Patient pathological data analysis and evaluation method for clinical nursing

Abstract

The invention provides a patient pathological data analysis and evaluation method for clinical nursing, and relates to the technical field of nursing informatization. Comprising the steps of collecting biochemical and pathological data of a patient and standardizing units and reference values, analyzing the dynamic trend of detection values and classifying change directions and amplitudes, matching pathological labels and establishing association with nursing items, comparing differences between nursing records and the pathological data and establishing a transverse index relationship, and judging whether the nursing task difference exceeds a threshold value or not and adjusting a nursing path generation scheme. By integrating multi-source pathological data and standardizing units and reference intervals, data consistency and comparability are ensured, the index fluctuation trend is dynamically tracked, the change direction and amplitude are quantitatively analyzed, disease course evolution is accurately captured, direct mapping between pathological categories and nursing items is established, logic binding is clear, intervention precision is improved, and the method is suitable for clinical application. Differences between nursing records and pathological data are transversely compared, nursing dynamic adaptation is optimized, a path scheme is rapidly adjusted through threshold judgment, and the lagging or misjudgment risk is reduced.

Description

Technical Field

[0001] The present invention relates to the field of nursing information technology, and in particular to a method for analyzing and evaluating patient pathology data for clinical nursing. Background Art

[0002] The field of nursing information technology includes the technical content of the entire process of information collection, management, processing and analysis in nursing services. Its core lies in the systematic recording, storage, analysis and call of data generated in the clinical nursing process through information technology to assist nursing decision-making, optimize nursing processes and improve nursing quality. This technical field mainly covers electronic nursing records, nursing process management, nursing decision support systems, patient health record maintenance, and the integration and interoperability of nursing data. Nursing information technology emphasizes the deep integration of nursing business and information systems, improves data utilization efficiency through structured data formats, standard interfaces and intelligent processing technologies, and promotes the scientific and standardized nursing work.

[0003] Among them, the patient pathology data analysis and evaluation method for clinical nursing refers to a technical solution for collecting, classifying, statistically modeling and evaluating and analyzing patient pathology data in clinical nursing scenarios. The patent subject mainly focuses on the centralized collection of biochemical test data such as blood routine, liver and kidney function, electrolyte levels, and tissue pathology section data generated by patients during the nursing process, and through the set data mapping relationship, the pathological indicators from different sources are numerically quantified and graded according to the unified evaluation standards, and then an evaluation time series table is constructed in combination with the time nodes of disease progression to assist nursing staff in status judgment and priority sorting when formulating nursing plans. This process generally completes the construction of the evaluation model by setting the indicator reference interval, establishing the indicator scoring function, and designing multiple logical judgment criteria, and the statistical program completes the output of the evaluation results.

[0004] Existing technologies rely on unified assessment standards for numerical quantification during the data integration phase, but fail to explicitly address the standardization of multi-source testing units and reference intervals. This results in the need for additional manual calibration when integrating data across platforms or institutions, increasing operational complexity and the risk of error. Pathology data analysis focuses on static indicator classification and scoring function design, lacking dynamic trend tracking of continuous test values. This makes it difficult to capture indicator fluctuation patterns and potential signs of deterioration, impacting the timeliness of nursing interventions. A structured relationship is not established between nursing response items and pathology data, resulting in nursing plans relying on empirical judgment and lacking a basis for prioritizing pathological changes. This can lead to irrational resource allocation or delayed intervention. Comparison of differences between nursing records and pathology test data is limited to one-way assessment and lacks a cross-dimensional framework for analyzing horizontal indicator differences. This makes it impossible to systematically verify the synergistic effects of nursing interventions and pathological progression, reducing the targeted nature of quality improvements. Existing technologies rely on fixed thresholds to determine nursing pathway adjustments and lack a dynamic threshold optimization mechanism. This makes it difficult to adapt to individualized changes in the disease course, easily leading to delayed pathway adjustments or excessive intervention, impacting the effective use of nursing resources. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method for analyzing and evaluating patient pathology data for clinical nursing.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for analyzing and evaluating patient pathology data for clinical nursing, comprising:

[0008] S1: Obtain the patient's biochemical test and histopathological data, collect blood routine, renal function, electrolyte levels and pathological diagnosis data, organize the data according to the medical record number and sampling time, unify the standardized test units and reference value intervals, and generate a pathology data compilation table;

[0009] S2: Based on the pathology data arrangement table, a dynamic trend analysis is performed on the continuous detection values of each pathology item, the change direction and amplitude are recorded, and the values are classified according to the degree of change to generate a classification record of pathology trend changes;

[0010] S3: According to the classification records of the pathological trend changes, each pathological category label is matched with the corresponding nursing response item, the relationship between the nursing assessment item and the pathological data item is established, and a nursing data association list is generated;

[0011] S4: Based on the nursing data association list, the nursing record data is compared with the pathological test data to establish a horizontal indicator difference relationship, and a pathology nursing evaluation comparison table is generated based on the evaluation differences and trend changes;

[0012] S5: According to the pathology nursing assessment comparison table, analyze the differences between the nursing tasks and the pathology data to determine whether they exceed the set threshold. If so, adjust the nursing path and generate a nursing path adjustment plan.

[0013] Optionally, the pathology data collation table includes routine blood data, renal function data, electrolyte level data, pathology diagnosis data, medical record number classification records and sampling time sorting, the pathology trend change classification records include change direction records, change amplitude records, and change degree classification results, the nursing data association list includes pathology category labels, nursing response items, and association relationship mapping tables, the pathology nursing evaluation comparison table includes horizontal indicator difference relationships, evaluation difference matrices, and trend change comparison modules, and the nursing path adjustment plan includes nursing task difference analysis results, threshold judgment criteria, and path adjustment strategies.

