Clinical nursing personnel management method and system
By acquiring nursing staff and demand data, using neural networks to predict new demands and optimize scheduling plans, the problem of irrational resource allocation in traditional scheduling methods is solved, accurate matching of nursing staff and patients is achieved, and the quality and flexibility of nursing care are improved.
Patent Information
- Application Number
- CN202510682376.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-26
- Publication Date
- 2025-10-10
AI Technical Summary
The existing nursing staff scheduling method lacks the ability to dynamically respond to changes in actual nursing needs, is difficult to adapt to fluctuations in clinical nursing load, and does not fully consider the nursing staff's proficiency in nursing different types of diseases, resulting in irrational resource allocation and reduced nursing quality.
By acquiring nursing staff data and nursing demand data, using long-short-term memory neural networks to predict new demands, combined with nursing staff proficiency, a scheduling recommendation model is used for dynamic optimization, and the scheduling plan is periodically adjusted through a dynamic feedback module to achieve accurate matching of nursing staff and patients.
It improves the pertinence and quality of nursing services, increases the flexibility and adaptability of scheduling, reduces the shortage and irrational allocation of nursing resources, and reduces the probability of medical accidents.
Smart Images
Figure CN120766893A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of clinical nursing staff scheduling management, and in particular to a clinical nursing staff management method and system. Background Art
[0002] As the workload of clinical nursing continues to grow, the importance of nursing staff scheduling management has become increasingly prominent. Reasonable scheduling can not only improve the quality of nursing services, but also effectively relieve the workload of nursing staff and reduce the probability of medical accidents.
[0003] However, existing nursing staff scheduling methods mostly rely on the experience of managers or simple rule-setting, and have the following major problems: traditional scheduling methods are usually based on a fixed shift system, lack the ability to dynamically respond to changes in actual nursing needs, and find it difficult to adapt to fluctuations in clinical nursing load in a timely manner. When facing daily changes in patients, there is a lack of effective prediction and adjustment mechanisms, which can easily lead to shortages or unreasonable allocation of nursing resources; secondly, the current scheduling process often fails to fully consider the proficiency of nursing staff in nursing different types of diseases, which can easily lead to a mismatch between nursing skills and patient needs, thereby affecting the quality of care.
[0004] Therefore, there is an urgent need for a nursing staff management method that can intelligently predict and dynamically optimize nursing staff characteristics, patient disease types, and future changes in nursing needs, so as to achieve more flexible, reliable, and adaptive nursing staff scheduling management. Summary of the Invention
[0005] The purpose of the present invention is to provide a clinical nursing staff management method and system, which can achieve more flexible, reliable and adaptive nursing staff scheduling management.
[0006] The present invention is achieved through the following technical solutions: A clinical nursing staff management method comprises the following steps: Acquiring clinical nursing staff data, including all nursing staff, the corresponding historical average daily number of nursing patients, and nursing proficiency for each disease type; Obtain clinical nursing demand data for the next scheduling cycle from the current database, wherein the clinical nursing demand data includes all objects to be cared for and corresponding disease types; Predicting new clinical nursing demand data in the next scheduling cycle using a change prediction model for patients to be cared for, wherein the new clinical nursing demand data includes the predicted daily increase in patients to be cared for by disease type; Obtain clinical nursing staff arrangements for the next scheduling cycle through a scheduling recommendation model based on clinical nursing staff data, clinical nursing demand data, and newly added clinical nursing demand data; The scheduling suggestion model is periodically optimized through the dynamic feedback module according to the working conditions after scheduling.
[0007] Preferably, the method for obtaining the historical average daily number of nursing objects is to obtain the number by obtaining the work records within a historical time period ending at the current time.
[0008] Preferably, the method for obtaining the nursing proficiency of the nursing staff for each disease type field is to perform the following operations for each nursing staff: According to the work record of the nursing staff within a historical time period up to the present, the number of patients with various disease types that the nursing staff has cared for in historical experience is obtained; Obtain the ratio of the number of patients to be cared for with each disease type cared for by the i-th caregiver to the total number of patients to be cared for: , ; in, is the ratio of the number of patients with the jth disease type cared for by the i-th caregiver to the total number of patients cared for, and represents the number of patients with the jth and j'th disease types cared for by the i-th caregiver, where N is the number of disease types; Normalize the relative values to get the nursing proficiency of the i-th nurse for the j-th disease type. : .
