Arnold J. Moleman

Process mining in healthcare: data challenges when answering frequently posed questions

In hospitals, huge amounts of data are recorded concerning the diagnosis and treatments of patients. Process mining can exploit such data and provide an accurate view on healthcare processes and show how they are really executed. In this paper, we describe the different types of event data found in current Hospital Information Systems (HISs). Based on this classification of available data, open problems and challenges are discussed that need to be solved in order to increase the uptake of process mining in healthcare.

From Requirements via Colored Workflow Nets to an Implementation in Several Workflow Systems

Hospitals and other healthcare organizations need to support complex and dynamic workflows. Moreover, these processes typically invoke a number of medical disciplines. This makes it important to avoid the typical disconnect between requirements and the actual implementation of the system. In this paper we apply a development approach where an Executable Use Case (EUC) and a Colored Workflow Net (CWN) are used to close the gap between a given requirements specification and the realization of these requirements based on workflow technology. In order to do so, we describe a large case study where the diagnostic process of the gynecological oncology care process of the Academic Medical Center (AMC) hospital is used as a candidate process. The process consists of hundreds of activities. These have been modeled and analyzed using an EUC and a CWN. Moreover, based on the CWN, the process has been implemented using four different workflow systems. In this way, we demonstrate the general application of the approach and its applicability to distinct technology systems.

Process-Aware Information System Development for the Healthcare Domain - Consistency, Reliability, and Effectiveness

Optimal support for complex healthcare processes cannot be provided by a single out-of-the-box Process-Aware Information System and necessitates the construction of customized applications based on these systems. In order to allow for the seamless integration of the new technology into the existing operational processes of a healthcare organization, ensuring the correct operation and reliability of the developed system are of the utmost importance. This paper proposes an approach in which the same model is used for specifying, developing, testing and validating the operational performance of a new system. The benefits of using the same model for different purposes are decreased potential for loss of user requirements and increased confidence in reliability and correct operation of the resultant system before its deployment. This approach has been applied to a schedule-based workflow system developed for the AMC hospital in Amsterdam.

Simulation to Analyze the Impact of a Schedule-aware Workflow Management System

Today’s workflow management systems (WfMSs) offer workitems to users through specific work-lists. Users select the workitems they will perform without having a schedule in mind. However, in many environments work needs to be scheduled and performed at particular times. For example, in hospitals many workitems are linked to appointments, e.g., a doctor cannot perform surgery without reserving an operating theater and making sure that the patient is present and ready. One of the problems when applying workflow technology in such domains is the lack of calendar-based scheduling support. In collaboration with the Academic Medical Center (AMC), a large hospital in the Netherlands, we developed a schedule-aware WfMS that supports the seamless integration of unscheduled (flow) and scheduled (schedule) tasks. However, before deployment of the resultant system in the hospital, a seamless integration with AMC’s running healthcare processes needs to be guaranteed. Therefore, for a large and complex healthcare process, we apply computer simulation to validate and to investigate, for different configurations of the system, the operational performance for a selected healthcare process when supported by the schedule-aware workflow management system. One of the important characteristics of our approach is the tight coupling between the simulation model and the actual implemented system. While performing simulation experiments, parts of the system may be simulated using CPN Tools while connected to the actual system components. Our simulation experiments demonstrate that the developed schedule-aware WfMS can be safely applied in the AMC hospital.

Lightweight interacting patient treatment processes

Processes concerning the diagnosis and treatment of patients cannot be straightjacketed into traditional production-like workflows. They can be best characterized as weakly-connected interacting light-weight workflows where tasks reside at different levels of granularity, and for each individual patient a doctor proceeds in a step-by-step way deciding what next steps be taken. Classical workflow notations fall short in supporting these patient processes as they have been designed to support monolithic processes. Classical notations (WF-nets (work flow nets), BPMN (Business Process Model and Notation), EPCs (Electronic Prescriptions for Controlled Substances), etc.) assume that a workflow process can be modeled by specifying the life-cycle of a single case in isolation. To address these problems, the authors present an extension of the Proclets framework which allows for dividing complex entangled processes into simple autonomous fragments. Additionally, increased emphasis is placed on interaction related aspects such that fragment instances for individual patients can cooperate in any desired way. The authors describe an implementation of the Proclets framework. Proclets have been added to the open-source Workflow Management System YAWL to better support inter-workflow support functionalities.

Model-Based Development and Testing of Process-Aware Information Systems

The development of Process-Aware Information Systems (PAISs) heavily relies on the notion of software configuration rather than classical software development approaches based on coding. To date, there is no accepted development strategy for the configuration of PAIS technology and construction of specific applications added to these systems tends to proceed on an ad-hoc basis rather than following a prescribed approach. Similar, there has been minimal consideration for testing and validating the capabilities of the PAISs. This paper proposes a domain-oriented system testing approach where a conceptual model (expressed in terms of a Colored Petri Net) is used for specification, simulation, and testing of the developed PAIS. This approach has been applied to a schedule-based workflow system developed for the AMC hospital in Amsterdam. One of the important characteristics of our approach is the tight coupling between the conceptual model and the actual implemented system. While testing and validating, parts of the system may be simulated using CPN Tools while connected to the actual system components.