short in providing flexibility [10, 11] which seriously hampers the application of workflow technology in the healthcare domain. In addition to this, support is needed for the cross-departmental nature of healthcare processes [12]. Currently, administrative workflows are typically limited to single departments [19]. Successful implementation of workflow man-agement exists but widespread adoption and dissemination is the exception rather than the rule [14]. It is expected that the use of workflow technology by healthcare institutions will grow dramatically in the future [14] and it is likely that it will become a core component in future healthcare systems [7].
Despite all these efforts, no work has been performed on the combination of appointment scheduling and workflow management systems, i.e., existing approaches are either focusing on planning with little consideration for workflow aspects or are focusing on workflow while ignoring that much work is done via appointments rather than worklists. We are not aware of any research looking at the mixture of flow and schedule tasks.
For various systems, CPNs have been used to formalize and validate functional require-ments. For example, the formalization of the design of the so-called worklet service, which adds flexibility and exception handling capabilities to the YAWL workflow system [3], for-malizing the implementation of a healthcare process in a workflow management system [13], and presenting a model-based approach to requirements engineering for reactive systems, in which CPNs are used for validating the functional requirements [8]. Related to this is [17], in which CPNs are used for specifying the operational semantics ofnewYAWL, a business process modeling language founded on the well-known workflow patterns5.
7 Experiences and Conclusions
In this paper, we have discussed the design and implementation of a workflow management system offering planning and monitoring facilities. As approach, we started with a workflow language, followed by a conceptual model in CPNs and finally a concrete implementation
of the system. The conceptual model consists of 27 distinct nets, 377 transitions, 169 places and over 1000 lines of ML code. The construction of the whole model took around three months of work. These figures indicate that a workflow system augmented with planning facilities is a fairly complex system and the task of developing it is far from trivial.
One of the main benefits of building the conceptual model in CPNs is that it can be executed in the CPN Tools offering. In this way, it allows for experimentation during which comprehensive insights can be obtained about the design and behavior of the system to be realized which probably would not have been possible to obtain by pursuing other approaches to designing the system. Parts of the system can be tested early in the development process, thus enabling early detection of design errors. The costs of repairing these errors in this phase of the development process is far less than would be the case in a later phase. For example, when experimenting with the subnet of the planning service we identified errors with regard to the correct planning of appointments.
Another advantage of modeling the conceptual model in CPNs is that it completely specifies the behavior of the system to be implemented while abstracting from implemen-tation details and language specific issues. So, for the conceptual model we only needed to worry about the behavior of the system, while for the implementation we focused on the realization. In this way, these kinds of issues are distinguished, allowing for a separation of concerns. The importance of this distinction can probably best be illustrated by the fact that it took more than 3 months to build the conceptual model, and just 3 months to imple-ment the whole system. For the impleimple-mentation of the system it was necessary to produce over 8000 lines of code by hand. Although the main functionality of the system was fully implemented during the implementation phase, a significant amount of time still needs to be spent on component selection, coding, and dealing with residual implementation issues.
The fact that we started completely from scratch ending up with a concrete imple-mentation of the system with the proposed functionality shows both the applicability and feasibility of our approach. However, the developed system has only been tested in a limited set of scenarios. As future work, we plan to systematically test parts of the system by “re-placing” one or more components in the conceptual model by a complete implementation for it, based on third party software, allowing for the testing of thousands of scenarios. In this way by simply executing the CPN model, we are able to identify errors in the components which probably would not have been found with using a scenario based approach of testing.
In addition to this, we plan to use the conceptual model for evaluating alternative planning approaches using various performance indicators.
References
1. W.M.P. van der Aalst and K.M. van Hee. Workflow Management: Models, Methods, and Systems. MIT press, Cambridge, MA, 2002.
2. W.M.P. van der Aalst and A.H.M. ter Hofstede. YAWL: Yet Another Workflow Language.
Information Systems, 30(4):245–275, 2005.
3. M.J. Adams. Facilitating Dynamic Flexibility and Exception Handling for Workflows. PhD thesis, Faculty of Information Technology, Queensland University of Technology, 2007.
4. L. Ardissono, A.D. Leva, G. Petrone, M. Segnan, and M. Sonnessa. Adaptive medical workflow management for a context-dependent home healthcare assistance service. Electronic Notes in Theoretical Computer Science, 146(1):59–68, 2006.
