User-based electronic medical record data quality evaluation method

Abstract

The invention discloses a user-based electronic medical record data quality evaluation method. The method comprises the following steps: step 1, constructing an electronic medical record data quality evaluation index DQAI; 2, dimension division is carried out on the electronic medical record data quality evaluation index DQAI, and an electronic medical record data quality evaluation index system is constructed; 3, comprehensively evaluating the electronic medical record data quality evaluation index system; step 4, based on different user types, respectively performing preference analysis on the electronic medical record data quality evaluation index DQAI from an index level and a dimension level to obtain an analysis result; and completing user-based electronic medical record data quality evaluation. The technical effects of improving the accuracy and correlation of electronic medical record data quality evaluation, enhancing the user satisfaction and data availability of an electronic medical record system, providing a user-oriented improvement direction for electronic medical record data quality management and the like can be achieved.

Description

Technical Field

[0001] The present invention relates to an electronic medical record data quality assessment method, in particular to an electronic medical record data quality assessment method based on users. Background Art

[0002] This section merely provides background information related to the present disclosure and is not necessarily prior art.

[0003] With the increasing application of technologies such as the Internet of Things, big data, and artificial intelligence in the field of medical health, electronic medical record data, as the main source of health and medical big data, has played an increasingly important role in its related fields. As an important carrier of medical information resources, the quality of electronic medical record data is related to the accuracy of medical decision-making, the effectiveness of patient care, the reliability of medical research, and the application and development of digital health care. High-quality electronic medical record data is the basis and prerequisite for conducting medical information analysis and medical decision-making. In order to improve the quality of electronic medical record EHR data, active research has been carried out on the formulation and evaluation of electronic medical record data quality standards, and even standard documents have been issued. However, overall, the quality problem of electronic medical record data is still severe.

[0004] Data quality reflects the "fitness for use" of data, that is, data that can meet the specific needs of specific users is high-quality data. Different methods and ideas have been applied to the research on electronic medical record data quality assessment, but they still show certain limitations: ① In the construction of the electronic medical record data quality assessment index system, expert models are mostly used, quantitative methods are rarely used, and there is a lack of scientific evaluation of the constructed system; ② In the construction process, the expected use purpose (i.e. "fitness") of the electronic medical record EHR system and its data by different users is rarely considered. In the few studies that attempt to take into account the "fitness" of electronic medical record data, the secondary use of data is mostly used as the purpose, while the original goal of using the electronic medical record system, such as assisting clinical care and medical management, is ignored; ③ In the application of data quality assessment in different usage scenarios, there is a lack of discussion on its differences from different types of perspectives and the views of direct users of electronic medical record EHR on relevant indicators, so it is difficult to make targeted suggestions for system optimization.

[0005] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0006] Purpose of the invention: The technical problem to be solved by the present invention is to provide a user-based electronic medical record data quality assessment method in view of the deficiencies in the prior art.

[0007] In order to solve the above technical problems, the present invention discloses a user-based electronic medical record data quality assessment method, comprising the following steps:

[0008] Step 1: Construct the electronic medical record data quality assessment index DQAI;

[0009] Step 2, dividing the electronic medical record data quality assessment index DQAI into dimensions and constructing an electronic medical record data quality assessment index system;

[0010] Step 3, comprehensively evaluating the electronic medical record data quality assessment index system;

[0011] Step 4: Based on different user types, conduct preference analysis on the electronic medical record data quality indicator DQAI at the indicator level and dimension level to obtain analysis results; complete the user-based electronic medical record data quality assessment.

[0012] Furthermore, the electronic medical record data quality assessment index DQAI described in step 1 includes:

[0013] Accuracy, completeness, timeliness, consistency, sophistication, standardization, uniqueness, credibility, rationality, traceability, portability, availability, accessibility, relevance, applicability and understandability.

[0014] Furthermore, the dimensions are: structural dimension and relational dimension; wherein,

[0015] The structural dimension includes 9 indicators: accuracy, completeness, timeliness, consistency, precision, standardization, uniqueness, credibility and rationality;

[0016] The relational dimension includes seven indicators: traceability, portability, availability, accessibility, relevance, applicability and understandability.

[0017] Furthermore, the comprehensive evaluation of the electronic medical record data quality assessment index system described in step 3 includes:

[0018] Step 3-1, collecting user scoring samples for the electronic medical record data quality assessment index DQAI;

[0019] Step 3-2, using the entropy weight method to calculate and sort the electronic medical record data quality assessment index DQAI;

[0020] Step 3-3, use fuzzy comprehensive evaluation method to calculate the comprehensive evaluation score.

