J Med Syst (2015) 39: 73 DOI 10.1007/s10916-015-0257-8

TRANSACTIONAL PROCESSING SYSTEMS

Research and Development of Semantics-based Sharable Clinical Pathway Systems Hua-Qiong Wang 1 & Tian-Shu Zhou 1 & Yi-Fan Zhang 1 & Li Chen 2 & Jing-Song Li 1

Received: 8 November 2014 / Accepted: 2 June 2015 / Published online: 13 June 2015 # Springer Science+Business Media New York 2015

Abstract The clinical pathway (CP) as a novel medical management schema is beneficial for reducing the length of stay, decreasing heath care costs, standardizing clinical activities, and improving medical quality. However, the practicability of CPs is limited by the complexity and expense of adding the standard functions of electronic CPs to existing electronic medical record (EMR) systems. The purpose of this study was to design and develop an independent clinical pathway (ICP) system that is sharable with different EMR systems. An innovative knowledge base pattern was designed with separate namespaces for global knowledge, local knowledge, and real-time instances. Semantic web technologies were introduced to support knowledge sharing and intelligent reasoning. The proposed system, which was developed in a Java integrated development environment, achieved standard functions of electronic CPs without modifying existing EMR systems and integration environments in hospitals. The interaction solution between the pathway system and the EMR system simplifies the integration procedures with other hospital information systems. Five categories of transmission information were summarized to ensure the interaction process. Detailed procedures for the application of CPs to patients and managing This article is part of the Topical Collection on Transactional Processing Systems * Jing-Song Li [email protected] 1

Engineering Research Center of EMR and Intelligent Expert System, Ministry of Education, Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, College of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou, China

2

Department of General Surgery, The Second Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou, China

exceptional alerts are presented by explicit data flow analysis. Compared to embedded pathway systems, independent pathway systems feature greater feasibility and practicability and are more advantageous for achieving the normalized management of standard CPs. Keywords Clinical pathways . Knowledge base construction . Electronic medical record systems . Interactive process

Introduction Health care costs are increasing in many countries across the world. In addition, preventable adverse medical events can have fatal consequences or seriously injure patients. Thus, greater focus is now being placed on delivering cost-effective, efficient health care and at the same time improving medical quality. The clinical pathway (CP) as a medical management schema can reduce the length of hospital stays, decrease health care costs, standardize clinical activities, and optimize patient outcomes [1–7]. However, few attempts have been made to research and develop a pathway system that could be practically utilized in hospitals. To promote the standardization of pathway management in China, the Ministry of Health has issued paperbased CP specifications for more than 300 diseases covering 22 departments over the last 5 years [8]. Despite an extensive study of methods to model and computerize paper-based CPs [9–12], only a few studies have focused on the practical aspects of pathway execution in the healthcare setting [13]. The utility rate of CPs remains unsatisfactory [14, 15]. Clinical guidelines, which are based on a systematic review of clinical evidence, are focused on supporting clinical decision making by providing recommendations for physicians [16, 17].

73 Page 2 of 11

However, the details regarding implementing therapeutic interventions are not included. In addition, when and how to perform the interventions are also undefined, which hinders the electronic implementation of clinical guidelines in practical healthcare settings [18]. Recent studies have reported the effectiveness of CPs and have identified CPs as standard care processes to support the implementation of clinical guidelines and protocols [19–22]. Taken together, these studies indicate that electronic pathways are of increasing significance due to the construction of digital hospitals and concerns regarding medical quality. Embedding pathway functions in the electronic medical record (EMR) system is a common electronic pathway implementation mode [19, 23–25]. The functions of electronic pathways are embedded in the application of the EMR system as a CP module. However, several drawbacks make this solution difficult to popularize in medical institutions. First, embedded pathway functions as a child module of the EMR system must be designed and developed during the EMR development process. The processes of creating, editing, and executing a CP should be simultaneously considered during the pre- and post-design of the EMR system to allow effective real-time information sharing and exchange between the EMR system and the pathway module. However, most hospitals have already adopted EMR systems [26–29] that do not integrate the functions of CPs and in which no interface for CPs was reserved in the development process. Under such circumstances, existing EMR systems must be replaced with new EMR systems to implement embedded electronic pathways. Replacing an existing EMR system is tremendously costly and is a substantial obstacle to the computerization of pathways. Second, the knowledge of standard CPs embedded in the EMR system is changeable because the pathway module is developed in combination with the EMR system. During the implementation of embedded pathways, clinical physicians are allowed to adjust the detailed tasks of pathways when necessary, which is not conducive to the normalized management of the standard pathways. To address the problems outlined above, a sharable, independent clinical pathway (ICP) system was proposed in this study. Without replacing or modifying the existing EMR systems, the following functions of electronic pathways was achieved using the ICP system: (1) provide standard care plans for specific diseases; (2) customize personalized pathways for different patients; and (3) manage the entire medical process. Therefore, implementation of the ICP system is beneficial for regulating medical behavior and enhancing medical quality. Compared to the embedded pathway module, the ICP system is sharable and becomes more intelligent. To support the sharable nature of the ICP system, semantic web technologies [30–32] were introduced to construct the knowledge base. A novel pattern of the knowledge base was designed with

