Hospital nursing resource intelligent allocation method, apparatus and device, and medium

Abstract

The invention relates to a hospital nursing resource intelligent allocation method and device, equipment and a medium. The method comprises the following steps: generating a nurse skill assessment report based on nursing operation records, training data and patient feedback; secondly, analyzing a nursing demand dependency relationship and a time constraint in the electronic medical record, and constructing a patient demand constraint graph; further quantitatively matching the nurse skill items with the patient demand items, and generating a nurse-demand sorting result sorted according to the matching degree; and finally, through deviation analysis of actual execution data and a prediction result, continuously optimizing the nurse skill evaluation model. According to the method, the technical defects that dynamic changes of nurse skills are not captured in real time, a complex nursing demand dependency relationship is not systematically analyzed, man-post matching lacks a quantitative basis and execution feedback does not drive model optimization in traditional manual distribution are overcome, and the nursing resource utilization rate and the clinical demand response accuracy are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of nursing resource management, and in particular to a hospital nursing resource intelligent allocation method, device, equipment and medium. BACKGROUND

[0002] In the field of hospital nursing management, the rational allocation of nursing resources is the core link to ensure the quality of medical services. This allocation method involves dynamically matching nursing staff with corresponding qualifications according to the clinical needs of patients, while taking into account multiple factors such as nurse skill level, workload, and task urgency. Traditional allocation patterns usually rely on manual scheduling and experience-based judgment, with nurse qualifications and patient needs recorded through paper or simple electronic systems, and then tasks are assigned by management personnel through subjective coordination.

[0003] However, the existing technology has significant defects: static nurse ability assessment cannot capture the dynamic changes of skills over time, resulting in high-difficulty nursing tasks being assigned to nurses with degraded skills; the multi-dimensional nursing needs derived from complex patient conditions are difficult to be systematically analyzed and prioritized; the human-post matching process lacks quantitative basis, leading to resource mismatch and response delay; and there is a lack of closed-loop feedback mechanism for the execution effect and prediction deviation, making it impossible to continuously optimize the allocation strategy. These defects result in low utilization of nursing resources, high rate of specialist operation mismatch, and delayed emergency response, among other problems. SUMMARY

[0004] Based on this, the purpose of the present application is to provide a hospital nursing resource intelligent allocation method that can dynamically evaluate nurse skills, accurately analyze clinical needs, intelligently match human-post resources, and continuously optimize in a closed loop.

[0005] The purpose of the present application is achieved by the following scheme:

[0006] In a first aspect, the present application provides a hospital nursing resource intelligent allocation method, comprising the following steps:

[0007] S1: processing the nursing operation records, nurse training records and patient nursing feedback data stored in the hospital information system, calculating the operation skill change trend of the nurses based on a preset nursing skill evaluation model, and generating a nurse skill evaluation report;

[0008] S2: processing the nurse skill evaluation report and the obtained patient electronic medical record, analyzing the patient nursing needs and constructing a time-constrained dependency graph, and generating a patient demand constraint graph;

[0009] S3: Process the nurse skill evaluation report and the patient demand constraint graph, match and calculate the operation skill items of the nurses with the nursing demand items of the patients, generate a nurse-demand ranking result and send it to the nurse terminal, the nurse-demand ranking result contains a nurse list arranged in descending order of matching degree and a corresponding skill matching value;

[0010] S4: Based on the nursing feedback data of the nurse terminal and the nurse-demand ranking result, calculate the prediction result of the nursing operation, compare the actual operation data of the nursing feedback data with the prediction result to calculate the execution deviation, and update the nursing skill evaluation model based on the execution deviation.

[0011] In one of the embodiments, the hospital nursing resource intelligent allocation method provided by the application specifically comprises the following steps:

[0012] S11: Process the nursing operation records stored in the hospital information system, calculate the dynamic skill weight according to the nursing operation frequency and time interval, and generate a skill decay curve;

[0013] S12: Process the nurse training records stored in the hospital information system, calculate the certification ability value of the nurses in combination with the certification status and examination results of the operation certificates, and generate a certification ability vector;

[0014] S13: Process the skill decay curve and the certification ability vector based on the preset nursing skill evaluation model, combine the operation quality score normalized and weighted fusion calculated based on the patient nursing feedback data stored in the hospital information system, generate a nurse skill evaluation report, and the nurse skill evaluation report is used to indicate the quantitative value of each skill level of the nurses.

[0015] In one of the embodiments, the hospital nursing resource intelligent allocation method provided by the application specifically comprises the following steps:

[0016] S21: Obtain the electronic medical record of the patient, process the diagnosis information in the electronic medical record, analyze the nursing operation dependency relationship associated with the complication through the preset medical knowledge base, and generate a demand dependency relationship;

[0017] S22: Process the medical order text in the electronic medical record, convert the natural language description into a standardized nursing operation item, and generate a nursing demand item set;

[0018] S23: Graph structure modeling is performed on the demand dependency relationship and the nursing demand item set, a time window constraint attribute required by the medical order is added, a patient demand constraint graph is generated, and the patient demand constraint graph is used to indicate the dependency relationship and execution time requirement between the nursing operations.

[0019] In one of the embodiments, the hospital nursing resource intelligent allocation method provided by the application specifically comprises the following steps:

[0020] S31: Perform skill item extraction processing on the nurse skill assessment report, filter the skill items and ability values related to nursing needs, and generate an available skill matrix;

[0021] S32: Perform demand item analysis processing on the patient demand constraint graph, extract nursing operation items and their time constraint conditions, and generate a demand feature vector;

[0022] S33: Perform similarity calculation processing on the available skill matrix and the demand feature vector, calculate the matching score of the nurse skill value and the patient demand standard, and generate a nurse-demand ordering result in descending order of score.

