The application of data mining technology in hospital performance management under the background of big data

doi:10.21203/rs.3.rs-3118501/v1

Download PDF

Research Article

The application of data mining technology in hospital performance management under the background of big data

https://doi.org/10.21203/rs.3.rs-3118501/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

With the rapid development of computer and information technology, all walks of life have realized informatization, and the industries and fields covered by informatization technology almost cover the daily life. In all aspects, the medical industry is paying more and more attention to informatization construction. From the initial manual documents to the later electronic prescriptions, from the previous focus on equipment, to the gradual focus on software, from passive acceptance in the past to the active use of software to cooperate with management, there are various signs that the medical industry has gone through It has entered the stage of "information electronization", and has entered an important period of relying on information technology. With the gradual development of the effect of business intelligence on hospital management decision support, the expansion of hospital information systems, the accumulation of data, and the upgrading of technology, new contents, functions and applications are continuously introduced. Through the construction of data mining, it helps the hospital to initially realize the transformation from the information system provider to the decision support assistant, and becomes the starting point and effect feedback point of the hospital management decision. This paper analyzes the application history and current situation of data mining in hospitals, as well as the main advantages of data mining in hospital performance management for reference.

Big data

data mining

hospital performance management

With the continuous development and progress of electronic information technology, more and more hospitals have established their own hospital information systems. an important measure of service levels. But in the actual use process, more and more drawbacks began to appear. Nowadays, the reports and related data provided by the hospital information system have always been manually operated, and the internal relationship between the various systems cannot be fully reflected. In the process of statistics of important hospital data, the expansion mechanism and system framework are relatively backward, resulting in inconsistent statistical results and increased error rate. At the same time, the electronic statistics background is relatively backward, and there is a lack of systematic and standardized electronic statistics. In addition, the information and data of different departments cannot be effectively linked, the data is single, and the departments and departments, data and data cannot be shared, and there is a weak decision support management and comprehensive information system, which cannot provide a relatively complete hospital., accurate and efficient decision-making prompts and data analysis.

Since the 1980s, the digitization of hospital information has gone through a process of about 30 years, and has experienced a process of gradual development and expansion, from the simple digitization of hospital financial systems, the digitization of local pharmacy drugs and material management to electronic medical records, and then To the integrated medical image archiving and transmission system (PACS), inspection information system (LIS), etc., the hospital information system (HIS), solutions and functions have been gradually improved and improved ^[1]. At present, the scale of hospital data is getting larger and larger, and the annual growth of business data is increasing year by year, so the analysis and reuse of data is becoming increasingly important. However, there are certain limitations in hospital informatization, and it is difficult to support cross-domain comprehensive analysis in the distributed storage of data. The existing query and statistical data of the hospital come from their respective reports in each business system. It is difficult for the hospital to grasp the overall picture of the operation in a timely manner, and the processing of some reports manually delays the decision-making cycle., performance monitoring and tracking system ^[2]. How to meet the needs of management decision-making and support knowledge discovery is a problem that every hospital has to face. As shown in Fig. 1, long-term, systematic and comprehensive data can be obtained from the data of the decentralized business system, so as to conduct deeper mining and obtain abundant auxiliary decision-making information to improve the quality and level of hospital performance management.

Based on this background, this paper helps hospitals initially realize the transformation from information system providers to decision support assistants through data mining, which becomes the starting point and effect feedback point of hospital management decisions. At the same time, the application history and current situation of data mining in hospitals, and the main advantages of data mining in hospital performance management are analyzed for reference.

data mining application in hospitals can completely realize closed-loop management from insight, decision-making to execution, and realize continuous improvement and micro-management of hospital management, that is, to realize business intelligence through the accumulation of data in the application system. The application of data mining to hospitals can realize the transformation of the following management modes: from experience management to process management supported by data; from recording transaction data to using data for control and analysis to support decision-making; from scattered and inconsistent business data to business data Centralized and unified standardized management; from extensive manual management to refined, microscopic and intelligent management based on data; to help the hospital information system initially realize the transformation from an information system provider to a decision support assistant, becoming a The starting point and effect feedback point of hospital management decision ^[3].