[0009] Preferably, the method for predicting the new clinical nursing demand data in the next scheduling cycle by using the change prediction model for the objects to be cared for is: Get the current Day Tiandi The number of new patients requiring care for each disease type , , , forming a time series of newly added numbers, where N is the number of disease types; By setting the sliding window length of the model to And set the prediction step to 1, and predict the next shift schedule in each day based on the long short-term memory neural network. The number of people in need of care for each disease type will increase.
[0010] Preferably, the scheduling suggestion model includes: An input module, configured to receive the clinical nursing staff database, the clinical nursing demand data, and the newly added clinical nursing demand data; A first feature extraction module is used to extract features based on the newly added clinical nursing demand data to obtain a first feature; A second feature extraction module is used to extract features based on the clinical nursing staff database and the clinical nursing demand data to obtain a second feature; A decision module is used to obtain the clinical nursing staff arrangement for the next scheduling period based on the first feature and the second feature.
[0011] Preferably, the method for extracting the first feature from the clinical nursing staff database is: Get the next shift cycle Tianzhongdi The number of new patients requiring care under the disease type , , The number of days in the next scheduling cycle, , V, N is the number of disease types; Get the first Average daily number of caregivers for patients with different disease types Demand and average length of stay ; Get the next shift cycle Tianzhongdi Estimated additional nursing staff demand for each disease type : ; Among them, if is a truth function; Constructing the Matrix As the first feature, the element in the uth row and jth column of the first feature is .
[0012] Preferably, the method for extracting features from the clinical nursing staff database and the clinical nursing demand data to obtain the second feature is: The second feature includes a nursing proficiency matrix describing the proficiency of nursing staff for each disease type and an estimated current nursing staff number requirement matrix describing the daily nursing staff demand for each disease type; Obtain the nursing proficiency matrix of the nursing staff , Nursing Proficiency Matrix The element in row i and column j of , represents the nursing proficiency of the i-th nurse in the field of the j-th disease type; Get the next shift cycle Tianzhongdi Estimated current nursing staff demand matrix for disease types , the element in the uth row and jth column of the matrix is : ; in, Representative The remaining hospitalization days of the xth nursing staff of the disease type in the next scheduling cycle, For the The total number of people currently waiting for care for each disease type.
[0013] Preferably, the decision-making method of the decision-making module is: Step S801: Initialization , i=1,2,…,L, L is the total number of nursing staff, Representative The number of remaining days of care provided by the nursing staff mentioned above, For the The average number of persons to be cared for by the nursing staff on a daily basis in history; Step S802: Initialize u=1, where u represents the first sky; Step S803: Initialization , s represents the number of the nursing object, and the order of the nursing object number is arranged from high to low according to the urgency of the nursing object's condition; Step S804: Obtain the disease type of the patient to be cared for s and the number of caregivers required daily ; Get the first Evaluation parameters of the nursing staff for the nursing object s : ; ;
[0014] in, and are the first training weight and the first training bias respectively, and are the second training weight and the second training bias respectively, and is the intermediate parameter; On day u, The largest value The nursing staff is assigned to the nursing object s, and the number of the assigned nursing staff is ; Step S805: Update the first value: ; ; ; …; ; Step S806: If , Y is the total number of objects to be cared for, then return to step S804, otherwise perform the second value update: u=u+1, and proceed to the next step; Step S807: If Then return to step S803, otherwise terminate the process.
[0015] Preferably, the method for optimizing the scheduling suggestion model according to the working conditions after scheduling by the dynamic feedback module is: Establish an optimized objective function based on the actual schedule : ; ; ; Among them, max represents the maximization function Represents the caregiver for the care recipient s on day u in the next shift cycle The disease type of the patient to be cared for proficiency, The number of days in the next shift cycle. is the total number of the objects to be cared for on the uth day in the next shift cycle, e is a natural constant, is the exponential decay parameter, is the intermediate parameter, For the The number of people to be cared for arranged by the nursing staff on the uth day in the next scheduling cycle, For the The average number of persons to be cared for by the nursing staff on a daily basis in history.
