Designing a software interface for healthcare requires thorough domain knowledge, and effective user research and benchmark analysis. This thesis examines the requirements for an improved patient information management system for emergency medicine and describes the iterative process of designing and evaluating the system. I conducted observational study of Emergency Department (ED) physicians’ workflow and information needs, from which I derived a set of functional requirements, created scenarios, performed hierarchical task analysis, and developed a preliminary user model for the patient information management system. Based on these, I developed an interface prototype and evaluated the design with a sample of ED physicians. I review the user testing and design iterations carried out and report on the design improvements made based on the user feedback.

This thesis investigates the associations between crash outcomes, the existence and type of driver distraction as well as driver’s age. The crash outcomes considered in this thesis consist of the type of crash as well as the injury severity sustained by occupants involved in the crash. An ordered logit model was built to predict the likelihood of severe injuries and a multinomial model was developed to predict the likelihood that a driver will be involved in one of three common crash types: singular, angular, and rearend. In these models, various factors (e.g., weather, driver’s gender, and speeding) have been statistically controlled for, but the main focus was on the interaction of driver’s age and distraction type. The findings of this thesis have implications for policy making and prioritizing capabilities of distraction-related safety systems.

Secondary infusions are a common and convenient method to administer intermittent infusions unattended through a single IV access using infusion pump technology. Previous studies have indicated that clinicians have a high frequency of committing operation errors while administering secondary infusions, which can cause patient harm. The purpose of this study was to evaluate the safety of secondary infusion practice by identifying and analyzing potential failure modes when delivering secondary infusions on five different smart infusion pumps. Healthcare Failure Mode and Effect Analysis (HFMEA) was used to prioritize potential failure modes that are considered high-risk for each pump. Results showed that four of the five pumps were not able to mitigate physical set-up errors. As well, each pump contributed differently to programming errors due to difference in interface design. Recommendations from this study focused on outlining desired infusion pump features and mitigation strategies to help alleviate high-risk secondary infusion failure modes.

Research suggests that interruptions are ubiquitous in healthcare settings and have a negative impact on patient safety. However, there is a lack of solutions to reduce harm arising from interruptions. Therefore, this research aimed to design and test the effectiveness of interventions to mitigate the effects of interruptions during medication administration. A three-phased study was conducted. First, direct observation was conducted to quantify the state of interruptions in an ambulatory unit where nurses routinely administered high-risk medications. Secondly, a user-centred approach was used to design interventions targeting errors arising from these interruptions. Finally, the effectiveness of these interventions was evaluated through a high-fidelity simulation experiment. Results showed that medication administration error rates decreased significantly on 4 of 7 measures with the use of interventions, compared to the control condition. Results of this work will help guide the implementation of interventions in nursing environments to reduce medication errors caused by interruptions.

Designing a software interface for healthcare requires thorough domain knowledge, and effective user research and benchmark analysis. This thesis examines the requirements for an improved patient information management system for emergency medicine and describes the iterative process of designing and evaluating the system. I conducted observational study of Emergency Department (ED) physicians’ workflow and information needs, from which I derived a set of functional requirements, created scenarios, performed hierarchical task analysis, and developed a preliminary user model for the patient information management system. Based on these, I developed an interface prototype and evaluated the design with a sample of ED physicians. I review the user testing and design iterations carried out and report on the design improvements made based on the user feedback.

This thesis investigates the associations between crash outcomes, the existence and type of driver distraction as well as driver’s age. The crash outcomes considered in this thesis consist of the type of crash as well as the injury severity sustained by occupants involved in the crash. An ordered logit model was built to predict the likelihood of severe injuries and a multinomial model was developed to predict the likelihood that a driver will be involved in one of three common crash types: singular, angular, and rearend. In these models, various factors (e.g., weather, driver’s gender, and speeding) have been statistically controlled for, but the main focus was on the interaction of driver’s age and distraction type. The findings of this thesis have implications for policy making and prioritizing capabilities of distraction-related safety systems.

Secondary infusions are a common and convenient method to administer intermittent infusions unattended through a single IV access using infusion pump technology. Previous studies have indicated that clinicians have a high frequency of committing operation errors while administering secondary infusions, which can cause patient harm. The purpose of this study was to evaluate the safety of secondary infusion practice by identifying and analyzing potential failure modes when delivering secondary infusions on five different smart infusion pumps. Healthcare Failure Mode and Effect Analysis (HFMEA) was used to prioritize potential failure modes that are considered high-risk for each pump. Results showed that four of the five pumps were not able to mitigate physical set-up errors. As well, each pump contributed differently to programming errors due to difference in interface design. Recommendations from this study focused on outlining desired infusion pump features and mitigation strategies to help alleviate high-risk secondary infusion failure modes.