[0014] Optionally, the specific steps of S1 are:

[0015] S101: Obtain the patient's biochemical test data and tissue pathology data, collect three types of test indicators: blood routine, renal function, and electrolyte levels, integrate the pathology diagnosis data, and preliminarily classify the original data by medical record number to generate a multi-source pathology data set;

[0016] S102: Based on the multi-source pathology dataset, extract the sampling time data under each medical record number, arrange the detection indicators in chronological order, match the detection values of different time nodes under the same medical record number, and generate a time series index dataset;

[0017] S103: Calling the time series index data set, uniformly converting the detection units of the three indicators of blood routine, renal function, and electrolyte levels, comparing them with the standard reference value interval, adjusting the value range to a unified benchmark, and generating a pathology data summary table.

[0018] Optionally, the specific steps of S2 are:

[0019] S201: extracting continuous test values of each pathology item based on the pathology data arrangement table, evaluating the relationship between the degree of difference between the data at adjacent time points and the initial value, and generating a dynamic trend analysis result;

[0020] S202: Calling the dynamic trend analysis result, identifying the positive and negative directions of the difference, quantifying the difference ratio, integrating the direction and ratio into a unified description item, and generating change direction amplitude data;

[0021] S203: Based on the change direction amplitude data, the direction and proportion combination is divided into three categories of labels: rising, falling, and stable according to preset classification rules, and the items under the same label are merged to generate a classification record of pathological trend changes.

[0022] Optionally, the specific calculation formula for evaluating the relationship between the degree of difference between data at adjacent time points and the change in the initial value is:

[0023]

[0024] Where ΔT h Represents the difference fluctuation characteristic value between the hth detection value and the previous detection value, V h Represents the value of a pathological item in the hth test, V h-1 Represents the value of the item in the h-1th test, It represents the sum of the squares of the mean of all test values minus the squares of the squares of each test value from the 2nd to the zth time, μ V Represents the arithmetic mean of all test values of pathological items from the 1st to the zth time, Represents the variance of all test values of pathological items from the 1st to the zth time, The denominator adjustment factor is the square root of the product of the current detection value squared and the detection variance.

[0025] Optionally, the specific steps of S3 are:

[0026] S301: Based on the abnormal indicators, change intervals, and trend directions extracted from the pathology trend change classification records, pathology category labels are identified, trend directions and interval changes are combined for judgment, pathology labels are verified and numbered, and the number and trend combination content are called to generate trend association features;

[0027] S302 calls the corresponding response items in the nursing response item library based on the trend correlation feature, extracts the nursing assessment content, compares the time and frequency features with the trend correlation feature, identifies and classifies the matching segments, and generates matching segment distribution characteristics;

[0028] S303: Based on the matching segment distribution characteristics and combined with the label numbers in the trend association features, a mapping relationship between the pathology label and the nursing assessment content is established, and the identification group is screened according to the repetition frequency and independent situation of the content identification to generate the corresponding intensity value of the nursing item.

[0029] Optionally, the specific calculation formula for comparing the time and frequency features with the trend correlation features is:

[0030]

[0031] Among them, R t Represents the trend matching deviation indicator, T a represents the characteristic time value recorded in the ath time period, F a represents the corresponding occurrence frequency in the ath time period, C brepresents the current observation value of the b-th trend-related feature, Δ b Represents the change value of the trend correlation feature of item b in the current period, V k Represents the trend response value within the k-th sample segment, It represents the average value of trend response values in all sample segments, u is the number of time-frequency samples, m is the number of trend feature items, and l is the total number of trend response sample segments.

[0032] Optionally, the specific steps of S4 are:

[0033] S401: Based on the nursing data association list, extract the patient number, time node, and test item content from the nursing records and pathology tests, aggregate the nursing frequency and pathology test results of the same patient at different time nodes, align the nursing frequency and pathology abnormality marks according to the time nodes, and obtain a time-aligned difference group of nursing and pathology data;

[0034] S402: Based on the time alignment difference group of nursing and pathology data, samples with time differences within a set interval are screened, the deviation amplitudes of nursing frequency and pathology values are calculated, and the deviation amplitudes corresponding to the nursing event types are aggregated to obtain the total deviation amplitudes corresponding to the nursing events;

[0035] S403: Call the total amount of offset amplitude corresponding to the nursing events, set the offset amplitude reference range for each type of nursing events, determine whether there is a nursing event type that exceeds the range, classify and mark the corresponding values, and generate a pathological nursing assessment comparison table.

[0036] Optionally, the specific calculation formula for determining whether there is a nursing event type that is out of range is:

[0037]

[0038] in, Represents the deviation abnormal intensity index of the i-th type of nursing event, δ ij represents the offset amplitude value of the jth record in the i-th type of nursing event, w ij represents the risk weighting coefficient of the jth event of category i, n i represents the total number of records of the i-th type of nursing events, represents the arithmetic mean of the deviation amplitude values of the i-th type of nursing event, θ i Represents the correction factor for the change in the offset amplitude corresponding to the i-th type of nursing event, μ i Represents the historical average offset intensity of the i-th type of nursing events.