[0021] Furthermore, the weight calculation and sorting described in step 3-1 include:

[0022] Step 3-1-1, establish the original data evaluation matrix K, as follows:

[0023]

[0024] Among them, k ij represents the score value of the jth indicator of the i-th scoring sample, n represents the number of scoring samples, and m represents the number of indicators;

[0025] Step 3-1-2, normalize the original data evaluation matrix K to obtain the normalized evaluation matrix K ′ , as follows:

[0026]

[0027] Among them, k i ′ j is the normalized evaluation matrix K ′ Elements in

[0028] Step 3-1-3, calculate the proportion p of the evaluation score of the i-th sample under the j-th indicator ij , as follows:

[0029]

[0030] Step 3-1-4, calculate the information entropy of each indicator, as follows:

[0031] Meet e j ≥0

[0032] Among them, e j represents the information entropy of the jth indicator;

[0033] Step 3-1-5, calculate the weight of each indicator, as follows:

[0034]

[0035] Among them, w j Represents the weight of the j-th indicator.

[0036] Furthermore, the fuzzy comprehensive evaluation method described in step 3-2 is used to calculate the comprehensive evaluation score, including:

[0037] Step 3-2-1, establish factor set U, as follows:

[0038] U={u 1 ,u 2 ,…,u j ,…,u m}

[0039] Among them, the element u j Represents the jth factor that affects the evaluation object;

[0040] Step 3-2-2, establish the evaluation set V, as follows:

[0041] V={v 1 ,v 2 , …, v h , …, v l}

[0042] Among them, the element v h represents the hth evaluation result, l represents the number of evaluation result types, which is determined by the design method of the scoring sample;

[0043] Step 3-2-3, establish the weight set W and determine the factor weight vector:

[0044] W = {w 1 , w 2 , ..., w j , ..., w m}

[0045] Each factor u j Corresponding to a weight w j , which is the weight w in step 3-1-5 ij , used to indicate its importance, the weights in the weight set W satisfy the following conditions:

[0046]

[0047] Step 3-2-4, establish the fuzzy relationship matrix R, as follows:

[0048]

[0049] Among them, r jh is the membership degree of the jth evaluation index in the factor set U to the hth evaluation level in the evaluation set V;

[0050] Step 3-2-5, fuzzy comprehensive evaluation, is as follows:

[0051] Through fuzzy changes, the weight set W is related to the fuzzy relationship matrix R to conduct a comprehensive evaluation of the evaluation indicators, and the fuzzy evaluation vector B is obtained, which is expressed as follows:

[0052] B=W·R

[0053] Step 3-2-6, calculate the comprehensive evaluation score, as follows:

[0054] According to the assignment of the evaluation set V and the fuzzy evaluation vector B, the comprehensive score S of the electronic medical record data quality evaluation index system is obtained as follows:

[0055] S=B·V T

[0056] Where T represents the matrix transpose.

[0057] Further, the preference analysis described in step 4 includes:

[0058] The preferences of different user categories for the electronic medical record data quality assessment indicator system are analyzed from the indicator level and dimension level, as follows:

[0059] Step 4-1, classifying the users corresponding to the scoring samples;

[0060] Step 4-2, calculate the mean score of each category of users on different indicators, and arrange them in descending order to obtain the preference order of different user categories on the electronic medical record data quality assessment indicator system at the indicator level;

[0061] Step 4-3, calculate the score of each scoring sample in the structural dimension and relational dimension, draw a grouped scatter plot for visual analysis, and obtain the preferences of different user categories for the electronic medical record data quality assessment index system at the dimensional level.

[0062] Furthermore, the user's scoring sample for the electronic medical record data quality assessment index DQAI in step 3-1 includes:

[0063] A scoring sheet was designed based on the electronic medical record data quality assessment indicator system.

[0064] Furthermore, the scoring table described in step 3-1 adopts the Likert five-point scoring method to express the user's opinion on the importance of the indicator.

[0065] Furthermore, the evaluation set V described in step 3-2-2 is set according to the scoring table described in step 3-1, as follows:

[0066] v 1 is very unimportant, v 2 Not important, v 3 For general, v 4 For important, v 5 is very important, and the evaluation set V is assigned accordingly.

[0067] Beneficial effects:

[0068] The user-based electronic medical record data quality assessment index system and assessment method provided by the present invention can propose electronic medical record EHR system optimization suggestions that are more suitable for each user group from the perspective of different types of users, thereby improving the quality of electronic medical record EHR data. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments, and the above and / or other advantages of the present invention will become more clear.

[0070] Figure 1 It is a general step flow chart of the present invention.

[0071] Figure 2 It is the construction method of the present invention.

