Theresa K. Guarrera

Emergency Department Status Boards: A Case Study in Information Systems Transition:

Patient status boards play an important role in coordinating and communicating about patient care in hospital emergency departments (EDs). Status boards are transitioning from dry-erase whiteboards to electronic systems. Although electronic systems may preserve some surface features of the manual artifacts, important affordances of the manual technology are not always maintained. We compared information on manual and electronic status boards in an ED. Photographs of the manual board and screen shots of the electronic system were obtained before and after a hospital transitioned between systems. Displayed information as well as detailed content regarding patient chief complaints, clinical plans, and dispositions were coded and analyzed to understand the type and function of information present, as well as the use of features such color. Results indicated that although categories of information found were similar, the frequency with which some types of information appeared on the two system displays was substantially different. In particular, information used to coordinate aspects of patient treatment was more frequently found in the manual system. Results suggest that in the design of new information technologies, simply matching the format or information fields available on an existing system may not be sufficient to sustain current work practices or to prevent unanticipated shifts in use.

Cognitive Artifacts in Transition: An Analysis of Information Content Changes between Manual and Electronic Patient Tracking Systems

Todays emergency departments (ED) could not function without a patient tracking system of some kind, manual or electronic. Manual patient tracking systems such as “whiteboards” are large, dry erase, manually updated status boards used as information tracking devices in most EDs. Although it is expected that manual whiteboard systems will completely transition to electronic patient tracking systems with increasing availability of technological solutions, it is not clear if these technological solutions will sufficiently address the information tracking needs of providers. This study documents the changes in the use of a manual whiteboard versus an electronic patient tracking system in an ED to compare types of information and meanings of the represented information. Results show that both systems were used to represent information serving a variety of functional roles. In addition, an analysis of patient chief complaint entries indicates that manual whiteboards are used more dynamically than electronic systems. Differences in functional uses of the systems and the consequences of these changes are discussed.

Development of a simulation environment to study emergency department information technology

Introduction: This article presents a simulation architecture for a patient tracking system simulator to study caregiver performance in emergency departments (EDs). The architecture integrates discrete event simulation modeling with clinical patient information. Evaluation components for electronic patient tracking system displays are also described. Methods: A simulation of an ED electronic whiteboard was developed to study situation awareness metrics. Dynamic process data from an actual ED was used to generate simulation parameters including patient arrivals at various hours, distribution of severities, times required to treat the ED patients, and ancillary turnaround times (laboratory and radiology). A team of industrial engineers and ED physicians contributed demographic and clinical information for simulator patients. ED simulation parameters were combined with clinical information resulting in an event timeline database. Event timelines were used to populate a front-end patient-tracking system display simulation. Results: The resulting patient-tracking system display simulation consists of underlying software, desktop and large-screen displays, a phone call/pager system, and typical tasks that enhance the realism of the simulation experience. The system can evaluate the impact of display parameters and ED operations on user performance. Conclusions: Modular design of the patient-tracking system display simulation helps adaptation for different studies to support various interface features and interaction types. The methodology described in this work exploits the benefits of discrete event simulation to iteratively design and test technologies such as electronic patient tracking systems and allows assessment of human performance measures.

The impact of interoperability of electronic health records on ambulatory physician practices: a discrete-event simulation study

BACKGROUND The effect of health information technology (HIT) on efficiency and workload among clinical and nonclinical staff has been debated, with conflicting evidence about whether electronic health records (EHRs) increase or decrease effort. None of this paper to date, however, examines the effect of interoperability quantitatively using discrete event simulation techniques. OBJECTIVE To estimate the impact of EHR systems with various levels of interoperability on day-to-day tasks and operations of ambulatory physician offices. METHODS Interviews and observations were used to collect workflow data from 12 adult primary and specialty practices. A discrete event simulation model was constructed to represent patient flows and clinical and administrative tasks of physicians and staff members. RESULTS High levels of EHR interoperability were associated with reduced time spent by providers on four tasks: preparing lab reports, requesting lab orders, prescribing medications, and writing referrals. The implementation of an EHR was associated with less time spent by administrators but more time spent by physicians, compared with time spent at paper-based practices. In addition, the presence of EHRs and of interoperability did not significantly affect the time usage of registered nurses or the total visit time and waiting time of patients. CONCLUSION This paper suggests that the impact of using HIT on clinical and nonclinical staff work efficiency varies, however, overall it appears to improve time efficiency more for administrators than for physicians and nurses.