J Med Syst (2015) 39: 73

separate namespaces for global knowledge, local knowledge, and real-time instances, which were represented using web ontology language (OWL) [33, 34]. By predefining the mapping rules between the global knowledge and local knowledge, the ICP system can be shared between different EMR systems. Despite this knowledge sharing, the terms and rules in the global knowledge are under unified management, which is advantageous for normalized implementation of standard CPs. In addition, semantic reasoning provides a novel approach to improve the intelligence of the ICP system. Therefore, under the current implementation of hospital information systems, the independent pathway system is more practical and intelligent than the embedded pathway module and does not require modification of the original EMR system.

Methods Pattern design of the knowledge base To represent pathway knowledge in a standard manner, this paper describes CP concepts and relations using OWL and defines pathway execution rules using the Jena semantic framework. The knowledge base of the ICP system is a pivotal foundation for improving the drawback of embedded CPs. The ontology-based knowledge base is sharable, ensuring the sharing of the ICP system with different EMR systems. Additionally, the terms and rules defined in the knowledge base are under unified management and maintenance, ensuring the normalized implementation of standard CPs. A novel knowledge base pattern was designed with three parts, as shown in Fig. 1. The global knowledge includes the standard execution procedures of CPs, common terms, medical rules and pathway rules. The terms and rules in global knowledge are typically established by domain experts and are maintained unchanged after being issued to medical institutions. The local knowledge includes hospital-specific terms. The terms in the local knowledge are established by the healthcare information manager of each hospital, are represented by knowledge engineers in a semi-automatic manner, and are finally stored as a part of the knowledge base semantically. Mapping between the local terms of each hospital and the terms within the global knowledge ensures knowledge connection and sharing between the ICP system and other hospital information systems. Real-time instances include clinical data, such as the patient’s real-time status, the operational condition of medical equipment, clinical activities in standard pathways, the execution results of pathway tasks, and so on. The real-time instances are updated during the CP execution process. The content, source, stability, and establishment of the three parts of the knowledge base are listed in Table 1.

J Med Syst (2015) 39: 73

Page 3 of 11 73

Fig. 1 Pattern design of the knowledge base

In previous studies, there has been no clear division between global knowledge, local knowledge, and real-time instances. All of the terms, rules, and instances are usually in one namespace, which limits the reasoning efficiency and is difficult for both management and maintenance. To address these problems, according to the knowledge stability we proposed an innovative knowledge base pattern with three separate namespaces, which can effectively improve system performance and control knowledge normalization. However, the three parts are interconnected. Terms in local knowledge usually have mapping relationships with terms in global knowledge, and terms and instances in different parts also have a close connection. As shown in Fig. 1, within the local knowledge, each hospital has its own set of terms, which are independent of each other and can be mapped to the common terms in the global knowledge. Three steps are necessary to construct local

Table 1

knowledge before adopting the ICP system at a specific hospital. (1) The content of local terms must be established according to the data dictionary tables in the hospital information systems, primarily in EMR systems. (2) The local terms must be translated to a semantic format in a semi-automatic manner using semantic transformation tools, such as D2RQ [35], and stored in a separate namespace. (3) Relationships with the predefined common terms in global knowledge must be mapped using the ontology editor tool Protégé [36]. The mapping step enables the standardization of hospital-specific local terms to facilitate information sharing between the ICP system and other clinical information systems. System architecture of the ICP system In addition to the knowledge base as the core of the ICP system, two other modules are included to achieve system

Three parts of the clinical pathway knowledge base Content

Source

Stability

Established by

Global knowledge

Global terms Global rules Pathway process

Stable

Official institutions, such as the Ministry of Health

Local knowledge

Local terms Local rules Real-time data

SNOMED CT Specifications recommended by the Ministry of Health of China; clinical guidelines; expert suggestions EMR systems Physicians Clinical practice