[0072] Figure 3 It is a schematic diagram of a model for providing optimization suggestions in the present invention. DETAILED DESCRIPTION

[0073] The overall idea of ​​the present invention is as follows: As an important foundation of health and medical big data, users of electronic medical record data can be hospital managers, clinical researchers, doctors, nurses and other data producers, health institutions or industry partners, etc. In the hospital system, users of different job categories have different needs for DQ. Different users have different motivations for using the electronic medical record system, which will directly affect the quality of the data and decision-making. Based on the "applicability" of electronic medical record data quality, the present invention systematically conducts literature review analysis and uses quantitative statistical methods to sort out 16 high-frequency indicators of electronic medical record data quality and give them definitions and explanations; then, a questionnaire survey of EHR users is used, combined with the exploratory factor analysis method to construct two dimensions, thereby forming an electronic medical record data quality indicator system DQAI. Then, the system is comprehensively evaluated and scientifically verified by combining the entropy weight method with the fuzzy comprehensive evaluation method. Finally, combined with the empirical data from questionnaire surveys of the four most representative EHR user groups, namely hospital managers, doctors, nurses, and clinical researchers, the different preferences for electronic medical record data quality indicators are analyzed from different user perspectives at the indicator level and dimensional level, and targeted suggestions are put forward for the optimization of the electronic medical record system, thereby promoting the improvement of electronic medical record data quality.

[0074] The purpose of the present invention is to provide a user-oriented EHR data quality assessment indicator system and method, which can evaluate the EHR data quality from the perspective of different user groups and put forward targeted suggestions for optimizing the EHR system and improving data quality.

[0075] To achieve the above object, the technical solution of the present invention provides a method for evaluating the quality of electronic medical record data, comprising the following steps:

[0076] Step 1: Use quantitative methods to extract electronic medical record data quality indicators. Based on key literature, list all data quality assessment indicators; check these indicators according to the "fitness-for-use" principle, and those that meet this principle will be included, otherwise they will be removed; count the frequency of all listed indicators, retain high-frequency ones, and remove low-frequency ones. Finally, get the final indicator.

[0077] Step 2: Verify the rationality of the indicators through a questionnaire survey. Design the questionnaire "Electronic Medical Record Data Quality Assessment Indicator Questionnaire", use the Likert five-point scoring method to examine the electronic medical record users' understanding and importance preference of each indicator. Based on the questionnaire results, conduct reliability tests, internal consistency tests, difference analysis, weight analysis, structural validity analysis and construct dimensions.

[0078] Step 3: Combine entropy weight with fuzzy comprehensive evaluation and use a quantitative and qualitative method to conduct a comprehensive evaluation of the electronic medical record data quality assessment index system.

[0079] Step 4: Based on different user types, conduct preference analysis on the electronic medical record data quality assessment indicators at the indicator level and dimension level. According to the analysis results, make targeted suggestions for optimizing the electronic medical record system and improving data quality.

[0080] The 16 electronic medical record data quality indicators DQAI include: accuracy, completeness, timeliness, consistency, precision, standardization, uniqueness, credibility, rationality, traceability, portability, availability, accessibility, relevance, applicability, and understandability. The definitions and examples of each indicator are as follows:

[0081] ● Accuracy: The patient information recorded in the electronic medical record is consistent with the patient's actual situation. For example, whether the patient's name, age, etc. are registered accurately;

[0082] ●Completeness: The patient information recorded in the electronic medical record is detailed and complete. For example, whether the patient's personal information and medical condition information are complete;

[0083] ● Timeliness: The patient's condition recorded in the electronic medical record system is timely and effective. For example, when a patient's condition changes from a mild to a severe condition, relevant information is updated in a timely manner;

[0084] ·Consistency: The internal and external consistency of electronic medical record data should meet the consistency indicators claimed by the manager. For example, the patient's electronic medical record data should be consistent within the department and between joint diagnosis and treatment departments, medical insurance and other internal and external departments;

[0085] ● Precision: The qualitative or quantitative accuracy of the data in the electronic medical record should meet the level claimed by the data maker. For example, the location of breast lumps should be completed based on the combination of clock faces and quadrants, and the height should be recorded accurately to centimeters, etc.

[0086] ● Standardization: Electronic medical record data is stored, processed and circulated in a standardized format. For example, in the electronic medical record, information is registered uniformly according to the department's digital code and the patient's ID number;

[0087] ● Uniqueness: Electronic medical record data records are non-repetitive; some data must remain unique. For example, patient medical record data registered with ID number as the primary keyword must remain unique and avoid duplication;

[0088] ● Credibility: Electronic medical record data comes from professional institutions; the data is often reviewed. For example, the diagnosis data in the electronic medical record needs to be filled in by the attending physician of a professional hospital;

[0089] ● Reasonableness: whether the values ​​of electronic medical record data are reasonable, etc. For example, whether the patient's age, condition and other data are in line with common sense;

[0090] ● Traceability: Ensure the auditability of electronic medical record data access traces and electronic medical record data change traces. For example, electronic medical record data can track information such as diagnosis and treatment institutions and doctors;

[0091] ● Portability: The degree to which electronic medical record data can be stored, replaced, or transferred from one system to another and maintain the existing quality should be consistent with the level claimed by the data maker. For example, electronic medical record data can be copied and stored from medical institutions to medical insurance institutions in accordance with certain specifications to provide support for medical insurance settlement;

[0092] ● Availability: The degree to which electronic medical record data can be used should be consistent with the level claimed by the data manager. For example, the attending physician can effectively call up the patient's previous physical examination data from the electronic medical record system to diagnose the current condition;

