Nursing measure recommendation method based on evidence-based nursing decision and related equipment

Abstract

The invention discloses a nursing measure recommendation method based on evidence-based nursing decision and related equipment, and relates to the field of intelligent medical treatment, and the method comprises the steps: obtaining the health data information of a target patient; monitoring a nursing trigger instruction corresponding to the target patient; under the condition that a nursing triggering instruction is received, risk label information of the target patient is determined based on the health data information on the basis of a rule engine and a clustering algorithm; generating nursing measure recommendation information based on the risk tag information; sending the nursing measure recommendation information to the target patient to obtain feedback information of the target patient; and generating a target nursing measure based on the feedback information and the nursing measure recommendation information. The intelligent, precise and individualized levels of the nursing path are remarkably improved, the nursing resource configuration and intervention effect are optimized, and the satisfaction degree and rehabilitation efficiency of the patient are improved.

Description

Technical Field

[0001] This specification relates to the field of smart healthcare; more specifically, this application relates to a method and related equipment for recommending nursing interventions based on evidence-based nursing decision-making. Background Technology

[0002] Current clinical nursing decisions rely primarily on nurses' experience and limited clinical guidelines, lacking systematic data support and real-time evidence-based evidence, making it difficult to meet the demands of modern medicine for efficient and precise nursing care. In existing nursing management models, problems such as information silos, fragmented data, and poor communication are prevalent, severely restricting the scientific rigor and timeliness of nursing decisions. Faced with increasingly complex conditions and more individualized patient needs, traditional nursing methods show significant shortcomings in response speed, intervention precision, and individualized treatment plan suitability. Furthermore, frontline clinical nursing generally lacks intelligent decision-making tools, making it difficult for nurses to access the latest evidence-based medicine and real-time patient data in a timely manner. This leads to delayed and biased nursing decisions, affecting intervention effectiveness and potentially prolonging patient recovery periods and increasing the risk of complications. Evidence-based nursing emphasizes optimizing nursing practice based on scientific evidence, but existing systems have not yet achieved deep integration and dynamic utilization of multi-source data, including patient medical records, physiological monitoring information, and clinical research data.

[0003] Therefore, there is an urgent need to develop a method and system for recommending nursing interventions based on evidence-based nursing decision-making, in order to improve the quality and efficiency of nursing care and promote the improvement of the overall treatment effect for patients. Summary of the Invention

[0004] The summary section introduces a series of simplified concepts, which will be further explained in detail in the detailed description section. This summary section is not intended to limit the key and essential technical features of the claimed technical solution, nor is it intended to determine the scope of protection of the claimed technical solution.

[0005] Firstly, this application proposes a method for recommending nursing interventions based on evidence-based nursing decision-making, including:

[0006] Obtain health data information of the target patients;

[0007] Monitor the nursing trigger commands corresponding to the target patient;

[0008] Upon receiving the aforementioned nursing trigger instruction, the risk label information of the target patient is determined based on the aforementioned health data information using a rule engine and clustering algorithm.

[0009] Based on the above risk label information, nursing intervention recommendations are generated;

[0010] The above nursing intervention recommendations will be sent to the target patients to obtain their feedback.

[0011] Based on the above feedback information and the above nursing intervention recommendations, target nursing interventions are generated.

[0012] In one feasible implementation, the acquisition of the target patient's health data information includes:

[0013] Multi-source data acquired from hospital information systems, electronic medical records, and IoT devices;

[0014] The above multi-source data is parsed using natural language to obtain the aforementioned health data information.

[0015] In one feasible implementation, the aforementioned nursing triggering instructions are generated based on one or more of the following methods: data-driven, semantic matching, and human system collaboration.

[0016] In one feasible implementation, the determination of the risk label information of the target patient based on the aforementioned health data information using a rule engine and clustering algorithm includes:

[0017] A three-dimensional risk labeling system is constructed based on the above health data information, which includes the patient's pathological data, physiological parameters, and socio-psychological information.

[0018] The rule engine is used to match the above patient health data to identify the risk label information of the target patients.

[0019] If the rule engine fails to match the risk label information, a clustering algorithm is used to analyze the feature space containing data from multiple patients, grouping patients with similar risk characteristics into the same category, and automatically labeling the target patients with their corresponding risk label information.

[0020] In one feasible implementation, the above-mentioned conditional matching of the patient health data information using a rule engine to identify the tag content corresponding to a specific risk rule includes:

[0021] A rule knowledge base is constructed, which includes multiple nursing risk identification rules. Each rule consists of prerequisites, triggering logic, and label output. The prerequisites are generated based on the patient's structured health information and text mining results.

[0022] Define a conditional expression language, wherein the conditional expression language adopts a rule expression syntax based on a domain-specific language, and supports nested logic, numerical comparison, range judgment, missing value fault tolerance, and multi-field joint triggering mechanism;

[0023] Load the aforementioned health data information of the target patient;

[0024] The rule matching process involves comparing each precondition in each rule one by one. If all preconditions are met, the rule is triggered, and the risk label set in the rule is added to the risk label set of the target patient.

[0025] In one feasible implementation, the above method further includes:

[0026] Obtain information on label usage frequency, measure compliance, and intervention results;

[0027] The rule knowledge base is dynamically adjusted based on the above-mentioned label usage frequency information, above-mentioned measure compliance information, and above-mentioned intervention results.

[0028] In one feasible implementation, the aforementioned recommendation information includes candidate nursing interventions corresponding to multiple risk labels. Each of the aforementioned nursing interventions includes an intervention description, implementation method, recommendation strength, level of evidence, and applicable conditions.

[0029] The feedback information mentioned above includes the target patients' acceptance of the recommended measures, willingness to implement them, past adherence, individual preferences, and history of adverse reactions.

[0030] The above-mentioned target nursing interventions are generated based on the above feedback information and the above-mentioned nursing intervention recommendations, including:

[0031] Based on the above feedback information and the above nursing intervention recommendations, an adaptability analysis was conducted to obtain compliance scores, expected outcome scores, and risk tolerance scores.

[0032] Based on the above compliance score, above expected effect score, and above risk tolerance score, candidate measures were screened and ranked, and all candidate nursing measures were comprehensively evaluated to obtain the above target nursing measures.