[0039] Optionally, the specific steps of S5 are:

[0040] S501: Based on the nursing tasks in the pathology nursing assessment comparison table and the data in the pathology collection record, extract the monitoring value and operation frequency corresponding to each task, compare the nursing task set value with the actual monitoring value, calculate the difference set between the two, and generate an offset value distribution;

[0041] S502: Based on the multiple differences in the offset value distribution, the offset threshold set in the nursing task scheduling benchmark is compared, the task items exceeding the threshold are screened, and the path structure is adjusted in combination with the position and time characteristics in the nursing path to generate a nursing path offset node group;

[0042] S503: Combine the task numbers and execution periods in the nursing path offset node group, extract the associated path sequence and resource configuration, correct the operation cycle and the interval between tasks, update the path execution sequence and resource allocation arrangement, and generate a nursing path adjustment plan.

[0043] Compared with the prior art, the advantages and positive effects of the present invention are:

[0044] In the present invention, by integrating multi-source pathology data and standardizing units and reference intervals, the consistency and comparability of data are ensured, the fluctuation trend of indicators is dynamically tracked, the direction and magnitude of changes are quantitatively analyzed, the evolution of the disease course is accurately captured, a direct mapping between pathology categories and nursing items is established, logical binding is clarified, the accuracy of intervention is improved, the differences between nursing records and pathology data are compared horizontally, the dynamic adaptation of nursing is optimized, the path plan is quickly adjusted through threshold judgment, and the risk of lag or misjudgment is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic diagram of the steps of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0047] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0048] See also Figure 1 The embodiment of the present invention provides a method for analyzing and evaluating patient pathology data for clinical nursing, comprising:

[0049] S1: Obtain the patient's biochemical test and histopathological data, collect blood routine, renal function, electrolyte levels and pathological diagnosis data, organize the data according to the medical record number and sampling time, unify the standardized test units and reference value intervals, and generate a pathology data compilation table;

[0050] S2: Based on the pathology data collation table, perform dynamic trend analysis on the continuous test values of each pathology item, record the direction and amplitude of the change, and classify them according to the degree of change to generate a classification record of pathology trend changes;

[0051] S3: Classify the records based on the pathological trend changes, match each pathological category label with the corresponding nursing response item, establish the relationship between the nursing assessment item and the pathological data item, and generate a nursing data association list;

[0052] S4: Based on the nursing data association list, the differences between nursing record data and pathological test data are compared to establish the horizontal indicator difference relationship, and a pathology nursing evaluation comparison table is generated based on the evaluation differences and trend changes;

[0053] S5: According to the pathology nursing assessment comparison table, analyze the differences between nursing tasks and pathology data to determine whether they exceed the set threshold. If so, adjust the nursing pathway and generate a nursing pathway adjustment plan.

[0054] The pathology data collation table includes routine blood data, renal function data, electrolyte level data, pathology diagnosis data, medical record number classification records and sampling time sorting. The pathology trend change classification records include change direction records, change amplitude records, and change degree classification results. The nursing data association list includes pathology category labels, nursing response items, and association relationship mapping tables. The pathology nursing evaluation comparison table includes horizontal indicator difference relationships, evaluation difference matrices, and trend change comparison modules. The nursing pathway adjustment plan includes nursing task difference analysis results, threshold judgment criteria, and pathway adjustment strategies.

[0055] The specific steps of S1 are:

[0056] S101: Obtain the patient's biochemical test data and tissue pathology data, collect three types of test indicators: blood routine, renal function, and electrolyte levels, integrate the pathology diagnosis data, and preliminarily classify the original data by medical record number to generate a multi-source pathology data set;

[0057] First, the hospital information system and laboratory information system must be connected. Using an interface or data extraction program, three types of biochemical indicators, including routine blood tests, renal function tests, and electrolyte levels, must be extracted within a specified timeframe. Routine blood tests include hemoglobin, white blood cell count, and platelet count; renal function tests include creatinine, urea nitrogen, and glomerular filtration rate; and electrolyte levels focus on sodium, potassium, and chloride ion concentrations. Extraction should be performed using ICD codes to filter for cases diagnosed with chronic kidney disease to ensure data targeting. A typical filter might be: "Filter all medical records matching the ICD-10 code N18 within a specified timeframe and match them to relevant laboratory records." Pathology data originates from pathology systems or electronic medical record text. Information such as pathological diagnosis, tissue source, and staining markers must be extracted from structured fields. Text keyword recognition and standard format matching are used to ensure accurate correspondence between each medical record. To achieve data integration, the medical record number is used as the core identification field, and various test results are matched by time and grouped into the same record. When creating multi-source pathology data tables, it is recommended to use a table format, where each row identifies a medical record number and corresponds to all test data at a specific time point. Data tables can be constructed using data analysis software, merged through table joins, grouped by medical record number, and the corresponding data items simultaneously classified to form a structured multi-source pathology data set for subsequent indexing and processing.

[0058] S102: Based on the multi-source pathology dataset, extract the sampling time data under each medical record number, arrange the detection indicators in chronological order, match the detection values of different time nodes under the same medical record number, and generate a time series index dataset;

[0059] Based on the integrated multi-source pathology dataset, a time extraction operation is performed on all records under each medical record number. The extracted field is usually the sampling time or test report time, and the records are sorted from earliest to latest using a sorting function. After chronological arrangement, the structure needs to be reorganized so that the same test items for the same patient at different time points can be listed in a unified table to facilitate subsequent comparative analysis. To clarify the differences in tests between time points, a data pair structure is established, with each group containing two different time nodes and corresponding test values. This is usually handled by enumerating all time combinations and finding their corresponding test values. During the construction process, if a time point has a missing test value, it should be recorded as a blank value or forward-filled with data from adjacent time points. The choice of interpolation depends on the characteristics of the test item. The goal of the entire process is to construct a time series index structure based on the medical record number and test item, containing multiple test time points and the values at each time point. The resulting structure can support comparisons and derivative operations along the time dimension.