2.1 The development status of hospital performance management

The concept of performance management was first proposed by American management scientist Aubrey Daniels in the 1970s. In 1842, the US federal government implemented performance evaluation for civil servants. The U.S. federal government has conducted a comprehensive performance evaluation of thousands of federal programs ^[4]. Performance management is a complete system, usually composed of measurement, evaluation, and management sequence. In this system, employees' performance is planned in advance, and they are monitored, guided, and evaluated in the process, and ultimately achieved by improving individual performance. Corporate strategic goals. The performance of medical institutions is similar to the performance management of enterprises, and they often use the methods of enterprise performance management for performance management and performance appraisal ^[5]. The United States Joint Evaluation Committee of Medical Organizations (JACHO) defines the "performance" of acute medical facilities as: "the ability of an individual, group or organization to perform some degree or step to increase the expected result", we can consider performance Is the result of the actions, activities, procedures, and actions of an institution and its members that the institution and members expect to achieve with a minimum of resources.

At present, hospital performance management mainly refers to KPI (Key Performance Indicator), BSD (Balanced Scorecard) and other methods ^[6]. The BSD method formulates management objectives from four aspects, namely, finance, customers, internal operations, and learning and growth, and sets corresponding assessment indicators for different specialties and levels. The original intention of KPI is to encourage project participants such as owners, contractors, and suppliers to accurately evaluate their own performance. Based on the "two-eight principle", the purpose of controlling and improving the overall situation is achieved by controlling a small number of important indicators. Good performance management can improve hospital management capabilities and operational efficiency, and promote high-quality development of hospitals ^[7].

Domestic hospitals generally adopt two methods for performance appraisal: revenue and expenditure balance and KPI. The former is more intuitive, and the economic benefits can be seen at a glance by subtracting expenditures from income and multiplying the coefficient. The latter evaluates medical workers by formulating a small number of indicators, and the operation is relatively simple. However, the former easily leads hospitals to ignore social responsibilities, while the latter is often misunderstood that the smaller the better. To solve these problems, hospitals headed by many tertiary hospitals have basically carried out and completed the first round of performance reform. Not only through hiring consulting companies, on the basis of income as the dominant factor, some KPI indicators have been added, and a set of relatively mainstream performance evaluation system has been formed through the combination of economic indicators and assessment indicators ^[8].

2.2 Data mining technology

Data mining (DM) is to extract hidden, unknown, but potentially useful data from a large amount of incomplete, noisy, fuzzy and random data. process of information and knowledge ^[9]. After years of development, data mining has involved many fields. With the development of computer science, people are more comfortable with data processing, and there are more and more ways to acquire knowledge

Current data mining can generally be divided into three steps: data preparation, data mining, and interpretation of results ^[10].

Data preparation is generally carried out when building a data warehouse, and this stage can be subdivided into: data integration, data selection, and data preprocessing. The data preparation stage is to obtain a large amount of data from the various heterogeneous systems of the hospital, and perform data association, cleaning, and standardization, so as to organize and simplify the data into standardized data that is easy to identify and has clear semantics ^[11]. If the data mining is carried out on the basis of the data warehouse, this part of the data will be stored in the data warehouse. When creating a data warehouse, it is often accompanied by the step of data preparation^[12].

Data mining, as the most core part of data mining, requires participants to have knowledge in multiple fields, so as to help them decide to select a specific one from many data mining algorithms as the solution for this knowledge acquisition. Therefore, data mining is the most critical and difficult part of the entire data mining process. As shown in Fig. 2, the collection link in the hospital data processing process is to extract data related to decision-making goals from multiple databases ^[13]. Using the backup database provided by historical data, data extraction is performed from the backup database, and the extraction of clinical data is realized by the "data extraction-load-transformation" (ELT) technology ^[14].