Relatively stable

Hospitals

Real-time upgrade

Clinical practice

Real-time instances

73 Page 4 of 11

functionality: the engine module for reasoning and execution and the interface module for human interaction, as shown in Fig. 2. Based on Jena’s inference engine, the reason engine consists of an OWLReasoner and creates an inference model from the OWL ontology. The reason engine executes the semantic reasoning for pathway personalization and real-time dynamic adjustment. The execution engine, consisting of the Jena Ontology API, accesses the knowledge base using a simple protocol and PDF query language (SPARQL) [37] and edits the OWL ontology through the Jena ModelFactory. By reading and writing the ontology in the clinical pathway knowledge base, the execution engine achieves interface control of the knowledge base. The interface module, which includes a pathway editor interface and a pathway execution management interface, is designed and developed within the Java-integrated development environment Eclipse. The interfaces for managing pathway execution include the doctor’s workstation, the nurse’s workstation, and the patient’s view interface. Implementation solution for the ICP system Medical information constructs in hospitals have been developed for decades in many other countries. Hospital information systems, such as EMR systems, pharmacy systems, laboratory information systems (LIS), and radiology information systems (RIS), have been widely used. The integration environment has also been fully developed using data table interfaces or integrative platforms. When the ICP system is implemented into the hospital information environment as a new independent system, it is possible to exchange and share information with other hospital information systems. To improve the adaptability of the ICP system to different medical environments, the ICP system is designed to only exchange information with the EMR system and to share information with other systems indirectly via Fig. 2 System architecture of ICPs

J Med Syst (2015) 39: 73

the EMR system. This interaction solution will significantly simplify the exchange procedures between the ICP system and other systems. The interaction process is the main content of the implementation solution for the ICP system, and the most important step to ensure the interaction process is the identification of the interaction information. Based on the clinical requirements, the transmission of information between ICP systems and EMR systems are summarized, as listed in Table 2. As an independent system, the ICP system has the complete functionality of pathway creation and management. When a pathway is applied to a patient, the ICP system first captures the patient’s information from the EMR system. Then, the ICP system saves the pathway information in the knowledge base, including the patient’s admission and procedure entry information, and synchronizes this information with the EMR database. The EMR system records patient transfer information, procedure execution results, and vital sign values, which are received in real time by the ICP system. Using semantic reasoning with the procedure execution results and vital sign values in the knowledge base, the ICP system monitors the medical process and generates exception alerts for physicians. The clinical pathway data recorded in the knowledge base of the ICP system are available for statistical analysis and are useful for optimizing the pathway model. Above all, the ICP system can achieve pathway creation and analysis independently and can provide complete medical process management due to the interaction solution with the EMR systems.

Results Realization of the knowledge base As mentioned above, the knowledge base of the ICP system includes global knowledge, local knowledge, and real-time

J Med Syst (2015) 39: 73 Table 2 Transmission of information between ICP systems and EMR systems

Page 5 of 11 73

Transmission Information

Source

Target

Description

Patient information Basic information

EMRs

ICPs

EMRs

ICPs

Patient’s unique identifier, name, gender, birthday, blood type, allergy history, medical history, etc. Diagnostic type, diagnostic no, diagnostic name, ICD-10 code, etc.

ICPs

EMRs

EMRs

ICPs

Procedure information Procedure entry information

ICPs

EMRs

Execution status Execution results

EMRs EMRs

ICPs ICPs

EMRs

ICPs

Patients’ vital signs recorded by nurses

ICPs

PC clients, mobile clients

Order repeat, execution delay, allergy, exceptional vital signs, etc.

Diagnosis information Patient hospital information Admission information Transfer or discharge records

Vital sign information Vital sign records Exception information Exception alerts

instances. To retain the portability and maintenance of the knowledge base, 3 separate namespaces are constructed for global ontology (globalOnto), local ontology (localOnto), and real-time instances (eventInst), as shown in Fig. 3. In Fig. 3, 3 super classes (globalOnto:Person, globalOnto:Comm, and globalOnto:ClinicalPathway) are defined in the global ontology to express the global common knowledge. Class globalOnto:Person records a person’s demographic information. As the demographic information does not change at different hospitals, class globalOnto:Person is defined as part of the global ontology. Class globalOnto:ClinicalPathway records the essential steps in the care of patients with a specific clinical problem and primarily lists the essential procedures, documents, and observations related to the patient. Class globalOnto:Comm, with 6 subclasses (CommDisease, CommDocument, CommFinding, CommObservation, CommProcedure, and CommConstant), records all of the global clinical terms that have been normalized by the systematized nomenclature of human and veterinary medicine (SNOMED) [38]. Class CommProcedure records the terms of the clinical procedures, such as examination orders, laboratory orders, operation orders, injection orders, prescription orders, and so on. For instance, a laboratory order Bblood cell count (CBC)^ with SNOMED ID P3-30500 is an individual of class CommLaboratoryOrder, which is a subclass of CommProcedure.