[0060] S103: Calling the time series index data set, uniformly converting the detection units of the three indicators of blood routine, renal function, and electrolyte levels, comparing them with the standard reference value interval, adjusting the value range to a unified benchmark, and generating a pathology data summary table;

[0061] When accessing the collated time-series index dataset, the units for three indicators—routine blood tests, renal function, and electrolyte levels—must first be standardized. Because some source data have varying units, such as hemoglobin, which may be recorded in grams per deciliter (g / dL) or grams per liter (g / L), a unit conversion table is needed to convert the original units to standard units. For example, a value of 15 g / dL should be converted to 150 g / L. After standardization, all values are categorized as low, normal, and high using standard medical reference intervals. For example, the normal range for creatinine for men is generally 53 to 106 micromoles per liter (μmol / L). A value of 120 μmol / L is considered high. Further standardization is needed for the numerical ranges of each test item, using a unified baseline value to adjust deviating values. The baseline value for creatinine is set at 80 μmol / L. Any values that are above the baseline range are adjusted proportionally to the baseline range. This adjustment is performed by multiplying the current value by a scaling factor determined based on the ratio of the current value to the baseline value to obtain the adjusted value. When multiple tests need to be processed at the same time, they can be processed in batches, and unit conversion and value adjustment operations can be performed by grouping the test items. When dividing the test result interval, the numerical boundaries should be set according to the clinical significance of each item. For example, a potassium ion value below 3.5 is considered a low value, a value above 5.5 is a high value, and the middle is normal. All settings need to refer to authoritative medical standards or expert consensus documents. The data after sorting should be organized into a standardized table structure according to the medical record number and time node to facilitate export and further analysis. The entire process ensures that the values at each time node are comparable and are output in a standard table format.

[0062] The specific steps of S2 are:

[0063] S201: Based on the pathology data arrangement table, extract the continuous test value of each pathology item, evaluate the relationship between the degree of difference of data at adjacent time points and the initial value, and generate dynamic trend analysis results;

[0064] The specific calculation formula for evaluating the relationship between the degree of difference between data at adjacent time points and the change in the initial value is:

[0065]

[0066] Where ΔT h Represents the difference fluctuation characteristic value between the hth detection value and the previous detection value, V h Represents the value of a pathological item in the hth test, V h-1 Represents the value of the item in the h-1th test, It represents the sum of the squares of the mean of all test values minus the squares of the squares of each test value from the 2nd to the zth time, μ V Represents the arithmetic mean of all test values of pathological items from the 1st to the zth time, Represents the variance of all test values of pathological items from the 1st to the zth time, The denominator adjustment factor is the square root of the product of the current test value squared and the test variance;

[0067] Formula parameter settings and sources:

[0068] V h : Taken from actual medical records, assuming the test was performed at the 5th time; V5 = 102 units, V h-1 : The 4th test, V4=98 units.

[0069] z is set according to the project schedule, assuming that the total number of inspections z = 10.

[0070] V c For example, the 2nd to 10th detection values are obtained based on actual records.

[0071] μ V :V1 to V z The mean is calculated by dividing the sum of all test values by the number of tests. If the V value is 95, 97, 99, 98, 102, 105, 107, 103, 101, 100, then μ V =100.7.

[0072] Variance, calculated as the sum of all V values and μ V The sum of the squares of the differences divided by z. Using the above data,

[0073] Formula calculation steps:

[0074] Calculate (V h +V h-o ) 2 :

[0075] (V5+V4) 2 =(102+98) 2 =200 2 =40000;

[0076] Calculates the sum of the squared differences and the sum of the squared differences from the mean

[0077]

[0078] Calculates the square root value in the denominator

[0079]

[0080] Calculate the overall formula:

[0081]

[0082] Result Interpretation: The result ΔT5 = 39.05 represents the differential fluctuation characteristic between the fifth and fourth tests. This value indicates the magnitude of the change between the two test results and the degree of deviation from the overall trend. Dynamic trend analysis of consecutive pathology test values can reveal the stability or volatility of pathology over time, thereby assisting medical experts in better understanding the pathology and making subsequent treatment decisions.

[0083] S202: Call the dynamic trend analysis results, identify the positive and negative directions of the difference, quantify the difference ratio, integrate the direction and ratio into a unified description item, and generate change direction amplitude data;

[0084] According to the analysis results, the dynamic trend analysis results will be further used to judge the direction of the difference. According to the change value of each set of continuous time points, the direction of the difference can be judged. If the difference is positive, it is judged as "rising", if it is negative, it is "falling", and if it is zero, it is "flat". At the same time, the difference direction and its corresponding ratio are integrated into a unified description item. For example, if the detection value of a pathology item changes from 30U / L to 60U / L, the change value is 30, and the change ratio is 100%. At this time, the description item is "rising 100%". If an abnormal and drastic fluctuation occurs (such as the proportion change exceeds the set upper limit), an abnormal shielding mechanism can be set and marked as "abnormal change". The description item is composed of a direction and a change ratio, which is used to clearly express the trend change of each pathology data. In this way, the trend changes of all pathology items can be recorded in detail and displayed uniformly through standardized description items, which is convenient for subsequent further analysis and processing.