Interface Design Characteristics of a Popular Emergency Department Information System

Integration of information technology (IT) in the medical field has increased with the patient safety movement of the past decade, but without an optimization of user interface (UI) design. In particular, health IT such as an emergency department information system (EDIS) is becoming a common tool in EDs across the US. While there are published standards for the design of medical devices requiring usability testing, health IT systems do not have the same requirements. Although the EDIS was created to support work previously accomplished through a dry erase board, its impact on patient safety has yet to be studied. This study explores usability issues in an existing EDIS application and their potential impact on patient safety.

Usability evaluation and assessment of a novel emergency department IT system developed using a cognitive systems engineering approach

This paper presents the results of a usability evaluation conducted of an electronic Emergency Department information system (EDIS) prototype that was designed using a cognitive system engineering (CSE) approach. Participants were asked to complete tasks using the EDIS prototype, while thinking aloud about their interactions with the displays. Participants also completed subjective assessments of the system that related to 1) cognitive performance support objectives, 2) usability, usefulness, and frequency of system use, and 3) qualitative feedback. Mean scores were calculated for cognitive performance support objectives as well as usability, usefulness, and frequency of use for each display. Mean scores for all cognitive performance support objectives were six or higher. The mean usability score was five or higher, and the mean usefulness score was six or higher. Displays that provided information on individual patient status had the highest scores. Results from this evaluation regarding positive ratings for the cognitive support objectives provide evidence that CSE design methods can be used to understand the goals and objectives of medical work domains.

Better Pairing of Providers and Tools Development of an Emergency Department Information System using Cognitive Engineering Approaches

Health IT systems are often designed without a sufficient understanding of the clinical activities they are intended to support; thus desired benefits in quality of care, safety, and efficiency may not accrue. We present the second part of a multi-phase study which utilizes cognitive systems engineering (CSE) methods to design and test novel user interfaces for a hospital emergency department (ED) information system. Using the work domain analysis results from phase one, display components were developed for a system which can better support work activities and information needs of ED clinicians and staff.

Engineering Better Health IT: Cognitive Systems Engineering of a Novel Emergency Department IT System

Often, health IT systems are designed without a sufficient understanding of the clinical activities they are intended to support; thus desired benefits in quality of care, safety, and efficiency may not accrue. We present the first part of a multi-phase study which utilizes cognitive systems engineering methods to design and test novel user interfaces for a hospital emergency department information system. A work domain analysis was performed and the resultant model was used to identify information requirements and display components for a system which can better support work activities of ED clinicians and staff.

Characterising the effect of interoperability on healthcare work: a novel framework

Successful design and implementation of interoperable health IT requires an understanding of specific technological capabilities of systems, as well as how these systems impact clinical workflow. Several existing frameworks classify interoperability levels, but none focus on the impact on clinical work, particularly at the task level. A synthesis of existing interoperability frameworks from select interoperability frameworks in the literature about both medical and non-medical systems is presented and a new, 7-level framework in order to characterise the effect of varying levels of interoperability on the users’ work based upon qualitative data collected in a field study of the use of health information exchange in 12 ambulatory practices is proposed. The lowest level describes paper-based tasks completed with no access to electronic information from other institutions; the highest level describes interoperable systems in which data elements from other institutions are integrated into the patients record seamlessly and in a computable format.

Evaluating the Creation and Interpretation of Causal Influence Models

Bayesian networks (BNs) are probabilistic models frequently used to capture domain knowledge for use in computational systems that can reason about states, causes, and effects. While BNs have many advantages, their complexity can hamper the process of knowledge elicitation and encoding. First, domain experts may not have expertise in artificial reasoning or probabilistic models, and that lack of understanding may complicate the elicitation of probabilities relevant to BN model structure. In addition, BNs require the definition of a priori, conditional probabilities: for complex models, this requires eliciting large numbers of complex probabilities. Multiple “canonical modeling” approaches, such as Causal Influence Models (CIMs), have been developed to address these complexities. However, little progress has been made towards human-in-the-loop evaluation of such approaches - specifically, their accessibility and usability, their related user interfaces, and how they enable a user to correctly create and interpret variables and probabilistic relationships. In this study, we evaluated the CIM approach (implemented in a software application) to determine the effect on user task performance. Results indicate that the model complexity has an adverse effect on performance when users are interpreting an existing model; that semantics of a model may impact performance; and that users were generally successful in creating new models of different situations.