Admission date, time, department, ward and bed, diagnosis, doctor, nurse, etc. Patients’ transferred-out and transferred-in records, discharge records Order type, order name, execution time, execution department, etc. Execution status of patients’ order Outcome, laboratory report, radiation report, etc.

Correspondingly, the local ontology, which is an independent namespace, records hospital-specific clinical terms. Class localOnto:Local, which corresponds to class globalOnto:Comm in the global ontology, records local terms according to the existing terms utilized in the hospital information systems. In addition, the 6 subclasses LocalDisease, LocalDocumet, LocalFinding, LocalObservation, LocalProcedure, and LocalConstant have a one-to-one correspondence with the 6 subclasses of globalOnto:Comm, and subclasses LocalDepartment and LocalStaff are added to record hospital-specific department information and staff information. A super class LocalOnto:Hospital is defined to record basic hospital information. A set of object properties is defined to achieve a connection between the local ontology and the global ontology. Instances recorded in the knowledge base can be divided into two parts: (1) Implementation information for CPs. eventInst:EventClinicalPathway, which is a class defined in the namespace of real-time instances, corresponds to globalOnto:ClinicalPathway in the global ontology via an object property eventInst:corresPathway. Several properties, such as hasPatientId, hasStatus, hasStartDate, and hasVariance, are defined to record detailed information about the execution of a CP. (2) Transmission of information between the ICP system and the EMR system. As shown in Table 2, the transmission information includes the patient’s information, hospital information, order information, vital sign information, and exception information.

73 Page 6 of 11

J Med Syst (2015) 39: 73

Fig. 3 Realization of the CP knowledge base

System functions of the ICP system The system was developed via the Eclipse Java integrated development environment in a Java development environment. The Jena semantic web framework was imported to construct ontology models and to support semantic reasoning. Wakamiya and Yamauchi classified the standard functions of electronic CPs into six categories: displaying, recording, ordering, editing, variance, and statistics [23]. In the ICP system, these standard CP functions are achieved via the following six modules. (1) Patient list. Via the module interface of the patient list, existing patient information in the knowledge base can be queried and viewed. For a new patient who has not been recorded in the knowledge base, the patient’s identifier is captured by reading the patient’s clinic card.

Then, detailed patient information is obtained from EMR systems and is recorded into the knowledge base. (2) Patient overview. As shown in Fig. 4, the detailed implementation situation for the CPs is displayed in the module interface within the patient overview. By doubleclicking on one record in the left table of the patient list, the basic information and treatment procedures (usually a CP) for the selected patient will be displayed in the widgets on the right. The patient’s treatment procedures are displayed according to the procedures’ execution time. Selecting an item from the treatment procedures allows detailed procedure information to be displayed in the bottom widget. Creating a new CP or a new order for the patient is achieved by using the buttons in the bottom right corner. (3) Diagnosis management. The diagnosis management module provides physicians with the convenience of

J Med Syst (2015) 39: 73

Page 7 of 11 73

Fig. 4 A primary interface for the ICP system

managing patient diagnostic information within the ICP system without switching to the EMR system interface. (4) Pathway management. Existing definitions of CPs can be viewed in this module. Managers with the necessary credentials can edit and verify the definitions of CPs. To ensure the normalized management of standard CPs, the credentials are strictly controlled. (5) Variance management. By using the module interface within variance management, physicians can query all previous variances and verify the occurrence of the current variance. The variance statistical results are also provided to strengthen medical quality management. (6) Knowledge management. Global ontology and local ontology terms can be viewed in this module. The statistical results from the CPs and the variance are displayed in a knowledge format.

management of CPs as an auxiliary system. The ICP system interacts with the EMR system via the predefined terms in the local ontology. The orders, documents, and vital sign observation items in the EMR system are encapsulated within the patterns of the CPs. However, the execution and management of the orders, documents, and vital sign observation items are achieved by means of the existing healthcare information systems at the hospital via the integration environment. The procedures for creating a CP for a patient via an ICP system are shown in Fig. 5. &

& Execution cases of the ICP system EMR systems accomplish patient diagnosis and regular treatment, whereas the ICP system completes the creation and

The ICP system captures a patient’s unique identifier via the card reader attached to the ICP workstation and subsequently obtains the patient’s basic information and diagnosis information from the EMR system (1a/1b). The patient information obtained is automatically saved to the CP knowledge base (2a) and semantically reasoned according to the predefined input rules of the CPs (2b). Based on the reasoning result, an applicable CP is provided on the ICP workstation (2c).