The expression and interpretation of the results is to distinguish the most valuable part from the output part of the algorithm according to the needs of the end user, which is used to judge the quality of the completion of the above two parts. If the result presentation cannot satisfy the user, it is possible to re-prepare the data, re-select the data mining algorithm, etc., and so on for many rounds^[15].

The ultimate goal of data mining is the acquisition of useful knowledge, so if satisfactory knowledge cannot be obtained, the whole mining process will be iterative until the result is satisfactory or the research fails. At each stage of the data mining process, various factors may hinder the final result, such as filtering useful information during data cleaning, insufficient data collection samples; using inappropriate data in the data mining stage Algorithms and models; failure to identify important output results when the results are expressed, etc., may cause the mining task to fail and start over^[16].

2.3 The main advantages of applying data mining to hospital performance management

The quality of a hospital performance system directly affects the loyalty and satisfaction of employees, thereby affecting the cohesion and competitiveness of the hospital. Therefore, performance management is the top priority of hospital management. The performance system is an important indicator system for evaluating the contribution of the organization and employees ^[17]. Business intelligence BI (hospital data mining and application analysis platform) is based on the cycle of performance management (performance planning-process management-performance evaluation-result application), and realizes the evaluation and evaluation of the work performance of the organization and employees through a variety of evaluation methods. Goal management and process management with employee performance. Link hospital strategic management and performance management, guide managers through process application to improve the standardization of hospital internal operations, and link hospital strategic management and performance management^[18].

2.3.1 More detailed data reports

The hospital performance management system should be based on the SOA architecture and provide a variety of modular functions. These modular functions include dimensional reports and dashboards, scorecards, analysis, business event management, and unified metadata. Data and security management. All functions are provided on top of a single Web service architecture for easy implementation and deployment for scientific performance management ^[19].The performance management system can quickly and flexibly realize various reporting requirements of the hospital, and provide the ability of multi-dimensional data analysis. Hospitals can use data mining technology to drill through to detailed data after discovering anomalies, providing data support for hospitals. In the overall solution of the performance management software used by the hospital, the functions of all performance management system modules can be brought into full play, such as drill-through, complex format report production, etc.

2.3.2 The overall management is fast and efficient

The report on the multi-dimensional analysis Cube can be obtained in seconds through data mining, but the same report cannot be obtained due to timeout after running on the past software for more than 2 hours. Compared with the original implementation of the hospital, data mining is faster and more efficient. The platform software based on business intelligence can realize medical cost analysis, performance appraisal, operation analysis, medical quality analysis, clinical analysis, profit analysis, medical accident analysis, medical quality control, medication safety management, patient relationship in batches according to the needs of the hospital. Management and other needs to meet the diverse needs of different departments and different business personnel ^[20].

The tasks of data mining include data extraction, classification discovery, clustering and association rules. Among them, association rules are mainly used in transactional databases. Through the mining and analysis of factual data, the association relationship between various indicators is established, so as to obtain valuable information. Generally, according to its mode of action, it can be divided into two categories: predictive mode and descriptive mode. Since the development of data mining, it has penetrated into many disciplines, and produced many mining techniques and models, each with its own strengths. Figure 3 shows a prototype of a data mining system model.

3.1 Regression Analysis Algorithms

Traditional statistics is a science based on empirical data. It uses data samples for analysis and prediction to find certain laws of numbers, and then uses these laws to analyze certain objective laws. Statistics has very important applications in many fields such as enterprise application, scientific research, Web mining, medical treatment, finance and so on. The linear regression model can be observed intuitively through the scatter plot (as shown in Fig. 4). It can be seen from the figure that the straight line simulated by the univariate linear regression equation is not a straight line passing through all the points, but an "optimal straight line", that is, the sum of the squares of the distances between it and all the points is the smallest ^[21].