[0085] S203: Based on the change direction and amplitude data, according to the preset classification rules, the direction and ratio combination is divided into three categories: rising, falling, and stable, and the items under the same label are merged to generate a classification record of pathological trend changes;

[0086] Based on the direction and magnitude of change, trends are further labeled according to pre-set classification rules. The rules are as follows: changes within ±10% are labeled "stable," while those exceeding ±10% are labeled "increasing" or "decreasing," depending on the positive or negative direction of the change. For example, a +12% change is classified as "increasing," a -15% change is "decreasing," and a change between ±5% and ±10% is "stable." During the classification process, all pathology items with the same label are grouped together. For example, a patient's ALT and AST may be labeled "increasing," while ALB may be labeled "decreasing." Each classification group contains all pathology items matching the trend label, along with relevant information such as the total number of items, maximum change, and timeframe. This process also involves adjusting thresholds based on the specific data type. For example, for sensitive indicators (such as creatinine), stricter thresholds may be set, while for other indicators (such as the inflammatory marker CRP), more relaxed thresholds may be used. The resulting classification records of pathological trend changes provide data support for subsequent clinical decision-making and also facilitate group analysis of pathological data.

[0087] The specific steps of S3 are:

[0088] S301: Based on the abnormal indicators, change intervals, and trend directions extracted from the pathology trend change classification records, the pathology category labels are identified, the trend direction and interval changes are combined for judgment, the pathology labels are verified and numbered, and the number and trend combination content are called to generate trend association features;

[0089] Abnormal indicators extracted from pathology trend classification records typically include continuously monitored data such as body temperature, blood pressure, heart rate, respiratory rate, and blood sugar. The timeline variation range of these data can be determined item by item based on a recent recording period, such as seven consecutive days of temperature records. After arranging these data in chronological order, the magnitude of the change can be calculated daily to determine the overall trend direction. For example, if the temperature value continuously rises from 36.8°C to 38.7°C from day one to day seven, the trend is considered "increasing" and the range of change is recorded as 36.8°C to 38.7°C. Once the trend direction is determined, it is compared with predefined pathology types. For example, if the temperature exceeds 38°C and continues to rise, it qualifies as a "fever type" pathology, designated PT-01. The trend direction, range, and number are combined to form a trend-associated feature. In practice, if a patient in intensive care shows a significant increase in temperature for several consecutive days, this method can be used to quickly classify them as "fever type," completing the recording process of combining trend numbers with data features.

[0090] S302 calls the corresponding response items in the nursing response item library based on the trend correlation feature, extracts the nursing assessment content, compares the time and frequency features with the trend correlation feature, identifies the matching segments and classifies and labels them, and generates matching segment distribution characteristics;

[0091] The specific calculation formula for comparing time and frequency characteristics with trend correlation characteristics is:

[0092]

[0093] Among them, R t Represents the trend matching deviation indicator, T a represents the characteristic time value recorded in the ath time period, F a represents the corresponding occurrence frequency in the ath time period, C b represents the current observation value of the b-th trend-related feature, Δ b Represents the change value of the trend correlation feature of item b in the current period, V k Represents the trend response value within the k-th sample segment, represents the average value of trend response values in all sample segments, u is the number of time-frequency samples, m is the number of trend feature items, and l is the total number of trend response sample segments;

[0094] Parameter description and value acquisition method:

[0095] T a The data are automatically collected through the nursing activity recording system. According to the ICU Nursing Activity Assessment Scale (NAS) standard, the time value of a single nursing activity is between 5 and 15 minutes.

[0096] F a Statistics from the nursing record system show that, based on nursing quality control indicators, the average daily frequency of a single nursing activity is usually between 3 and 10 times.

[0097] C b Obtained in real time via patient monitoring devices.

[0098] Δ b It is calculated as the difference between the current observation and the previous period's observation.

[0099] V k Derived from historical data statistics.

[0100] The calculation method is all V k The arithmetic mean of .

[0101] u represents the total number of statistical time periods.

[0102] m represents the total number of trend-related features counted.

[0103] l represents the total number of sample segments counted.

[0104] Specific numerical settings:

[0105] u=3: statistics are collected over three time periods.

[0106] m=2: Two trend correlation features are counted.

[0107] l=4: 4 sample segments are counted.

[0108] T1 = 10 minutes, F1 = 5 times;

[0109] T2 = 12 minutes, F2 = 4 times;

[0110] T3=8 minutes, F3=6 times.

[0111] C1=120mmHg,Δ1=5mmHg;

[0112] C2=80mmHg, Δ2=-3mmHg.

[0113] V1=130, V2=125, V3=128, V4=132;

[0114]

[0115] Formula calculation process:

[0116] Calculate the molecular part:

[0117]

[0118] Numerator sum: 193.064 + 144.4 = 337.464

[0119] Calculate the denominator:

[0120] u·m=3×2=6;

[0121]

[0122] Total of the denominators: 6 + 43.4 = 49.4

[0123] Final calculation:

[0124]

[0125] The results show that the trend matching deviation index R t The value of 6.83 indicates the degree of match between the time and frequency characteristics of current nursing activities and the trend-related characteristics. This value can be used to identify and classify matching segments, thereby generating distribution characteristics for matching segments.

[0126] S303: Based on the matching segment distribution characteristics and the label numbers in the trend association features, a mapping relationship between the pathology label and the nursing assessment content is established, and the label group is screened according to the repetition frequency and independence of the content label to generate the corresponding intensity value of the nursing item;

[0127] After establishing a correspondence between the nursing responses in the matching segments and the pathology numbers in the trend features, a set of identification mapping structures can be sorted out, each of which includes a trend number and one or more matching segments. For example, a trend feature numbered PT-01 corresponds to three matching segments located on different dates. It is necessary to check whether these segments come from different source records. If so, they are classified as independent record groups; if multiple records occur in the same nursing document, they are classified as duplicate record groups. Then, the number of response behaviors in each record group is counted. For example, if a group has a total of 5 records per day, and the corresponding number of nursing items is 3, it can be concluded that the response intensity of this type of nursing under this pathology trend is medium. This intensity division method can be used to reflect the coverage density of nursing behaviors in the development of pathology trends, thereby completing the response intensity definition process between nursing items and pathology labels.

