A Model for Workarounds Management Lifecycle

Figure 1 depicts the proposed workaround management lifecycle model, which attempts to articulate workarounds management as an approach to continuous improvement of healthcare information systems. The logic of the workaround management lifecycle model draws upon the lens of structuration theory and adaptive structuration. Adaptive structuration theory (DeSanctis and Poole 1994), originated from structuration theory(Giddens 1984), recognizes the mutual influence of user actions and system structure in organizations. System structure refers to the structural features and spirit of a system that serve as templates for accomplishing system-related tasks (DeSanctis and Poole 1994). During the system design and implementation phases, designers and/or organizational management often create the expected system structure, which influence the actions performed by system users. However, system users’ actions are not

The workaround management lifecycle model is developed based on research and observation of HITs. We only consider workarounds that are induced by HITs. Other types of workaround (such as change of personnel, policy, etc) will not be considered in this paper.

Workarounds are a post-implementation phenomenon widespread in healthcare organizations.

Alongside the increasing computerization of healthcare services, there is a parallel increase in HIT-related workarounds (Murray, 2001). Workaround management is a particularly important issue in healthcare context as healthcare professionals are extremely sensitive to potential negative consequences (such as risks) that are caused by workarounds. The workaround

91

management lifecycle model asserts that HIT development is an iterative process in which the system should be continuously improved based on workarounds that happens in system usage.

IT implementation leads to three types of workaround triggers. Based on social decisions, workarounds can be handled by either managerial control or material handling mechanisms (workflow-based or case-based). Material handling mechanisms lead to technical enhancement to the system. Next, we will explain the major stages (numbered arrows in Figure 1) in the workaround management lifecycle.

While HITs aim to standardizing and automating healthcare routines via formal workflows, they may cause many problems which also enable workarounds. After a new system is implemented, users may have difficulties in using the system for reasons such as inefficient process design, poor system interface, and inadequate user training (Halbesleben et al., 2008;

Vogelsmeier et al., 2008). For example, if a system requires verification steps that are considered unnecessary by users, people want to save time and skip the verification tasks. Workarounds may be also caused by exceptions. Exceptions are usually caused by unexpected events (e.g. special requirements or emergent cases) that are not anticipated by the design of the system workflow.

Given the ad hoc nature of exceptional events, this type of workaround does not happen frequently in routines. System deficiencies may also lead to workarounds. For example, people have to go with manual process when the system is down or has runtime errors. For the path from IT usage to work around triggers, we argue that

Arrow 1: The use of newly implemented HIT functions can lead to three types of workaround triggers: user motivation, exceptional cases and system deficiencies.

After workarounds occur, managerial decisions should be made based on the power structure of participants and management policies. Power structure means the power distribution between a workaround actor and the affected actors. For example, physicians inherently possess higher power over nurses in hospitals. Hence, hospital management may tolerant physicians’

workarounds even when these workarounds burden nurses’ work. The other way may not work.

Management policies mean organization’s regulations and rules. For instance, in hospitals, the management can rely on certain rules to decide the workarounds handling methods: 1) whether the reported workarounds are executed by large amount of users and can benefit the healthcare process; 2) whether certain exceptional cases have the indications that could potentially trigger mass breakouts, e.g., SARS. Thus, we argue that

Arrow 2: The workaround handling mechanism will be decided by managerial decisions that are made based on power structure and management policies.

Workaround can be managed with either technical (material perspective) or non-technical (social perspective) approaches. The social control mainly uses managerial intervention that does not require any changes to the system. Sometimes, managers may decide to do nothing (e.g., be tolerant about physicians’ system delegation). Most of the time, managerial decision for social control will lead to creation of new management policies (e.g., provide monetary incentives to users who do not perform workarounds). Material handling mechanism tries to improve the system that can better manage the workarounds. If a workaround is not appropriately managed, the potential risks may rise. The objective of workaround management is to avoid

92

workaroundswith negative impact and enable workarounds with positive impact.

Arrow 3:Social mechanisms that does not involve technical enhancement can be used for workaround handling.

Depending on the triggers of workarounds, two major material mechanisms can be used to handle workarounds, cased-based handling and workflow-based handling.Workflow-based handling relies on contemporary workflow management system (WfMS) that enables the modeling, execution, and monitoring of business processes(Georgakopoulos et al. 1995). When an actor (nurses or physicians) is working on a particular process step (i.e., activity), typically, in WfMS-based applications, only data needed for executing this activity is visible to the actor, but no other workflow data. This handling paradigm is typically strong in control but weak in flexibility. Case-based handling is different from workflow-based handling(Madhusudan et al.

2004; Mutschler et al. 2008;Van der Aalst et al. 2005). The central concept behind case-based handling is that every individual case and its process are stored in the system. Usually, case-based handling presents all data about a case at any time to all the users. Case-case-based handling orchestrates the execution of activities based on processes of similar cases. This handling paradigm is strong in flexibility and weak in control. Three major steps of case-based handling and workflow-based handling are listed in Figure 1. (“Direct workaround tasks” are defined as the tasks that have workaround; “tasks affected by workaround” are tasks that don’t have workaround but the execution of these tasks are changed by workarounds.) The steps for both handling mechanisms are conducted sequentially and allow retrospective backtracking. Due to page limit, the details of the workaround handling steps are omitted.

Arrow 4&5. Two material mechanisms can be used to handle workarounds, depending on the triggers of workarounds: cased-based handling and workflow-based handling.

 User motivation triggered workarounds require workflow-based handling

 System error triggered workarounds require case-based handling

 Exceptional case triggered workarounds can be handled in either way, depending on the frequency of exceptional cases and expected trend of frequency in the future.

User motivation caused workarounds requires workflow-based handling. User motivation usually does NOT change for the same workflow. If the healthcare system stays the same, user motivation triggered workarounds are very likely to happen again. The changes to workflow includes the ways of task execution (e.g., new system user interface), task assignment (inclusion of duplicate users), and/or process logic(e.g., sequence of taking a medication order). System error caused workaround requires case-based handling. System errors happen rarely and each occurrence can be treated as a case. Exceptional case caused workaround may be handled in either way, depending on the frequency of exceptional cases and expected trend of frequency in the future. For exceptional cases that happen more and more, workflow-based handling should be used. Workflow based handling will need to change the configuration of system and redesign the workflow. Case-based handling will need to add the case and rules to the repository. Both of these handling methods require the development and implementation of new IT artifacts. With several rounds of workaround management, the system can be continuously improved.

93

Arrow 6. Material workaround handling mechanisms lead to technical enhancement and implementation of new IT artifacts.

4. Conclusions

In this paper, we propose a model of workarounds management lifecycle based on the theories of organizational routines and business process management. We model workarounds management as an approach to continuous improvement of healthcare information systems. It sets forth a theoretical framework for further empirical studies to investigate workaround management. In this study, we incorporate managerial and technical principals into the workarounds management lifecycle. The workaround lifecycle model advances the IS design and post-adoption literature by focusing on the deviated IS use behaviors and their contribution to system improvement. Future work includes developing falsifiable hypotheses based on the workaround management lifecycle model and collecting empirical data to validate framework.

References are available upon request.

94