!DOCTYPE html>
9. Intensive Care Unit Collaboration and Workflow in the Information Age
Keywords
Information technologyWorkflowCommunicationElectronic medical records9.1 Complexity of Collaboration and Teamwork in the ICU
Distributed cognition can be used to describe how a group of individuals may process shared cognitive tasks while taking into account the social and material setting [1, 2]. Compared to the cognition of an individual, knowledge management across a group is much more complex, with knowledge artifacts spanning various mediums, and interdependent processes occurring throughout the cycle of the day. In the case of an ICU team, members are beholden to their own administrative responsibilities and scope of practice. Yet, despite their disparate physical locations and short-term goals, team members share in the essential overall tasks of an ICU: triage, admission, diagnosis, treatment, monitoring, and discharge of multiple sick patients. Increasingly, HIT is the infrastructure through which knowledge artifacts are carried between members of a team and between shifts. Models of workflow in the ICU focusing on distributed cognition and knowledge management [3, 4] may become useful tools in the development of effective HIT. They may also be useful to identify potential breakpoints to reduce the possibility of human errors and communication failures [5].
The study of interruptions in critical care and their impact on workflow and collaboration has been especially active, with the intent of identifying opportunities for avoiding errors [6, 7]. A significant portion of ICU clinicians’ time are spent in communication [8] (29 communication events an hour in one study) [9]. Forty-two percent of communication events consist of an interruption [6, 9]. And in general, interruptions consist of communication events (through telephone or pagers) [10]. It is not disputed that the ICU working environment is highly interruptive and requires multitasking of its members [11–13]. Given that ICU work is heavily based on communication, collaboration, and rapid incorporation of new information, it is likely that interruptions are inescapable or even necessary and have suffered from pejorative labelling [14, 15]. Interruptions that should appropriately be targeted for reduction can be classified by the harm potential of the halted primary task (e.g., during procedures or medication retrieval/administration) [16–20]. HIT and communication tools should be thoughtfully adopted and implemented with an important goal of enhancing communication while reducing interruptions.
The recent decrease in resident working hours has increased the need for handoffs between clinicians and across disciplines in the care of patients. This transition in care also introduces a point in the workflow that can introduce opportunity for error. There is a need to understand the information needs of both the staff member on service as well as the one taking over care. HIT offers the ability to support communication between clinicians as part of the handoff process.
9.2 Impact of IT on Workflow and Communication
Hospitals and other medical facilities have been deploying information systems at a rapid rate. The use of HIT to improve medical and cost outcomes has exploded over the last 20 years. The speed and utilization will only increase as the implementation of electronic medical records and other clinical information systems have been tied to government reimbursement for medical services [21]. This is complicated by the fact that health information systems failures have been widely documented and reports identify the problem issues around social, technical, and organizational factors [22–24]. The reason that so many systems fail lies in the sociocultural fit of HIT systems. We do not understand how to deliver information to healthcare providers when they need it, in a format they can use, and in a way that fits transparently into their workflow and into the daily lives of patients.
The 2001 report “Crossing the Quality Chasm,” the Institute of Medicine (Committee on Quality of Health Care in America [25]) highlighted improved information systems as a means for achieving quality in healthcare. A major challenge has been to unravel the many effects the HIT has on work activities and processes [24, 26–28]. Socio-technical systems theory posits that HIT is a social system as well as a technical system and should be studied as such. There are a growing number of examples of this systems approach in health services research including the redesign of bar code medication administration in the Veterans Affairs Medical Centers [29], monitoring equipment in the operating room [30], and computerized clinical guidelines [31]. Influential work in the field has identified positive examples of HIT being used to support clinical workflow. These examples include simplification of data entry [32], change in perception of organizational relationships with other professionals [27], elimination of redundant processes [33], positive perception by physicians of the impact on practice [34], and the provision of health maintenance alerts regarding future patient care [35]. However, the study of HIT’s impact on workflow has also uncovered a number of issues presenting challenges to the cognitive work of clinicians. First, misalignment between workflow and system design has been well documented as a cause for HIT failure [22, 36, 37]. Secondly, HIT is often poorly automated [38, 39] because of a simplified understanding of cognitive tasks. Thirdly, HIT often introduces unanticipated workflow and safety concerns [22, 40].
9.2.1 Electronic Medical Records
HIT in the form of electronic medical records was introduced to solve a number of problems including data presentation and care documentation. With the ability to present data came the problem of data overload or overcompleteness of data presentation [41]. The inverse problem that also emerged was the “keyhole effect”—when only a very small sample of the data is displayed such that it skews the interpretation of the full data set [42]. Documenting care with electronic medical records was designed to replace the paper record and with it the limitations of handwritten notes that are only available in one location by one user. However, the usefulness of electronic notes is diminished by templated notes, and not helped by the phenomenon of cut-and-paste that arises as a function of efficiency. Thus informal paper artifacts persist that clinicians find integral to their patients’ care; cognitive work is supported with memory prompts on paper known as “shadow” or “ghost” charts which are largely discarded at the end of a shift [43, 44]. While HIT offers users the opportunity to integrate data display and real-time information sharing, the formalized requirements of documentation dissuade users from transitioning their shadow charts online.
Care must be taken during design and implementation of the electronic medical record to support actual ICU workflow. The physical setup of the interaction with technology (as compared with paper) impedes the ability of the attending physician to lead rounds or even other team members from contributing naturally to discussion [45]. The electronic medical record in the ICU, much like in the outpatient setting, alters the social interaction between team members; however, users can modify these changes to their interaction if given the opportunity to make technical changes to the system [46].
9.2.2 Usability and Usefulness of HIT
The usability and usefulness of HIT is key for success in implementation. Nielsen and colleagues have established key characteristics that make user interfaces usable and useful [47, 48]. Providing sufficient feedback, visibility of system status and supporting memory are key interaction heuristics that need to be optimized in the design of HIT. Providing feedback to users is crucial for them to understand the functionality of the system. In the time-sensitive environment of the ICU, it is necessary for system functions to be clearly identifiable to not only limit mistakes but also direct them to the most efficient method to complete tasks. A similar heuristic is to maintain the visibility of system status at all times. It is necessary for clinicians to identify what data is available and what is unknown.
9.2.3 Communication
Communication has been defined as the transfer of information between two entities with a goal. Clinicians have been observed to prefer synchronous communication (texting, conversations face to face, conversations by phone, paging), resulting in frequent multitasking [9, 49]. Synchronous methods of communication allow greater mobility and efficiency but result in multitasking, are interruptive in nature, and possibly have a negative impact on interprofessional relationships [50, 51]. Physicians and nurses are both the culprits and the receivers of these interruptions. While interruptions cause a break in concentration for clinicians, they also offer the opportunity for a break in the cognitive flow that may offer an alternative view of situations and events that can get clouded when users fixate on a solution to a problem too soon [52, 53]. There is room for more quantitative evaluations of communication tools and HIT influence on effective communication [54].
9.3 Conclusion
There is increasingly a consideration of human factors and human-computer interaction, as it is recognized that collaboration and workflow are yielding to technology rather than the ideal reverse. A few points are clear: we must understand collaboration and teamwork systems locally in order to optimize communication and workflow. By recognizing areas for potential missteps with HIT, we can design and implement better systems, make standards for orienting new staff, and make meaningful steps to improve patient safety.
