared with the prior art, the beneficial effects of the present invention are: This invention achieves standardized collection and fusion of multi-dimensional data from IgA nephropathy patients by constructing a multi-source clinical data acquisition device and data quality assessment framework. Based on a dynamic disease course identification engine, it comprehensively utilizes glomerular filtration rate change sequence analysis, proteinuria load accumulation curve, and immune activity composite index assessment to classify patients into disease stages. Simultaneously, it generates differentiated data management strategies for each disease stage and automatically executes collection frequency adjustment, storage level migration, and multi-database synchronization mode switching based on a distributed collaborative engine, forming a closed-loop evaluation and strategy iteration of management effectiveness. This effectively improves the completeness and accuracy of multi-source heterogeneous medical data collection and significantly enhances the automation level and execution efficiency of cross-database collaborative management. Attached Figure Description

[0015] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Example: like Figure 1 As shown, a multi-database data management method for the treatment of IgA nephropathy includes multi-source clinical data acquisition, dynamic identification of disease stages, generation of differentiated management strategies, execution of multi-database collaborative management, and evaluation and iteration of management effectiveness.

[0018] S1. Multi-source clinical data acquisition: Basic diagnostic and treatment data, laboratory test data, and pathological data of IgA nephropathy patients are acquired from multiple clinical data sources using data acquisition equipment; the specific implementation process is as follows: Through multi-source clinical data acquisition devices, the system automatically identifies the type of database to be connected by using pre-stored feature identifiers of commonly used clinical databases; it establishes stable data connections with clinical databases of different architectures based on multiple built-in data interaction protocols; and it filters out invalid data entries based on preset basic data validity verification rules. A data quality scoring framework was constructed based on the clinical characteristics of IgA nephropathy. The framework is defined as follows: ; in, Indicates the data quality score. The completeness score is calculated as the ratio of the actual number of required fields collected to the total number of required fields for IgA nephropathy. The score indicates timeliness, based on the time when the data was generated; the longer the time interval between data collection, the lower the score. The consistency score is obtained by cross-validation across data sources for overlapping data items from the same patient at the same time point. This score represents the specific score for IgA nephropathy, based on the contribution of data to the determination of the course of IgA nephropathy. The core indicators include serum IgA, complement C3, and quantitative proteinuria; 25 points are deducted for each missing core indicator. These are the influencing weighting factors for the integrity score, timeliness score, consistency score, and IgA nephropathy-specific score, respectively. Scope of multi-source clinical data sources and data collection scope: Electronic medical record database: As the core data source for basic medical data, the basic medical data collected from it specifically includes: patient demographic information, medical history information, treatment plan records, clinical symptom records and follow-up records; Laboratory Test Database: As a dedicated data source for laboratory test data, the database collects laboratory test data including routine test data and characteristic test data for IgA nephropathy. It also associates the test time, test instrument model, test personnel information, and abnormal test result annotations for each test data. Pathology Information Database: As the core data source for pathology data, the pathology data collected from it specifically includes pathological morphological data, immunofluorescence detection data, and electron microscopy data of renal biopsy tissues. It also collects pathological diagnosis conclusions, pathological slide numbers, pathological examination times, and pathologists' review opinions. Additionally, it collects Oxford classification MEST-C score data, including mesangial cell proliferation score, capillary proliferation score, segmental sclerosis score, renal tubular atrophy and interstitial fibrosis score, and crescent score. Medication Management Database: Supplement the collection of patients' medication lifecycle data, including prescription drug purchase records, drug dispensing records, medication adherence monitoring data, and adverse drug reaction records, to achieve cross-validation and data complementarity with the treatment plan records in the electronic medical record database; Patient follow-up database: Supplement the collection of dynamic data during the outpatient follow-up process, including home blood pressure monitoring records, home blood glucose monitoring records, outpatient re-examination test results, lifestyle questionnaire data and patient self-symptom score data, to improve the data coverage of the entire course of the patient's disease; Multi-source data fusion acquisition and conflict resolution: After establishing a stable connection with various clinical data sources through a preset adapter interface, the data acquisition device adopts a combination of timed incremental acquisition and real-time triggered acquisition mode: for routine data in electronic medical record databases, laboratory test databases, and pathology information databases, a timed acquisition window is set for the early morning of each day, and only data entries newly added and updated within the previous natural day are collected; for medication dispensing records in the medication management database and follow-up records in the patient follow-up database, a real-time triggered acquisition mode is adopted. When a new data record is generated, the acquisition device is triggered to perform an immediate acquisition operation through the database change log monitoring mechanism. Configure a cross-data source conflict detection engine to automatically resolve data conflicts for the same patient across different data sources. The conflict resolution rules are as follows: Time priority rule: When multiple data source records exist for the same data item, the record with the latest timestamp is used first; Data source authority rules: Establish a data source authority weight table, with the weights sorted from high to low as follows: laboratory test database, pathology information database, electronic medical record database, medication management database, and patient follow-up database; when timestamps are the same, the data source with the higher authority weight is used. Clinical rationality rules for IgA nephropathy: Construct a knowledge base of clinically reasonable ranges for core indicators of IgA nephropathy, including: reasonable ranges for serum creatinine, reasonable ranges for serum IgA, and reasonable ranges for 24-hour proteinuria quantification; When the collected data exceeds the clinically reasonable range, a manual review process is triggered, and the review results are fed back to the data quality score for iterative optimization; During the data acquisition process, the data acquisition device attaches a unique acquisition identifier to each acquired data entry; the acquisition identifier consists of the data source code, acquisition timestamp, and random sequence; at the same time, a Data Quality Score (DQS) label is attached, forming a structured data entry with quality labeling; Based on preset basic data validity verification rules, data entries lacking key information, such as those without a unique patient identifier or collection time, are removed. At the same time, data entries with data quality scores below the preset range are removed, and data entries with data quality scores within the preset range are marked as pending review. This forms a preliminary multi-source clinical dataset.