[0023] In one embodiment, the present application provides a hospital nursing resource intelligent allocation method S4, which specifically comprises the following steps:

[0024] S41: Process the nursing feedback data fed back by the nurse terminal, extract the actual execution time and operation completion degree indicators, and generate an actual nursing quality data set;

[0025] S42: Perform nursing operation prediction based on the nurse-demand ordering result, calculate the expected execution effect by combining the nursing operation records stored in the hospital information system, and generate a nursing operation prediction result;

[0026] S43: Process the actual nursing quality data set and the nursing operation prediction result, calculate the time deviation rate and quality deviation degree of the actual operation data and the prediction result, and generate a multi-dimensional execution deviation matrix;

[0027] S44: Perform model parameter adjustment processing on the multi-dimensional execution deviation matrix, update the weight coefficients and decay factors of the skill assessment model, and generate an updated nursing skill assessment model.

[0028] In a second aspect, the present application provides a hospital nursing resource intelligent allocation device, which is configured with the following modules:

[0029] A nurse skill assessment module for processing the nursing operation records, nurse training records and patient nursing feedback data stored in the hospital information system, calculating the operation skill change trend of the nurses based on a preset nursing skill assessment model, and generating a nurse skill assessment report;

[0030] A patient demand analysis module for processing the nurse skill assessment report and the obtained patient electronic medical record, analyzing the patient nursing needs and constructing a time-constrained dependency graph, and generating a patient demand constraint graph;

[0031] The nurse-demand matching sorting module is configured to process the nurse skill assessment report and the patient demand constraint graph, match the operation skill items of the nurses with the nursing demand items of the patients, generate a nurse-demand matching sorting result, and send the result to the nurse terminal. The nurse-demand sorting result includes a nurse list arranged in descending order of matching degree and a corresponding skill matching value.

[0032] The nursing model updating module is configured to calculate a prediction result of the nursing operation based on the nursing feedback data of the nurse terminal and the nurse-demand matching sorting result, calculate an execution deviation by comparing actual operation data of the nursing feedback data with the prediction result, and update the nursing skill assessment model based on the execution deviation.

[0033] In a third aspect, the present application provides a computer device including a memory and a processor. The memory stores a computer program, and the processor implements any of the above hospital nursing resource intelligent allocation methods when executing the computer program.

[0034] In a fourth aspect, the present application provides a computer readable storage medium having a computer program stored thereon. The computer program is executed by a processor to implement any of the above hospital nursing resource intelligent allocation methods.

[0035] In summary, the hospital nursing resource intelligent allocation method provided by the present application can accurately capture the real-time change trend of the skills of nurses by dynamically tracking the skill assessment report generated by the nursing operation record, training data and patient feedback, and solve the problem that skill degradation is not identified due to traditional static assessment. Based on the deep analysis of the electronic medical record by the medical knowledge base, the complex nursing demand dependency relationship can be structured and the time constraint can be quantitatively modeled, overcoming the technical defect that the complications associated operations cannot be systematically processed by artificial experience. The nurse-demand sorting result generated by the feature vector similarity calculation is used to realize the objective quantitative decision of the person-post matching, avoiding the resource mismatch risk caused by subjective allocation. Finally, the model is updated by the deviation between the actual execution data and the prediction result, which can establish a continuous self-optimizing learning mechanism to ensure that the allocation strategy evolves dynamically with clinical practice.

[0036] For better understanding and implementation, the present application is described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 A flowchart of a hospital nursing resource intelligent allocation method provided by an embodiment of the present application;

[0038] Figure 2 A structural diagram of a hospital nursing resource intelligent allocation device provided by another embodiment of the present application. DETAILED DESCRIPTION

[0039] For the purpose of promoting the understanding of the present application, a more complete description of the application will be provided below with reference to the accompanying drawings. The preferred embodiments of the present application are shown in the drawings. However, the present application can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the application more thorough and comprehensive.

[0040] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0041] In one embodiment, as shown in Figure 1 A hospital nursing resource intelligent allocation method is provided, and the embodiment is exemplified by the method applied to a terminal. It can be understood that the method can also be applied to a server, and can also be applied to an apparatus including a terminal and a server, and is implemented through the interaction of the terminal and the server. In the embodiment, the method includes the following steps:

[0042] S1: processing the nursing operation records, nurse training records and patient nursing feedback data stored in the hospital information system, calculating the operation skill change trend of the nurses based on a preset nursing skill evaluation model, and generating a nurse skill evaluation report.

[0043] Specifically, the system accesses the hospital information system and reads the stored nursing operation records, nurse training records and patient nursing feedback data. The nursing operation records cover the names, execution times, operation objects, operation results, and whether complications occur during the operation of various nursing operations performed by the nurses, etc. The nurse training records include training project names, training start and end times, training contents, examination methods, examination scores, training teacher evaluations, etc. The patient nursing feedback data involves patient satisfaction scores, feedback opinions, nursing effect evaluations, etc.