[0016] The present invention also provides a clinical nursing staff management system, which is applied to the above-mentioned clinical nursing staff management method, comprising: A clinical nursing staff data acquisition module is used to acquire clinical nursing staff data, wherein the clinical nursing staff data includes all nursing staff, the corresponding historical average daily number of nursing objects, and the nursing proficiency for each disease type; A clinical nursing demand data acquisition module is used to obtain clinical nursing demand data for the next scheduling cycle from the current database, wherein the clinical nursing demand data includes all objects to be cared for and corresponding disease types; A prediction module is used to predict new clinical nursing demand data in the next scheduling cycle using a change prediction model for patients to be cared for. The new clinical nursing demand data includes the predicted daily increase in patients to be cared for by disease type. The scheduling suggestion module is used to obtain the clinical nursing staff arrangement for the next scheduling cycle through the scheduling suggestion model based on the clinical nursing staff data, clinical nursing demand data and newly added clinical nursing demand data; The model optimization module is used to periodically optimize the scheduling suggestion model according to the work status after scheduling through the dynamic feedback module.
[0017] The technical solution of the present invention has at least the following advantages and beneficial effects: Based on the historical average daily number of nursing recipients and nursing proficiency information of nursing staff, the present invention conducts a comprehensive analysis in combination with nursing needs, thus achieving a precise match between nursing staff and patients with different disease types, and improving the relevance and quality of nursing services. The present invention uses a change prediction model for the patients to be cared for to estimate in advance the situation of the patients to be cared for in the next scheduling cycle, so as to facilitate the targeted reservation of appropriate manpower through subsequent data processing, making the scheduling more forward-looking and proactive; The present invention evaluates the scheduling execution status by periodically calling the dynamic feedback module and continuously optimizes the scheduling suggestion model to form a closed-loop management mechanism, thereby achieving continuous improvement and intelligent evolution of nursing scheduling. When scheduling, the present invention not only takes into account the current situation of the patients to be cared for, but also takes into account the dynamic changes of the patients to be cared for in the future, thus having greater flexibility and adaptability. The present invention is rationally designed and can effectively address the deficiencies of traditional nursing scheduling methods in terms of demand adaptation, skill matching, and dynamic optimization, thereby achieving an efficient, intelligent, and sustainable management solution. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1 A schematic diagram of a flow chart of a clinical nursing staff management method provided in Example 1 of the present invention; Figure 2 This is a schematic diagram of the principle of the scheduling suggestion model provided in Example 1 of the present invention; Figure 3This is a schematic diagram of the principles of a clinical nursing staff management system provided in Example 2 of the present invention. DETAILED DESCRIPTION
[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations.
[0020] Example 1 This embodiment provides a clinical nursing staff management method, see Figure 1 , including the following steps: Acquiring clinical nursing staff data, including all nursing staff, the corresponding historical average daily number of nursing patients, and nursing proficiency for each disease type; Obtain clinical nursing demand data for the next scheduling cycle from the current database, wherein the clinical nursing demand data includes all objects to be cared for and corresponding disease types; Predicting new clinical nursing demand data in the next scheduling cycle using a change prediction model for patients to be cared for, wherein the new clinical nursing demand data includes the predicted daily increase in patients to be cared for by disease type; Obtain clinical nursing staff arrangements for the next scheduling cycle through a scheduling recommendation model based on clinical nursing staff data, clinical nursing demand data, and newly added clinical nursing demand data; The scheduling suggestion model is periodically optimized through the dynamic feedback module according to the working conditions after scheduling.