The univariate linear regression model is:

$$\left\{ {\begin{array}{*{20}{c}} {Y={\beta _0}+{\beta _1}x+ε } \\ {ε \sim N(0,{\sigma ^2})} \end{array}} \right.$$

The XGBoost algorithm is a classic method in the integrated technology Boosting. The idea of Boosting algorithm is to form a strong classifier by integrating a large number of weak classifiers. XGBoost uses the CART regression tree. The CART regression tree is a hypothesis tree binary tree that can continuously split the features. If the tree node is split based on the mth eigenvalue in the data, if the eigenvalue is less than s, it is divided into the left subtree, on the contrary, if the eigenvalue is greater than s, it is divided into the right subtree:

$$\begin{array}{*{20}{c}} {{R_1}(m,s)=\{ x|{x^m} \leqslant s\} }&{and}&{{R_2}(m,s)=\{ x|{x^m}>s\} } \end{array}$$

The XGBoost algorithm grows into a tree by continuously adding trees and feature divisions. Each time it is added, it learns a new function to better fit the residuals of the last prediction. After the training of K trees, the sample can be predicted according to the characteristics of the sample. During the training process, each tree will have a score of sub-nodes. By comparing the scores of these sub-nodes That adds up to the final predicted value for that sample.

$$\widehat {{{y_i}}}=\sum\nolimits_{{i=1}}^{K} {{f_i}({x_i}),{f_i} \in F}$$

$$F=\{ f(x)={w_q}(x)\} (q:{R^m} \in T,w \in {R^T})$$

Formula (3) represents the strong classification prediction model, and formula (4) represents the weak classifier model that needs to be integrated. where F represents all CART regression trees, q represents the structure of the tree mapping each sample, and T represents the number of leaf nodes. In addition to the residual value, the loss function of the XGBoost model also introduces a regularization term to increase its generalization ability. The final objective function is expressed as:

$$Obj(\theta )=\sum\nolimits_{{i=1}}^{n} {l({y_i},\widehat {{{y_i}}})} +\sum\nolimits_{{i=1}}^{K} {\Omega ({f_k})}$$

$$\Omega ({f_k})=\gamma T+\frac{1}{2}||w|{|^2}$$

The right part of formula (5) represents the regularization term, which represents the complexity of the model to ensure the final learning result and avoid overfitting. The left part represents the loss function, which ensures the difference between the predicted value and the actual value. difference. Here, the objective function is optimized by the additive training distribution, that is, the first tree of CART is optimized first, and then the second tree is optimized, and so on, until the end of optimizing the Kth tree:

$${\widehat {{{y_i}}}^{(t)}}=\sum\nolimits_{{i=1}}^{t} {{f_k}({x_i})=} {\widehat {{{y_i}}}^{(t - 1)}}+{f_t}({x_i})$$

After the optimization of the Kth tree, the obtained CART tree()tifx is optimal, which minimizes the objective function on the basis of 1()tifx−, namely:

$$\begin{gathered} Ob{j^{(t)}}=\sum\nolimits_{{i=1}}^{n} {l({y_i},{{\widehat {{{y_i}}}}^{(t)}})} +\sum\nolimits_{{i=1}}^{K} {\Omega ({f_k})} \hfill \\ =\sum\nolimits_{{i=1}}^{n} {l({y_i},{{\widehat {{{y_i}}}}^{(t - 1)}}+{f_t}({x_i}))} +\Omega ({f_t})+constant \hfill \\ \end{gathered}$$

When the structure of the t-th tree in the CART regression tree is determined, the optimal value of each leaf node is obtained as:

$$w_{j}^{*}= - \frac{{{G_j}}}{{{H_j}+\lambda }}$$

The value of the objective function is

$$Ob{j^*}=\frac{1}{2}\sum\nolimits_{{j=1}}^{T} {\frac{{G_{j}^{2}}}{{{H_j}+\lambda }}} +\gamma T$$

The XGBoost algorithm has been improved on the CART base classifier, so that it has the function of supporting linear classifiers, which can not only deal with classification problems, but also deal with regression problems, so that it has more uses and scenarios; Through automatic learning, XGBoost can predict the splitting direction of sample data with missing values, instead of simply skipping, reducing data loss and improving data quality and quantity.