[0044] Exemplarily, the system processes the read data, identifies and eliminates invalid data and abnormal data according to preset rules. The invalid data includes data with format errors and missing key information, and the abnormal data includes scores or operation results that do not conform to common sense. After data cleaning, the system performs data standardization, converts time data in different formats into a unified timestamp format, and converts text feedback information into quantifiable feature values through natural language processing technology. Preferably, the system can call a preset nursing skill assessment model. The model is a time series prediction model based on gradient boosting tree. The model receives the cleaned and standardized nursing operation records, nurse training records and patient care feedback data as input features, learns from historical data to construct a prediction function of the change of nurse operation skills over time, calculates the change trend of nurses in different skill dimensions, and finally generates a nurse skill assessment report. The report includes the basic information of the nurse, the current score of each skill, the skill change curve in the past period of time and the skill development trend prediction in the future period of time.

[0045] S2: Process the nurse skill assessment report and the obtained patient electronic medical record, analyze the patient care needs and construct a time-constrained dependency graph to generate a patient demand constraint graph.

[0046] Specifically, the system obtains the nurse skill assessment report and the patient electronic medical record. The nurse skill assessment report provides the skill reserve and level of the nurse, which serves as the skill benchmark for subsequent matching. The patient electronic medical record contains the patient's basic information, chief complaint, present illness history, past medical history, physical examination results, laboratory examination results, imaging examination results, clinical diagnosis, treatment plan and other information.

[0047] Preferably, when analyzing the patient care needs, the system can apply information extraction technology based on natural language processing to extract specific care items from the patient electronic medical record and determine the urgency of each care need and the required nurse skill level. The system constructs a time-constrained dependency graph with the resolved patient care needs as nodes and determines the directed edges between the nodes according to the clinical specifications and logical relationships of nursing operations. The weight of the directed edge represents the time constraint information, such as the execution of a certain care need within a certain time after the completion of another care need, or the execution of a certain care need at a fixed time point. The dependency graph constructed through the above process shows the sequence and time association between the care needs. The system generates a patient demand constraint graph that contains all care need nodes, dependency relationships between nodes and corresponding time constraint parameters.

[0048] S3: Process the nurse skill assessment report and the patient demand constraint graph, match the operation skill items of the nurses with the nursing demand items of the patients, generate a nurse-demand ranking result and send it to the nurse terminal, the nurse-demand ranking result contains a nurse list arranged in descending order of matching degree and corresponding skill matching values.

[0049] Specifically, the system processes the nurse skill assessment report and the patient demand constraint graph, extracts the scores and skill level information of each operation skill of the nurses from the nurse skill assessment report, and extracts the skill level and time constraint conditions required for each nursing demand from the patient demand constraint graph. Preferably, the system can use a weighted scoring method for matching calculation, set weights for the skills required for each nursing demand, and the weights are determined according to the importance and difficulty of the nursing demand. Then, the scores of the nurses on the corresponding skills are multiplied by the weights and summed to obtain the skill matching value of the nurse and the nursing demand. For nursing demands with time constraints, the system combines the nurse's work schedule to filter out nurses who can meet the constraint conditions in terms of time, and then calculates the matching value of the filtered nurses to generate a nurse-demand ranking result. The nurse-demand ranking result contains the name, job number, skill matching value, etc. of the nurses, and is sorted in descending order of skill matching value. The system sends the ranking result to the nurse terminal through the hospital's internal communication network, and the nurse terminal receives and displays the result through a special application program.

[0050] S4: Calculate the predicted result of the nursing operation based on the nursing feedback data of the nurse terminal and the nurse-demand ranking result, calculate the execution deviation by comparing the actual operation data of the nursing feedback data with the predicted result, and update the nursing skill assessment model based on the execution deviation.

[0051] Specifically, the system receives the nursing feedback data uploaded by the nurse terminal, which includes actual operation data such as actual execution time, execution process record, operation result, problems encountered during operation and handling methods, etc. The system calculates the predicted result of the nursing operation based on the nurse-demand ranking result, uses the skill matching value in the nurse-demand ranking result as the basis, calls historical nursing operation data, extracts nursing effect parameters under similar matching conditions, and predicts the expected result of this nursing operation through a regression analysis model. The expected result includes the expected degree of improvement of the patient's symptoms and the operation completion time, etc.

[0052] Preferably, the system compares the actual operation data in the nursing feedback data with the predicted results, calculates the execution deviation by the mean square error method, compares the indicators corresponding to the predicted results in the actual operation data, and calculates the mean square error between the two as the execution deviation. The system updates the nursing skill evaluation model based on the execution deviation. If the execution deviation exceeds the preset threshold, the system adjusts the weight parameters in the preset nursing skill evaluation model through the back propagation algorithm, so that the model more accurately reflects the relationship between the nurse's skill and the nursing operation effect, and improves the accuracy of subsequent nurse skill evaluation.

[0053] In summary, the hospital nursing resource intelligent allocation method provided by the present application can accurately capture the real-time change trend of the nurse's skill by generating a skill evaluation report through dynamically tracking nursing operation records, training data and patient feedback, solving the problem of skill degradation not being identified caused by traditional static evaluation. Based on the deep analysis of the electronic medical record by the medical knowledge base, the structured expression of the dependency relationship of complex nursing needs and the quantitative modeling of time constraints can be realized, overcoming the technical defect that the correlation operation of complications cannot be systematically processed by artificial experience. The nurse-demand sorting result generated by the feature vector similarity calculation is used to realize the objective quantitative decision of the person-post matching, avoiding the resource mismatch risk caused by subjective allocation. Finally, the model is updated by the deviation between the actual execution data and the predicted results, which can establish a continuous self-optimizing learning mechanism to ensure that the allocation strategy evolves dynamically with clinical practice.

[0054] In one of the embodiments, the S1 of the hospital nursing resource intelligent allocation method provided by the present application specifically includes the following steps:

[0055] S11: processing the nursing operation records stored in the hospital information system, calculating the dynamic skill weight according to the nursing operation frequency and time interval, and generating a skill decay curve.