[0019] S2. Dynamic Identification of Disease Stages: Based on the collected clinical data, the disease stage of IgA nephropathy patients is dynamically identified and classified using a disease stage identification engine; the specific implementation process is as follows: The system retrieves characteristic biomarker datasets for IgA nephropathy patients from the laboratory testing database via a data interface. These datasets include serum creatinine, serum blood urea nitrogen, 24-hour proteinuria, urine red blood cell count, serum IgA level, and serum complement C3 level. Additionally, it retrieves galactose-deficient IgA1 level, soluble transferrin receptor level, and urine kidney injury molecule-1 level as supplementary data for IgA nephropathy-specific biomarkers. During the extraction process, the data format is standardized, and non-uniform unit data output by different testing instruments are converted into preset standard units. At the same time, abnormal outliers that exceed the clinically reasonable range are removed. Using the date of the patient's first diagnosis of IgA nephropathy as the baseline time point, continuous time-series observation windows were divided with a fixed span. For each time-series observation window, the three most recent test results of the patient within the window were extracted, and the window representative value of each biomarker was calculated using a weighted average method. The weighting coefficient of each biomarker was determined based on the time interval between the test time and the end of the window, with a higher weight for shorter time intervals. Serum creatinine values ​​based on each time-series observation window Through formula The glomerular filtration rate was calculated. ,in, This represents a sex-specific constant, with a value of 0.9 for males and 0.7 for females. Indicates age, This represents the gender coefficient constant, with a value of 1.000 for males and 1.018 for females; Glomerular filtration rate (GFR) was extracted from continuous time-series observation windows to construct a GFR sequence. The GFR sequences from adjacent time-series observation windows were then compared to obtain the change in GFR, thus generating a GFR change sequence. The Daubechies-4 wavelet basis was used to perform three-level wavelet decomposition on the glomerular filtration rate change sequence to separate the trend component A3, the medium-term fluctuation component D3, the short-term fluctuation component D2, and the noise component D1. The trend component A3 reflects the long-term evolution trend of renal function, the intermediate fluctuation component D3 reflects the intermediate changes related to seasonality or treatment cycle, and the short-term fluctuation component D2 reflects test error and physiological fluctuation. Based on the linear regression slope of the trend component A3 within all time series observation windows over a year, the annual... rate of change; If the year If the rate of change is greater than a preset range, it is marked as a stable state; if the annual rate of change is greater than a preset range, it is marked as a stable state. If the rate of change is within a preset range, it is marked as a slow decline; if the annual... If the rate of change is less than a preset range, it is marked as a rapid progress state; for example, in the current year... A rate of change < -5 mL / min / 1.73 m² / year is marked as a rapid progression state; The peak value and mean value of 24-hour proteinuria quantification within each time series observation window are obtained, and the proteinuria load value of the time series observation window is obtained by combining the weighted fusion formula. The cumulative proteinuria load values ​​of each time-series observation window are summed based on the time sequence to obtain the cumulative proteinuria load curve. The horizontal axis of the curve is the time-series window number, and the vertical axis is the cumulative proteinuria load value. Based on the slope of the cumulative proteinuria load curve between adjacent time-series observation windows, when the slope of the cumulative proteinuria load curve in two consecutive time-series observation windows is greater than the slope of the previous adjacent time-series observation window, it is judged as a worsening trend of proteinuria; when the slope of the cumulative proteinuria load curve in two consecutive time-series observation windows is less than the preset threshold, it is judged as stable proteinuria; the rest are judged as a fluctuating state of proteinuria. The standardized IgA value is obtained by comparing the difference between the mean serum IgA of patients and the mean serum IgA of healthy individuals with the standard deviation of serum IgA of healthy individuals. The C3 standardized value is obtained by comparing the difference between the mean serum C3 of healthy individuals and the serum C3 of patients with the standard deviation of serum C3 of healthy individuals. The standardized value of Gd-IgA1 is obtained by comparing the difference between the mean Gd-IgA1 of patients and the mean Gd-IgA1 of healthy individuals with the standard deviation of Gd-IgA1 of healthy individuals. Based on the standardized values ​​of IgA, C3, and Gd-IgA1, and combined with a weighted formula, the composite index of immune activity in IgA nephropathy was obtained. When the composite index of immune activity in IgA nephropathy is greater than the preset range, it is determined to be an active immune state; when the composite index of immune activity in IgA nephropathy is within the preset range, it is determined to be a borderline immune state; when the composite index of immune activity in IgA nephropathy is less than the preset range, it is determined to be a quiescent immune state. Construct a decision matrix for disease progression to comprehensively determine disease progression stages: Define input variables: renal function status, proteinuria trend, and immune activity status; Define a fuzzy rule base: (a) When renal function is stable, proteinuria trend is stable, and immune activity is quiescent, the current stage of the disease is determined to be the stable phase. (b) When renal function is in a state of slow decline, proteinuria trend is in a state of fluctuation, and immune activity is in a state of quiescence or borderline, the current stage of disease is determined to be the slow progression stage. (c) When renal function is in a rapidly progressive state, proteinuria trend is in a fluctuating state, and immune activity is in an active state, the current stage of the disease is determined to be the active progressive stage; (d) When the glomerular filtration rate is greater than the preset threshold, the current stage of the disease is determined to be the renal failure stage; Based on renal function status, proteinuria trend and immune activity status, after normalization, the confidence score for disease stage determination is obtained by combining a weighted formula; when multiple rules are activated at the same time, the disease stage with the highest confidence score for disease stage determination is taken as the output.

