Search

CN-121999962-A - Clinical medication data management system based on DRGs

CN121999962ACN 121999962 ACN121999962 ACN 121999962ACN-121999962-A

Abstract

The invention relates to the technical field of medical information and discloses a clinical medication data management system based on DRGs. The system comprises a data acquisition module, a mode construction module, a deviation identification module, a cause association module and a strategy generation module. The system constructs a typical medication pattern map as a behavior benchmark by extracting and integrating the medication usage details and medication time track data under each DRG group. In real-time monitoring, the system performs multidimensional comparison on the new case data and the map, and identifies abnormal medication sequence, dosage and cost. The system intelligently attributing deviation based on clinical medication knowledge graph, traversing medicine interaction, treatment specification and cost influence path, forming a structured cause association network, and generating medication adjustment suggestion according to the result. The scheme realizes closed-loop management from clinical medication behavior modeling to deviation depth analysis and intervention strategy generation, and improves the fineness of medication monitoring and the effectiveness of decision support.

Inventors

  • SU DAN

Assignees

  • 常州市第二人民医院
  • 医顺通信息科技(江苏)有限公司

Dates

Publication Date
20260508
Application Date
20260130

Claims (10)

  1. 1. A DRGs-based clinical medication data management system, the system comprising: The data acquisition module is used for extracting a structured medication record set corresponding to each group from an electronic medical record database of a target medical institution according to the diagnosis related group identification, wherein the structured medication record set comprises medicine use detail data, medication time track data and associated medical expense item data; the mode construction module is used for carrying out joint analysis on the drug use detail data and the drug use time track data to construct a typical drug use mode map of each case group; the deviation recognition module is used for carrying out multidimensional comparison on the new case medication data acquired in real time and the typical medication pattern map, recognizing medication sequence deviation, dose abnormal deviation and cost increase abnormal deviation, and generating a medication deviation report; The factor association module is used for traversing a medicine interaction path, a treatment standard path and a cost influence factor path which are associated with the deviation item in a preset clinical medication knowledge graph based on the medication deviation report to generate a deviation factor association network; the strategy generation module is used for searching a preset medication optimization rule base according to the deviation factor association network, obtaining a medication adjustment suggestion set aiming at a specific deviation type through matching, and sending the medication adjustment suggestion set to a clinical decision support terminal.
  2. 2. The DRGs-based clinical medication data management system of claim 1, wherein the pattern construction module performs a joint analysis of the medication usage detail data and medication time trace data to construct a typical medication pattern map, comprising: the typical medication pattern map comprises a medicine use sequence relation, a dose adjustment node and a charge accumulation track; performing medicine classification coding mapping processing on the medicine use detail data, and mapping various medicine names into standard medicine classification codes to form coded medicine use sequences; performing time window division processing on the medication time track data, and dividing the whole medication time track into a plurality of continuous treatment stage time windows according to treatment stages; In each treatment stage time window, carrying out frequent pattern mining processing on the coded medicine use sequence, and identifying a medicine combination pattern which appears together with high frequency in the treatment stage time window as a typical medicine combination of a treatment stage; Connecting typical medicine combinations identified in each treatment stage according to the time flow direction according to the sequence of the time window, and marking the adding and deleting change conditions of medicine use among stages on the connecting edge to form a preliminary medication mode directed graph; And extracting medicine cost data in each treatment stage time window from the medical cost item data, calculating cost contribution degree of the typical medicine combination, marking the cost contribution degree as a weight attribute on a corresponding node in the medication pattern directed graph, and completing construction of the typical medication pattern map.
  3. 3. The DRGs-based clinical medication data management system of claim 1, wherein the deviation identification module performs a multi-dimensional comparison of the new case medication data collected in real time with the typical medication pattern map to identify deviations, comprising: analyzing the medicine data of the new case to obtain a medicine use sequence of the new case and a medicine use time track of the new case; Aligning and matching the medication time track of the new case with a treatment stage time window in the typical medication pattern map, and determining the current treatment stage of the new case; Acquiring a typical medicine combination corresponding to the current treatment stage in the typical medicine pattern map, comparing the medicine use sequence of the new case with the typical medicine combination, and recording the medicine use sequence of the new case as medicine sequence deviation if the medicine use sequence of the new case contains medicines in atypical medicine combinations; Calculating the actual use dose of each medicine in the current treatment stage of the new case, comparing the actual use dose with a preset standard dose range of the treatment stage, and recording the actual use dose as abnormal deviation of the dose if the actual use dose exceeds the standard dose range; Accumulating the total cost of the medicines from the treatment start to the current stage of the new case, comparing the total cost with the historical average cost of the nodes in the same treatment stage in the typical medication pattern map, and recording the current total cost as a cost increase abnormal deviation if the current total cost exceeds a historical average cost preset proportion threshold; and integrating the medication sequence deviation, the dose abnormal deviation and the cost increase abnormal deviation to generate a structured medication deviation report.
  4. 4. The DRGs-based clinical medication data management system of claim 3, wherein the cause association module traverses a generated deviation cause association network in a preset clinical medication knowledge graph based on the medication deviation report, comprising: Extracting atypical medicines corresponding to the medication order deviation, specific medicines and excessive numerical values corresponding to the dose abnormality deviation and the expense hyperbranched medicine category corresponding to the expense growth abnormality deviation from the medication deviation report; Taking the atypical medicines, the specific medicines and the expense hyperbranched medicines as query entities, executing entity retrieval in the clinical medication knowledge graph, and positioning node positions of the query entities in the knowledge graph; Reading all preset type relation paths taking nodes as starting points from the clinical medication knowledge graph, wherein the preset type relation paths comprise medicine substitution relation paths, medicine compatibility tabu relation paths, treatment guideline recommendation relation paths and medicine cost benefit relation paths; performing multi-hop traversal along each read relation path, and collecting all associated intermediate entity nodes and relation edges on the paths to form a plurality of subgraphs taking query entities as root nodes; And merging all collected subgraphs, removing repeated nodes and edges, and constructing a deviation cause association network which is connected with all deviations in the medication deviation report and potential clinical and cost factors.