[0033] Secondly, the present invention also proposes a nursing intervention recommendation device based on evidence-based nursing decision-making, comprising:

[0034] The first acquisition unit is used to acquire the health data information of the target patient;

[0035] The monitoring unit is used to monitor the nursing trigger commands corresponding to the target patient;

[0036] The determining unit is used to determine the risk label information of the target patient based on the health data information, a rule engine, and a clustering algorithm, upon receiving the aforementioned nursing trigger instruction.

[0037] The first generation unit is used to generate nursing intervention recommendation information based on the aforementioned risk label information;

[0038] The second acquisition unit is used to send the above-mentioned nursing intervention recommendation information to the above-mentioned target patient in order to obtain the feedback information of the above-mentioned target patient.

[0039] The second generation unit is used to generate target nursing measures based on the above feedback information and the above nursing measure recommendation information.

[0040] Thirdly, the present invention also proposes an electronic device comprising: a memory and a processor, characterized in that the processor is configured to execute a computer program stored in the memory to implement the steps of the method for recommending nursing interventions based on evidence-based nursing decisions as described in any of the first aspects.

[0041] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the method for recommending nursing interventions based on evidence-based nursing decision-making as described in any one of the first aspects.

[0042] In summary, this invention overcomes the problems of information silos and data fragmentation in existing systems in terms of data integration and patient profiling. By unifying the access and natural language processing of multi-source heterogeneous data collected from Hospital Information Systems (HIS), Electronic Medical Record Systems (EMR), and IoT devices, it not only achieves the fusion processing of structured and unstructured information but also significantly improves the system's comprehensive perception of individual patient conditions. Compared to traditional methods relying on subjective observation and limited medical history data, the comprehensive health profile constructed by this solution lays a solid data foundation for accurate diagnosis and intervention. Regarding the triggering mechanism for diagnosis and intervention, this invention proposes a triple triggering path of data-driven, semantic matching, and human collaboration, overcoming the problems of delayed diagnostic response and single triggering logic in existing nursing processes. The system can not only automatically identify changes in key vital signs and chief complaint keywords but also support clinical staff to subjectively trigger the diagnostic process based on experience, with the system retrieving supporting data in real time, greatly improving the speed and coverage of diagnostic response. In terms of risk identification mechanism, this invention constructs a dual-pathway risk label generation strategy that integrates a rule engine and a clustering algorithm. The rule engine module incorporates structured nursing knowledge rules, enhancing the interpretability and consistency of the reasoning process. The clustering algorithm, based on historical big data, mines and analyzes the patient feature space, enabling intelligent identification and automatic labeling of risk tags that are not explicit or not yet covered by the rule base. This solves the existing system's problems of "insufficient rule coverage" and "cold start," enhancing the system's adaptability to different populations and scenarios. Regarding the measure recommendation and optimization mechanism, the system relies on an evidence-based nursing knowledge base to provide corresponding nursing measures for each risk tag, along with meta-information such as evidence level, recommendation strength, and applicable conditions, ensuring clear scientific basis and transparency in the recommendations. By incorporating patient feedback information (such as willingness to accept, difficulty of implementation, and adverse reactions) for suitability analysis, the system not only avoids "one-size-fits-all" interventions but also dynamically adjusts recommendation results based on compliance scores, expected effects, and risk tolerance, achieving truly "patient-centered" personalized nursing strategy development. In terms of closed-loop feedback and continuous optimization, this solution constructs a closed-loop process from data collection, diagnosis triggering, risk identification, measure recommendation to feedback collection and re-evaluation. The system can automatically record the execution process and intervention results of each nursing pathway, and use the effect evaluation data to feed back into the knowledge base, enabling dynamic optimization of nursing measures and continuous evolution of the knowledge system. This effectively supports hospital nursing quality management, indicator achievement monitoring, and scientific research analysis. In summary, compared with traditional nursing decision-making methods that rely on experience and static guidelines, this invention significantly improves the intelligence, precision, and individualization of nursing pathways. It not only optimizes the allocation of nursing resources and intervention effects but also helps improve patient satisfaction and rehabilitation efficiency, demonstrating good clinical promotion value and application prospects.

[0043] Other advantages, objectives and features of this application will be apparent in part from the description which follows, and in part from what those skilled in the art will understand through study and practice of this application. Attached Figure Description

[0044] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0045] Figure 1 A schematic diagram of a nursing intervention recommendation method based on evidence-based nursing decision-making provided in this application embodiment;

[0046] Figure 2 A structural schematic diagram of a nursing intervention recommendation device based on evidence-based nursing decision-making provided in this application embodiment;

[0047] Figure 3 This is a schematic diagram of an electronic device structure provided in an embodiment of this application. Detailed Implementation

[0048] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. The technical solutions of the embodiments of this application will now be clearly and completely described in conjunction with the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them.

[0049] Please see Figure 1 This is a flowchart illustrating an evidence-based nursing decision-making method for recommending nursing interventions, as provided in an embodiment of this application. Specifically, it may include:

[0050] Firstly, this application proposes a method for recommending nursing interventions based on evidence-based nursing decision-making, including:

[0051] S110. Obtain health data information of the target patient;

[0052] S120, Monitor the nursing trigger instructions corresponding to the target patient;

[0053] S130. Upon receiving the aforementioned nursing trigger instruction, the risk label information of the target patient is determined based on the aforementioned health data information using a rule engine and clustering algorithm.

[0054] S140. Generate nursing intervention recommendations based on the above risk label information;

[0055] S150. Send the above nursing intervention recommendations to the target patients to obtain their feedback.

[0056] S160. Generate target nursing measures based on the above feedback information and the above nursing measure recommendation information.

[0057] For example, this embodiment provides a nursing intervention recommendation method based on evidence-based nursing decision-making. By constructing an intelligent closed-loop recommendation path, it achieves personalized and dynamically optimized nursing intervention delivery. Its core process includes six steps, as follows:

[0058] First, in step S110, the system acquires the target patient's health data. This health data comes from multiple data channels, including the Hospital Information System (HIS), Electronic Medical Record (EMR), and Internet of Things (IoT) devices (such as wearable devices). The data includes both structured and unstructured content such as the patient's basic demographic information, physiological measurement parameters (such as blood pressure, blood oxygen, and heart rate), pathological diagnosis records, nursing records, scale assessment results, and chief complaints. Natural Language Processing (NLP) technology is used to perform entity recognition and semantic extraction on the unstructured text, further enriching the patient's health profile.