[0020] The decision result based on the disease stage decision matrix outputs the final disease stage of the target patient and stores all data of each decision result and intermediate calculation results of the matrix.

[0021] S3. Generation of Differentiated Management Strategies: Based on the dynamic identification results of disease stages, corresponding differentiated data management strategies are generated for different disease stages, taking into account data characteristics and clinical needs. The specific implementation process is as follows: Based on multi-source clinical data and disease stage identification results, combined with clinical guidelines for IgA nephropathy and multi-center clinical research data, a three-dimensional feature-demand mapping matrix was constructed: Data feature dimensions include data update frequency, data sensitivity, and data correlation strength; Clinical needs include real-time monitoring, retrospective analysis, and multi-center collaboration. Individual characteristics dimension: including patient baseline data, genetic risk characteristics, treatment response history, and comorbidities; Differentiated control strategies based on disease stage: For the stable period, each quarter is set as a collection cycle, and the data items to be collected are limited to the core indicator set, including serum creatinine value, glomerular filtration rate and 24-hour proteinuria quantitative value. Other non-core indicators are suspended from routine collection. The system obtains the sequence of the three most recent valid observations of the core indicator set and calculates the coefficient of variation based on the ratio of the standard deviation to the mean of the valid observations. If the coefficient of variation is lower than the preset threshold for two consecutive collection periods, the collection period is automatically extended to once every six months. If the coefficient of variation is higher than the preset threshold for two consecutive collection periods, the collection period is shortened to once a month. The collection method is a combination of timed automatic collection and lightweight supplementary collection based on patient self-reporting. The current quarter's data is stored in the SATA hard drive warm storage layer, while historical data older than one year is migrated to the low-cost object storage cold storage layer after lossless compression. The hot storage layer only retains the summary index of the cold storage data, including the data collection time, the average of core indicators, and the data storage path. Configure the data synchronization strategy during the stable period as an asynchronous delayed synchronization mode, set the synchronization frequency to once a week, the synchronization time window to early Sunday morning, and the synchronization delay tolerance window between multiple databases to 72 hours; the scope of synchronized data is limited to the update records of core indicators, and the synchronization of non-core indicators is suspended, with incremental synchronization only triggered as needed during retrospective analysis; access is only authorized to the attending physician of the patient and the department data administrator, and access operation logs are retained; For the slow progression phase, a monthly data collection cycle is set, with data consisting of a core indicator set plus an extended monitoring set. The extended monitoring set includes blood pressure values, medication adherence data, and complication indicators; data is collected using a formula. The trend deviation is calculated, where, This indicates the total number of data items collected. Indicates the first The benchmark value of the indicator. Indicates the first The latest measured value of the indicator; Indicates the first The influence weighting factor of each indicator; When the trend deviation is greater than the preset threshold, or the deviation of any collected data item parameter from the corresponding benchmark value is greater than the preset maximum allowable deviation, enhanced collection is triggered, and the next collection point is brought forward. Data from the slow-progressing period