[0093] ● Accessibility: Whether the electronic medical record data can be easily accessed and extracted, and whether the interface is user-friendly. For example, the electronic medical record data provides a professional website for attending doctors and patients to access and obtain conveniently across time and space boundaries;

[0094] ● Relevance: There should be some kind of relationship between electronic medical record data. For example, the electronic medical record patient ID number is associated with diagnosis, treatment and medical reimbursement information;

[0095] ● Applicability: The content recorded in the electronic medical record is applicable to the patient's health management and disease diagnosis and treatment; the extracted data is applicable to the research conducted or the diagnosis made. For example, the patient's physical examination information recorded in the electronic medical record provides relevant support for the diagnosis of the disease;

[0096] Understandability: The degree of preview and explanation of electronic medical record data should be consistent with the level claimed by the data collector. For example, the information explained by the electronic medical record data can be understood by doctors and patients and achieve the expected use value.

[0097] The "Electronic Medical Record Data Quality Assessment Index Questionnaire" consists of two parts: basic information and electronic medical record data quality assessment index survey (16 questions in total). The basic information part includes 10 questions such as gender, age, job category, education, professional title, department, years of clinical work, years of using the electronic medical record system, and application level of the electronic medical record system. The electronic medical record data quality assessment index survey part is based on the 16 items to construct the electronic medical record data quality index DQAI, forming 16 questions including the name, definition, and examples of each indicator. The Likert five-point method is used to describe the EHR users' views on the importance of these indicators. 1 point is very unimportant, 2 points are unimportant, 3 points are generally important, 4 points are important, and 5 points are very important.

[0098] Sample questionnaire questions:

[0099] 1. In your daily work, how important do you think the accuracy of electronic medical record data is?

[0100] (Accuracy: The patient information recorded in the electronic medical record is consistent with the patient's actual situation. For example, whether the patient's name, age, etc. are registered accurately)

[0101] ○1 Very unimportant ○2 Not important ○3 Average ○4 Important ○5 Very important

[0102] The principal component analysis was used for the survey results of the questionnaire. By fixing two factors and orthogonal rotation, two DQAI dimensions were defined, namely: structural dimension (including 9 indicators: accuracy, completeness, timeliness, consistency, precision, standardization, uniqueness, credibility, and rationality) and relational dimension (including 7 indicators: traceability, portability, availability, accessibility, relevance, applicability, and understandability). That is, the electronic medical record data quality assessment index system described in the present invention consists of 16 indicators and 2 dimensions.

[0103] For the 16 electronic medical record data quality indicators DQAI, the entropy weight method is used for weight calculation and sorting, which is used as a reference for different user preference analysis and a weight set for fuzzy comprehensive evaluation. The entropy weight method is an objective value assignment method, and entropy can measure the degree of disorder of the system. Generally speaking, the smaller the information entropy of an indicator, the greater the degree of variation of the indicator, the greater its role in the comprehensive evaluation, and the greater its weight. The calculation process of the entropy weight method is as follows:

[0104] (1) Establishing the original data evaluation matrix;

[0105] (2) Normalization processing;

[0106] (3) Calculate the proportion of the i-th sample value under the j-th indicator to the indicator;

[0107] (4) Calculate the information entropy of each indicator;

[0108] (5) Calculate the weight of each indicator.

[0109] For the constructed electronic medical record data quality evaluation index system, the weight set obtained by the entropy weight method is used, the entropy weight method is combined with fuzzy comprehensive evaluation, and a qualitative and quantitative method is used for comprehensive evaluation to ensure the scientific nature of the index system. Fuzzy comprehensive evaluation is a method that converts qualitative evaluation into quantitative evaluation based on fuzzy mathematical membership theory. The general steps are as follows:

[0110] (1) Establish a factor set;

[0111] (2) Establishing an evaluation set;

[0112] (3) Establish a weight set and determine the factor weight vector;

[0113] (4) Establish fuzzy relationship matrix;

[0114] (5) Fuzzy comprehensive evaluation;

[0115] (6) Calculate the comprehensive evaluation score.

[0116] For the electronic medical record data quality assessment index system that has been evaluated as excellent by fuzzy comprehensive evaluation, we further analyze the preferences of different user groups for the DQAI system from the indicator level and dimension level.

[0117] ●Tendency analysis of indicator level:

[0118] By calculating and visualizing the mean values ​​of different types of EHR users on each indicator, we can intuitively discover the importance of different indicators held by different user groups. For example, clinical researchers place the credibility of the data as the most important aspect, while hospital managers only have high requirements for the accuracy of the data, and so on.

[0119] ●Tendency analysis at the dimension level:

[0120] By visually analyzing the scores of the two factors, we can intuitively observe that, for example, hospital managers prefer quantitative structural dimensions; doctors emphasize both relational and structural dimensions, but some extremely low outliers can be observed, and so on.

[0121] Explanations are provided for these observations to make targeted recommendations for EHR system optimization.

