We are examining how the design features and characteristics of healthcare chatbots should be employed for effective and efficient health education and patient learning. These chatbots based on Large Language Models (LMMs) are increasingly being used for different internet-based tasks, yet the complex interactions with these automation services needs to be studied and evaluated.
We are examining how the processes and the design of spaces and tools impact the care delivered to pediatric patients with mental and behavioral health issues.
We are examining the way that the anesthesia workstations within operating rooms are designed and arranged to support the work within anesthesia care. This has included task analysis, workspace assessments, and examining the movement of supplies and materials around in the workspace during a surgical case.
Despite decades of research and improvements in safety, the risk of patient harm in the practice of anesthesia remains. The failure of efforts to effectively address patient harm represents a lack of consideration of the cognitive processes and diverse work practices employed in anesthesia delivery. We are using vignettes (hypothetical scenarios) as interview tools to elucidate how anesthesia providers may diverge from one another in their decision-making processes. We have identified variability in anesthesia providers’ decision-making strategies which must be considered in the development of future efforts to address patient harm, to support anesthesia providers’ critical resilience.
The use of chatbots is growing in telehealth applications, yet few studies have examined how design characteristics influence the effectiveness of chatbots in the healthcare domain. We are examining how chatbot characteristics including language (i.e., reading level), personas (i.e., doctor, nurse, or nursing student), and the use of avatars influence the effectiveness, usability, and trust in healthcare chatbots.
In-air gestures are a promising input modality as they are expressive, easy to use, quick to use, and natural for users. They may be particularly useful for domains like anesthesia, as the use of gestures could bypass the bacterial transmission that may occur when touching devices. However, gestural systems can be challenging to develop as gesture choice is dependent on context, and intuitive gestures may not be constant for every user (for example, an expert may find different gestures intuitive when compared to a novice). We are studying how expertise and exposure influence gestural behaviors, and whether we can develop Bayesian statistical models that can accurately predict how users would choose intuitive gestures.
Patient portals are websites or mobile applications that are designed to help patients access their Electronic Health Records (EHR), health summaries, pay bills, schedule appointments and, in some cases, interact with care providers. The use of patient portals has been associated with generating positive healthcare outcomes in recent studies. We are interested in examining the characteristics of EHR patient portal users, how they use patient portals, how EHR patient portal users access health information online, and how these factors impact users’ trust in their patient portals.
Individuals with chronic diseases are a unique user population in terms of their potential use of online health information in the self-management of their health. Crohn’s disease and ulcerative colitis are collectively referred to as inflammatory bowel disease (IBD), a chronic condition that affects the intestines, colon, and bowel. It is a complex, incurable disease that can result in long-term disability or mortality, and its highest incidence occurs in younger adults. A recent study suggested that the incidence of IBD has seen a dramatic increase to over 0.3% in North America and many European countries, and the incidence of IBD is expected to increase. We are interested in examining which factors may influence individuals with IBD to search the internet for healthcare-related information, what information they search for, and how to better support individual’s in their self-care of their IBD.
Discharge instructions (DI) are documents and information provided to patients when they complete a procedure or exam and need to follow up information or self-care tasks. The design of these information tools is a critical aspect of preventing complications and rehospitalizations. We are examining the design characteristics of DI using eye trackers to study how individuals engage with and read DIs. We are working to provide design recommendations and strategies for efficient and effective discharge instructions.
We have examined how intuitive gestures could be used in the driving domain for specific in-vehicle infotainment systems. It is important these gestures be intuitive so that they map naturally to users’ expectations for usable and safe interactions with the infotainment systems.