of 2 years is stored in the SAS hard disk warm storage layer, supporting minute-level data retrieval; a data tiered storage mode is adopted, with core indicator data retaining its original accuracy and extended indicator data being stored based on daily averages and peak values; Configure the data synchronization strategy for the slow progress period as a scheduled batch synchronization mode, with a synchronization frequency of midnight on the last working day of each month and a synchronization delay tolerance window set to 24 hours; authorize attending physicians, department directors, and clinical pharmacists to access the data and retain access operation logs. During the active phase, the basic data collection cycle is set to once a week. When patients receive immunosuppressive therapy, the collection cycle is automatically adjusted to once every 3 days based on the treatment intensity. The collected data items are a full set of indicators, including the core indicator set, the extended monitoring set, and the treatment response dataset. The treatment response dataset includes the blood concentration of immunosuppressants, treatment side effect scores, and the degree of symptom improvement. Based on the preset expected efficacy curve, if the actual collected value of any parameter in the full set of indicators deviates from the maximum allowable deviation range of the expected efficacy curve at multiple consecutive collection points, an early warning will be triggered and the medical team will be notified. All data during the event's progress will be stored on high-speed SSD storage media; original test data and original records from testing instruments will be retained. Configure the data synchronization strategy during the activity progress period as real-time synchronization mode, enable the database change data capture mechanism, and synchronize all related databases after any database data is written or updated; the synchronization priority is sorted based on treatment response data, core indicator data, and extended monitoring data, with high-priority data occupying an independent synchronization channel; access is authorized to attending physicians, department directors, clinical pharmacists, and members of multi-center research collaboration groups, and encrypted transmission and real-time data sharing are supported. For patients in the renal failure stage, data collection is set to continuous monitoring mode, and dialysis-related parameters and key vital signs are continuously and frequently collected. Dialysis-related parameters include blood flow, dialysate flow, and ultrafiltration volume, while key vital signs include heart rate, blood pressure, and blood oxygen saturation. At the same time, electrolyte levels and acid-base balance indicators are collected. Set a life support safety threshold. If any parameter exceeds the life support safety threshold or the signal is interrupted, an early warning will be triggered in real time and the medical team will be notified. At the same time, the system will automatically switch to the backup acquisition link and generate an event timeline. A redundant hot storage mode is adopted, with data stored simultaneously on multiple physically isolated SSD storage nodes, and data is backed up in real time between nodes; the original high-frequency data is retained for 72 hours, and subsequent data is automatically converted into 5-minute aggregated data and retained for a 3-month hot storage period before being migrated to the archive storage layer; access permissions are bound to the current emergency team list, and the permissions of non-core members are automatically revoked after the emergency is completed. Differentiated management strategy parameters for each stage of the disease process are encapsulated into structured strategy configuration packages, which include disease stage identifiers, collection strategies, storage strategies, synchronization strategies, and access strategies. Each structured strategy configuration package is assigned a unique strategy version number and automatically stored in the strategy management database after being associated with the corresponding disease stage identifier.