3.2 Clustering Analysis Algorithm

Clustering is the process of using a computer to automatically divide object data into groups according to a certain standard. Objects between groups have similar attributes or relationships. When classifying, only rely on the attributes of the analyzed objects themselves. to distinguish between objects and their similarities. At present, cluster analysis is widely used in speech recognition, image segmentation and other fields, and it can also effectively distinguish consumer groups with similar consumption patterns in business analysis. Simply put, a clustering algorithm is to classify the population in order to find structure in the data, as similar as possible within the classes, and as different as possible between the classes. The clustering algorithm generally goes through four steps: feature acquisition, similarity calculation, grouping, and result display. There are many clustering algorithms, which can be divided into different categories. The more common types are: division method, grid method, hierarchical method and model method ^[22].

The K-means algorithm belongs to the division method. The K-means clustering algorithm divides the data set into several limited clusters through cluster analysis, and the data in the clusters have similar characteristics. Since it was proposed, K-means clustering algorithm has been applied to all aspects of social life due to its simplicity and flexibility. In the process of using the algorithm, the shortcomings of the K-means clustering algorithm are gradually revealed. The main shortcomings of K-means clustering algorithm are: 1. The value K must be determined in advance before the experiment starts; 2. The effect of clustering is easily affected by the initial cluster center; 3. When the classification attribute is not obvious, the classification effect is poor And it is easy to obtain the local optimal solution; Fourth, when dealing with large-scale data sets, the amount of calculation is large and the time overhead is large.

The K-means algorithm is the simplest and most commonly used clustering algorithm. In the K-means algorithm, the distance is used as the evaluation index of similarity, that is, the closer the distance is, the higher the similarity is. At the same time, the clustering criterion function is the error sum of squares function:

$$V=\sum\nolimits_{{i=1}}^{k} {\sum\nolimits_{{{x_{j \in {S_i}}}}} {{{({x_j} - {u_i})}^2}} }$$

The objective function of the classical K-means algorithm is shown in formula (11), which aims to find the optimal partitioning scheme. When the weight is introduced into the K-means algorithm, its objective function will change. The new objective function is as follows:

$$V(P,\lambda )=\sum\nolimits_{{i=1}}^{K} {\sum\nolimits_{{l=1}}^{p} {(\lambda _{ε }^{l}\sum\nolimits_{{{x_{j \in {S_i}}}}} {{{(x_{j}^{l} - u_{i}^{l})}^2}} )} }$$

where the centroid is

$$u=(u_{i}^{1},u_{i}^{2}, \cdots ,u_{i}^{p}),i=1,2, \cdots ,K$$

The calculation method of each attribute weight, let

$${φ _l}=\sum\nolimits_{{{x_{j \in {S_i}}}}} {{{({x_j} - {u_i})}^2}}$$

formula can be converted into formula

$$V(P,\lambda )=\sum\nolimits_{{i=1}}^{K} {\sum\nolimits_{{l=1}}^{p} {(\lambda _{ε }^{l}{φ _l})} }$$

When the sum of the distances between the samples in each class and the mass points is the smallest and does not change, V(p) has a minimum value, so the Lagrangian function method can be used to calculate the weights under the constraints. Therefore, the derivation formula of the available attribute weights is as follows:

$$\lambda _{ε }^{l}=\frac{u}{{{φ _l}}}=\frac{{{{(\prod\limits_{{h=1}}^{p} {(\sum\nolimits_{{{x_{j \in {S_i}}}}} {{{(x_{j}^{h} - u_{i}^{h})}^2}} )} )}^{\frac{1}{p}}}}}{{\sum\nolimits_{{{x_{j \in {S_i}}}}} {{{({x_j} - {u_i})}^2}} }}$$

The basic process of the improved K-means is as follows:

(1) randomly select K samples from the overall data as the centroids;

(2) calculate the Euclidean distance between each sample and each centroid, find the nearest centroid, and set the It is classified into this class (j);

(3) Calculate the weight of each attribute of each class according to formula (16);

(4) After obtaining the attribute weight of each class, recalculate the centroid of each class;

(5) Calculate the value of the evaluation index according to the clustering quasi-measurement function, if it is the value of the minimum accuracy, end; if not, turn to 2), and perform related operations.