73 Page 8 of 11

J Med Syst (2015) 39: 73

Fig. 5 Procedures for creating a CP for a patient via an ICP system

&

&

&

The orders, documents, and vital sign observations predefined by the CPs are modified and confirmed by the physicians. A personalized CP is created considering the patient’s specific situation (3). The orders, documents, and vital sign observations from the CP created in the ICP system are saved to the EMR database via the web service (4a). The EMR workstation automatically displays the orders, documents, and vital sign observations saved in the EMR database (4b). Physicians can manage and execute these items in the EMR system (4c). For orders such as medication, examinations, and laboratory work, the EMR system exchanges information between the pharmacy systems, the LIS, and the RIS using the existing integrated environment, and achieves task assignment and result retrieval (4c’). Ultimately, the execution results of different information systems are recorded to the EMR database. The ICP system captures the resulting data from the EMR system by monitoring updates of the EMR database using a background process (4d). The resulting data include vital sign values, execution status, execution results, and reports of orders and documents.

Variance management is a significant function module of ICPs. The rules for variance alerts are predefined in the global ontology of the CP knowledge base. The variance information primarily includes the following three parts. 1) Vital sign values that are outside a specified range: To implement CPs, a patient’s vital signs usually must be relatively stable within a normal range. For instance, a high body temperature will terminate the execution of most CPs. 2) Exceptional alterations of standard CPs: Procedures may change based on the practical situation in pathway execution. For instance, physicians may disable a laboratory order for a CBC at the EMR workstation if a patient’s fever is reduced more rapidly than expected. The ICP system will record these changes to support CP statistics and analysis. 3) Exceptional execution results: If the execution results of the laboratory/examination orders defined in the CP are not in accordance with the expected outcomes, the pathway execution will be hampered or possibly terminated. The procedures for managing variance information are shown in Fig. 6.

J Med Syst (2015) 39: 73

Page 9 of 11 73

Fig. 6 Procedures for managing variance information

&

&

& &

As described in the procedures for creating a CP for a patient via the ICP system, the ICP system monitors record updates in the EMR system and captures relative data, which include observation values, execution status, execution results, and reports (4d). The captured observation values, execution statuses, and execution results are saved in the CP knowledge base (5a) and are determined according to the exception rules (5b). Execution reports are saved in document files named with the patients’ identifiers and can be viewed on the ICP workstation by physicians. If an exceptional situation occurs, the ICP knowledge base returns the variance information in the ICP workstation (5c). Physicians confirm the exceptional alerts in the ICP workstation and can decide to stop the pathway based on the practical situation (6). To alert the physicians in charge in a timely manner, the variance information, which includes exceptional items and the reasons for their occurrence, can be transmitted to PC or mobile clients (7). The information on the PC or mobile clients is predefined in the local ontology by the physicians using the ICP system.

As described in Figs. 5 and 6, the interaction procedures between the ICP system and the EMR system only require the EMR system to provide access and write permission to the data tables related to the CPs rather than permission to modify the original EMR system program. As shown in Table 2, the number of pathway-related data tables is not large. The ICP system proactively monitors the data tables in the EMR system to obtain necessary information. Therefore, the functions of the ICP system can be totally achieved without modifying the original EMR system.