[0022] S4. Multi-database collaborative management execution: Based on a differentiated management strategy, collaborative management operations are performed on IgA nephropathy diagnosis and treatment data distributed across multiple databases; the specific implementation process is as follows: Configuration package extraction and loading: Based on the structured strategy configuration package generated by S3, the acquisition strategy parameters, storage strategy parameters, synchronization strategy parameters and access parameters corresponding to each stage of the disease process are extracted; the acquisition strategy parameters include acquisition period, data items and verification threshold; the storage strategy parameters include storage level, media type and compression ratio; the synchronization strategy parameters include synchronization mode, latency tolerance time and priority. The validity of each extracted parameter is validated, including the compliance of the parameter format, the rationality of the numerical range, and the logical consistency between the parameters of each strategy. After the validation is passed, the parameter set is loaded into the distributed data collaboration engine to generate a list of management execution tasks corresponding to each stage of the disease. The task list is associated with the patient's unique identifier and the disease stage identifier. Data acquisition frequency adjustment: Based on the loaded acquisition strategy parameters, the distributed data collaboration engine sends acquisition cycle adjustment instructions to the acquisition terminals corresponding to each clinical data source. The instructions include the target acquisition cycle, triggering method, and verification rules. After receiving the instructions, the acquisition terminal modifies the trigger cycle of the scheduled acquisition task through the local configuration update interface, and at the same time associates the patient's disease stage tag to ensure that the data acquisition tasks for patients at different disease stages are executed in a classified manner. For scenarios with shortened collection cycles, a supplementary collection operation is automatically triggered after the command is issued. The collection scope is the set of missing core indicators within the cycle change interval. After the supplementary collection is completed, it is integrated with the historical collection data by sorting by timestamp. During the data collection process, data verification rules are executed in real time. The newly collected data is compared with the historical average. If the data exceeds the preset deviation threshold, a supplementary data collection reminder is triggered according to the strategy configuration. The supplementary data collection record is synchronously written to the collection log, including the collection time, data items, deviation value and supplementary data collection result. Data storage hierarchy migration and optimization: Based on storage strategy parameters, the storage management device performs data tiered migration operations: For data that needs to be migrated to cold storage, it is first compressed using the LZ4 lossless compression algorithm. After compression, it is written to the object storage cold storage medium, and at the same time, a summary index is generated in the SATA hard drive warm storage layer. The index includes the data's unique identifier, collection time, average core indicators, and cold storage path; For data that needs to be migrated to hot storage, it is sorted according to the data access priority queue, and each data item is preloaded from warm and cold storage to the SSD hot storage cache in turn. Capacity monitoring is performed on data at each storage level. When the cold storage utilization rate exceeds the preset threshold, historical data archiving and cleanup is triggered, retaining core data within the last 5 years, while non-core data that has exceeded the expiration date is migrated to offline storage media after being encrypted and archived. The warm and hot storage levels monitor the data access frequency in real time and automatically downgrade non-core data that has not been accessed for 3 consecutive months to the next lower storage level. Multi-database synchronization mode switching and collaboration: The synchronization management device configures the synchronization channels between databases based on synchronization policy parameters: For real-time synchronization mode, a database change data capture listener is enabled to capture data write, update, or delete events in real time. The change events and complete data content are broadcast to other related databases through a high-priority message queue. After synchronization is completed, confirmation responses are received from each database. For scheduled batch synchronization mode, the trigger time window and single batch processing limit of the scheduled batch synchronization task are configured. During synchronization, the MD5 checksum of the data to be synchronized is extracted first. Incremental change data is filtered by checksum comparison. After batch transmission is completed, a full checksum comparison is performed. If there is a discrepancy, incremental retry is triggered. For asynchronous delayed synchronization mode, data change records are accumulated within the delay tolerance window and synchronized in batches based on a fixed periodic time; during the synchronization process, core indicator data is transmitted first, and non-core indicator data is paused for synchronization. For strong consistency synchronization mode, distributed transactions are executed based on the two-phase commit protocol to ensure that data writing is confirmed only after all databases are synchronized. If synchronization fails, subsequent write operations are blocked and an alarm is triggered. During the synchronization process, transmission bandwidth is allocated based on the priority configured by the strategy, and high-priority data occupies an independent transmission channel. All collaborative management operations, including strategy loading, data collection and adjustment, storage migration, and synchronous execution, generate standardized operation logs. The logs include operation timestamps, operation types, associated patient identifiers, disease stages, operation parameters, and execution results, and are stored in a separate log database.