[0056] Specifically, the system accesses the hospital information system and reads the stored nursing operation records through a preset interface. The nursing operation records cover the identity of the nurse, the name of the operation performed, the start and end time of the operation, the patient identifier involved in the operation, the operation result identifier, the description of abnormal events recorded during the operation, and other information. Illustratively, the system pre-processes the read nursing operation records. First, it verifies the completeness of each record. For records missing key information, the system marks them and temporarily stores them in the temporary data area. After the missing information is supplemented, they are included in the processing flow. For records with complete fields, the system groups them by nurse identity and collects all operation records of the same nurse into the same data set.

[0057] For each data set, the system extracts the execution record of each nursing operation, counts the execution times of each operation in a set time period, and forms operation frequency data. At the same time, the system sorts the execution time of the same operation in chronological order, calculates the difference between the execution times of adjacent two times, forms time interval data, and calculates the dynamic skill weight based on the operation frequency and time interval data. In the calculation process, the operation frequency and the dynamic skill weight are positively correlated, and the time interval and the dynamic skill weight are negatively correlated. The system takes the time axis as the horizontal axis and the dynamic skill weight as the vertical axis, connects the dynamic skill weights of the same operation at different time points in sequence, forms a skill decay curve, and the curve reflects the change trend of the dynamic skill weight with time under the condition of decreasing operation frequency or increasing time interval.

[0058] S12: Process the nurse training records stored in the hospital information system, calculate the certification ability value of the nurse in combination with the authentication state of the operation certificate and the examination score, and generate a certification ability vector.

[0059] Specifically, the system retrieves the nurse training records through the database query function of the hospital information system. The nurse training records include nurse identity, training project name and code, nursing operation type and code corresponding to the training, operation certificate number, issue date, expiration date, current authentication state identifier, raw score of theoretical examination, raw score of practical examination, signature record of the examiner, and other information.

[0060] Exemplarily, the system sorts the retrieved training records, classifies the records according to the nursing operation type code, and forms independent data groups for the training records of the same operation type. For each data group, the system reads the current authentication state identifier of the operation certificate, assigns corresponding base scores to different states according to the preset rules, and the base score of the record with a valid authentication state is higher than that of the record with an expired or pending authentication state. The system converts the raw scores of the theoretical examination and the practical examination into a calculable numerical form, calculates them through a preset weighting formula, obtains a comprehensive examination score, and performs a multiplication operation between the base score and the comprehensive examination score. The result is used as the certification ability value of the nurse for the corresponding nursing operation. The system arranges the certification ability values in a preset order according to the nursing operation type code, and forms a certification ability vector.

[0061] S13: Process the skill decay curve and the certification ability vector based on a preset nursing skill evaluation model, normalize and weight the operation quality score based on the patient nursing feedback data stored in the hospital information system, and generate a nurse skill evaluation report. The nurse skill evaluation report is used to indicate the quantitative value of the skill level of the nurse.

[0062] Specifically, the system calls a preset nursing skill assessment model, which includes a time series analysis module and a multi-dimensional weighted calculation module. The system inputs the generated skill decay curve into the time series analysis module, and the module extracts dynamic skill parameters of each nursing operation through trend analysis of the curve. Meanwhile, the system inputs the generated authentication ability vector into the multi-dimensional weighted calculation module, and the module analyzes each authentication ability value in the vector as a static parameter.

[0063] The system obtains patient care feedback data through the interface of the hospital information system, including patient identification, corresponding nursing operation name, operation quality score raw data, nursing effect text description, problem feedback record, etc. Preferably, the system processes the operation quality score raw data, which can convert different source and scale score data to a unified numerical range by using a normalization algorithm, eliminate the influence of scale difference on calculation, and input the dynamic skill parameter, static parameter and normalized operation quality score into the multi-dimensional weighted calculation module. The module performs weighted summation on the three types of parameters according to the preset weight distribution rule to obtain the comprehensive quantitative value of each nursing operation. The system associates each comprehensive quantitative value with the corresponding nursing operation name and nurse identity, forms structured data, and converts the structured data into a nurse skill assessment report through a report generation function. The report lists each nursing operation name and corresponding comprehensive quantitative value, and presents the skill level of the nurse.

[0064] In one of the embodiments, the hospital nursing resource intelligent allocation method provided by the application specifically comprises the following steps:

[0065] S21: Obtain the electronic medical record of the patient, process the diagnosis information in the electronic medical record, analyze the nursing operation dependency relationship associated with the complication through the preset medical knowledge base, and generate the demand dependency relationship.

[0066] Specifically, the system obtains the electronic medical record of the patient through the interface of the hospital information system, and the electronic medical record contains the basic information, diagnosis result, course record, examination report and past medical history of the patient. The system extracts the diagnosis information in the electronic medical record, and performs structured processing on the extracted diagnosis information to convert non-standardized diagnosis terms into preset standardized diagnosis codes, ensuring the consistency of the diagnosis information, wherein the diagnosis information includes text description of primary diagnosis, secondary diagnosis and complication diagnosis.

[0067] The system calls a preset medical knowledge base. The preset medical knowledge base stores data of association relationships between various diseases, complications, and nursing operations, including common complications corresponding to each diagnosis, and information of the order and coordination relationship of nursing operations required by the complications. The system matches the standardized diagnosis code with the data in the medical knowledge base, locates the nursing operation items corresponding to the diagnosis and related complications, and analyzes the dependency relationship between the nursing operations, for example, a certain type of complication requires performing a vital sign monitoring operation first, and then performing a symptomatic nursing operation; or two nursing operations need to be performed simultaneously to avoid mutual interference. Based on the above processing, the system stores the analyzed dependency relationship in a structured data form to form a requirement dependency relationship, which includes the name of the nursing operation, the associated complication diagnosis, the name of the preceding operation, the name of the subsequent operation, and the like.