Discussion The ICP system proposed in this study achieves the functions of electronic CPs. The ICP system can provide a standard care plan for specific diseases and effective medical process management as well as attaining the objectives of normalizing medical action and improving medical quality. The knowledge-based clinical pathways are maintained in a unified matter, are stable, and are advantageous for the normalization of medical actions. A novel pattern for the knowledge base with separate namespaces was designed to promote semantic reasoning and improve system performance. The

73 Page 10 of 11

interaction solution between the ICP system and the EMR system simplifies system integration, achieving process management of the pathway data. Compared to the common pathway systems that are embedded in the EMR system, the ICP system has the following advantages: (1) The ICP system adds to the functionality of electronic CPs without modifying the existing EMR system and integration environment in the hospital. Introducing a new, independent CP system to the previous system is less expensive than replacing the existing EMR system. (2) In the knowledge base, the terms and processes of CPs, which are described by OWL in the global ontology, cannot be modified by clinical physicians. This restriction ensures that the global knowledge will be sharable under normalized management and unified maintenance and, consequently, that the electronic CPs will be created and executed regularly. Once a modification occurs, it will be recorded as a variance. Recording variance information is significantly important to CP analysis and optimization. (3) Compared to the embedded CP system, the independent pathway system has greater portability and expandability. Semantic web technologies are introduced to support knowledge sharing and term mapping. Mapping the relations between global ontology and different EMR systems enables the centralized management of patientcentered information. With respect to patient privacy, clinical physicians only have permission to access global knowledge and local knowledge. Access to instances in the knowledge base is restrained by the authority limits defined in the local knowledge of the ICP system. The permission rules are consistent with the rules for the EMR system. The ICP system is designed to avoid modifying the original EMR system. Records in the EMR system are passively captured by the ICP system rather than actively provided by the EMR system, and records in the ICP system are actively transferred to the EMR system and saved in the EMR database. The process of information transport is led by the ICP system, whereas the EMR system does not actively provide any information. The EMR system must grant permission to access and edit the related tables in the EMR database. In addition, exceptional alerts cannot emerge in the EMR system directly. Therefore, the messages in PC clients or mobile clients are designed to be used to remind physicians to focus on exceptional situations. As a tool to support clinical decision making, the intelligence of the pathway system is very important. The knowledge base of the ICP system is constructed using semantic web technologies, which feature greater intelligence than the common database. In addition, the statistical analysis results for existing data were provided to make the standard care plans defined by standard CPs more explicit and detailed. By merging the treatment experience acquired from the clinical data, more detailed and complete decision support is

J Med Syst (2015) 39: 73

provided for clinical physicians to develop more efficient and intelligent clinical procedures [39]. A limitation of this study is that natural text is not included for intelligence reasoning. Natural language processing for unstructured text is computationally expensive. Few approaches to analyze and categorize text efficiently have been identified. Based on a comprehensive consideration of system performance, the ICP system executes the reasoning engine on the structured instances in the knowledge base.

Conclusion Standard functions of electronic pathways are designed and achieved in the ICP system without replacing or modifying the original EMR system and integration environment in hospitals. Applying semantic web technologies to support knowledge base construction and semantic reasoning ensures knowledge sharing and optimizes system intelligence. A novel pattern for the knowledge base was designed with separate namespaces for global knowledge, local knowledge and realtime instances. By predefining the mapping rules between global knowledge and the local knowledge, the ICP system is sharable with different EMR systems. Despite this knowledge sharing, the terms and rules in the global knowledge are under unified management, which is advantageous for the normalized implementation of standard CPs. In addition, the interaction solution between the ICP system and the EMR system simplifies system integration. Compared to embedded pathway systems, independent pathway systems present greater feasibility and portability and are more beneficial to achieve the normalized management of terms and procedures defined in standard CPs. Acknowledgments This work was supported by the National Natural Science Foundation (Grant No. 61173127), National High-tech R&D Program (No. 2013AA041201, No. 2015AA020109) and the Fundamental Research Funds for the Central Universities.

References 1.

2.

3.

Smith, K. A., Matthews, T. W., Dube, M., Spence, G., and Dort, J. C., Changing practice and improving care using a Low-risk tracheotomy clinical pathway. JAMA Otolaryngol. Head Neck Surg. 140(7):630–634, 2014. doi:10.1001/jamaoto.2014.921. Deng, Y. M., Jiao, Y. H., Hu, R. R., Wang, Y. L., Wang, Y. J., and Zhao, X. Q., Reduction of length of stay and costs through the implementation of clinical pathways for stroke management in china. Stroke 45(5):E81–E83, 2014. doi:10.1161/strokeaha.114. 004729. Tarin, T., Feifer, A., Kimm, S., Chen, L., Sjoberg, D., Coleman, J., and Russo, P., Impact of a common clinical pathway on length of hospital stay in patients undergoing open and minimally invasive

J Med Syst (2015) 39: 73

4.

5.

6.

7.