  5. 5. The DRGs-based clinical medication data management system of claim 4, wherein the policy generation module retrieves a preset medication optimization rule base matching to obtain a medication adjustment suggestion set according to the deviation factor association network, comprising: Carrying out structural decomposition processing on the deviation factor association network, and splitting the deviation factor association network into a plurality of independent deviation factor graphs, wherein each deviation factor graph describes a main deviation type and an associated cause path thereof; extracting the deviation type and key cause entity of the core of each deviation cause graph, and combining the deviation type and key cause entity into a query statement; performing pattern matching in the medication optimization rule base by using the query statement, wherein the medication optimization rule base stores rules in a format of deviation condition-cause condition-adjustment suggestion; When the deviation type and the key cause entity in the query statement are matched with the deviation condition and the cause condition in a certain rule, triggering the rule and acquiring a corresponding adjustment suggestion; Summarizing and deduplicating the adjustment suggestions triggered and acquired for all deviation factor graphs to form the medication adjustment suggestion set for the new case.
  6. 6. The DRGs-based clinical medication data management system of claim 5, wherein the policy generation module sends the set of medication adjustment suggestions to a clinical decision support terminal, comprising: Prioritizing each adjustment suggestion in the set of medication adjustment suggestions, the prioritization calculated based on a severity of a deviation associated with the adjustment suggestion on treatment safety and cost impact; Generating a corresponding execution context specification for each adjustment suggestion, the execution context specification including applicable treatment phases, related drugs, expected adjustment effect summaries and reference bases; packaging the ordered medication adjustment suggestions and the execution context description thereof into a decision support message in a specific format; pushing the decision support message to a clinical decision support terminal of a target clinician through a medical data exchange interface, and highlighting the decision support message in a designated interface area of the terminal.
  7. 7. The DRGs-based clinical medication data management system of claim 2, wherein the frequent pattern mining process is performed on the encoded sequence of drug use during each treatment phase time window to identify patterns of drug combinations that co-occur at high frequency during the treatment phase time window as typical drug combinations for a treatment phase, comprising: acquiring coded medicine use sequences of all cases in a current treatment stage time window, wherein each sequence comprises a plurality of standard medicine classification codes ordered according to medicine use time; Setting a minimum support threshold, wherein the minimum support threshold represents the minimum frequency of occurrence of the drug combination in all case sequences; Traversing all the medicine use sequences by adopting a frequent pattern growth algorithm, counting the occurrence frequency of each medicine code, and generating a head table according to the descending order of the frequency; projecting the medicine use sequence of each case based on the head table to construct a condition pattern base; recursively mining frequent item sets on the condition mode base, and marking the medicine combination as a high-frequency combination when the support degree of the item sets reaches a minimum support degree threshold value; Sorting all the identified high-frequency medicine combinations in descending order according to the support degree, and selecting a plurality of medicine combinations with the highest support degree as typical medicine combinations of a current treatment stage time window; And binding and storing the typical medicine combination and the corresponding treatment stage time window for subsequent construction of a medication pattern directed graph.
  8. 8. The DRGs-based clinical medication data management system of claim 3, wherein said aligning the medication time trace of the new case with the treatment phase time window in the typical medication pattern map, determining the treatment phase in which the new case is currently located, comprises: Analyzing the medication time track of the new case, and extracting a medication start time point, each medicine use time point and a current time point; Acquiring a predefined treatment phase time window dividing rule from a typical medication pattern map, wherein the rule comprises a duration range and a sequence relation of each treatment phase; Taking the medication start time point of the new case as a time axis zero point, and calculating the offset of each medicine use time point relative to the zero point; dividing a plurality of continuous treatment phase time windows on a time axis according to the duration range of the treatment phase time windows; Performing time sequence matching on the medicine use time point sequence of the new case and the divided treatment stage time windows, and counting the number of medicine use events in each time window; selecting a treatment phase time window with the largest number of drug use events as a treatment phase in which a new case is most likely to be in at present; when the number of the medicine using events with a plurality of time windows is equal, selecting the time window with the highest time density as the current treatment stage by combining the time density distribution characteristics of the medicine using; and outputting the finally determined treatment stage identification for subsequent drug use sequence comparison analysis.
  9. 9. The DRGs-based clinical medication data management system of claim 1, further comprising a medication pattern update module for dynamically updating the typical medication pattern based on feedback data for new cases, comprising: Receiving feedback data from a clinical decision support terminal, wherein the feedback data records the medication scheme actually adopted by a doctor for a new case and the execution condition of the medication adjustment suggestion set; Extracting a drug pattern fragment which is verified to be effective from the adopted drug regimen; the similarity between the medication pattern fragment and the node corresponding to the treatment stage in the typical medication pattern map is calculated, if the similarity is higher than a preset threshold value, the weight of the node in the map is enhanced, and if the similarity is lower than the preset threshold value but the medication result is evaluated well, the medication pattern fragment is used as a new candidate pattern branch to be added into the map; And re-calculating and optimizing the weight and the branch structure of each node in the typical medication pattern map based on the feedback data of all new cases in a time period, and generating an updated typical medication pattern map version.
  10. 10. The DRGs-based clinical medication data management system of claim 9, wherein the medication pattern update module adds a medication pattern clip to the map as a new candidate pattern branch, comprising: Analyzing the difference points of the medication pattern segments and the existing map nodes in the aspects of medicine composition, medication sequence and dosage, and positioning target nodes which are the same as the treatment stage of the medication pattern segments in the typical medication pattern map; Creating a new child node to represent the medication pattern segment by taking the target node as a father node; establishing a directed edge between the father node and the new child node, and marking specific difference attributes between the father node and the new child node on the directed edge; initializing weights for the new child nodes, the child node initializing weights being calculated based on the number of cases in which the medication pattern fragment was generated and the medication outcome evaluation score.