[0128] The specific steps of S4 are:

[0129] S401: Based on the nursing data association list, extract the patient number, time node, and test item content from the nursing records and pathology tests, aggregate the nursing frequency and pathology test results of the same patient at different time nodes, align the nursing frequency and pathology abnormality marks according to the time nodes, and obtain a time-aligned difference group of nursing and pathology data;

[0130] The nursing records in the nursing data association list need to extract the patient number, recording time and nursing event type, and at the same time match the patient number, test time and various test contents in the pathology test table to build a unified identification based on "patient number + time node". When processing data at multiple time points, it is necessary to complete the preliminary matching through the time sequence combined with the set threshold of time distance. Assuming that the nursing time is March 1, 3, and 5, and the pathology test time is March 2 and 4, the minimum time difference is used for alignment, that is, March 1 matches March 2, and March 3 matches March 4. The time interval for each set of matching data needs to be recorded, and a reasonable allowable range is set, such as no more than 3 days. The interval value is calculated for all matching data and the symbol is retained. Matches that meet the conditions are selected, and the frequency of nursing events on each matching day is then counted. For example, there are 3 nursing records on March 1 and 2 on March 3, which are used as the nursing frequency on that day. Abnormal markers are then extracted from pathological tests. Abnormal judgment is based on the reference value range of the test items. For example, the normal range of white blood cell count is 4.0–10.0. If the test value is 12.5, it is marked as abnormal. The control time points are organized into record groups consisting of nursing frequency and pathological abnormality markers, and the difference in days between the nursing and pathological times is recorded. For example, if the nursing occurs on March 1 and the test occurs on March 2, the difference is 1 day. A unified data set containing patient identification, nursing frequency, abnormal markers and time intervals is constructed to provide a matching structure basis for subsequent analysis.

[0131] S402: Based on the time alignment difference group of nursing and pathology data, samples with time differences within a set interval are screened, the deviation amplitudes of nursing frequency and pathology values are calculated, and the deviation amplitudes corresponding to the nursing event types are aggregated to obtain the total deviation amplitudes corresponding to the nursing events;

[0132] All matching records in the constructed data set are screened, and the allowable range of time difference is set, for example, no more than 3 days. Matching groups with time intervals outside the range are eliminated, and the nursing frequency and the corresponding pathological test value are extracted from the retained records. The difference between the pathological test value and the standard value in each record is calculated, and the offset difference is calculated with reference to the middle position of the normal value of the pathological test item. For example, the normal value of creatinine is 88 with a fluctuation of 12, so 88 is used as the standard value. If the test result is 110, the offset difference is 22. The offset is recorded and associated with the nursing frequency of the day. In multiple nursing records Classify and count event types. For example, if a patient receives multiple infusion operations at the same time point, the pathological offset differences associated with all infusion operations are collected and the average offset is calculated. For example, if the pathological offset values associated with the patient under five infusion care are 5, 10, 8, 6, and 9, respectively, the average offset of this nursing event type is calculated to be 7.6. Similarly, all offset values of different nursing event types are summarized to form an average offset mapping table corresponding to the nursing event type, which lays the foundation for subsequent classification and labeling. The mapping table will display the average offset value corresponding to each nursing type to distinguish differences.

[0133] S403: Calling the total amount of deviations corresponding to nursing events, setting a reference range of deviations for each type of nursing event, determining whether there are nursing event types that exceed the range, and classifying and labeling the corresponding values to generate a pathological nursing assessment comparison table;

[0134] The specific calculation formula for determining whether there are out-of-range nursing event types is:

[0135]

[0136] in, Represents the deviation abnormal intensity index of the i-th type of nursing event, δ ij represents the offset amplitude value of the jth record in the i-th type of nursing event, w ij represents the risk weighting coefficient of the jth event of category i, n i represents the total number of records of the i-th type of nursing events, represents the arithmetic mean of the deviation amplitude values of the i-th type of nursing event, θ i Represents the correction factor for the change in the offset amplitude corresponding to the i-th type of nursing event, μ i represents the historical average offset intensity of the i-th type of nursing events;

[0137] The specific steps of S5 are:

[0138] S501: Based on the nursing tasks in the pathology nursing assessment comparison table and the data in the pathology collection record, extract the monitoring value and operation frequency corresponding to each task, compare the nursing task set value with the actual monitoring value, calculate the difference set between the two, and generate the offset value distribution;

[0139] In the pathology nursing assessment comparison table, all records containing nursing tasks are listed, and the monitoring indicators and operation frequency set for each task are extracted item by item. For example, a task is set to record body temperature six times a day, and the target value is 36.5℃. Then, from the pathology collection record, the actual collection data and actual collection frequency are extracted according to the patient ID, task ID and recording time. The data integrity is judged against the set target. For example, if the body temperature should be recorded six times but only four times, it should be marked as missing data and the completion rate should be calculated. Then, the difference between the actual monitoring value and the task setting value is extracted, and the offset value of each record is listed item by item. For example, if the setting value is 36.5℃, record it. If the recorded value is 37.8℃, the offset value is 1.3℃. After collecting all the offset values, their mean, maximum and minimum values, and dispersion are counted to establish a monitoring data offset set. At the same time, the operation frequency offset is analyzed. For example, if it is set to 6 times a day and the actual number is 4 times, the frequency offset is -2. The frequency and distribution of this type of offset over a period of time are calculated, and the statistical results of each task offset are sorted out through structured tables, including numerical distribution, frequency offset, offset duration and interval trend. Data analysis tools can be used to display the distribution item by item, and the central trend and dispersion of various offset indicators can be presented in combination with histograms and box plots, thus forming a complete set of offset value distribution.