8. 9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

kidney surgery. J. Urol. 191(5):1225–1230, 2014. doi:10.1016/j. juro.2013.11.030. Burgers, P., Van Lieshout, E. M. M., Verhelst, J., Dawson, I., and de Rijcke, P. A. R., Implementing a clinical pathway for hip fractures; effects on hospital length of stay and complication rates in five hundred and twenty six patients. Int. Orthop. 38(5):1045–1050, 2014. doi:10.1007/s00264-013-2218-5. Gonenc, M., Dural, A. C., Celik, F., Akarsu, C., Kocatas, A., Kalayci, M. U., Dogan, Y., and Alis, H., Enhanced postoperative recovery pathways in emergency surgery: a randomised controlled clinical trial. Am. J. Surg. 207(6):807–814, 2014. doi:10.1016/j. amjsurg.2013.07.025. Markar, S. R., Schmidt, H., Kunz, S., Bodnar, A., Hubka, M., and Low, D. E., Evolution of standardized clinical pathways: refining multidisciplinary care and process to improve outcomes of the surgical treatment of esophageal cancer. J. Gastrointest. Surg. 18(7): 1238–1246, 2014. doi:10.1007/s11605-014-2520-6. Chu, S., Computerised clinical pathway as process quality improvement tool. In: Patel, V. L., Rogers, R., and Haux, R. (Eds.), Proceedings of the 10th world congress on medical informatics, vol. 84. Ios Press, London, pp. 1135–1139, 2001. China clinical pathway web. http://www.ch-cp.org.cn/. Okada, O., Ohboshi, N., Kuroda, T., Nagase, K., and Yoshihara, H., Electronic Clinical path system based on semistructured data model using personal digital assistant for onsite access. J. Med. Syst. 29(4): 379–389, 2005. doi:10.1007/s10916-005-5896-8. Ye, Y., Jiang, Z., Diao, X., Yang, D., and Du, G., An ontologybased hierarchical semantic modeling approach to clinical pathway workflows. Comput. Biol. Med. 39(8):722–732, 2009. doi:10.1016/ j.compbiomed.2009.05.005. Okada, O., Ohboshi, N., and Yoshihara, H., Clinical path modeling in XML for a Web-based benchmark test system for medication. J. Med. Syst. 29(5):539–553, 2005. doi:10.1007/s10916-005-61108. Du, G., Jiang, Z., Yao, Y., and Diao, X., Clinical pathways scheduling using hybrid genetic algorithm. J. Med. Syst. 37(3):1–17, 2013. doi:10.1007/s10916-013-9945-4. Hurley KF, Abidi SSR (2007) Ontology Engineering to Model Clinical Pathways: Towards the Computerization and Execution of Clinical Pathways. Paper presented at the Proceedings of the 20th IEEE International Symposium on Computer-Based Medical Systems, Maribor, Slovenia, 2007 Jun 20–27 Hindle, D., and Yazbeck, A. M., Clinical pathways in 17 european union countries: a purposive survey. Aust. Health Rev.: Publ. Aust. HospAssoc. 29(1):94–104, 2005. Tao, H., Liu, P., Liang, J., Ke, X., Guo, S., and Qu, H., Survey and analysis of the hospitals implementing clinical pathway. Chin. Hosp. Manag. 30(2):28–30, 2010. Isern, D., and Moreno, A., Computer-based execution of clinical guidelines: a review. Int. J. Med. Inform. 77(12):787–808, 2008. doi:10.1016/j.ijmedinf.2008.05.010. Peleg, M., Tu, S., Bury, J., Ciccarese, P., Fox, J., Greenes, R. A., Hall, R., Johnson, P. D., Jones, N., Kumar, A., Miksch, S., Quaglini, S., Seyfang, A., Shortliffe, E. H., and Stefanelli, M., Comparing computer-interpretable guideline models: a case-study approach. J. Am. Med. Inform. Assn. 10(1):52–68, 2003. doi:10.1197/jamia. M1135. Gooch, P., and Roudsari, A., Computerization of workflows, guidelines, and care pathways: a review of implementation challenges for process-oriented health information systems. J. Am. Med. Inform. Assn. 18(6):738–748, 2011. doi:10.1136/amiajnl-2010-000033. Li, W. Z., Liu, K. C., Yang, H. Q., and Yu, C. G., Integrated clinical pathway management for medical quality improvement - based on a semiotically inspired systems architecture. Eur. J. Inform. Syst. 23(4):400–417, 2014. doi:10.1057/ejis.2013.9.

Page 11 of 11 73 20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34. 35. 36. 37. 38. 39.