[0021] The embodiment can first obtain information of all current clinical nursing staff from a human resource database, including historical daily nursing object quantity of each nursing staff, and their proficiency in various disease nursing tasks, wherein the historical daily nursing object quantity reflects their work load bearing capacity, and the proficiency reflects the disease field they are good at, and the above data is used as the basic input for nursing resource capacity assessment. Then, the existing nursing demand data of the next scheduling period (such as one week or two weeks) can be extracted from the hospital database, including the basic information of all nursing objects and their corresponding disease types (such as diabetes, heart disease, postoperative rehabilitation, etc.), to form the clinical nursing demand data. Then, the nursing object change prediction model predicts the number of new nursing objects that may appear in the next scheduling period, so as to reserve human resources for the predicted new objects in the subsequent processing, and improve the dynamic adaptability of the whole system. Based on the above data, the scheduling suggestion model can be called to generate the nursing staff scheduling suggestion of the next scheduling period. In addition, the dynamic feedback module can periodically evaluate the actual scheduling effect, and optimize the parameters of the scheduling suggestion model accordingly, to improve the accuracy and rationality of future scheduling.
[0022] As a preferred solution, the method for obtaining the historical average daily nursing nursing object quantity is to obtain the work records within a historical time period ending at the present.
[0023] On the other hand, the method for obtaining the nursing proficiency of the nursing staff for each disease type field is to perform the following operations for each nursing staff respectively: According to the work records of the nursing staff within a historical time period ending at the present, the number of nursing objects of various disease types cared for by the nursing staff in historical experience is obtained; The proportion of the number of nursing objects of each disease type cared for by the i-th nursing staff to the total number of nursing objects cared for is obtained respectively: , ; Wherein, is the proportion of the number of nursing objects of the j-th disease type cared for by the i-th nursing staff to the total number of nursing objects cared for, and represent the number of nursing objects of the j-th and j'-th disease types cared for by the i-th nursing staff, and N is the number of disease types; The relative value is normalized to obtain the nursing proficiency of the i-th nursing staff for the j-th disease type field : .
[0024] The above two sets of data are obtained based on the work records within a historical time period up to the present, and the latest data status can be obtained. This embodiment first normalizes the actual amount of care provided by nursing staff in different disease types, which can fully reflect their relative proficiency in multiple disease areas, making it easier for the system to perform more reasonable scheduling and task matching based on actual capabilities. The proficiency is then calculated in the form of a relative proportion, which effectively avoids the deviation caused by evaluating nursing experience only by absolute numbers, and improves the scientific nature and objectivity of the proficiency assessment. Inputting the proficiency index of this embodiment into the subsequent scheduling recommendation model can effectively improve the model's intelligent matching ability in the task allocation process, making nursing work arrangements more accurate and efficient.
[0025] Next, the method for predicting the new clinical nursing demand data in the next scheduling cycle by using the change prediction model for the nursing objects is as follows: Get the current Day Tiandi The number of new patients requiring care for each disease type , , , forming a time series of newly added numbers, where N is the number of disease types; By setting the sliding window length of the model to And set the prediction step to 1, and predict the next shift schedule in each day based on the long short-term memory neural network. The number of people in need of care for each disease type will increase.
[0026] Then, see Figure 2 , the scheduling suggestion model preferably includes: An input module, configured to receive the clinical nursing staff database, the clinical nursing demand data, and the newly added clinical nursing demand data; A first feature extraction module is used to extract features based on the newly added clinical nursing demand data to obtain a first feature; A second feature extraction module is used to extract features based on the clinical nursing staff database and the clinical nursing demand data to obtain a second feature; A decision module is used to obtain the clinical nursing staff arrangement for the next scheduling period based on the first feature and the second feature.
[0027] In the above scheme, the first feature extraction module extracts key features based on the predicted new nursing demand data, which can respond to potential changes in nursing needs in advance and improve the initiative and adaptability of scheduling results. The second feature extraction module generates optimized features based on the comprehensive nursing staff ability data and current nursing needs, providing more targeted feature information to the decision-making module, thereby improving the scientificity and accuracy of scheduling recommendations. Combining the first and second features, while ensuring the reasonable and high-quality allocation of current nursing staff, it is possible to flexibly reserve spare human resources based on characteristics such as the number and disease type of new nursing objects that may be added in the next half-cycle.