[0023] S5. Management Effectiveness Evaluation and Iteration: Evaluate the effectiveness of multi-database collaborative management and dynamically iterate and optimize differentiated management strategies based on the evaluation results; the specific implementation process is as follows: The data collection completeness rate is obtained by comparing the total number of data items that should be collected at each stage of the disease course within the assessment period with the number of data items that were actually successfully collected by extracting data collection logs. The total number of data change records within the evaluation period is extracted from the data synchronization log and compared with the number of change records that are synchronized within the preset time limit to obtain the data synchronization timeliness rate. The storage resource utilization rate is obtained by comparing the pre-allocated storage space quota of each storage level with the actual used storage space obtained by the storage management device. The response time of all data query requests within the evaluation period is extracted by querying the standardized operation logs. The percentage of query requests in each interval is calculated based on the preset response time interval. If the response time is greater than the maximum value of the preset response time interval, it is judged as a slow query, and the corresponding query data type, associated disease stage and query user role are recorded. Based on the data collection completeness rate, data synchronization timeliness rate, storage resource utilization rate, and response time, a comprehensive evaluation score is obtained by combining the weighted fusion formula; When the comprehensive evaluation score is greater than the preset threshold, it is determined that the management effect meets the standard, and the current management strategy is maintained; when the comprehensive evaluation score is less than the preset threshold, it is determined that the management effect does not meet the standard; the policy iteration optimization process is executed; Compare the data collection completeness rate, data synchronization timeliness rate, storage resource utilization rate, and response time with their corresponding preset thresholds respectively, and perform policy iteration optimization based on the comparison results: For the data collection process: Optimize the improvement requirements existing in the collection link, including: If the main failure reason is interface connection exception, increase the collection retry times and optimize the collection task execution time window; if the main failure reason is data format incompatibility, update the protocol parsing rules of the data source adaptation layer and add an automatic conversion protocol for abnormal format data; For the data synchronization process: Optimize the improvement requirements existing in the synchronization link, including: For timeouts caused by large-volume data, implement a data sharding transmission strategy, split a single large-volume data into several shards for synchronization, and reorganize in the target database after synchronization; for batch synchronization task timeouts, reduce the amount of data processed in a single batch based on the timeout frequency; For the data storage process: Optimize the improvement requirements existing in the storage link, including: Migrate non-core data in cold storage that has been created for more than 5 years and has no access records in the last 3 years to an offline tape library for archiving after AES-256 encryption; for non-core data in warm and hot storage that has no access records for 3 consecutive months, automatically downgrade it to the cold storage layer and update the data downgrade trigger conditions in the storage policy parameters; For the query response process: Optimize the improvement requirements existing in the query performance link, including: Based on the core index data involved in the high-frequency query conditions in the corresponding record data of slow queries, construct a composite index; for high-frequency query data in the active progress stage and renal failure stage, configure a query result cache preloading rule, and the cache validity period is adjusted based on the data update frequency; Based on the optimized policy parameters, construct a temporary test configuration package, select typical patient data at each disease course stage for simulation operation, use one evaluation period as the test period, if after the test ends, the comprehensive evaluation score is still less than the preset threshold, then retrospectively analyze the parameters of the optimization measures, and readjust the unqualified parameters until the management effect meets the standard; After the test is passed, the iteratively optimized strategy parameters will be synchronously updated to the strategy configuration package of the corresponding disease stage, the configuration package version number will be updated, and the optimization and adjustment content will be recorded. At the beginning of the next management execution cycle, the distributed data collaboration engine will automatically load the updated strategy configuration package to replace the original configuration package to perform management operations.

[0024] A multi-database data management system for the treatment of IgA nephropathy, comprising: Multi-source clinical data acquisition module: Through multi-source clinical data acquisition devices including data source identification, data adaptation, preliminary verification and quality assessment units, it pre-stores clinical database feature identifiers and various data interaction protocols, connects to multiple data sources, and collects basic diagnosis and treatment, laboratory, pathology, medication and follow-up data; it adopts a collection mode that combines timed incremental and real-time triggering, and resolves cross-data source conflicts through time priority, data source authority and clinical rationality rules, attaches a unique collection identifier and quality score label to each data, and eliminates invalid and low-quality data to form a structured dataset; The dynamic identification module for disease stages retrieves patient-specific biomarkers and supplementary data, performs format standardization and outlier removal, divides the time-series observation window based on the date of initial diagnosis, derives glomerular filtration rate and annual rate of change through serum creatinine values, analyzes the composite index of proteinuria trend and immune activity, constructs a decision matrix containing renal function status, proteinuria trend and immune activity status, dynamically determines the disease stage by combining a fuzzy rule base, and associates and stores relevant data and intermediate calculation results. Differentiated management strategy generation module: Based on data characteristics, clinical needs and individual characteristics, a three-dimensional feature-demand mapping matrix is ​​constructed to formulate differentiated strategies for different disease stages; the strategy parameters are encapsulated into a structured strategy configuration package with a unique version number and stored in the strategy management database; Multi-database collaborative management execution module: Extracts the collection, storage, synchronization and access parameters from the structured strategy configuration package, loads them into the distributed data collaboration engine after validity verification, generates a management task list associated with patient and disease course identifiers; adjusts the collection cycle and triggers supplementary collection, performs data hierarchical migration and capacity optimization, configures multi-mode synchronization channels, allocates transmission bandwidth based on priority, records standardized logs of all collaborative operations and stores them in an independent log database.