The clustering algorithm can automatically group the samples without classification basis, which is consistent with the work of the hospital to classify the departments. First of all, the hospital has a variety of data to evaluate the departments, but the evaluation items are too many to make it impossible to accurately sort the departments. Second, some departments can artificially and accurately distinguish their levels. For example, the work intensity and technical content of the ICU have been unanimously recognized by everyone. However, most of the departments are relatively close subjectively, and cannot be accurately judged. Third, the data evaluated by the department can be quantified, and it is easy to be grouped from the analysis of each feature, which is suitable for classification by the K-means algorithm. Combining the above three points, this paper uses the K-means algorithm in the clustering algorithm to classify the departments without grading, and then combines the subjective judgment of professionals to identify the iconic departments in the clustering, and finally grade the departments.

Data mining technology can extract useful data through the data warehouse of the hospital information system and use various data mining tools to perform statistics, analysis and classification, and discover the data information in the medical performance indicator system. Provide hospital performance management. decision reference.

4.1 Experimental methods

The research object selected the data of inpatients with a certain disease in a tertiary hospital from 2012 to 2018. Select the data on the front page of medical records from different data sources of the same batch of patients: (1) Extract the data on the front page of the electronic medical records filled in by doctors from the special disease database of the tertiary hospital, hereinafter referred to as the data of the doctor group or the doctor group; (2) After the medical record administrator and the medical department Data on the first page of medical records after quality control, hereinafter referred to as medical record group or medical record group data. The specific information of data collection items includes: medical record number, gender, age, occupation, number of hospitalizations, admission route, admission time, discharge time, main diagnosis name and disease code, other diagnosis name and disease code, pathological diagnosis, surgical operation name and operation Code, total cost and comprehensive service fee, diagnosis and treatment fee, drug fee, consumable fee, payment method, attending doctor and admission department, and the specific data dictionary is organized into standard field descriptions for data collection, etc. Preprocessing of data is the foundation of any data analysis and mining work. According to the data extraction rules, data extraction, data integration, data cleaning and data mapping are performed on the homepage data of the electronic medical records of the doctor group and the medical record group. Analyze data quality and discuss the reasons for differences in data quality. The tools used are: the application software Microsoft Excel2007 for data preprocessing, SPSSStatistics24.0 for statistical analysis, and SPSSModerler for decision tree grouping.

4.2 Comparison and analysis of experimental results

From the analysis of the daily work of doctors, the main items include surgery, treatment, billing, handling medical records, ward rounds, outpatient clinics, etc. Therefore, the main workload of doctors can be simply divided into two parts: billing execution items and Treatment of discharged/outpatient patients. The former data can be collected from the HIS system, which includes the billing and execution items of the whole hospital. The latter can be obtained from the medical record management system. According to the requirements of the higher-level health system, the medical record system needs to fill in the health statistics form on a monthly basis. The performance of thoracic surgery is analyzed through data mining technology, and the results shown in Fig. 5 can be obtained.

Figure 6 shows the changes in the number of inpatients in the department corresponding to the special disease obtained through data mining in two years. Univariate analysis was carried out with the cost of hospitalization as the dependent variable, and gender, age, admission route, way of leaving the hospital, number of hospitalizations, use of medical insurance, surgery, and degree of comorbidities as independent variables. The normal distribution test showed that the total cost of hospitalization showed a skewed distribution, as shown in Fig. 7. The KruskalWallis test was used for multi-sample comparisons, and the Wilcoxon test was used for two-group comparisons.