[0059] Next, in step S120, the system monitors whether there are nursing trigger instructions that trigger nursing diagnoses and interventions. These trigger instructions can be generated in three ways: first, data-driven, automatically triggered when the system detects an abnormality in a patient's indicator (such as blood pressure exceeding a threshold); second, semantic matching, matching diagnostic logic by identifying keywords in the patient's complaints (such as "chest pain" or "shortness of breath"); and third, a human-system collaborative approach, where a nurse manually initiates a diagnostic request, and the system uses real-time patient data to provide decision support.

[0060] In step S130, after the system detects a nursing trigger command, it enters the risk identification phase. Based on previously acquired health data, the system analyzes the risk label information of the target patient through a rule engine and a clustering algorithm. The rule engine uses preset nursing risk identification rules to match patient data according to logical expressions. For example, "age > 75 years and Braden (Braden Scale for Predicting Pressure Sore Risk) score < 12" triggers the "high risk of fall" label. If the rule engine fails to match effectively, the clustering algorithm module is activated. Through similarity analysis of historical patient feature spaces, the patient is classified into a specific risk group and inherits the typical risk labels of that group, thereby achieving automatic label generation and completion.

[0061] Subsequently, in step S140, the system extracts associated recommended measures from the nursing intervention knowledge base based on the identified risk label information. Each nursing intervention is accompanied by its level of evidence (e.g., systematic review, clinical guidelines), strength of recommendation (e.g., strong recommendation, conditional recommendation), and applicable conditions, forming structured nursing intervention recommendation information.

[0062] In step S150, the system displays or pushes the above-mentioned nursing intervention recommendations to the target patient's or nurse's end for the patient's reference and feedback. Patients can express their acceptance of the recommended interventions, their willingness to implement them, their past experiences, and adverse reactions through the system's interactive interface, or nurses can enter the feedback on their behalf.

[0063] Finally, in step S160, the system performs a comprehensive fit assessment based on patient feedback and the original recommendations. The system screens and ranks candidate interventions based on multiple factors, including compliance scores, expected outcome scores, and risk tolerance, ultimately determining the most suitable target nursing intervention for the current patient. If the original recommended intervention presents implementation obstacles or risks, the system will automatically recommend the optimal alternative, ensuring that the nursing intervention is both safe and effective while meeting the patient's individualized needs.

[0064] In summary, this invention overcomes the problems of information silos and data fragmentation in existing systems in terms of data integration and patient profiling. By unifying the access and natural language processing of multi-source heterogeneous data collected from Hospital Information Systems (HIS), Electronic Medical Record Systems (EMR), and IoT devices, it not only achieves the fusion processing of structured and unstructured information but also significantly improves the system's comprehensive perception of individual patient conditions. Compared to traditional methods relying on subjective observation and limited medical history data, the comprehensive health profile constructed by this solution lays a solid data foundation for accurate diagnosis and intervention. Regarding the triggering mechanism for diagnosis and intervention, this invention proposes a triple triggering path of data-driven, semantic matching, and human collaboration, overcoming the problems of delayed diagnostic response and single triggering logic in existing nursing processes. The system can not only automatically identify changes in key vital signs and chief complaint keywords but also support clinical staff to subjectively trigger the diagnostic process based on experience, with the system retrieving supporting data in real time, greatly improving the speed and coverage of diagnostic response. In terms of risk identification mechanism, this invention constructs a dual-pathway risk label generation strategy that integrates a rule engine and a clustering algorithm. The rule engine module incorporates structured nursing knowledge rules, enhancing the interpretability and consistency of the reasoning process. The clustering algorithm, based on historical big data, mines and analyzes the patient feature space, enabling intelligent identification and automatic labeling of risk tags that are not explicit or not yet covered by the rule base. This solves the existing system's problems of "insufficient rule coverage" and "cold start," enhancing the system's adaptability to different populations and scenarios. Regarding the measure recommendation and optimization mechanism, the system relies on an evidence-based nursing knowledge base to provide corresponding nursing measures for each risk tag, along with meta-information such as evidence level, recommendation strength, and applicable conditions, ensuring clear scientific basis and transparency in the recommendations. By incorporating patient feedback information (such as willingness to accept, difficulty of implementation, and adverse reactions) for suitability analysis, the system not only avoids "one-size-fits-all" interventions but also dynamically adjusts recommendation results based on compliance scores, expected effects, and risk tolerance, achieving truly "patient-centered" personalized nursing strategy development. In terms of closed-loop feedback and continuous optimization, this solution constructs a closed-loop process from data collection, diagnosis triggering, risk identification, measure recommendation to feedback collection and re-evaluation. The system can automatically record the execution process and intervention results of each nursing pathway, and use the effect evaluation data to feed back into the knowledge base, enabling dynamic optimization of nursing measures and continuous evolution of the knowledge system. This effectively supports hospital nursing quality management, indicator achievement monitoring, and scientific research analysis. In summary, compared with traditional nursing decision-making methods that rely on experience and static guidelines, this invention significantly improves the intelligence, precision, and individualization of nursing pathways. It not only optimizes the allocation of nursing resources and intervention effects but also helps improve patient satisfaction and rehabilitation efficiency, demonstrating good clinical promotion value and application prospects.

[0065] In one feasible implementation, the acquisition of the target patient's health data information includes:

[0066] Multi-source data acquired from hospital information systems, electronic medical records, and IoT devices;

[0067] The above multi-source data is parsed using natural language to obtain the aforementioned health data information.

[0068] For example, in one feasible implementation, the system integrates multiple data channels to achieve comprehensive collection and processing of health data information of the target patient, ensuring that subsequent nursing risk identification and measure recommendation have sufficient data foundation and context awareness capabilities.

[0069] Specifically, the system first acquires structured and unstructured health information about patients during their hospital stay and throughout their medical history, based on multiple data sources including the Hospital Information System (HIS), Electronic Medical Record (EMR), and IoT devices. The HIS primarily provides basic demographic data, hospital registration, medication administration records, and medication history. The EMR includes clinical data such as diagnoses, nursing records, surgical records, progress notes, and laboratory reports. IoT devices transmit real-time dynamic physiological parameters, such as heart rate, blood pressure, blood oxygen saturation, body temperature, and respiratory rate, providing a continuous data stream for the system to construct the patient's current health status.

[0070] To fully utilize the aforementioned multi-source data, the system incorporates Natural Language Processing (NLP) technology to perform semantic parsing on unstructured text. Taking nursing records as an example, these documents typically contain free text such as chief complaints, observation descriptions, and assessment results. The system employs entity recognition and contextual analysis models (such as BERT or BiLSTM-CRF) to identify key health indicators, abnormal symptoms, nursing procedures, and compliance behaviors within the text. Simultaneously, the system can structure these identification results into standardized fields, further supplementing the information missing from the patient's structured data.