Description

Clinical medication data management system based on DRGs Technical Field The invention relates to the technical field of medical information, in particular to a clinical medication data management system based on DRGs. Background In the medical management mode of DRGs payment, the fine management and control of clinical medication behavior has become a core requirement of medical institutions. The prior art scheme mainly relies on statistical analysis and threshold monitoring of historical cost data, and is mainly used for calculating macroscopic indexes such as total cost, average cost or medicine ratio of medicines in each DRG group. The method can realize the traceability and the exceeding early warning of the financial result, but the management logic of the method stays at the level of the total amount of the cost and cannot go deep into the clinical medication process. A disadvantage of the prior art is that it lacks quantitative definition and description of standardized, rationalized medication behavior patterns under specific DRG groupings. Because the clinical behavior standard which fuses the medicine selection, the dosage and the time sequence relationship cannot be established, the system is difficult to distinguish reasonable cost fluctuation caused by complex disease conditions from resource consumption caused by irregular behaviors such as disordered medication sequence, improper dosage and the like. The management actions are thus lagging and not targeted enough. When the system monitors that the cost is abnormal or the medicine is not used in compliance, the prior proposal usually triggers the warning according to a preset single rule which is isolated from each other. A limitation of this approach is that it cannot systematically trace back and present a number of potential causes of possible interleaving behind a medication deviation. Decision support therefore stays at the surface phenomenon notification level, lacking in correlation analysis and depth interpretation of the deviation sources. Disclosure of Invention The invention aims to provide a clinical medication data management system based on DRGs, so as to solve the problems in the background technology. To achieve the above object, the present invention provides a DRGs-based clinical medication data management system, the system comprising: The data acquisition module is used for extracting a structured medication record set corresponding to each group from an electronic medical record database of a target medical institution according to the diagnosis related group identification, wherein the structured medication record set comprises medicine use detail data, medication time track data and associated medical expense item data; the mode construction module is used for carrying out joint analysis on the drug use detail data and the drug use time track data to construct a typical drug use mode map of each case group; the deviation recognition module is used for carrying out multidimensional comparison on the new case medication data acquired in real time and the typical medication pattern map, recognizing medication sequence deviation, dose abnormal deviation and cost increase abnormal deviation, and generating a medication deviation report; The factor association module is used for traversing a medicine interaction path, a treatment standard path and a cost influence factor path which are associated with the deviation item in a preset clinical medication knowledge graph based on the medication deviation report to generate a deviation factor association network; the strategy generation module is used for searching a preset medication optimization rule base according to the deviation factor association network, obtaining a medication adjustment suggestion set aiming at a specific deviation type through matching, and sending the medication adjustment suggestion set to a clinical decision support terminal. Preferably, the mode construction module performs a joint analysis on the drug usage detail data and the medication time track data to construct a typical medication mode map, including: the typical medication pattern map comprises a medicine use sequence relation, a dose adjustment node and a charge accumulation track; performing medicine classification coding mapping processing on the medicine use detail data, and mapping various medicine names into standard medicine classification codes to form coded medicine use sequences; performing time window division processing on the medication time track data, and dividing the whole medication time track into a plurality of continuous treatment stage time windows according to treatment stages; In each treatment stage time window, carrying out frequent pattern mining processing on the coded medicine use sequence, and identifying a medicine combination pattern which appears together with high frequency in the treatment stage time window as a typical medicine combination of a treatment stag