[0122] Example:

[0123] See also Figure 1 , Figure 1 It is a flowchart of the overall steps of a construction and evaluation method of an electronic medical record data quality evaluation index system provided by an embodiment of the present invention, and the method steps include:

[0124] Step 1: Use quantitative methods to extract electronic medical record data quality indicators. Based on key literature, list all data quality assessment indicators; check these indicators according to the "fitness-for-use" principle, and those that meet this principle will be included, otherwise they will be removed; count the frequency of all listed indicators, retain high-frequency ones, and remove low-frequency ones.

[0125] Finally, we get the final indicator

[0126] Step 2: Verify the rationality of the indicators through a questionnaire survey. Design the questionnaire "Electronic Medical Record Data Quality Assessment Indicator Questionnaire", use the Likert five-point method to score, and examine the electronic medical record users' understanding and importance preference of each indicator. According to the questionnaire results, Cronbach's α was used for reliability test, Kendall's W coefficient was used for internal consistency test, Kruskal–Wallis (KW) test was used to test the differences of different types of users on each indicator, and exploratory factor analysis was used to perform structural validity analysis and construct dimensions. The entropy weight method was used to weight the indicators.

[0127] Step 3: Combine entropy weight with fuzzy comprehensive evaluation and use a quantitative and qualitative method to conduct a comprehensive evaluation of the electronic medical record data quality assessment index system.

[0128] Step 4: Based on different user types, conduct preference analysis on the electronic medical record data quality assessment indicators at the indicator level and dimension level. According to the analysis results, make targeted suggestions for optimizing the electronic medical record system and improving data quality.

[0129] See also Figure 2 , Figure 2It is a method for constructing an electronic medical record data quality assessment index system based on applicability provided by an embodiment of the present invention. The present invention mainly refines electronic medical record data quality assessment indicators through three steps: the first step is to list all data quality assessment related indicators that are focused on in 30 articles; the second step is to verify these indicators according to the "fitness-for-use" principle, and the indicators that meet this principle are included, otherwise they are removed; the third step is to count the frequency of all listed indicators, retain high-frequency ones, and remove low-frequency ones. Finally, 16 indicators are obtained for subsequent research procedures. The 16 indicators are defined and examples are given. The 16 electronic medical record data quality indicators DQAI include: accuracy, completeness, timeliness, consistency, precision, standardization, uniqueness, credibility, rationality, traceability, portability, availability, accessibility, relevance, applicability, and understandability. The definitions and examples of each indicator are as follows:

[0130] ● Accuracy: The patient information recorded in the electronic medical record is consistent with the patient's actual situation. For example, whether the patient's name, age, etc. are registered accurately;

[0131] ●Completeness: The patient information recorded in the electronic medical record is detailed and complete. For example, whether the patient's personal information and medical condition information are complete;

[0132] Timeliness: The patient's condition recorded in the electronic medical record system is timely and effective. For example, when a patient's condition changes from a mild to a severe condition, relevant information is updated in a timely manner;

[0133] ·Consistency: The internal and external consistency of electronic medical record data should meet the consistency indicators claimed by the manager. For example, the patient's electronic medical record data should be consistent within the department and between joint diagnosis and treatment departments, medical insurance and other internal and external departments;

[0134] ● Precision: The qualitative or quantitative accuracy of the data in the electronic medical record should meet the level claimed by the data maker. For example, the location of breast lumps should be completed based on the combination of clock faces and quadrants, and the height should be recorded accurately to centimeters, etc.

[0135] ● Standardization: Electronic medical record data is stored, processed and circulated in a standardized format. For example, in the electronic medical record, information is registered uniformly according to the department's digital code and the patient's ID number;

[0136] ● Uniqueness: Electronic medical record data records are non-repetitive; some data must remain unique. For example, patient medical record data registered with ID number as the primary keyword must remain unique and avoid duplication;

[0137] Credibility: Electronic medical record data comes from professional institutions; the data is often reviewed. For example, the diagnosis data in the electronic medical record needs to be filled in by the attending physician of a professional hospital;

[0138] Reasonableness: whether the values ​​of electronic medical record data are reasonable, etc. For example, whether the patient's age, condition and other data are in line with common sense;

[0139] Traceability: Ensure the auditability of electronic medical record data access traces and electronic medical record data change traces. For example, electronic medical record data can track information such as diagnosis and treatment institutions and doctors;

[0140] Portability: The degree to which electronic medical record data can be stored, replaced, or transferred from one system to another and maintain the existing quality should be consistent with the level claimed by the data producer. For example, electronic medical record data can be copied and stored from a medical institution to a medical insurance agency in accordance with certain specifications to provide support for medical insurance settlement;

[0141] Availability: The degree to which electronic medical record data can be used should be consistent with the level claimed by the data manager. For example, the attending physician can effectively call up the patient's previous physical examination data from the electronic medical record system to diagnose the current disease;

[0142] Accessibility: Whether the electronic medical record data can be easily accessed and extracted, and whether the interface is user-friendly. For example, the electronic medical record data provides a professional website for attending doctors and patients to access and obtain conveniently across time and space boundaries;