[0140] S502: Based on the multiple differences in the offset value distribution, the offset threshold set in the nursing task scheduling benchmark is compared, the task items exceeding the threshold are screened, and the path structure is adjusted based on the position and time characteristics in the nursing path to generate a nursing path offset node group;

[0141] First, extract the key statistical indicators such as the offset mean, offset extreme value and offset dispersion of each task from the offset value distribution of the previous stage, and compare them item by item with the offset critical value set in the nursing task scheduling benchmark. Suppose a blood glucose monitoring task has a target of 5.5mmol / L and the offset critical value is set to 1.0mmol / L. If the average offset of the task in the offset value set is 1.3mmol / L, which exceeds the critical value, it is marked as an offset abnormal task. Then, extract the corresponding position and execution period in the nursing pathway according to the marked task. For example, if it is in stage three of the pathway and the execution period is from the second to the fourth day of hospitalization, this task is recorded as a pathway offset node. If multiple nodes deviate continuously, for example, all tasks from stage three to stage five are abnormal, they are summarized into continuous deviated path segments and integrated into a nursing path deviated node group. The deviated node group contains information such as task number, path sequence, and execution time period. The path structure view can be used to identify areas where abnormalities are concentrated, providing a basis for subsequent path structure adjustments. The setting of the deviated threshold is determined based on the fluctuation range of historical task completion data and safety monitoring requirements. The critical value is usually calculated based on the statistical mean of the past 30 days and the allowable fluctuation range. For example, if the normal deviated value does not exceed 0.8mmol / L, the critical value is set to 1.0mmol / L. Based on this setting, the deviated screening and node group generation process is completed.

[0142] S503: Combine the task numbers and execution periods in the nursing pathway offset node group, extract the associated pathway sequence and resource allocation, correct the operation cycle and the interval between tasks, update the pathway execution sequence and resource allocation arrangement, and generate a nursing pathway adjustment plan;

[0143] Extract each task number and corresponding execution time period from the offset node group. Combined with the task's path sequence, find the execution interval of the upstream and downstream tasks in the nursing pathway. For example, if a task needs to be executed every 6 hours, if the offset statistics show frequent deviations, the cycle should be adjusted to every 4 hours. After the cycle of the task is corrected, the execution time of subsequent tasks must be adjusted synchronously to prevent resource usage conflicts. The original manpower, equipment, and time usage of the task are checked through the resource allocation table. For example, if a task originally requires a nurse and a monitoring device, after the cycle is encrypted, more manpower needs to be invested, the resource priority needs to be raised to a high level, and the execution time period needs to be scheduled earlier. If the task needs to be moved forward due to the increased priority, the task order needs to be rearranged and placed before the previous task to form a new path sequence. The new execution time, required resource quantity, and allocated time period of each task are recorded in the new path configuration table. Combined with the resource scheduling diagram, the task timeline and resource demand changes before and after the update are displayed. The time interval between tasks should also be updated, and the time difference and resource call difference before and after the cycle change should be marked. Finally, the updated path sequence and resource allocation list are output.

[0144] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A method for analyzing and evaluating patient pathology data for clinical nursing, characterized in that: The following steps are involved: S1: Obtain the patient's biochemical test and histopathological data, collect blood routine, renal function, electrolyte levels and pathological diagnosis data, organize the data according to the medical record number and sampling time, unify the standardized test units and reference value intervals, and generate a pathology data compilation table; S2: Based on the pathology data arrangement table, a dynamic trend analysis is performed on the continuous detection values of each pathology item, the change direction and amplitude are recorded, and the values are classified according to the degree of change to generate a classification record of pathology trend changes; S3: According to the classification records of the pathological trend changes, each pathological category label is matched with the corresponding nursing response item, the relationship between the nursing assessment item and the pathological data item is established, and a nursing data association list is generated; S4: Based on the nursing data association list, the nursing record data is compared with the pathological test data to establish a horizontal indicator difference relationship, and a pathology nursing evaluation comparison table is generated based on the evaluation differences and trend changes; S5: According to the pathology nursing assessment comparison table, analyze the differences between the nursing tasks and the pathology data to determine whether they exceed the set threshold. If so, adjust the nursing path and generate a nursing path adjustment plan.

2. The method for analyzing and evaluating patient pathology data for clinical nursing according to claim 1, wherein: The pathology data collation table includes routine blood data, renal function data, electrolyte level data, pathology diagnosis data, medical record number classification records and sampling time sorting; the pathology trend change classification records include change direction records, change amplitude records, and change degree classification results; the nursing data association list includes pathology category labels, nursing response items, and association relationship mapping tables; the pathology nursing assessment comparison table includes horizontal indicator difference relationships, assessment difference matrices, and trend change comparison modules; the nursing pathway adjustment plan includes nursing task difference analysis results, threshold judgment criteria, and pathway adjustment strategies.

3. The method for analyzing and evaluating patient pathology data for clinical nursing according to claim 1, wherein: The specific steps of S1 are: S101: Obtain the patient's biochemical test data and histopathological data, collect three types of test indicators: blood routine, renal function, and electrolyte levels, integrate the pathological diagnosis data, and preliminarily classify the original data by medical record number to generate a multi-source pathology data set; S102: Based on the multi-source pathology dataset, extract the sampling time data under each medical record number, arrange the detection indicators in chronological order, match the detection values of different time nodes under the same medical record number, and generate a time series index dataset; S103: Calling the time series index data set, uniformly converting the detection units of the three indicators of blood routine, renal function, and electrolyte levels, comparing them with the standard reference value interval, adjusting the value range to a unified benchmark, and generating a pathology data summary table.