5. B. Cardoen, E. Demeulemeester, and J. Beli¨en. Operating room planning and scheduling: A literature review. FEB Research Report KBI 0807, Katholieke Universiteit Leuven, Leuven, 2008.
6. T. Cayirli and E. Veral. Outpatient scheduling in health care: a review of literature. Product Operations Management, 12(4):519–549, 2003.
7. A. Dwivedi, R. Bali, A. James, and R. Naguib. Workflow Management Systems: the Health-care Technology of the Future? In the 23rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society, volume 4, pages 3887–3890, 2001.
8. J.M. Fernandes, S. Tjell, and J.B. Jorgensen. Requirements Engineering for Reactive Systems with Coloured Petri Nets: the Gas Pump Controller Example. In K. Jensen, editor,Proceedings of the Eight Workshop and Tutorial on Practical Use of Coloured Petri Nets and the CPN Tools, pages 207–222, 2007.
9. K. Jensen, L.M. Kristensen, and L. Wells. Coloured Petri Nets and CPN Tools for Modelling and Validation of Concurrent Systems. STTT, 9(3-4):213–254, 2007.
10. M. Klein, C. Dellarocas, and A. Bernstein, editors. Adaptive Workflow Systems, Special Issue of Computer Supported Cooperative Work, 2000.
11. R. Lenz, T. Elstner, H. Siegele, and K. Kuhn. A Practical Approach to Process Support in Health Information Systems. JAMIA, 9(6):571–585, December 2002.
12. R. Lenz and M. Reichert. IT Support for Healthcare Processes - Premises, Challenges, Per-spectives. Data and Knowledge Engineering, 61:49–58, 2007.
13. R.S. Mans, W.M.P. van der Aalst, P.J.M. Bakker, A.J. Moleman, K.B. Lassen, and J.B. Jor-gensen. From Requirements via Colored Workflow Nets to an Implementation in Several Work-flow Systems. In K. Jensen, editor,Proceedings of the Eight Workshop and Tutorial on Practical Use of Coloured Petri Nets and the CPN Tools, pages 187–206, 2007.
14. M. Murray. Strategies for the Successful Implementation of Workflow Systems within Health-care: A Cross Case Comparison. InProceedings of the 36th Annual Hawaii International Con-ference on System Sciences, pages 166–175, 2003.
15. M. Poulymenopoulou and G. Vassilacopoulos. A Web-based Workflow System for Emergency Healthcare. InProceedings of the MIE 2002 conference, 2002.
16. S. Quaglini, M. Stefanelli, G. Lanzola, V. Caporusso, and S. Panzarasa. Flexible Guideline-based Patient Careflow Systems. Artificial Intelligence in Medicine, 22(1):65–80, 2001.
17. N.C. Russell, A.H.M. ter Hofstede, and W.M.P. van der Aalst. newYAWL: specifying a workflow reference language using coloured petri nets. In K. Jensen, editor, Proceedings of the Eight Workshop on the Practical Use of Coloured Petri Nets and CPN Tools (CPN 2006), volume 584 ofDAIMI, pages 107–126, Aarhus, Denmark, October 2007. University of Aarhus.
18. G. Russello, C. Dong, and N. Dulay. Consent-Based Workflows for Healthcare Management. In Proceedings of 2008 IEEE Workshop on Policies for Distributed Systems and Networks (Policy 08), pages 153–161, Palisades, NY, US, 2008.
19. X. Song, B. Hwong, G. Matos, and A. Rudorfer. Understanding and classifying requirements for computer-aided healthcare workflows. InCOMPSAC (1), pages 137–144. IEEE Computer Society, 2007.
20. M. Stefanelli. Knowledge and Process Management in Health Care Organizations.Methods Inf Med, 43:525–535, 2004.
21. S.W. Tu, M.A. Musen, R. Shankar, J. Campbell, K. Hrabak, J. McClay, S.M. Huff, R. McClure, C. Parker, and R. Rocha. Modeling Guidelines for Integration into Clinical Workflow. Studies in Health Technology and Informatics, 107:174–178, 2005.
22. I. Vermeulen, H. La Poutr´e, S.M. Bohte, S.G. Elkhuizen, and P.J. Bakker. Decentralized Online Scheduling of Combination-Appointments in Hospitals. InProceedings of ICAPS-2008, the International Conference on Automated Planning and Scheduling, Sydney, Australia, 2008.
AAAI Press.