[0143] · Relevance: There should be some kind of relationship between electronic medical record data. For example, the electronic medical record patient ID number is associated with diagnosis, treatment and medical reimbursement information;

[0144] Applicability: The content recorded in the electronic medical record is applicable to the patient's health management and disease diagnosis and treatment; the extracted data is applicable to the research conducted or the diagnosis made. For example, the patient's physical examination information recorded in the electronic medical record provides relevant support for the diagnosis of the disease;

[0145] Understandability: The degree of preview and explanation of electronic medical record data should be consistent with the level claimed by the data collector. For example, the information explained by the electronic medical record data can be understood by doctors and patients and achieve the expected use value.

[0146] The "Electronic Medical Record Data Quality Assessment Index Questionnaire" consists of two parts: basic information and electronic medical record data quality assessment index survey (16 questions in total). The basic information part includes 10 questions such as gender, age, job category, education, professional title, department, years of clinical work, years of using the electronic medical record system, and application level of the electronic medical record system. The electronic medical record data quality assessment index survey part is based on the 16 items to construct the electronic medical record data quality index DQAI, forming 16 questions including the name, definition, and examples of each indicator. The Likert five-point method is used to describe the EHR users' views on the importance of these indicators. 1 point is very unimportant, 2 points are unimportant, 3 points are generally important, 4 points are important, and 5 points are very important.

[0147] Sample questionnaire questions:

[0148] 1. In your daily work, how important do you think the accuracy of electronic medical record data is?

[0149] (Accuracy: The patient information recorded in the electronic medical record is consistent with the patient's actual situation. For example, whether the patient's name, age, etc. are registered accurately)

[0150] ○1 Very unimportant ○2 Not important ○3 Average ○4 Important ○5 Very important

[0151] The principal component analysis was used for the survey results of the questionnaire. By fixing two factors and orthogonal rotation, two DQAI dimensions were defined, namely: structural dimension (including 9 indicators: accuracy, completeness, timeliness, consistency, precision, standardization, uniqueness, credibility, and rationality) and relational dimension (including 7 indicators: traceability, portability, availability, accessibility, relevance, applicability, and understandability). That is, the electronic medical record data quality assessment index system described in the present invention consists of 16 indicators and 2 dimensions.

[0152] For the 16 electronic medical record data quality indicators DQAI, the entropy weight method is used for weight calculation and sorting, which is used as a reference for different user preference analysis and a weight set for fuzzy comprehensive evaluation. The entropy weight method is an objective value assignment method, and entropy can measure the degree of disorder of the system. Generally speaking, the smaller the information entropy of an indicator, the greater the degree of variation of the indicator, the greater its role in the comprehensive evaluation, and the greater its weight. The calculation process of the entropy weight method is as follows:

[0153] (1) Establish the original data evaluation matrix K:

[0154]

[0155] Among them, k ijRepresents the final evaluation score of the jth evaluation indicator of the i-th sample.

[0156] (2) Normalization processing:

[0157] For the evaluation matrix K, the indicators of different qualities are normalized. The calculation formula is:

[0158] If the indicator is a positive indicator (generally speaking, the indicators included in the field of electronic medical record data quality are all positive indicators):

[0159]

[0160] If the indicator is a negative indicator (if there is a negative indicator, it can also be calculated using the following formula):

[0161]

[0162] (3) Calculate the proportion of the i-th sample value under the j-th indicator to the indicator. The formula is:

[0163]

[0164] (4) Calculate the information entropy of each indicator. The information entropy calculation formula of the jth indicator is:

[0165]

[0166] (5) Calculate the weight of each indicator. The weight formula of the jth indicator is:

[0167]

[0168] For the constructed electronic medical record data quality evaluation index system, the weight set obtained by the entropy weight method is used, the entropy weight method is combined with fuzzy comprehensive evaluation, and a qualitative and quantitative method is used for comprehensive evaluation to ensure the scientific nature of the index system. Fuzzy comprehensive evaluation is a method that converts qualitative evaluation into quantitative evaluation based on fuzzy mathematical membership theory. The general steps are as follows:

[0169] (1) Establishing a factor set: A factor set is a set of elements consisting of various factors that affect the evaluation object, usually represented by U, where U = {u 1 ,u 2 ,…,u m}, where element u j Represents the jth factor that affects the evaluation object, and these factors usually have different degrees of fuzziness.

[0170] (2) Establishing an evaluation set: An evaluation set is a set of various possible results that the evaluator may make on the evaluation object, usually represented by V, where V = {v 1 ,v2 , …, v l}, where element v j represents the jth evaluation result. In the present invention, v 1 Very unimportant, v 2 Not important, v 3 For general, v 4 For important, v 5 is very important and is assigned a value of V = {55, 65, 75, 85, 95}.