[0071] For example, if a patient's EMR text describes the patient as having "a persistent low-grade fever and night sweats for the past two days, with body temperature maintained at around 37.8°C", the system can automatically extract the keywords "low-grade fever" and "night sweats" and combine them with temperature data from IoT devices to classify it as a "fever trend" feature. At the same time, if the nursing record mentions "the patient lives alone and is in low mood", the system can further identify labels related to "insufficient social support" and "risk of depression".

[0072] Based on this, the system integrates all collected and analyzed information to construct a health information profile of the target patient, and provides input basis for subsequent nursing risk identification, label allocation and nursing intervention recommendation.

[0073] In summary, this implementation method integrates structured data and unstructured text content from HIS, EMR, and IoT devices, and utilizes natural language processing technology to extract high-dimensional health features. This enables the full-dimensional acquisition and semantic enhancement of target patient health data, providing a complete, accurate, and scalable data foundation for intelligent nursing decision-making.

[0074] In one feasible implementation, the aforementioned nursing triggering instructions are generated based on one or more of the following methods: data-driven, semantic matching, and human system collaboration.

[0075] For example, in one feasible implementation, the nursing trigger instruction can be generated through one or more of the following methods: data-driven, semantic matching, and human system collaboration, to support timely identification and intelligent response to patient care needs in different clinical scenarios.

[0076] Specifically, the data-driven approach relies primarily on the system's monitoring and analysis of patients' real-time physiological parameters and historical indicators. When the system detects that a key indicator exceeds a preset threshold, it will automatically trigger corresponding nursing diagnostic prompts. For example, if a patient's blood pressure remains above 140 / 90 mmHg for two consecutive hours, the system can automatically trigger a nursing prompt for "potential hypertension risk"; if the Braden score is below 12, the system will automatically generate a nursing prompt request for "high risk of pressure ulcers." This approach is highly real-time and objective, suitable for high-frequency parameter monitoring and early risk identification.

[0077] In the semantic matching approach, the system utilizes natural language processing technology to analyze free text in patient complaints, nursing records, and doctor-nurse interaction logs. Through keyword extraction and semantic understanding models (such as BERT), the system can identify high-risk descriptive words in patient expressions and map them to a nursing diagnosis database. For example, when complaints include descriptions such as "dyspnea" or "nighttime awakening due to shortness of breath," the system can automatically trigger nursing instructions related to "oxygenation disorder." When nursing records repeatedly show "irritability" or "depressed mood," the system can provide preliminary diagnostic suggestions of "anxiety" or "risk of depression." This approach is particularly effective in processing unstructured text information, improving the ability to identify subjective feelings and potential problems.

[0078] The manual-system collaboration approach provides a trigger mechanism for collaborative intervention by nursing staff and intelligent system decision-making. In clinical practice, nurses may identify potential nursing problems in patients based on clinical experience or visual observation, and can then manually initiate a nursing intervention command. Upon receiving the command, the system automatically retrieves all of the patient's health data, including vital sign trends over the past 24 hours, nursing assessment results, and laboratory reports, and uses built-in diagnostic rules to assist nurses in confirming the diagnosis type and suggesting intervention pathways. This approach ensures the system's flexibility and interpretability, facilitating the handling of complex and individualized clinical scenarios.

[0079] Furthermore, to enhance adaptability, this implementation supports the combined use of the three triggering methods described above. For example, in certain high-risk scenarios (such as ICU or chronic disease management for the elderly), the system can simultaneously enable data-driven and semantic matching mechanisms for dual early warning, with manual intervention used for final confirmation, thereby achieving complementary synergy of triggering mechanisms and improving the accuracy and response efficiency of nursing diagnoses.

[0080] In summary, this implementation method improves the timeliness, scientific rigor, and individual adaptability of nursing interventions by constructing a nursing triggering mechanism that integrates data-driven approaches, semantic understanding, and human collaboration, adapting to the dynamic needs of different information types, risk types, and clinical processes.

[0081] In one feasible implementation, the determination of the risk label information of the target patient based on the aforementioned health data information using a rule engine and clustering algorithm includes:

[0082] A three-dimensional risk labeling system is constructed based on the above health data information, which includes the patient's pathological data, physiological parameters, and socio-psychological information.

[0083] The rule engine is used to match the above patient health data to identify the risk label information of the target patients.

[0084] If the rule engine fails to match the risk label information, a clustering algorithm is used to analyze the feature space containing data from multiple patients, grouping patients with similar risk characteristics into the same category, and automatically labeling the target patients with their corresponding risk label information.

[0085] For example, based on the health data of target patients, a rule engine and clustering algorithm are used to jointly identify individual care risk labels of patients to support accurate recommendations and personalized interventions for subsequent care measures.

[0086] Specifically, the system first integrates and models the health data of the target patient. This data includes three main categories: first, pathological data, such as past disease diagnosis records, chronic disease labels, and genetic susceptibility information; second, physiological parameters, such as real-time and static indicators obtained from IoT devices and manual input, including blood pressure, body temperature, heart rate, respiratory rate, Braden score, and BMI; and third, psychosocial information, including the patient's marital and residential status, economic resources, social support evaluation, and scores on psychological assessment scales. Based on the above data, the system constructs a three-dimensional risk labeling system to describe the patient's comprehensive risk characteristics across the "pathological-physiological-psychosocial" dimensions.

[0087] Subsequently, the system prioritizes calling the rule engine module to compare patient data item by item according to preset risk identification rules. Each rule consists of three parts: preconditions, triggering logic, and label output. For example, rule R1 is set as: "If age ≥ 75 years and Braden score < 12", then the label "high risk of fall" is output; rule R2 is set as: "If PHQ-9 score ≥ 10 in the past two weeks and living alone", then the dual label "depression + insufficient social support" is output. The rule expression is based on a domain-specific language (DSL), supporting nested logic, multi-field combination judgment, and fault tolerance mechanisms. If all preconditions of a rule are met, the system immediately labels the patient with the corresponding risk label and terminates further judgment of the current label.