Mammen, C., Matsell, D. G., and Lemley, K. V., The importance of clinical pathways and protocols in pediatric nephrology. Pediatr. Nephrol. 29(10):1903–1914, 2014. doi:10.1007/s00467-0132577-6. Nussbaum, D. P., Penne, K., Speicher, P. J., Stinnett, S. S., Perez, A., White, R. R., Clary, B. M., Tyler, D. S., and Blazer, D. G., The role of clinical care pathways: an experience with distal pancreatectomy. J. Surg. Res. 190(1):64–71, 2014. doi:10.1016/j.jss.2014.02.026. Combi, C., Gambini, M., Migliorini, S., and Posenato, R., Representing business processes through a temporal data-centric workflow modeling language: an application to the management of clinical pathways. IEEE Trans. Syst. Man Cybern. Syst. 44(9): 1182–1203, 2014. doi:10.1109/tsmc.2014.2300055. Wakamiya, S., and Yamauchi, K., What are the standard functions of electronic clinical pathways? Int. J. Med. Inform. 78(8):543–550, 2009. Yang, H., Li, W., Liu, K., and Zhang, J., Knowledge-based clinical pathway for medical quality improvement. Inf. Syst. Front. 14(1): 105–117, 2012. doi:10.1007/s10796-011-9307-z. Wang, H. Q., Li, J. S., Zhang, Y. F., Suzuki, M., and Araki, K., Creating personalised clinical pathways by semantic interoperability with electronic health records. Artif. Intell. Med. 58(2):81–89, 2013. doi:10.1016/j.artmed.2013.02.005. Li, J. S., Zhang, X. G., Wang, H. Q., Wang, Y., Wang, J. M., and Shao, Q. D., The meaningful use of EMR in Chinese hospitals: a case study on curbing antibiotic abuse. J. Med. Syst. 37(2):1–10, 2013. doi:10.1007/s10916-013-9937-4. Li, J. S., Zhang, X. G., Chu, J., Suzuki, M., and Araki, K., Design and development of EMR supporting medical process management. J. Med. Syst. 36(3):1193–1203, 2012. doi:10.1007/s10916010-9581-1. Blumenthal, D., and Tavenner, M., The Bmeaningful use^ regulation for electronic health records. N. Engl. J. Med. 363(6):501–504, 2010. doi:10.1056/NEJMp1006114. Yui, B.-H., Jim, W.-T., Chen, M., Hsu, J.-M., Liu, C.-Y., and Lee, T.-T., Evaluation of computerized physician order entry system - a satisfaction survey in Taiwan. J. Med. Syst. 36(6):3817–3824, 2012. doi:10.1007/s10916-012-9854-y. García-Castro, R., and Gómez-Pérez, A., Interoperability results for semantic web technologies using OWL as the interchange language. Web Semant. Sci. Serv. Agents World Wide Web 8(4):278– 291, 2010. doi:10.1016/j.websem.2010.08.008. Janev, V., and Vranes, S., Applicability assessment of Semantic Web technologies. Inf. Process. Manag. 47(4):507–517, 2011. doi:10.1016/j.ipm.2010.11.002. Khan, W., Khattak, A., Hussain, M., Amin, M., Afzal, M., Nugent, C., and Lee, S., An adaptive semantic based mediation system for data interoperability among health information systems. J. Med. Syst. 38(8):1–18, 2014. doi:10.1007/s10916-014-0028-y. Jing, X., Kay, S., Marley, T., and Hardiker, N., Integration of an OWL-DL knowledge base with an EHR prototype and providing customized information. J. Med. Syst. 38(9):1–14, 2014. doi:10. 1007/s10916-014-0075-4. Web Ontology Language (OWL). . D2RQ Accessing Relational Databases as Virtual RDF Graphs. . The Protégé Ontology Editor. . SPARQL Query Language for RDF. . SNOMED CT The Global Language of Healthcare. http://www. ihtsdo.org/snomed-ct. Wang, H. Q., Zhou, T. S., Tian, L. L., Qian, Y. M., and Li, J. S., Creating hospital-specific customized clinical pathways by applying semantic reasoning to clinical data. J. Biomed. Inform. 2014. doi:10.1016/j.jbi.2014.07.017.

Research and Development of Semantics-based Sharable Clinical Pathway Systems.

The clinical pathway (CP) as a novel medical management schema is beneficial for reducing the length of stay, decreasing heath care costs, standardizi...
3MB Sizes 1 Downloads 12 Views