[0171] (3) Establish a weight set and determine the factor weight vector:

[0172] This study uses the weight set obtained by the entropy weight method as the factor weight vector. j The importance of each is different, and each corresponds to a weight, usually represented by W: W = {w 1 , w 2 ,…,w m}, and satisfy:

[0173]

[0174] (4) Establish fuzzy relationship matrix:

[0175]

[0176] r jh (j=1, 2, ..., m; h=1, 2, ..., l) is the membership of the j-th evaluation indicator in the factor set U to the h-th evaluation level in the evaluation set V, reflecting the fuzzy relationship between the evaluation indicator and the evaluation level expressed by the membership degree.

[0177] (5) Fuzzy comprehensive evaluation:

[0178] Through fuzzy change, the weight set W of 16 indicators calculated by entropy weight method is correlated with fuzzy relationship matrix R to conduct comprehensive evaluation on the evaluation indicators, and the fuzzy evaluation vector B is obtained as follows:

[0179] B=W·R (6)

[0180] Then the comprehensive evaluation result is obtained according to the maximum membership principle.

[0181] (6) Calculate the comprehensive evaluation score:

[0182] According to the assignment of the evaluation set V and the fuzzy evaluation vector B, the comprehensive score S of the data quality evaluation index system constructed based on electronic medical record data is obtained as follows:

[0183] S=B·V T (7)

[0184] See also Figure 3 , Figure 3 This is a schematic diagram of a model for providing optimization suggestions for the electronic medical record system using the tendency analysis of user preferences provided by the embodiment of the present invention. For the electronic medical record data quality evaluation index system that has been evaluated as excellent by fuzzy comprehensive evaluation, the preferences of different user groups for the DQAI system are further analyzed from the index level and the dimension level.

[0185] ·Tendency analysis at the indicator level:

[0186] By calculating and visualizing (such as heat maps or density maps) the mean values ​​of different types of EHR users on each indicator, we can intuitively discover the importance of different indicators held by different user groups. For example, clinical researchers place the credibility of the data as the most important aspect, while hospital managers only have high requirements for the accuracy of the data, and so on.

[0187] ·Tendency analysis at the dimension level:

[0188] Draw a grouped scatter plot based on the scores of the two factors. The x-axis represents the relational indicator dimension, and the y-axis represents the structural indicator dimension. The larger the X and Y values ​​are, the higher the factor score is and the higher the importance attached to the factor is. Use different colors and shapes to distinguish the distribution of different job roles. It can be intuitively observed that, for example, hospital managers are more inclined to quantify the structural dimension; doctors emphasize the relational and structural dimensions, but some extremely low outliers can be observed, and so on.

[0189] Explanations are provided for these observations to make targeted suggestions for optimizing the EHR system. For example, for users who are clinical researchers, who are very concerned about data credibility at the indicator level, reliability and credibility can be improved in EHR data acquisition and source invention to gain clinical researchers' recognition of the EHR system, thereby promoting the possibility of improving EHR data quality. Similarly, clinical researchers prefer structural dimensions at the dimension level, and more consideration can be given to using structured data when designing the system, thereby increasing their recognition of the system and willingness to use it, thereby promoting the improvement of EHR data quality.

[0190] In a specific implementation, the present application provides a computer storage medium and a corresponding data processing unit, wherein the computer storage medium can store a computer program, and when the computer program is executed by the data processing unit, the invention content of a user-based electronic medical record data quality assessment method provided by the present invention and some or all of the steps in each embodiment can be executed. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0191] Those skilled in the art can clearly understand that the technical solutions in the embodiments of the present invention can be implemented by means of computer programs and their corresponding general hardware platforms. Based on such an understanding, the technical solutions in the embodiments of the present invention can be essentially or partly contributed to the prior art in the form of computer programs, i.e., software products, which can be stored in a storage medium and include several instructions for enabling a device including a data processing unit (which can be a personal computer, a server, a single-chip microcomputer, an MCU or a network device, etc.) to execute the methods described in various embodiments of the present invention or certain parts of the embodiments.

[0192] The present invention provides a method and idea for a user-based electronic medical record data quality assessment method. There are many methods and approaches to implement the technical solution. The above is only a preferred implementation of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the protection scope of the present invention. All components not specified in this embodiment can be implemented using existing technologies.

Claims

1. A user-based electronic medical record data quality assessment method, characterized in that: The following steps are involved: Step 1: Construct the electronic medical record data quality assessment index DQAI; Step 2, dividing the electronic medical record data quality assessment index DQAI into dimensions and constructing an electronic medical record data quality assessment index system; Step 3, comprehensively evaluating the electronic medical record data quality assessment index system; Step 4: Based on different user types, conduct preference analysis on the electronic medical record data quality indicator DQAI at the indicator level and dimension level to obtain analysis results; complete the user-based electronic medical record data quality assessment.

2. A user-based electronic medical record data quality assessment method according to claim 1, characterized in that: The electronic medical record data quality assessment indicators DQAI described in step 1 include: Accuracy, completeness, timeliness, consistency, sophistication, standardization, uniqueness, credibility, rationality, traceability, portability, availability, accessibility, relevance, applicability and understandability.