[0068] S22: Processing the medical order text in the electronic medical record to convert the natural language description into a standardized nursing operation item, and generating a nursing requirement item set.

[0069] Specifically, the system extracts the medical order text from the electronic medical record. The medical order text includes the nursing-related instructions issued by the doctor, presented in the form of natural language, covering the operation type, execution frequency, execution method, and the like. The system preprocesses the medical order text, performs word segmentation, part-of-speech tagging, and syntax analysis through natural language processing technology, and identifies the key information in the text, such as nursing operation verbs, objects, time adverbs, and the like.

[0070] Preferably, the system can call a standardized nursing operation terminology library, which contains preset nursing operation names, operation codes, operation definitions, and corresponding natural language expression examples. The system matches the key information in the medical order text with the expression examples in the terminology library, determines the closest standardized nursing operation item through semantic similarity calculation, and for the medical order text that cannot be directly matched, the system parses the logical relationship in the text through a rule engine, disassembles it into multiple matchable sub-items, and then maps them to standardized terminologies respectively. The system aggregates all the mapped standardized nursing operation items, removes the duplicates to form a nursing requirement item set, and each nursing requirement item contains structured information such as operation code, operation name, and execution requirement.

[0071] S23: Graph structure modeling of the requirement dependency relationship and the nursing requirement item set, adding a time window constraint attribute required by the medical order, generating a patient requirement constraint graph, and the patient requirement constraint graph is used to indicate the dependency relationship and execution time requirement between nursing operations.

[0072] Specifically, the system takes each nursing operation item in the nursing demand item set as a node of a graph structure, and the node attributes include basic information such as operation code, operation name, etc. The system establishes a directed edge between the corresponding nodes according to the pre-operation and post-operation association in the demand dependency relationship. The direction of the directed edge represents the sequence of operation execution, for example, from the pre-operation node to the post-operation node. The system extracts the time requirements related to the nursing operation in the electronic medical record order, including the first execution time, the last execution time, the execution interval period, etc. of the operation, and converts these information into time window constraint attributes, such as the execution of a certain operation within a certain time period on a specified date, or the execution within a specified time period after the completion of another operation.

[0073] The system adds the time window constraint attributes to the corresponding nodes and directed edges. The time attribute of the node represents the execution time range of the operation itself, and the time attribute of the directed edge represents the time interval requirement between the two operations. The system integrates the nodes, directed edges and attribute information through a graph structure modeling tool to generate a patient demand constraint graph. The graph visually displays the association relationship of all nursing operation items and their respective execution time requirements. The position of the node and the connection mode of the edge intuitively reflect the dependency relationship, and the attribute label marks the time constraint details.

[0074] In one of the embodiments, the hospital nursing resource intelligent allocation method provided by the present application specifically comprises the following steps:

[0075] S31: Skill item extraction processing is performed on the nurse skill assessment report to filter the skill items and capability values related to nursing demand and generate an available skill matrix.

[0076] Specifically, the system calls the nurse skill assessment report, which contains fields such as nurse identification, skill item name, capability value, skill applicable department code, and skill update time. The system extracts all skill item names and corresponding capability values through field parsing to form an initial skill list. Each record in the list is associated with a nurse identification. The system also obtains a nursing demand item set, which contains demand operation name, demand code, associated diagnosis code, and execution department code. The system compares the initial skill list with the nursing demand item set, and filters the directly corresponding skill items through string exact matching. For items that do not match directly, the system uses semantic association analysis to convert the skill item name and demand item name into a word vector, and determines the association degree by calculating the cosine similarity of the word vector. The calculation formula is:

[0077]

[0078] Wherein, A is the skill item word vector, B is the demand item word vector, A B is the vector dot product, ‖A‖, ‖B‖ is the vector length respectively. The system retains the skill item and its ability value whose correlation degree reaches the preset threshold, arranges horizontally according to the nurse identifier, arranges vertically according to the skill item name, constructs the available skill matrix M, the matrix element M ij represents the ability value of the i th nurse on the j th skill, and the element corresponding to the skill item not involved is filled with zero value.

[0079] S32: The demand item analysis processing is performed on the patient demand constraint graph, the nursing operation item and its time constraint condition are extracted, and the demand feature vector is generated.

[0080] Specifically, the system reads the storage file of the patient demand constraint graph, the file adopts the graph structure data format, and contains the node table and the edge table. The node table records the unique identifier, name, code, earliest start timestamp, latest end timestamp, upper limit value of execution time length, lower limit value of execution time length of the nursing operation item; the edge table records the pre-operation identifier, post-operation identifier and dependency type code. The system traverses all nodes according to the depth first traversal algorithm, extracts the nursing operation item name, code and time window constraint condition of each node, and stores them into the temporary data table. The system topologically sorts the operation items in the temporary data table according to the dependency relationship in the edge table, and generates the ordered operation sequence. The system constructs the demand feature vector X with the ordered operation sequence as the dimension, and the vector dimension is consistent with the number of operation items. For the numerical conversion of the time window constraint, the system adopts time normalization processing, converts the earliest start time into the latest end time into Wherein, t si is the earliest start timestamp of the i th operation, t ei is the latest end timestamp, t base is the starting timestamp of the day, t max is the end timestamp of the day. Each element of the demand feature vector X contains the operation code, the earliest start time x i1 and the latest end time x i2 , and the vector order is consistent with the topological sorting result.