[0088] If the rules engine fails to match any risk label under the current data conditions, the system will automatically invoke the clustering algorithm module to perform similar patient analysis. This module first transforms data samples from multiple historical patients into standardized feature vectors, constructing a high-dimensional feature space. Using a hybrid clustering strategy (such as a combination of K-Means and DBSCAN), the system identifies several stable risk feature groups and assigns a set of typical risk labels to each group. Next, the system projects the target patient's feature vector onto this feature space, calculates its similarity to each cluster center (e.g., using Euclidean or Mahalanobis distance), and assigns it to the closest group. The system then transmits the risk labels shared by this group to the target patient, achieving automatic risk label completion.

[0089] For example, if a patient does not trigger the "hypertension" or "fall" rules, but their health profile is highly similar to that of patients in the historical database who are marked as "sleep disorder + anxiety risk", the system will automatically assign the label and prompt the nurse to pay attention and intervene accordingly.

[0090] In summary, this implementation method, through a collaborative mechanism of "rule priority and clustering completion," ensures the interpretability of rule matching in highly deterministic scenarios. At the same time, it introduces a data-driven inference mechanism when data is ambiguous or rule coverage is insufficient, thereby improving the coverage, flexibility, and individual adaptability of nursing risk label identification. This provides a structured and comprehensive input foundation for subsequent precise recommendations of nursing measures.

[0091] In one feasible implementation, the above-mentioned conditional matching of the patient health data information using a rule engine to identify the tag content corresponding to a specific risk rule includes:

[0092] A rule knowledge base is constructed, which includes multiple nursing risk identification rules. Each rule consists of prerequisites, triggering logic, and label output. The prerequisites are generated based on the patient's structured health information and text mining results.

[0093] Define a conditional expression language, wherein the conditional expression language adopts a rule expression syntax based on a domain-specific language, and supports nested logic, numerical comparison, range judgment, missing value fault tolerance, and multi-field joint triggering mechanism;

[0094] Load the aforementioned health data information of the target patient;

[0095] The rule matching process involves comparing each precondition in each rule one by one. If all preconditions are met, the rule is triggered, and the risk label set in the rule is added to the risk label set of the target patient.

[0096] For example, by constructing a rule engine and combining it with domain-specific language (DSL) syntax, patient health data can be conditionally matched to identify corresponding nursing risk labels, thereby achieving high-confidence and highly interpretable nursing risk identification.

[0097] First, the system establishes a rule knowledge base containing multiple nursing risk identification rules developed by clinical experts and supported by literature evidence. Each rule consists of three parts: preconditions, triggering logic, and label output. The preconditions define the necessary data conditions for a given risk to occur, typically generated based on structured health data (such as blood pressure, Braden score, and age) and text mining results (such as keywords or chief complaints in nursing records). For example, a typical rule might have the preconditions: "Braden score < 12 and age ≥ 75 years and 'bedridden' appears in the nursing record," with an "AND" triggering logic and a label output of "high risk of pressure ulcers."

[0098] To achieve high flexibility and automated execution of rules, the system defines a conditional expression language. This language is designed based on a domain-specific language (DSL) and possesses expressive capabilities tailored to nursing scenarios. The expression language supports:

[0099] Nested logical judgments: Rules can be nested with multiple sub-conditions, such as "A AND (B OR C)";

[0100] Numerical comparison and interval judgment: Supports judgments such as "indicator value > a certain value" and "score ∈ [x,y]";

[0101] Missing field tolerance mechanism: It can still execute partial matching logic when some fields are missing;

[0102] Multi-field joint triggering: Supports combining multiple data source fields as judgment conditions, such as "body temperature + white blood cell count + C-reactive protein" to jointly assess infection risk.

[0103] Next, the system loads the target patient's health data into the rule engine. Data sources include electronic medical records, nursing records, assessment scales, and real-time physiological parameters collected by IoT devices. All data undergoes structured and standardized preprocessing before input, and missing fields are filled in to enhance rule adaptability.

[0104] During rule matching, the system iterates through all rules in the rule base, comparing the preconditions of each rule with the current data status of the target patient. If all preconditions of a rule are met, the rule is triggered, and the system immediately writes the risk label corresponding to that rule into the patient's risk label set. For example, if a patient meets the criteria of "Braden score = 10", "age = 82 years", and "living status = bedridden and living alone", then rule R21 will be triggered, outputting the label "pressure ulcer + insufficient social support", which will serve as an input reference for subsequent nursing intervention recommendations.

[0105] In addition, the rule matching process is scalable and dynamically manageable. The system supports adding, deleting or editing rules through the interface, and also supports priority sorting and confidence weighting of rule execution, which facilitates continuous optimization of knowledge base content and execution effect.

[0106] In summary, this implementation method constructs a rule engine with structured definition, semantic expression capabilities, and dynamic management functions, and combines it with the multidimensional health data of the target patients to quickly and accurately determine nursing risk labels, providing efficient and reliable decision support for subsequent evidence-based nursing interventions.

[0107] In one feasible implementation, the above method further includes:

[0108] Obtain information on label usage frequency, measure compliance, and intervention results;

[0109] The rule knowledge base is dynamically adjusted based on the above-mentioned label usage frequency information, above-mentioned measure compliance information, and above-mentioned intervention results.

[0110] For example, by continuously acquiring and analyzing feedback information such as label usage frequency, nursing intervention compliance, and actual intervention results, the risk identification rules in the rule knowledge base can be regularly evaluated and dynamically adjusted.

[0111] First, the system continuously collects three types of key feedback information during the execution of the nursing pathway:

[0112] Tag usage frequency information: Records the number of times each rule is triggered within a certain time window. For example, whether the "high risk of pressure ulcers" tag was frequently automatically applied by the system and confirmed by caregivers in the past month.

[0113] Adherence information: The system assesses patient compliance with recommended nursing interventions, such as whether turning over was actually performed, whether health education was completed, and whether dietary recommendations were adopted. This information can be derived from structured fields in nursing records or actively collected through feedback interfaces on the system's interactive platform.

[0114] Intervention outcome information: The system evaluates whether the nursing intervention triggered by the corresponding risk label has produced the expected effect by observing the achievement of nursing goals, such as whether indicators have improved and whether complications have decreased. For example, when the completion rate of preventive measures corresponding to the "fall risk" label is high and the patient has no fall events, the system records it as "intervention effective".

[0115] After a certain period of data accumulation, the system aggregates and analyzes the feedback information and evaluates the actual effect of each rule based on a statistical model or weighting function. The system introduces a rule evaluation function, for example:

[0116] W r =α×U f +β×C r +γ×O s

[0117] Among them: W r Indicates the current weight of the rule; U f Normalized values ​​for label usage frequency; C r This corresponds to the compliance score for nursing interventions; O s α represents the intervention effect score (such as the target achievement rate); α, β, and γ are the weighting coefficients set by experience or optimized by learning.