[0025] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A multi-database data management method for the treatment of IgA nephropathy, characterized in that, include: S1. Multi-source clinical data acquisition: Acquire basic diagnostic and treatment data, laboratory test data, and pathological data of IgA nephropathy patients from multiple clinical data sources through data acquisition equipment; S2. Dynamic identification of disease stage: Based on the collected clinical data, the disease stage of IgA nephropathy patients is dynamically identified and classified through the disease stage identification engine. S3. Generation of Differentiated Management Strategies: Based on the dynamic identification results of disease stages, generate corresponding differentiated data management strategies for different disease stages and clinical needs. S4. Multi-database collaborative management execution: Based on differentiated management strategies, perform collaborative management operations on IgA nephropathy diagnosis and treatment data distributed across multiple databases; S5. Management Effectiveness Evaluation and Iteration: Evaluate the effectiveness of multi-database collaborative management and dynamically iterate and optimize differentiated management strategies based on the evaluation results.

2. The multi-database data management method for IgA nephropathy treatment according to claim 1, characterized in that, The specific implementation process of S1 is as follows: Based on the feature identifiers of commonly used clinical databases pre-stored in multi-source clinical data acquisition devices, the database type is identified, and connections are established based on multiple built-in data interaction protocols. At the same time, invalid data entries are filtered out. Based on the clinical characteristics of IgA nephropathy, a data quality scoring framework was constructed to assess data quality; Basic diagnosis and treatment, laboratory test data, pathology information data, medication management data, and patient follow-up data are collected from electronic medical record databases, laboratory test databases, pathology information databases, medication management databases, and patient follow-up databases, respectively. A combination of timed incremental data collection and real-time triggered data collection is adopted, with routine data collected on a timed basis and new medication and follow-up records collected in real time. Configure a cross-data source conflict detection engine to resolve data conflicts based on time priority, data source authority, and clinical rationality rules for IgA nephropathy; attach a unique collection identifier and data quality score label to each collected data, remove data entries without key information and low-quality data, and form a structured dataset.

3. The multi-database data management method for IgA nephropathy treatment according to claim 1, characterized in that, The specific operation steps of S2 are as follows: We retrieved characteristic biomarker datasets and specific supplementary data from patients with IgA nephropathy, standardized the data format, and removed outliers. We divided the continuous time-series observation windows based on the date of the first diagnosis, extracted the three most recent test results within each window, and used a weighted average method to obtain the representative values ​​of each biomarker window. The glomerular filtration rate was obtained based on serum creatinine value. A glomerular filtration rate sequence was constructed and the annual rate of change of glomerular filtration rate was obtained to determine the renal function status. Based on the proteinuria load value of the time window, a cumulative proteinuria load curve is generated and the proteinuria trend is determined; The immune activity composite index is obtained by standardizing serum IgA, C3 and Gd-IgA1, and the immune activity status is determined. A decision matrix is ​​constructed that includes renal function status, proteinuria trend, and immune activity status. The disease stage is determined by combining the matrix with a fuzzy rule base. The determination result is output and all relevant data and intermediate calculation results are stored together.

4. The multi-database data management method for IgA nephropathy treatment according to claim 1, characterized in that, The specific operation steps of S3 include: Based on multi-source clinical data and disease stage identification results, combined with clinical guidelines and multi-center clinical research data, a three-dimensional feature and demand mapping matrix is ​​constructed, including data characteristics, clinical needs, and individual characteristics. To address the different data characteristics and clinical needs of the stable phase, slow progression phase, active progression phase, and renal failure phase, differentiated acquisition, storage, synchronization, and access strategies are developed. This includes generating acquisition cycles, data item ranges, acquisition triggering conditions, storage media types, storage levels and data retention rules, synchronization modes, synchronization frequencies and latency tolerance windows, access permissions, and operation log retention requirements for each phase. The differentiated management strategy parameters for each stage of the disease are encapsulated into a structured strategy configuration package that includes a unique version number and a disease stage identifier. After being associated with the corresponding disease stage identifier, the package is stored in the strategy management database.