[0081] S33: Similarity calculation processing is performed on the available skill matrix and the demand feature vector, the matching degree score of the nurse skill value and the patient demand standard is calculated, and the nurse-demand sorting result is generated in descending order of score.

[0082] Specifically, the system retrieves the available skill matrix M and the demand feature vector X from the storage unit, and converts each row of data of the matrix into a nurse skill vector Sk, wherein k is the nurse index. The system calculates the similarity between the nurse skill vector Sk and the demand feature vector X, and the calculation formula is:

[0083] s ki = |Sk,j X i,std | -1

[0084] t ki = 1 - |x k,t X i,t |

[0085] wherein s ki is the skill matching degree of the nurse k in the corresponding skill item i, S k,j is the ability value of the nurse k in the corresponding skill item j, X i,std is the demand standard value of the i-th operation, t ki is the time matching degree of the nurse k in the corresponding skill item i, x k,t is the normalized value of the nurse executable time, X i,t is the normalized value of the operation time window. The system calculates the total matching degree score Score k using a weighted summation formula:

[0086]

[0087] wherein n is the number of operation items, w s is the skill matching weight, w t is the time matching weight, and w s + w t = 1. The system ranks all the nurses' Score k in descending order to generate a nurse-demand ranking result, which contains the nurse identifier, Score k value, s ki and t ki values of each operation.

[0088] In one embodiment, the hospital nursing resource intelligent allocation method provided by the present application specifically comprises the following steps:

[0089] S41: processing the nursing feedback data fed back by the nurse terminal, extracting the actual execution time and operation completion degree indicators, and generating an actual nursing quality data set.

[0090] Specifically, the system receives the nursing feedback data uploaded by the nurse terminal, which contains information such as execution records of nursing operations, operation process notes, and patient state change records. The system pre-processes the received data, checks the standardization of the data format, and adjusts the data to a unified structure through a format conversion tool for data that does not conform to the preset format; for data missing key fields, the system marks and sends a supplement request to the nurse terminal until the data is complete.

[0091] The system extracts the actual execution time from the processed nursing feedback data, including the operation start timestamp and end timestamp, calculates the operation duration; at the same time, the operation completion degree index is extracted, which is determined based on the completion ratio of operation steps, the compliance degree of operation specification and the improvement of patient state. The system associates the actual execution time, operation duration, operation completion degree index with the corresponding nursing operation item, nurse identity, and arranges them in time sequence to form a structured data set, i.e. actual nursing quality data set.

[0092] S42: Based on the nurse-demand ranking result, the nursing operation prediction is carried out, the expected execution effect is calculated combined with the nursing operation records stored in the hospital information system, and the nursing operation prediction result is generated.

[0093] Specifically, the system calls the generated nurse-demand ranking result, extracts the identity of the nurse, the matching degree score and the corresponding nursing operation item in the ranking result, retrieves the historical nursing operation records through the hospital information system interface, filters out the historical records with the same nursing operation item and similar nurse skill matching degree, and extracts the execution time, operation completion degree, patient feedback and other data in these records. Preferably, the system statistically analyzes the historical nursing operation records, calculates the average execution time, average completion degree and fluctuation range of similar operations in the same matching degree interval, combines the dynamic skill parameters in the skill assessment report of the current nurse, and constructs a prediction model. The prediction model takes the nurse-demand matching degree as the input variable, outputs the expected execution time, expected completion degree and expected patient state improvement index of the nursing operation, integrates these expected indexes to generate the nursing operation prediction result, which includes the expected parameters of each nursing operation and the confidence interval of the parameters.

[0094] S43: Process the actual nursing quality data set and the nursing operation prediction result, calculate the time deviation rate and quality deviation degree of the actual operation data and the prediction result, and generate a multi-dimensional execution deviation matrix.

[0095] Specifically, the system extracts the actual execution time T act and the actual completion degree C act from the actual nursing quality data set, and extracts the expected execution time T pre and the expected completion degree C pre from the nursing operation prediction result. The system calculates the time deviation rate R t , the formula is:

[0096]

[0097] Where R t is positive when the actual execution time exceeds the expected, and is negative when it is shorter than the expected. The system calculates the quality deviation degree R c , the formula is:

[0098]

[0099] wherein R c is positive if actual completion is higher than expected, and negative if lower. The system constructs a multi-dimensional execution deviation matrix D with nurses identified as rows and operation codes as columns, with matrix elements D ij = [R t,ij ,R c,ij ], where R t,ij is the time deviation rate of the ith nurse performing the jth operation, and R c,ij is the corresponding quality deviation degree.

[0100] S44: Perform model parameter adjustment processing on the multi-dimensional execution deviation matrix, update the weight coefficients and decay factors of the skill assessment model, and generate an updated nursing skill assessment.

[0101] Specifically, the system calls the multi-dimensional execution deviation matrix D and the parameter set of the current nursing skill assessment model, which contains the weight coefficients w j and skill decay factors λ of each skill item. The system performs statistical analysis on the matrix D, calculates the average time deviation rate and the average quality deviation degree For the adjustment of the weight coefficients, the formula is:

[0102]

[0103] wherein α is the learning rate, N is the number of nurses, M ij is the available skill matrix element, and the adjusted weight w j ′ = w j + Δw j . For the update of the decay factor, based on the time deviation rate, the calculation is wherein β is the adjustment coefficient. The system writes the updated weight coefficients and decay factors into the model parameter file, overwrites the original parameters, generates an updated nursing skill assessment model, and stores the model version number and update timestamp.