[0028] Furthermore, the method for extracting the first feature from the clinical nursing staff database is: Get the next shift cycle Tianzhongdi The number of new patients requiring care under the disease type , , The number of days in the next shift cycle. , V, N is the number of disease types; Get the first Average daily number of caregivers for patients with different disease types Demand and average length of stay ; Get the next shift cycle Tianzhongdi Estimated additional nursing staff demand for each disease type : ; Among them, if is a truth function; Constructing the Matrix As the first feature, the element in the uth row and jth column of the first feature is .
[0029] The above feature extraction not only considers the number of newly added nursing patients per day, but also introduces the average length of hospital stay as a parameter, which effectively reflects the nursing resource utilization of patients over multiple consecutive days, helping to further improve the timeliness and accuracy of scheduling predictions. Among them, if(condition) is a truth function that returns 1 if the condition is met and 0 otherwise. This setting is used to determine whether the newly added patient is still within the expected hospitalization care period. That is, based on the average length of hospital stay Reflects whether the newly added nursing objects have been discharged on the corresponding date. If It is judged that the patient has been discharged from the hospital, so the demand for new nursing staff for this patient will become zero. At the same time, the average daily demand for nursing staff for each patient of each disease type is introduced. The average length of stay is used as the weight and window period basis, and the shift rotation of nurses within a single day is taken into account. When a nurse is assigned to a single person to be cared for each day, the result can be given in the form of a group of people who take care of the person to be cared for. The specific time scheduling can be flexibly arranged according to demand, so that the system can automatically adapt to the nursing demand characteristics of different diseases according to the long-term operating data of the hospital department.
[0030] On this basis, the method for extracting features from the clinical nursing staff database and the clinical nursing demand data to obtain the second feature is: The second feature includes a nursing proficiency matrix describing the proficiency of nursing staff for each disease type and an estimated current nursing staff number requirement matrix describing the daily nursing staff demand for each disease type; Obtain the nursing proficiency matrix of the nursing staff , Nursing Proficiency Matrix The element in row i and column j of , represents the nursing proficiency of the i-th nurse in the field of the j-th disease type; Get the next shift cycle Tianzhongdi Estimated current nursing staff demand matrix for disease types , the element in the uth row and jth column of the matrix is : ; in, Representative The remaining hospitalization days of the xth nursing staff of the disease type in the next scheduling cycle, For the The total number of people currently waiting for care for each disease type.
[0031] In the above scheme, the nursing proficiency matrix It helps to structure the storage and extraction of the distribution characteristics of the nursing capabilities of different nursing staff in different disease areas. Based on the current nursing object information (i.e., patients who have not yet been discharged), the daily nursing needs in the next scheduling cycle can be obtained to form an estimated current nursing staff demand matrix. The estimated daily current nursing staff demand matrix is obtained by traversing all the corresponding nursing objects for each disease type and taking into account their remaining hospitalization days. Similarly, the use of the if function above, The main consideration is whether the nursing staff has been discharged on the uth day. On this basis, the current nursing staff demand matrix is estimated By tracking the remaining hospitalization time of each patient to be cared for on a daily basis, the nursing load of the future scheduling cycle can be dynamically estimated, which is close to the actual medical care scenario and effectively supports the rationalization of daily manpower deployment. In other words, the nursing proficiency matrix Based on historical experience, the current nursing staff demand matrix is estimated Based on the current hospitalization status, the combination of the two enables the model to utilize historical experience and extract features based on current needs, thereby enhancing the adaptability and reliability of scheduling.
[0032] In summary, the first and second features comprehensively reflect the dynamic changes in the number of patients waiting for care during the scheduling cycle, taking into account not only discharges but also admissions. Therefore, the features extracted in this implementation have greater dynamic adaptability, helping the decision-making module to obtain more flexible and targeted scheduling recommendations.