5. A multi-database data management method for the treatment of IgA nephropathy according to claim 4, characterized in that, The specific operation steps of S3 also include: For the stable period, a combination of scheduled automatic data collection and self-service supplementary data collection is adopted. The core indicator set is collected on a quarterly basis. The collection cycle is adjusted based on the coefficient of variation of the core indicators. The data is stored in layers and asynchronously delayed for synchronization. Access is restricted to attending physicians and department administrators. For the slow progression period, the data collection cycle is monthly, collecting core indicator sets and extended monitoring sets. If the trend deviation or any parameter deviates from the corresponding benchmark value greater than the preset maximum allowable deviation, enhanced data collection is triggered. The data is stored in a hierarchical aggregated manner and synchronized in batches on a regular basis. Access is authorized to attending physicians, department directors and clinical pharmacists. During the progress phase of the activity, a weekly data collection cycle is used to collect all indicators. When patients receive immunosuppressive therapy, the collection cycle is automatically adjusted in relation to the treatment intensity. If the full indicator data deviates from the expected efficacy curve and triggers an alert, high-speed SSD storage and real-time synchronization are used, supporting encrypted sharing by multiple roles. For patients with renal failure, dialysis and vital signs data are collected continuously and at high frequency. If any parameter exceeds the threshold, a real-time warning is issued. Redundant hot storage is used and the emergency team's permissions are bound to the system. After the emergency treatment is completed, the permissions of non-core members are automatically revoked.

6. A multi-database data management method for the treatment of IgA nephropathy according to claim 1, characterized in that, The specific operation steps of S4 are as follows: Extract the collection, storage, synchronization and access parameters from the structured strategy configuration package, and after verifying the format compliance, numerical rationality and logical consistency, load them into the distributed data collaboration engine and generate a management task list associated with patient and disease course identifiers; The terminal acquisition cycle is adjusted based on the acquisition strategy. When the cycle is shortened, supplementary acquisition is triggered. The acquisition process verifies the data in real time and records the supplementary acquisition log. Data tiered migration is performed based on storage strategies. Cold storage data is generated into a summary index after lossless compression. The capacity of each tier is monitored in real time, and data degradation and archiving are performed. Based on the synchronization strategy, real-time, timed batch, asynchronous delay and strong consistency synchronization modes are configured, and transmission bandwidth is allocated based on priority, with high-priority data occupying an independent channel. All strategy loading, data collection and adjustment, storage migration and synchronous execution operations generate standardized logs including operation timestamps, types, association identifiers and results, which are stored in an independent log database.

7. A multi-database data management method for the treatment of IgA nephropathy according to claim 1, characterized in that, The specific operation steps of S5 are as follows: Extract data collection completeness rate, data synchronization timeliness rate, storage resource utilization rate and query response time for each stage of the disease course within the assessment period, statistically analyze slow query information and generate a comprehensive assessment score; If the comprehensive evaluation score meets the standard, the current management strategy will be maintained; if it does not meet the standard, iterative optimization will be carried out in stages: in the collection stage, the retry mechanism and adaptation rules will be optimized for interface anomalies or format incompatibility; in the synchronization stage, data sharding or reduction of single batch processing volume will be used to solve timeout problems; in the storage stage, encrypted archiving or degradation will be performed on non-core data that has not been accessed for a long time; and in the query stage, composite indexes will be built and cache preloading rules will be configured. Based on the optimized strategy parameters, a temporary test configuration package is constructed, and typical patient data from each stage of the disease are selected for simulation. One evaluation cycle is used as the test cycle. If the comprehensive evaluation score is still less than the preset threshold after the test, the parameters of the optimization measures are back-analyzed, and the parameters that have not met the standards are readjusted until the management effect meets the standards. Once the test is passed, the structured strategy configuration package version for the corresponding disease stage is updated and the optimization content is recorded. In the next management cycle, the distributed data collaboration engine automatically loads the updated configuration package.

8. A system applied to the multi-database data management method for IgA nephropathy treatment according to any one of claims 1-7, comprising: Multi-source clinical data acquisition module: Connects to multiple databases through multi-source clinical data acquisition devices, collects multi-source data, performs data quality scoring and conflict resolution, and forms a structured dataset; Dynamic disease stage identification module: Based on structured datasets, it divides time-series observation windows and analyzes them, constructs a decision matrix and combines it with a fuzzy rule base to dynamically determine the disease stage; Differentiated management strategy generation module: Based on the three-dimensional feature and demand mapping matrix, differentiated management strategies are formulated for each stage of the disease process, and a structured strategy configuration package is generated; Multi-database collaborative management execution module: loads and verifies strategy parameters, adjusts the collection cycle, performs hierarchical data storage migration, configures multi-mode synchronization channels, records all collaborative operation logs and stores them in an independent log database; Management effectiveness evaluation and iteration module: Based on multi-dimensional evaluation indicators, it generates a comprehensive score, judges the management effectiveness status, and performs iterative optimization.

Images