The study draws on domestic and foreign experience, and according to the grouping logic of DRGs, eight indicators of gender, age, admission route, discharge method, number of hospitalizations, whether to use medical insurance, whether to operate, and comorbidities are used as input variables. The total cost is used as the target variable to construct a decision tree model. Figure 8 shows the basic description and statistical visualization of the 8 groups of data constructed by the decision tree. From this figure, we can see the number of cases in each group and the total proportion. The results show that the DRGs1 sample is the largest compared to the other 7 groups, accounting for 34% of the total, and the other groups have little difference in the distribution of samples.

All patients are divided into DRGs groups. Through the weights and CMI, the groups can be compared, and the medical service level of the departments and doctors can be ranked. It can also carefully observe which departments are the dominant departments for treatment. Which doctors are more advantageous in specific treatment paths. According to statistics, Doctor C provides the most total medical services, and Doctor D provides the most technically difficult medical services. Managers can evaluate the level of medical services according to two evaluation dimensions, and guide the performance distribution of doctors. According to the current data demand situation, four data warehouses including medical quality database, decision support database, medical record uploading database, and basic data dictionary are initially established. According to the content of performance appraisal indicators, relevant data are extracted from the data warehouse to establish indicator data sets. Through the perfect data mining mechanism, improve the quality of hospital medical management, reduce medical costs, provide patients with a good medical environment, and promote the improvement of scientific management in hospitals. Taking the medical income in 2019 as an example, data mining is used to compare the performance relationship before and after The effect is shown in Fig. 9.

Data mining is the process of extracting hidden useful information and knowledge from a large amount of disordered and fuzzy random data. A large amount of data generated by the hospital information system is processed and integrated according to certain classifications and standards, forming a medical data warehouse that supports the decision-making process of the hospital management. Using data mining technology to carry out micro and macro statistics and synthesis of the integrated data of the data warehouse, it is possible to mine the key factors affecting the operation and development of the hospital, discover their interrelationships, and use the existing data to predict the future development of the hospital, so as to improve the quality of the hospital. Good service for the majority of patients. According to the data mining results of medical performance indicators, it can help hospital managers to formulate hospital development strategies, optimize the allocation of medical resources, and play an auxiliary decision-making role in the strategic development of hospitals. Data mining technology is very necessary and promising in the field of medical management.

Data availability

The figures used to support the findings of this study are included in the article.

Conflicts of interest

The authors declare that they have no conflicts of interest.

Funding statement

This work was supported by the Ningxia Medical University University-level scientific research project “Research on medical service performance evaluation of Yinchuan tertiary hospitals based on DRGs” (No. XM2021063).

Acknowledgements

The authors would like to show sincere thanks to those techniques who have contributed to this research.

Compliance with Ethical Standards

Funding: This study was funded by the Ningxia Medical University University-level scientific research project “Research on medical service performance evaluation of Yinchuan tertiary hospitals based on DRGs” (No. XM2021063).

Author Ying Li declares that she has no conflict of interest.

Author Kehua Wang declares that he has no conflict of interest.

Author Chuanchuan Xu declares that he has no conflict of interest.

Ethical approval: This article does not contain any studies with human participants or animals performed by any of the authors.