[0033] In this embodiment, the decision-making method of the decision module is: Step S801: Initialization , i=1,2,…,L, L is the total number of nursing staff, Representative The number of remaining days of care provided by the nursing staff mentioned above, For the The average number of persons to be cared for by the nursing staff on a daily basis in history; Step S802: Initialize u=1, where u represents the first sky; Step S803: Initialization , s represents the number of the nursing object, and the order of the nursing object number is arranged from high to low according to the urgency of the nursing object's condition; Step S804: Obtain the disease type of the patient to be cared for s and the number of caregivers required daily ; Get the first Evaluation parameters of the nursing staff for the nursing object s : ; ;
[0034] in, and are the first training weight and the first training bias respectively, and are the second training weight and the second training bias respectively, and is the intermediate parameter; On day u, The largest value The nursing staff is assigned to the nursing object s, and the number of the assigned nursing staff is ; Step S805: Update the first value: ; ; ; …; ; Step S806: If , Y is the total number of objects to be cared for, then return to step S804, otherwise perform the second value update: u=u+1, and proceed to the next step; Step S807: If Then return to step S803, otherwise terminate the process.
[0035] The decision module in this embodiment is based on the historical ability indicators of the nursing staff, the current workload and the actual nursing needs of the nursing recipients, combined with the comprehensive evaluation index ϑ calculated by the Sigmoid function in the neural network , realizing dynamic and accurate intelligent scheduling of nursing staff. Specifically, the comprehensive evaluation index ϑ Comprehensively introduce the nursing proficiency of nurses for target disease types and the current remaining number of nurses per day based on the historical load capacity of the nurses and the real-time allocation status At the same time, it also takes into account the remaining allocable resources with sufficiently high proficiency in specific disease types and the estimated number of new patients in need of care for the corresponding disease types, ensuring that the matched nursing staff are both professional enough and have a certain amount of spare capacity to provide for the predicted new patients in need of care, thereby improving the scientificity and adaptability of the scheduling. It can reflect the ratio of the current nursing resource demand of the number of patients with a specific disease type to the nursing resource demand of the newly added patients to be cared for within a specified day. The higher the ratio, the higher the score of matching the most skilled nurses for patients with this disease type, and improve the foresight and the quality of the allocation of newly added patients to be cared for. The patients to be cared for are sorted by the urgency of their condition, and resources can be allocated to those who are in urgent need of care first, thus improving the safety of patients. Each time a nurse is assigned, the current remaining number of nurses for each day is calculated. Instant updates prevent overload on the same nursing staff, embodying a user-friendly scheduling design that helps reduce nursing fatigue and errors. The scheduling process is based on a nested loop structure, with a two-tier progression based on day (u) and care recipient (s). This logically clear and transparent process can be iteratively optimized using methods such as reinforcement learning, demonstrating strong system scalability.
[0036] Finally, the method for optimizing the scheduling suggestion model according to the work status after scheduling by the dynamic feedback module is preferably: Establish an optimized objective function based on the actual schedule : ; ; ; Among them, max represents the maximization function Represents the caregiver for the care recipient s on day u in the next shift cycle The disease type of the patient to be cared for proficiency, The number of days in the next scheduling cycle, is the total number of the objects to be cared for on the uth day in the next shift cycle, e is a natural constant, is the exponential decay parameter, is the intermediate parameter, For the The number of people to be cared for arranged by the nursing staff on the uth day in the next scheduling cycle, For the The average number of persons to be cared for by the nursing staff on a daily basis in history.
[0037] The above scheme can be used as the optimization objective function to adjust the training weights and training biases in the decision module. The optimization objective function TA takes into account the nursing proficiency of the nurses in the assigned disease type. The degree of deviation from its workload , to maximize the two-way adaptation between nursing capabilities and task allocation, and fundamentally improve the rationality and scientificity of scheduling. It can effectively control the importance of scheduling fairness. When the workload is unbalanced compared to normal work, the attenuation value is significant, thus punishing the situation of over- and under-scheduling. When , it will represent the current workload of each person and represents the historical average daily nursing capacity By incorporating deviation values into the evaluation, the system effectively prevents individual nursing staff from being over-assigned tasks, thereby improving team satisfaction and execution stability.