[0118] Based on the function results, the system can perform the following dynamic adjustment operations: keep rules with long-term high weights active and increase their execution priority; reduce the weight of rules with low trigger frequency, poor compliance, or poor intervention effect; and suggest freezing or re-reviewing rules that have no triggers for several consecutive periods or have a concentration of negative feedback.

[0119] For example, if a rule regarding "incontinence-related skin integrity risk" is frequently triggered in multiple patients, but in most cases the patients do not receive relevant measures or do not see improvement after intervention, the system will automatically reduce the weight of the rule and lower its trigger priority in subsequent matching.

[0120] In summary, this implementation method, by integrating tag usage frequency, patient compliance, and intervention effect feedback, achieves dynamic adjustment and intelligent iteration of the nursing risk rule knowledge base. This not only enhances the system's self-learning ability but also improves the practical clinical applicability and accuracy of rule matching. This mechanism constructs a closed-loop path of "identification-execution-feedback-optimization," which is a key component of the system's intelligent evolution.

[0121] In one feasible implementation, the aforementioned recommendation information includes candidate nursing interventions corresponding to multiple risk labels. Each of the aforementioned nursing interventions includes an intervention description, implementation method, recommendation strength, level of evidence, and applicable conditions.

[0122] The feedback information mentioned above includes the target patients' acceptance of the recommended measures, willingness to implement them, past adherence, individual preferences, and history of adverse reactions.

[0123] The above-mentioned target nursing interventions are generated based on the above feedback information and the above-mentioned nursing intervention recommendations, including:

[0124] Based on the above feedback information and the above nursing intervention recommendations, an adaptability analysis was conducted to obtain compliance scores, expected outcome scores, and risk tolerance scores.

[0125] Based on the above compliance score, above expected effect score, and above risk tolerance score, candidate measures were screened and ranked, and all candidate nursing measures were comprehensively evaluated to obtain the above target nursing measures.

[0126] For example, by introducing a feedback-driven intelligent matching mechanism, the process of optimizing candidate nursing interventions into target nursing interventions is realized, thereby enhancing the personalization and clinical applicability of nursing interventions.

[0127] First, when generating nursing intervention recommendations, the system associates multiple candidate nursing interventions with each identified risk label. Each intervention includes: a description (e.g., "turn over every 2 hours"), a method of implementation (e.g., "requires nurse assistance" or "patient self-management"), a strength of recommendation (e.g., strong recommendation, conditional recommendation), a level of evidence (e.g., systematic review, RCT, expert consensus), and applicable conditions (e.g., "Braden score < 12" or "BMI > 28"). This information constitutes the set of nursing intervention recommendations that the system pushes to patients or nursing staff.

[0128] Subsequently, the system obtains relevant feedback information from patients, including: acceptance of each recommended measure (whether they are willing to try it), willingness to implement it (whether they can cooperate in the long term), historical compliance records, individual preferences (such as a preference for lifestyle interventions or measures implemented by nurses), and history of adverse reactions (such as previous anxiety, pain or other negative experiences caused by a certain measure).

[0129] Based on this, the system enters the personalized adaptation analysis stage, matching the patient's feedback information with each candidate nursing intervention one by one, and assigning scores from the following three dimensions:

[0130] Compliance score (B1): Based on the patient's feedback willingness, historical cooperation level, and cognitive acceptance, assess the probability of the measure being adopted;

[0131] Expected Outcome Score (B2): Based on the actual intervention results in similar patient populations and the current status of the target patients, predict its effectiveness;

[0132] Risk tolerance score (B3): Assess the acceptable level of risk based on the discomfort, adverse reactions, and the patient's sensitivity and psychological tolerance that the measure may cause.

[0133] The system substitutes the results from these three dimensions into the following comprehensive scoring model:

[0134] Stotal=λ1×B1+λ2×B2-λ3×B3

[0135] Wherein, λ1, λ2, and λ = are the weight parameters for compliance, expected effect, and risk tolerance, respectively, which can be obtained through clinical strategy setting or training and optimization using historical data.

[0136] Based on the comprehensive score, the system ranks the candidate nursing interventions and prioritizes the one with the highest score as the target nursing intervention. If the preferred intervention has an incompatibility barrier or risk warning, the system will automatically select the second-best alternative, and simultaneously indicate the substitution logic and scoring basis in the user interface.

[0137] The finalized target nursing interventions will be presented to nursing staff through a visual interface, along with details of the scoring dimensions, reasons for recommendation, and explanations of the level of evidence, to facilitate final confirmation or personalized adjustments by nursing staff.

[0138] In summary, this implementation method integrates nursing intervention attributes and patient feedback information to construct a multi-factor decision-making system that considers compliance, expected outcomes, and risk tolerance. It also uses a comprehensive scoring model to drive the selection of target nursing interventions, forming a closed-loop decision-making path from data-driven approaches to clinical consensus. This improves the scientific rigor, individual adaptability, and clinical adoption rate of the recommendations.

[0139] It should be noted that, in a further optimized implementation, the system also includes a nursing intervention type module for structured classification of recommended nursing interventions. Based on the operational attributes and purposes of the interventions, the system categorizes them into four main types: assessment and monitoring, nursing implementation, education and guidance, and management referral. Classification labels are added to the recommendation results to enable nurses to quickly identify and prioritize interventions based on the nature of the nursing task. The classification information can also be used to statistically analyze the implementation rates and effectiveness of different types of interventions, supporting the standardization of nursing pathways.

[0140] In addition, the system incorporates a nursing intervention source module, which records the knowledge sources upon which each nursing intervention is based, including national or international clinical guidelines, authoritative expert consensus, systematic review literature, or the hospital's own practical experience. This source information is displayed synchronously in the user interface and serves as an important parameter for determining the strength of the intervention recommendation.

[0141] The accompanying nursing intervention recommendation strength module assigns a recommendation level to each intervention based on the level of evidence and the credibility of the source, such as "strong recommendation," "relatively recommended," and "conditionally recommended." The system dynamically adjusts the recommendation strength label based on the quality of evidence, scope of application, and historical implementation results of the intervention, assisting nurses in quickly determining the priority and necessity of multiple options.