[0104] It should be understood that although the steps in the flowcharts involved in the embodiments described above are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts involved in the embodiments described above can include multiple steps or multiple stages, which are not necessarily executed at the same time but can be executed at different times, and the execution order of these steps or stages is not necessarily sequential but can be alternately or alternately executed with at least part of other steps or steps or stages in other steps.

[0105] Based on the same inventive concept, the embodiments of the present application also provide a hospital nursing resource intelligent allocation device for implementing the above-mentioned hospital nursing resource intelligent allocation method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more hospital nursing resource intelligent allocation device embodiments provided below can refer to the limitations of the hospital nursing resource intelligent allocation method described above, which will not be repeated here.

[0106] Preferably, as shown in the drawings, the present application provides a hospital nursing resource intelligent allocation device 500, which is configured with the following modules: Figure 2

[0107] The nurse skill assessment module 510 is configured to process the nursing operation records, nurse training records and patient care feedback data stored in the hospital information system, calculate the operation skill trend of the nurses based on a preset nursing skill assessment model, and generate a nurse skill assessment report.

[0108] The patient demand analysis module 520 is configured to process the nurse skill assessment report and the obtained patient electronic medical record, analyze the patient care demand and construct a time-constrained dependency graph, and generate a patient demand constraint graph.

[0109] The nurse-demand matching and sorting module 530 is configured to process the nurse skill assessment report and the patient demand constraint graph, match the operation skill items of the nurses with the nursing demand items of the patients, generate a nurse-demand matching and sorting result and send it to the nurse terminal, and the nurse-demand sorting result includes a nurse list arranged in descending order of matching degree and a corresponding skill matching value.

[0110] The nursing model updating module 540 is configured to calculate the prediction result of the nursing operation based on the nursing feedback data of the nurse terminal in combination with the nurse-demand matching and sorting result, perform bias calculation by comparing the actual operation data of the nursing feedback data with the prediction result, and update the nursing skill assessment model based on the execution bias.​

[0111] Preferably, the nurse skill assessment module 510 provided by the present application is configured with the following units:

[0112] A nursing operation skill weight calculation unit is configured to process the nursing operation records stored in the hospital information system, calculate the dynamic skill weight according to the nursing operation frequency and time interval, and generate a skill decay curve;

[0113] A training certification ability evaluation unit is configured to process the nurse training records stored in the hospital information system, calculate the certification ability value of the nurse in combination with the certification status and examination results of the operation certificate, and generate a certification ability vector;

[0114] A skill assessment report generation unit is configured to process the skill decay curve and the certification ability vector based on a preset nursing skill assessment model, normalize and weight the operation quality score in combination with the patient care feedback data stored in the hospital information system, and generate a nurse skill assessment report. The nurse skill assessment report is used to indicate the quantitative value of the skill level of the nurse.

[0115] Preferably, the patient demand analysis module 520 provided by the present application is configured with the following units:

[0116] A diagnosis information dependency analysis unit is configured to obtain the electronic medical record of the patient, process the diagnosis information in the electronic medical record, analyze the nursing operation dependency relationship associated with the complication through a preset medical knowledge base, and generate a demand dependency relationship;

[0117] A medical order text standardization unit is configured to process the medical order text in the electronic medical record, convert the natural language description into a standardized nursing operation item, and generate a set of nursing demand items;

[0118] A demand constraint graph construction unit is configured to model the demand dependency relationship and the set of nursing demand items in a graph structure, add a time window constraint attribute required by the medical order, and generate a patient demand constraint graph. The patient demand constraint graph is used to indicate the dependency relationship and execution time requirement between nursing operations.

[0119] Preferably, the nurse-demand matching and sorting module 530 provided by the present application is configured with the following units:

[0120] A nurse skill item extraction and screening unit is configured to perform skill item extraction processing on the nurse skill assessment report, screen the skill items and ability values related to the nursing demand, and generate an available skill matrix;

[0121] A patient demand item analysis and feature extraction unit is configured to perform demand item analysis processing on the patient demand constraint graph, extract the nursing operation item and its time constraint condition, and generate a demand feature vector;

[0122] The skill-demand matching degree sorting unit is configured to perform a similarity calculation process on the available skill matrix and the demand feature vector, calculate a matching degree score of the nurse skill value and the patient demand standard, and generate a nurse-demand sorting result in descending order of the score.

[0123] Preferably, the nursing model updating module 540 provided by the present application is configured with the following units:

[0124] The feedback data extraction and processing unit is configured to process the nursing feedback data fed back by the nurse terminal, extract the actual execution time and operation completion degree index, and generate an actual nursing quality data set.

[0125] The operation effect prediction unit is configured to perform nursing operation prediction based on the nurse-demand sorting result, calculate an expected execution effect in combination with the nursing operation records stored in the hospital information system, and generate a nursing operation prediction result.

[0126] The execution deviation calculation unit is configured to process the actual nursing quality data set and the nursing operation prediction result, calculate a time deviation rate and a quality deviation degree of the actual operation data and the prediction result, and generate a multi-dimensional execution deviation matrix.

[0127] The skill evaluation model updating unit is configured to perform model parameter adjustment processing on the multi-dimensional execution deviation matrix, update the weight coefficient and the attenuation factor of the skill evaluation model, and generate an updated nursing skill evaluation model.