3. A user-based electronic medical record data quality assessment method according to claim 2, characterized in that: The dimensions are: structural dimension and relational dimension; among them, The structural dimension includes 9 indicators: accuracy, completeness, timeliness, consistency, precision, standardization, uniqueness, credibility and rationality; The relational dimension includes seven indicators: traceability, portability, availability, accessibility, relevance, applicability and understandability.

4. A user-based electronic medical record data quality assessment method according to claim 3, characterized in that: The comprehensive evaluation of the electronic medical record data quality assessment index system described in step 3 includes: Step 3-1, collecting user scoring samples for the electronic medical record data quality assessment index DQAI; Step 3-2, using the entropy weight method to calculate and sort the electronic medical record data quality assessment index DQAI; Step 3-3, use fuzzy comprehensive evaluation method to calculate the comprehensive evaluation score.

5. A user-based electronic medical record data quality assessment method according to claim 4, characterized in that: The weight calculation and sorting described in step 3-1 include: Step 3-1-1, establish the original data evaluation matrix K, as follows: Among them, k ij represents the score value of the jth indicator of the i-th scoring sample, n represents the number of scoring samples, and m represents the number of indicators; Step 3-1-2, normalize the original data evaluation matrix K to obtain the normalized evaluation matrix K ′ , as follows: Among them, k i ′ j is the normalized evaluation matrix K ′ Elements in Step 3-1-3, calculate the proportion p of the evaluation score of the i-th sample under the j-th indicator ij , as follows: Step 3-1-4, calculate the information entropy of each indicator, as follows: Meet e j ≥0 Among them, e j represents the information entropy of the jth indicator; Step 3-1-5, calculate the weight of each indicator, as follows: Among them, w j Represents the weight of the j-th indicator.

6. A user-based electronic medical record data quality assessment method according to claim 5, characterized in that: The fuzzy comprehensive evaluation method described in step 3-2 is used to calculate the comprehensive evaluation score, including: Step 3-2-1, establish factor set U, as follows: U={u1,u2,…,u j ,…,u m } Among them, the element u j Represents the jth factor that affects the evaluation object; Step 3-2-2, establish the evaluation set V, as follows: V={v1,v2,…,v h ,…,v l } Among them, the element v h represents the hth evaluation result, l represents the number of evaluation result types, which is determined by the design method of the scoring sample; Step 3-2-3, establish the weight set W and determine the factor weight vector: W={w1,w2,…,w j ,…,In m } Each factor u j Corresponding to a weight w j , which is the weight w in step 3-1-5 ij , used to indicate its importance, the weights in the weight set W satisfy the following conditions: Step 3-2-4, establish the fuzzy relationship matrix R, as follows: Among them, r jh is the membership degree of the jth evaluation index in the factor set U to the hth evaluation level in the evaluation set V; Step 3-2-5, fuzzy comprehensive evaluation, is as follows: Through fuzzy changes, the weight set W is related to the fuzzy relationship matrix R to conduct a comprehensive evaluation of the evaluation indicators, and the fuzzy evaluation vector B is obtained, which is expressed as follows: B=W·R Step 3-2-6, calculate the comprehensive evaluation score, as follows: According to the assignment of the evaluation set V and the fuzzy evaluation vector B, the comprehensive score S of the electronic medical record data quality evaluation index system is obtained as follows: S=B·V T Where T represents the matrix transpose.

7. A user-based electronic medical record data quality assessment method according to claim 6, characterized in that: Conduct a preference analysis as described in step 4, including: The preferences of different user categories for the electronic medical record data quality assessment indicator system are analyzed from the indicator level and dimension level, as follows: Step 4-1, classifying the users corresponding to the scoring samples; Step 4-2, calculate the mean score of each category of users on different indicators, and arrange them in descending order to obtain the preference order of different user categories on the electronic medical record data quality assessment indicator system at the indicator level; Step 4-3, calculate the score of each scoring sample in the structural dimension and relational dimension, draw a grouped scatter plot for visual analysis, and obtain the preferences of different user categories for the electronic medical record data quality assessment index system at the dimensional level.

8. A user-based electronic medical record data quality assessment method according to claim 7, characterized in that: The user's scoring sample for the electronic medical record data quality assessment indicator DQAI described in step 3-1 includes: A scoring sheet was designed based on the electronic medical record data quality assessment indicator system.

9. A user-based electronic medical record data quality assessment method according to claim 8, characterized in that: The rating scale described in step 3-1 uses the Likert five-point scoring method to express users' views on the importance of the indicator.

10. A user-based electronic medical record data quality assessment method according to claim 9, characterized in that: The evaluation set V described in step 3-2-2 is set according to the scoring table described in step 3-1, as follows: v1 is very unimportant, v2 is unimportant, v3 is average, v4 is important, and v5 is very important. The evaluation set V is assigned values ​​accordingly.

Images