[0142] After the nursing intervention is completed, the system conducts a standardized evaluation through the effectiveness evaluation module. The evaluation indicators cover both the goal setting dimension (cognition, state, behavior) and the execution process dimension (time achievement rate, compliance feedback, vital sign change curve, etc.). The system supports the automatic archiving of this data and the creation of individual patient intervention response trend charts for subsequent pathway optimization and research analysis.

[0143] Through further integration and completion of the above modules, the system has constructed a complete knowledge loop from nursing knowledge basis to intervention classification, recommendation intensity control and efficacy feedback, realizing a more systematic, interpretable and high-precision nursing decision support process.

[0144] In one implementation, the common nursing diagnosis of "imbalanced nutrition: below the body's requirements" can be used as an example to comprehensively demonstrate the system's evidence-based decision-making capabilities throughout the entire process, including diagnosis triggering, risk identification, goal setting, measure recommendation, and effect feedback.

[0145] First, the system automatically collects patients' nutritional assessment information through an integrated multimodal data acquisition mechanism, including nutritional risk screening (such as NRS-2002 score), body mass index (BMI), skinfold thickness, serum albumin and prealbumin levels, dietary intake records, and chief complaints in nursing records. For text-based content, such as "poor appetite," "insufficient food intake," or "sudden weight loss," the system uses natural language processing (NLP) technology for keyword extraction and semantic recognition to construct a chain of diagnostic criteria.

[0146] After a diagnosis is established, the system further identifies the risks and related factors, models them according to standardized labels, such as: pathological factors (e.g., gastrointestinal diseases, metabolic disorders), physiological factors (e.g., old age, dysphagia, postoperative status), psychosocial factors (e.g., depression, financial difficulties, loneliness), etc., and numbers them in a structured form (e.g., R1-fasting state, R3-dysphagia, S1-living alone) for the system logic engine to call.

[0147] Based on the diagnosis, the system automatically pushes nursing goal suggestions, such as: "Maintain the patient's weight change to no more than ±1kg within 5 days," "Increase daily calorie intake to more than 90% of the target value within 3 days," "Improve the NRS score by 1 point within 48 hours," or "The patient can correctly describe the appropriate dietary structure and nutritional supplementation method." These goals all have SMART characteristics (Specific, Measurable, Achievable, Relevant, and Time-bound), and allow nurses to further adjust them to suit the individual patient's situation.

[0148] The system then recommends nursing intervention packages corresponding to the diagnosis and risk label. These interventions cover dietary education, diet adjustments, nutritional supplementation, eating guidance, nutritional support pathway development (e.g., enteral / parenteral nutrition), and monitoring indicator settings. Each intervention is clearly labeled with its source (e.g., the *Chinese Dietary Guidelines*, *ASPEN Nutritional Support Guidelines*), strength of recommendation (strong recommendation / conditional recommendation), level of evidence (Level I systematic review / Level II randomized controlled trial, etc.), and indications. For example, for the "dysphagia" factor, the system might recommend "adopting a semi-liquid diet and assisting the patient with eating," with the pathway code "M-2," and can automatically link to relevant nursing monitoring items, such as the "Daily Food Intake Record."

[0149] Feedback from patients or nurses, such as "patient does not accept supplements" or "patient's food intake is still insufficient," will be entered into the system's feedback mechanism, triggering a measure adaptation scoring mechanism. Based on compliance scoring, expected effect assessment, and risk tolerance models, the system will optimize or automatically replace the original recommended measures (such as replacing oral supplements with nasogastric feeding).

[0150] After the intervention is implemented, the system will automatically extract corresponding monitoring indicators (such as weight trend charts, calorie intake statistics, and NRS score changes) based on the set nursing goals for effect evaluation, and feed the results back to the knowledge base for model optimization. The entire process realizes intelligent recommendation, closed-loop execution, and continuous optimization of the nursing pathway, significantly improving the scientific nature, timeliness, and execution efficiency of personalized nutritional care.

[0151] Secondly, this invention also proposes a nursing intervention recommendation device based on evidence-based nursing decision-making, such as... Figure 2 As shown, it includes:

[0152] The first acquisition unit 21 is used to acquire the health data information of the target patient;

[0153] Monitoring unit 22 is used to monitor the nursing trigger commands corresponding to the target patient;

[0154] The determining unit 23 is used to determine the risk label information of the target patient based on the health data information, a rule engine and a clustering algorithm, upon receiving the above-mentioned nursing trigger instruction.

[0155] The first generation unit 24 is used to generate nursing intervention recommendation information based on the above-mentioned risk label information;

[0156] The second acquisition unit 25 is used to send the above-mentioned nursing intervention recommendation information to the above-mentioned target patient in order to obtain the feedback information of the above-mentioned target patient.

[0157] The second generation unit 26 is used to generate target nursing measures based on the above feedback information and the above nursing measure recommendation information.

[0158] In one feasible implementation, the aforementioned nursing triggering instructions are generated based on one or more of the following methods: data-driven, semantic matching, and human system collaboration.

[0159] In one feasible implementation, the determination of the risk label information of the target patient based on the aforementioned health data information using a rule engine and clustering algorithm includes:

[0160] A three-dimensional risk labeling system is constructed based on the above health data information, which includes the patient's pathological data, physiological parameters, and socio-psychological information.

[0161] The rule engine is used to match the above patient health data to identify the risk label information of the target patients.

[0162] If the rule engine fails to match the risk label information, a clustering algorithm is used to analyze the feature space containing data from multiple patients, grouping patients with similar risk characteristics into the same category, and automatically labeling the target patients with their corresponding risk label information.

[0163] In one feasible implementation, the above-mentioned conditional matching of the patient health data information using a rule engine to identify the tag content corresponding to a specific risk rule includes:

[0164] A rule knowledge base is constructed, which includes multiple nursing risk identification rules. Each rule consists of prerequisites, triggering logic, and label output. The prerequisites are generated based on the patient's structured health information and text mining results.

[0165] Define a conditional expression language, wherein the conditional expression language adopts a rule expression syntax based on a domain-specific language, and supports nested logic, numerical comparison, range judgment, missing value fault tolerance, and multi-field joint triggering mechanism;

[0166] Load the aforementioned health data information of the target patient;

[0167] The rule matching process involves comparing each precondition in each rule one by one. If all preconditions are met, the rule is triggered, and the risk label set in the rule is added to the risk label set of the target patient.