[0128] In one embodiment, the present application further provides a computer device, including a memory and a processor, the memory stores a computer program, and the processor implements the above-mentioned hospital nursing resource intelligent allocation method when executing the computer program.

[0129] In one embodiment, the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the above-mentioned hospital nursing resource intelligent allocation method.

[0130] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example" or "some examples" means that the specific features, structures, materials or characteristics described in combination with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples without contradiction.

[0131] For the device embodiment, since it basically corresponds to the method embodiment, the relevant part can be seen from the part of the method embodiment. The device embodiment described above is only schematic, wherein the components shown as separate components can or can not be physically separate, and the components shown as a unit can or can not be physical units, i.e. can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purpose of the present disclosure according to actual needs. Those skilled in the art can understand and implement it without creative labor.

[0132] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of various changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for intelligent allocation of hospital nursing resources, characterized in that: The following steps are involved: S1: Process the nursing operation records, nurse training records and patient nursing feedback data stored in the hospital information system, calculate the nurses' operation skill change trend based on the preset nursing skill evaluation model, and generate a nurse skill evaluation report; S2: Processing the nurse skill assessment report and the obtained patient electronic medical records, parsing the patient's nursing needs and constructing a dependency graph with time constraints to generate a patient demand constraint graph; S3: Processing the nurse skill assessment report and the patient demand constraint graph, matching the nurse's operational skill items with the patient's nursing demand items, generating a nurse-demand ranking result, and sending it to the nurse terminal. The nurse-demand ranking result includes a list of nurses sorted in descending order of matching degree and corresponding skill matching values; S4: Calculate the predicted results of nursing operations based on the nursing feedback data from the nurse terminal combined with the nurse-demand ranking results, compare the actual operation data of the nursing feedback data with the predicted results to calculate the execution deviation, and update the nursing skill evaluation model based on the execution deviation.

2. The method according to claim 1, characterized in that Said S1 comprises: S11: Processing the nursing operation records stored in the hospital information system, calculating the dynamic skill weight according to the nursing operation frequency and time interval, and generating a skill decay curve; S12: Processing the nurse training records stored in the hospital information system, calculating the nurse's certification capability value based on the certification status of the operation certificate and the assessment results, and generating a certification capability vector; S13: Based on the preset nursing skill assessment model, the skill attenuation curve and the certification capability vector are processed, and the operation quality score is normalized and weighted fusion calculation is performed in combination with the patient nursing feedback data stored in the hospital information system to generate a nurse skill assessment report. The nurse skill assessment report is used to indicate the quantitative value of the nurse's various skill levels.

3. The method according to claim 1, characterized in that The S2 includes: S21: Obtain the patient's electronic medical record, process the diagnostic information in the electronic medical record, analyze the nursing operation dependency associated with the complication through a preset medical knowledge base, and generate a demand dependency; S22: Processing the medical order text in the electronic medical record, converting the natural language description into standardized nursing operation items, and generating a set of nursing requirement items; S23: Graph structure modeling is performed on the demand dependency relationship and the set of nursing demand items, and a time window constraint attribute of the medical order requirement is added to generate a patient demand constraint graph, where the patient demand constraint graph is used to indicate the dependency relationship and execution time requirements between nursing operations.

4. The method according to claim 1, wherein The S3 includes: S31: extracting skill items from the nurse skill assessment report, screening skill items and capability values ​​related to nursing needs, and generating an available skill matrix; S32: performing demand item parsing processing on the patient demand constraint graph, extracting nursing operation items and their time constraints, and generating a demand feature vector; S33: Perform similarity calculation on the available skill matrix and the demand feature vector, calculate the matching score between the nurse skill value and the patient demand standard, and generate a nurse-demand ranking result by arranging in descending order of the scores.

5. The method according to any one of claims 1 to 4, characterized in that The S4 includes: S41: Processing the nursing feedback data fed back by the nurse terminal, extracting the actual execution time and operation completion index, and generating an actual nursing quality data set; S42: performing nursing operation prediction based on the nurse-demand ranking result, calculating the expected execution effect in combination with the nursing operation records stored in the hospital information system, and generating a nursing operation prediction result; S43: Processing the actual nursing quality data set and the nursing operation prediction result, calculating the time deviation rate and quality deviation between the actual operation data and the prediction result, and generating a multi-dimensional execution deviation matrix; S44: Performing model parameter adjustment processing on the multidimensional execution deviation matrix, updating the weight coefficient and attenuation factor of the skill evaluation model, and generating an updated nursing skill evaluation model.

6. A hospital nursing resource intelligent allocation device, characterized in that: The device comprises: The nurse skill assessment module is used to process the nursing operation records, nurse training records and patient nursing feedback data stored in the hospital information system, calculate the nurse's operation skill change trend based on the preset nursing skill assessment model, and generate a nurse skill assessment report; A patient demand analysis module is used to process the nurse skill assessment report and the obtained patient electronic medical records, analyze the patient's nursing needs and construct a dependency graph with time constraints to generate a patient demand constraint graph; a nurse-demand matching ranking module, configured to process the nurse skill assessment report and the patient demand constraint graph, match the nurse's operational skill items with the patient's nursing demand items, generate a nurse-demand matching ranking result, and send it to the nurse terminal. The nurse-demand matching ranking result includes a list of nurses sorted in descending order of matching degree and corresponding skill matching values; The nursing model update module is used to calculate the predicted results of nursing operations based on the nursing feedback data from the nurse terminal combined with the nurse-demand matching ranking results, compare the actual operation data of the nursing feedback data with the predicted results to calculate the execution deviation, and update the nursing skill evaluation model based on the execution deviation.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images