Agostinelli, S., Covino, F., D'Agnese, G., De Crea, C., Leotta, F., & Marrella, A. (2020). Supporting governance in healthcare through process mining: a case study. IEEE Access, 8, 186012-186025.
Bhusal, A. (2021). Predicting Student's Performance Through Data Mining. arXiv preprint arXiv:2112.01247.
Emre, I. E., Erol, N., Ayhan, Y. I., Özkan, Y., & Erol, Ç. (2019). The analysis of the effects of acute rheumatic fever in childhood on cardiac disease with data mining. International journal of medical informatics, 123, 68-75.
Feng, Y., Chai, L., Li, X., Zhang, S., & Song, B. (2018, September). A Distributed Procurement Cost Control Scheme of Medical Materials for Regional Medical Consortiums. In 11th EAI International Conference on Mobile Multimedia Communications (pp. 177-186).
Qin, J. (2017, March). Design and implementation of sports performance management system in higher vocational colleges based on data mining. In The 1st EAI International Conference on Multimedia Technology and Enhanced Learning.
Lu, H., Wang, R., & Huang, Z. (2022). Application of data mining in performance management of public hospitals. Mobile Information Systems, 2022, 1-10.
Smith, P. C., & Goddard, M. (2002). Performance management and operational research: a marriage made in heaven?. Journal of the Operational Research Society, 53(3), 247-255.
Liu, Y., Cai, H., Li, G., & Zhao, H. (2018, May). Research on application of data mining in hospital management. In 2018 International Conference on Robots & Intelligent System (ICRIS) (pp. 374-376). IEEE.
Baghapour, M. A., Shooshtarian, M. R., Javaheri, M. R., Dehghanifard, S., Sefidkar, R., & Nobandegani, A. F. (2018). A computer-based approach for data analyzing in hospital’s health-care waste management sector by developing an index using consensus-based fuzzy multi-criteria group decision-making models. International journal of medical informatics, 118, 5-15.
Mosadeghrad, A. M., Shahebrahimi, S. S., & Ghazanfari, M. (2018). Exploring the relationship between accreditation and hospital performance: using data mining approach. Journal of School of Public Health and Institute of Public Health Research, 16(1), 33-50.
Najadat, H., Alaiad, A., Alasal, S. A., Mrayyan, G. A., & Alsmadi, I. (2020). Integration of data envelopment analysis and clustering methods. Journal of Information & Knowledge Management, 19(01), 2040006.
Al-Radaideh, Q. A., & Al Nagi, E. (2012). Using data mining techniques to build a classification model for predicting employees performance. International Journal of Advanced Computer Science and Applications, 3(2).
Lv, S., Song, H., & Cheng, L. (2021). High Quality Development of Comprehensive University Medical Education In China: Problems And Path In The New Era.
Demartini, C., & Mella, P. (2014). Beyond feedback control: the interactive use of performance management systems. Implications for process innovation in Italian healthcare organizations. The International journal of health planning and management, 29(1), e1-e30.
Sherazi, S. W. A., Jeong, Y. J., Jae, M. H., Bae, J. W., & Lee, J. Y. (2020). A machine learning–based 1-year mortality prediction model after hospital discharge for clinical patients with acute coronary syndrome. Health informatics journal, 26(2), 1289-1304.
Wang, J. Q., Xia, K. J., Qian, Y. H., & Peng, X. D. (2018). Hospital Performance Management Research Based on Improved Data-Driven Data Envelopment Analysis Algorithm. Journal of Medical Imaging and Health Informatics, 8(3), 561-565.
Wei, F. (2022). Performance evaluation of tourism human resource management based on fuzzy data mining. Journal of Mathematics, 2022.
Dwyer, A. J. (2010). Medical managers in contemporary healthcare organisations: a consideration of the literature. Australian Health Review, 34(4), 514-522.
Xie, S. T. (2018). The impact of public hospital performance evaluation policies on the hospital internal perform-ance management: A comparative case study based on Beijing and Shanghai. Chinese Journal of Health Policy, 8-13.
Cheng, H., Yang, D., Lu, C., Qin, Q., & Cadasse, D. (2022). Intelligent oil production stratified water injection technology. Wireless Communications and Mobile Computing, 2022, 1-7.

Download PDF

Version 1

posted

You are reading this latest preprint version

The application of data mining technology in hospital performance management under the background of big data

Status:

Version 1

Abstract

Figures

1 Introduction

2 Materials and Methods

2.1 The development status of hospital performance management

2.2 Data mining technology

2.3 The main advantages of applying data mining to hospital performance management

2.3.1 More detailed data reports

2.3.2 The overall management is fast and efficient

3 Methodology

3.1 Regression Analysis Algorithms

3.2 Clustering Analysis Algorithm

4 The Result Analysis And Discussion

4.1 Experimental methods

4.2 Comparison and analysis of experimental results

5 Conclusion

Declarations

References

Status:

Version 1