[0168] In one feasible implementation, the above method further includes:

[0169] Obtain information on label usage frequency, measure compliance, and intervention results;

[0170] The rule knowledge base is dynamically adjusted based on the above-mentioned label usage frequency information, above-mentioned measure compliance information, and above-mentioned intervention results.

[0171] In one feasible implementation, the aforementioned recommendation information includes candidate nursing interventions corresponding to multiple risk labels. Each of the aforementioned nursing interventions includes an intervention description, implementation method, recommendation strength, level of evidence, and applicable conditions.

[0172] The feedback information mentioned above includes the target patients' acceptance of the recommended measures, willingness to implement them, past adherence, individual preferences, and history of adverse reactions.

[0173] The above-mentioned target nursing interventions are generated based on the above feedback information and the above-mentioned nursing intervention recommendations, including:

[0174] Based on the above feedback information and the above nursing intervention recommendations, an adaptability analysis was conducted to obtain compliance scores, expected outcome scores, and risk tolerance scores.

[0175] Based on the above compliance score, above expected effect score, and above risk tolerance score, candidate measures were screened and ranked, and all candidate nursing measures were comprehensively evaluated to obtain the above target nursing measures.

[0176] Thirdly, the present invention also proposes an electronic device 300, such as... Figure 3 As shown, it includes a memory 310, a processor 320, and a computer program 311 stored on the memory 310 and executable on the processor. When the processor 320 executes the computer program 311, it implements the steps of the method for recommending nursing interventions based on evidence-based nursing decisions as described in any of the first aspects.

[0177] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the method for recommending nursing interventions based on evidence-based nursing decision-making as described in any one of the first aspects.

[0178] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0179] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0180] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0181] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0182] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0183] This application also provides a computer program product, which includes computer software instructions. When the computer software instructions are executed on a processing device, the processing device performs the voice-based identity recognition process in the corresponding embodiment.

[0184] A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0185] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0186] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0187] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0188] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0189] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0190] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for recommending nursing interventions based on evidence-based nursing decision-making, characterized in that, include: Obtain health data information of the target patients; Monitor the nursing trigger commands corresponding to the target patient; Upon receiving the nursing trigger instruction, the risk label information of the target patient is determined based on the health data information using a rule engine and clustering algorithm; Based on the risk label information, nursing intervention recommendations are generated; The nursing intervention recommendation information is sent to the target patient to obtain feedback from the target patient; Target nursing interventions are generated based on the feedback information and the nursing intervention recommendation information.

2. The method for recommending nursing interventions based on evidence-based nursing decision-making according to claim 1, characterized in that, The acquisition of the target patient's health data information includes: Multi-source data acquired from hospital information systems, electronic medical records, and IoT devices; The multi-source data is parsed using natural language to obtain the health data information.

3. The method for recommending nursing interventions based on evidence-based nursing decision-making according to claim 1, characterized in that, The nursing trigger instructions are generated based on one or more of the following methods: data-driven, semantic matching, and human system collaboration.

4. The method for recommending nursing interventions based on evidence-based nursing decision-making according to claim 1, characterized in that, The step of determining the risk label information of the target patient based on the health data information using a rule engine and clustering algorithm includes: A three-dimensional risk labeling system is constructed based on the aforementioned health data information, wherein the health data information includes the patient's pathological data, physiological parameters, and socio-psychological information; The patient's health data is matched using a rule engine to identify the risk label information of the target patient. If the rule engine does not match the risk label information, the feature space containing multiple patient data is analyzed based on the clustering algorithm, patients with similar risk characteristics are clustered into the same category, and the target patient is automatically labeled with its corresponding risk label information.

5. The method for recommending nursing interventions based on evidence-based nursing decision-making according to claim 1, characterized in that, The step of performing conditional matching on the patient's health data information through a rule engine to identify the tag content corresponding to specific risk rules includes: A rule knowledge base is constructed, wherein the knowledge base includes multiple nursing risk identification rules, each rule consists of preconditions, triggering logic and label output, and the preconditions are generated based on the patient's structured health information and text mining results; Define a conditional expression language, wherein the conditional expression language adopts a rule expression syntax based on a domain-specific language, and the conditional expression language supports nested logic, numerical comparison, range judgment, missing value fault tolerance, and multi-field joint triggering mechanism; Load the health data information of the target patient; The rule matching process involves comparing each precondition in each rule one by one. If all preconditions are met, the rule is triggered, and the risk label set in the rule is added to the risk label set of the target patient.

6. The method for recommending nursing interventions based on evidence-based nursing decision-making according to claim 5, characterized in that, The method further includes: Obtain information on label usage frequency, measure compliance, and intervention results; The rule knowledge base is dynamically adjusted based on the frequency of tag usage, compliance with the measures, and the intervention results.

7. The method for recommending nursing interventions based on evidence-based nursing decision-making according to claim 1, characterized in that, The recommended information includes candidate nursing interventions corresponding to various risk labels. Each nursing intervention includes a description of the intervention, implementation method, recommendation strength, level of evidence, and applicable conditions. The feedback information includes the target patient's acceptance of the recommended measures, willingness to implement them, past adherence, individual preferences, and history of adverse reactions. The generation of target nursing interventions based on the feedback information and the nursing intervention recommendation information includes: Based on the feedback information and the nursing intervention recommendations, an adaptability analysis was performed to obtain compliance scores, expected outcome scores, and risk tolerance scores. Candidate measures are screened and ranked based on the compliance score, the expected effect score, and the risk tolerance score. All candidate nursing measures are then comprehensively evaluated to obtain the target nursing measure.

8. A nursing intervention recommendation system based on evidence-based nursing decision-making, characterized in that, include: The first acquisition unit is used to acquire the health data information of the target patient; The monitoring unit is used to monitor the nursing trigger commands corresponding to the target patient; The determining unit is configured to, upon receiving the nursing trigger instruction, determine the risk label information of the target patient based on the health data information using a rule engine and a clustering algorithm; The first generation unit is used to generate nursing intervention recommendation information based on the risk label information; The second acquisition unit is used to send the nursing intervention recommendation information to the target patient in order to obtain the feedback information of the target patient; The second generation unit is used to generate target nursing measures based on the feedback information and the nursing measure recommendation information.

9. An electronic device, comprising: The memory and processor are characterized in that the processor, when executing a computer program stored in the memory, implements the steps of the method for recommending nursing interventions based on evidence-based nursing decision-making as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for recommending nursing interventions based on evidence-based nursing decision-making as described in any one of claims 1-7.

Images