[0038] Example 2 This embodiment provides a clinical nursing staff management system, which is applied to a clinical nursing staff management method of the above embodiment. Figure 3 ,include: A clinical nursing staff data acquisition module is used to acquire clinical nursing staff data, wherein the clinical nursing staff data includes all nursing staff, the corresponding historical average daily number of nursing objects, and the nursing proficiency for each disease type; A clinical nursing demand data acquisition module is used to obtain clinical nursing demand data for the next scheduling cycle from the current database, wherein the clinical nursing demand data includes all objects to be cared for and corresponding disease types; A prediction module is used to predict new clinical nursing demand data in the next scheduling cycle using a change prediction model for patients to be cared for. The new clinical nursing demand data includes the predicted daily increase in patients to be cared for by disease type. The scheduling suggestion module is used to obtain the clinical nursing staff arrangement for the next scheduling cycle through the scheduling suggestion model based on the clinical nursing staff data, clinical nursing demand data and newly added clinical nursing demand data; The model optimization module is used to periodically optimize the scheduling suggestion model according to the work status after scheduling through the dynamic feedback module.
[0039] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.
Claims
1. A clinical nursing staff management method, characterized in that: The following steps are involved: Acquiring clinical nursing staff data, including all nursing staff, the corresponding historical average daily number of nursing patients, and nursing proficiency for each disease type; Obtain clinical nursing demand data for the next scheduling cycle from the current database, wherein the clinical nursing demand data includes all objects to be cared for and corresponding disease types; Predicting new clinical nursing demand data in the next scheduling cycle using a change prediction model for patients to be cared for, wherein the new clinical nursing demand data includes the predicted daily increase in patients to be cared for by disease type; Obtain clinical nursing staff arrangements for the next scheduling cycle through a scheduling recommendation model based on clinical nursing staff data, clinical nursing demand data, and newly added clinical nursing demand data; The scheduling suggestion model is periodically optimized through the dynamic feedback module according to the working conditions after scheduling.
2. A clinical nursing staff management method according to claim 1, characterized in that: The method for obtaining the historical average daily number of nursing objects is to obtain the number by obtaining the work records within a historical time period ending at the current time.
3. A clinical nursing staff management method according to claim 1, characterized in that: The method for obtaining the nursing proficiency of the nursing staff for each disease type field is to perform the following operations for each nursing staff: According to the work record of the nursing staff within a historical time period up to the present, the number of patients with various disease types that the nursing staff has cared for in historical experience is obtained; Obtain the ratio of the number of patients to be cared for with each disease type cared for by the i-th caregiver to the total number of patients to be cared for: ， ； in, is the ratio of the number of patients with the jth disease type cared for by the i-th caregiver to the total number of patients cared for, and represents the number of patients with the jth and j'th disease types cared for by the i-th caregiver, where N is the number of disease types; Normalize the relative values to get the nursing proficiency of the i-th nurse for the j-th disease type. : 。 4. A clinical nursing staff management method according to claim 1, characterized in that: The method for predicting the new clinical nursing demand data in the next scheduling cycle by using the change prediction model for the nursing objects is as follows: Get the current Day Tiandi The number of new patients requiring care for each disease type , , , forming a time series of newly added numbers, where N is the number of disease types; By setting the sliding window length of the model to And set the prediction step to 1, and predict the next shift schedule in each day based on the long short-term memory neural network. The number of people in need of care for each disease type will increase.
5. A clinical nursing staff management method according to claim 1, characterized in that: The scheduling recommendation model includes: An input module, configured to receive the clinical nursing staff database, the clinical nursing demand data, and the newly added clinical nursing demand data; A first feature extraction module is used to extract features based on the newly added clinical nursing demand data to obtain a first feature; A second feature extraction module is used to extract features based on the clinical nursing staff database and the clinical nursing demand data to obtain a second feature; A decision module is used to obtain the clinical nursing staff arrangement for the next scheduling period based on the first feature and the second feature.