Failure to recognize changes in a patient’s clinical condition is a barrier to the effectiveness of CCRT outreach programs. The development of a vital signs capture and decision system could alert care providers and CCRTs when a patient’s clinical condition deteriorates. However, point-of-care vital signs capture and documentation is often problematic in clinical practice. Ethnographic research was conducted to understand the difficulties of replacing pen and paper charts and barriers to electronic nursing documentation systems. Analysis of workflows directed the design of two solutions; 1) Apple iPhone facilitated manual vital signs entry, 2) Motorola MC55 enabled automatic data capturing from physiological monitors. Nurses participated in high-fidelity usability testing, comparing the traditional method of paper documentation with the two electronic solutions. As a result of user-centered design process, both solutions were comparable to the efficiency of paper methods, were found acceptable to nurses, and could be successfully incorporated into current workflows.

Haptic technology enables computer users to touch and/or manipulate virtual objects in virtual environments (VEs). Similar to other human-in-the-loop applications, haptic applications require interactions between humans and computers. Thus, human-factors studies are required to recognize the limitations and capabilities of the user. This thesis establishes human-factors criteria to improve various haptic applications such as perception-based haptic compression techniques and haptic-enabled computer-aided design (CAD). Today, data compression plays a significant role in the transmission of haptic information since the efficient use of the available bandwidth is a concern. Most lossy haptic compression techniques rely on the limitations of human force perception, and this is used in the design of perception-based haptic compression techniques. Researchers have studied force perception when a user is in static interaction with a stationary object. This thesis focuses on cases where the human user and the object are in relative motion. The limitations of force perception are quantified using psychophysical methods, and the effects of several factors, including user hand velocity and sensory adaptation, are investigated. The results indicate that fewer haptic details need to be calculated or transmitted when the user's hand is in motion. In traditional CAD systems, users usually design virtual prototypes using a mouse via their vision system only, and it is difficult to design curved surfaces due to the number, shape, and position of the curves. Adding haptics to CAD systems enables users to explore and manipulate virtual objects using the sense of touch. In addition, human performance is important in CAD environments. To maintain the accuracy, active haptic manipulation of the user response can be incorporated in CAD applications. This thesis investigates the effect of forces on the accuracy of movement in VEs. The results indicate that factors such as the base force intensity and force increment/decrement can be incorporated in the control of users' movements in VEs. In other words, we can pull/push the users' hands by increasing/decreasing the force without the users being aware of it.

Wayfinding is a complex skill and the lack of tools supporting the specific sub-types of navigation hinders performance in large-scale virtual environments and consequently can slow the adoption of virtual technology for training. The VTrail System is designed to support virtual training by providing trainers (trailblazers) with the ability to create trails to guide users (trail followers) during training simulations. Without an effective interface to assist with creating trails, the task of trailblazing remains difficult. The objective of this research was to design a default interface for the VTrail System that adheres to the basic human factors engineering guidelines of simplicity, universality, and that does not interfere with primary task performance. Two studies (trailblazing, trail following), with a total of four experiments, were performed to evaluate and modify the proposed interfaces. The first experiments in each study determined that the proposed default interfaces are simple enough to use so as to not interfere with primary task performance. The second set of experiments found that, aside from the interface components included in the default interface, novice trailblazers and trail followers did not make use of any additional wayfinding aids when users were provided with the ability to create a custom interface. Secondary benefits included; the development of a novel approach for measuring spatial knowledge acquisition (called the SKAT), a set of criteria for qualitative analysis of trail quality in the form of the Trail Quality Questionnaire (referred to as TQQ), and improved understanding of the role individual differences, such as gender and spatial ability, in wayfinding performance. The high correlation between spatial ability score and performance on the SKAT suggests that the test provides a valid means of measuring spatial knowledge acquisition in a virtual environment. A measurable difference in the trail quality between males and females indicates that the TQQ can distinguish between trails of variable quality. Finally, there are measurable gender performance differences, despite similar levels in spatial ability between the genders. With the proposed interface designs the VTrail is closer to being ready to be incorporated as a support tool into virtual training programs. In addition, the designs for the VTrail System can be adapted for other platforms to support trailblazing in a range of applications, from use in military operations to providing an enhanced tourism experience. This research also serves as a starting point for future research projects on topics ranging from improving the design of the SKAT measure to understanding the effect of expertise on trailblazing performance.