4. The method for analyzing and evaluating patient pathology data for clinical nursing according to claim 3, wherein: The specific steps of S2 are: S201: extracting continuous test values of each pathology item based on the pathology data arrangement table, evaluating the relationship between the degree of difference between the data at adjacent time points and the initial value, and generating a dynamic trend analysis result; S202: Calling the dynamic trend analysis result, identifying the positive and negative directions of the difference, quantifying the difference ratio, integrating the direction and ratio into a unified description item, and generating change direction amplitude data; S203: Based on the change direction amplitude data, the direction and proportion combination is divided into three categories of labels: rising, falling, and stable according to preset classification rules, and the items under the same label are merged to generate a classification record of pathological trend changes.

5. The method for analyzing and evaluating patient pathology data for clinical nursing according to claim 4, characterized in that: The specific calculation formula for evaluating the relationship between the degree of difference between data at adjacent time points and the change in the initial value is: Where, ΔT h Represents the difference fluctuation characteristic value between the hth detection value and the previous detection value, V h Represents the value of a pathological item in the hth test, V h-1 Represents the value of the item in the h-1th test, It represents the sum of the squares of the mean of all test values minus the squares of the squares of each test value from the 2nd to the zth time, μ V Represents the arithmetic mean of all test values of pathological items from the 1st to the zth time, Represents the variance of all test values of pathological items from the 1st to the zth time, The denominator adjustment factor is the square root of the product of the current detection value squared and the detection variance.

6. The method for analyzing and evaluating patient pathology data for clinical nursing according to claim 4, wherein: The specific steps of S3 are: S301: Based on the abnormal indicators, change intervals, and trend directions extracted from the pathology trend change classification records, pathology category labels are identified, trend directions and interval changes are combined for judgment, pathology labels are verified and numbered, and the number and trend combination content are called to generate trend association features; S302 calls the corresponding response items in the nursing response item library based on the trend correlation feature, extracts the nursing assessment content, compares the time and frequency features with the trend correlation feature, identifies and classifies the matching segments, and generates matching segment distribution characteristics; S303: Based on the matching segment distribution characteristics and combined with the label numbers in the trend association features, a mapping relationship between the pathology label and the nursing assessment content is established, and the identification group is screened according to the repetition frequency and independent situation of the content identification to generate the corresponding intensity value of the nursing item.

7. The method for analyzing and evaluating patient pathology data for clinical nursing according to claim 6, wherein: The specific calculation formula for comparing the time and frequency characteristics with the trend correlation characteristics is: Among them, R t Represents the trend matching deviation indicator, T a represents the characteristic time value recorded in the ath time period, F a represents the corresponding occurrence frequency in the a-th time period, C b represents the current observation value of the b-th trend-related feature, Δ b Represents the change value of the trend correlation feature of item b in the current period, V k Represents the trend response value within the k-th sample segment, It represents the average value of trend response values in all sample segments, u is the number of time-frequency samples, m is the number of trend feature items, and l is the total number of trend response sample segments.

8. The method for analyzing and evaluating patient pathology data for clinical nursing according to claim 6, wherein: The specific steps of S4 are: S401: Based on the nursing data association list, extract the patient number, time node, and test item content from the nursing records and pathology tests, aggregate the nursing frequency and pathology test results of the same patient at different time nodes, align the nursing frequency and pathology abnormality marks according to the time nodes, and obtain a time-aligned difference group of nursing and pathology data; S402: Based on the time alignment difference group of nursing and pathology data, samples with time differences within a set interval are screened, the deviation amplitudes of nursing frequency and pathology values are calculated, and the deviation amplitudes corresponding to the nursing event types are aggregated to obtain the total deviation amplitudes corresponding to the nursing events; S403: Call the total amount of offset amplitude corresponding to the nursing events, set the offset amplitude reference range for each type of nursing events, determine whether there is a nursing event type that exceeds the range, classify and mark the corresponding values, and generate a pathological nursing assessment comparison table.

9. The method for analyzing and evaluating patient pathology data for clinical nursing according to claim 8, characterized in that: The specific calculation formula for determining whether there is a nursing event type that is out of range is: in, Represents the deviation abnormal intensity index of the i-th type of nursing event, δ ij represents the offset amplitude value of the jth record in the i-th type of nursing event, w ij represents the risk weighting coefficient of the jth event of category i, n i represents the total number of records of the i-th type of nursing events, represents the arithmetic mean of the deviation amplitude values of the i-th type of nursing event, θ i Represents the correction factor for the change in the offset amplitude corresponding to the i-th type of nursing event, μ i Represents the historical average offset intensity of the i-th type of nursing events.

10. The method for analyzing and evaluating patient pathology data for clinical nursing according to claim 9, characterized in that: The specific steps of S5 are: S501: Based on the nursing tasks in the pathology nursing assessment comparison table and the data in the pathology collection record, extract the monitoring value and operation frequency corresponding to each task, compare the nursing task set value with the actual monitoring value, calculate the difference set between the two, and generate an offset value distribution; S502: Based on the multiple differences in the offset value distribution, the offset threshold set in the nursing task scheduling benchmark is compared, the task items exceeding the threshold are screened, and the path structure is adjusted in combination with the position and time characteristics in the nursing path to generate a nursing path offset node group; S503: Combine the task numbers and execution periods in the nursing path offset node group, extract the associated path sequence and resource configuration, correct the operation cycle and the interval between tasks, update the path execution sequence and resource allocation arrangement, and generate a nursing path adjustment plan.

Images