6. A clinical nursing staff management method according to claim 5, characterized in that: The method for extracting the first feature from the clinical nursing staff database is: Get the next shift cycle Tianzhongdi The number of new patients requiring care under the disease type , , The number of days in the next shift cycle. , V, N is the number of disease types; Get the first Average daily number of caregivers for patients with different disease types Demand and average length of stay ; Get the next shift cycle Tianzhongdi Estimated additional nursing staff demand for each disease type : ； Among them, if is a truth function; Constructing the Matrix As the first feature, the element in the uth row and jth column of the first feature is .
7. A clinical nursing staff management method according to claim 6, characterized in that: The method for extracting features from the clinical nursing staff database and the clinical nursing demand data to obtain the second feature is: The second feature includes a nursing proficiency matrix describing the proficiency of nursing staff for each disease type and an estimated current nursing staff number requirement matrix describing the daily nursing staff demand for each disease type; Obtain the nursing proficiency matrix of the nursing staff , Nursing Proficiency Matrix The element in row i and column j of , represents the nursing proficiency of the i-th nurse in the field of the j-th disease type; Get the next shift cycle Tianzhongdi Estimated current nursing staff demand matrix for disease types , the element in the uth row and jth column of the matrix is : ； in, Representative The remaining hospitalization days of the xth nursing staff of the disease type in the next scheduling cycle, For the The total number of people currently waiting for care for each disease type.
8. A clinical nursing staff management method according to claim 7, characterized in that: The decision-making method of the decision-making module is: Step S801: Initialization , i=1,2,…,L, L is the total number of nursing staff, Representative The number of remaining days of care provided by the nursing staff mentioned above, For the The average number of persons to be cared for by the nursing staff on a daily basis in history; Step S802: Initialize u=1, where u represents the first sky; Step S803: Initialization , s represents the number of the nursing object, and the order of the nursing object number is arranged from high to low according to the urgency of the nursing object's condition; Step S804: Obtain the disease type of the patient to be cared for s and the number of caregivers required daily ; Get the first Evaluation parameters of the nursing staff for the nursing object s : ； ； in, and are the first training weight and the first training bias respectively, and are the second training weight and the second training bias respectively, and is the intermediate parameter; On day u, The largest value The nursing staff is assigned to the nursing object s, and the number of the assigned nursing staff is ; Step S805: Update the first value: ; ; ; …; ; Step S806: If , Y is the total number of objects to be cared for, then return to step S804, otherwise perform the second value update: u=u+1, and proceed to the next step; Step S807: If Then return to step S803, otherwise terminate the process.
9. A clinical nursing staff management method according to claim 1, characterized in that: The method for optimizing the scheduling suggestion model according to the working conditions after scheduling by the dynamic feedback module is: Establish an optimized objective function based on the actual schedule : ； ； ； Among them, max represents the maximization function Represents the caregiver for the care recipient s on day u in the next shift cycle The disease type of the patient to be cared for proficiency, The number of days in the next shift cycle. is the total number of the objects to be cared for on the uth day in the next shift cycle, e is a natural constant, is the exponential decay parameter, is the intermediate parameter, For the The number of people to be cared for arranged by the nursing staff on the uth day in the next scheduling cycle, For the The average number of persons to be cared for by the nursing staff on a daily basis in history.
10. A clinical nursing staff management system, applied to a clinical nursing staff management method according to any one of claims 1 to 9, characterized in that: include: A clinical nursing staff data acquisition module is used to acquire clinical nursing staff data, wherein the clinical nursing staff data includes all nursing staff, the corresponding historical average daily number of nursing objects, and the nursing proficiency for each disease type; A clinical nursing demand data acquisition module is used to obtain clinical nursing demand data for the next scheduling cycle from the current database, wherein the clinical nursing demand data includes all objects to be cared for and corresponding disease types; A prediction module is used to predict new clinical nursing demand data in the next scheduling cycle using a change prediction model for patients to be cared for. The new clinical nursing demand data includes the predicted daily increase in patients to be cared for by disease type. The scheduling suggestion module is used to obtain the clinical nursing staff arrangement for the next scheduling cycle through the scheduling suggestion model based on the clinical nursing staff data, clinical nursing demand data and newly added clinical nursing demand data; The model optimization module is used to periodically optimize the scheduling suggestion model according to the work status after scheduling through the dynamic feedback module.