From ELIZA to Alexa, Conversational Agents (CAs) have been deliberately designed to elicit or project empathy. Although empathy can help technology better serve human needs, it can also be deceptive and potentially exploitative. In this work, we characterize empathy in interactions with CAs, highlighting the importance of distinguishing evocations of empathy between two humans from ones between a human and a CA. To this end, we systematically prompt CAs backed by large language models (LLMs) to display empathy while conversing with, or about, 65 distinct human identities, and also compare how different LLMs display or model empathy. We find that CAs make value judgments about certain identities, and can be encouraging of identities related to harmful ideologies (e.g., Nazism and xenophobia). Moreover, a computational approach to understanding empathy reveals that despite their ability to display empathy, CAs do poorly when interpreting and exploring a user's experience, contrasting with their human counterparts.
doi.org/10.1145/3613904.3642336
Today's AI systems for medical decision support often succeed on benchmark datasets in research papers but fail in real-world deployment. This work focuses on the decision making of sepsis, an acute life-threatening systematic infection that requires an early diagnosis with high uncertainty from the clinician. Our aim is to explore the design requirements for AI systems that can support clinical experts in making better decisions for the early diagnosis of sepsis. The study begins with a formative study investigating why clinical experts abandon an existing AI-powered Sepsis predictive module in their electrical health record (EHR) system. We argue that a human-centered AI system needs to support human experts in the intermediate stages of a medical decision-making process (e.g., generating hypotheses or gathering data), instead of focusing only on the final decision. Therefore, we build SepsisLab based on a state-of-the-art AI algorithm and extend it to predict the future projection of sepsis development, visualize the prediction uncertainty, and propose actionable suggestions (i.e., which additional laboratory tests can be collected) to reduce such uncertainty. Through heuristic evaluation with six clinicians using our prototype system, we demonstrate that \system enables a promising human-AI collaboration paradigm for the future of AI-assisted sepsis diagnosis and other high-stakes medical decision making.
doi.org/10.1145/3613904.3642343
We introduce a multi-step reasoning framework using prompt-based LLMs to examine the relationship between social media language patterns and trends in national health outcomes. Grounded in fuzzy-trace theory, which emphasizes the importance of “gists” of causal coherence in effective health communication, we introduce Role-Based Incremental Coaching (RBIC), a prompt-based LLM framework, to identify gists at-scale. Using RBIC, we systematically extract gists from subreddit discussions opposing COVID-19 health measures (Study 1). We then track how these gists evolve across key events (Study 2) and assess their influence on online engagement (Study 3). Finally, we investigate how the volume of gists is associated with national health trends like vaccine uptake and hospitalizations (Study 4). Our work is the first to empirically link social media linguistic patterns to real-world public health trends, highlighting the potential of prompt-based LLMs in identifying critical online discussion patterns that can form the basis of public health communication strategies.
doi.org/10.1145/3613904.3642117
Recent advances in AI combine large language models (LLMs) with vision encoders that bring forward unprecedented technical capabilities to leverage for a wide range of healthcare applications. Focusing on the domain of radiology, vision-language models (VLMs) achieve good performance results for tasks such as generating radiology findings based on a patient's medical image, or answering visual questions (e.g., ``Where are the nodules in this chest X-ray?''). However, the clinical utility of potential applications of these capabilities is currently underexplored. We engaged in an iterative, multidisciplinary design process to envision clinically relevant VLM interactions, and co-designed four VLM use concepts: Draft Report Generation, Augmented Report Review, Visual Search and Querying, and Patient Imaging History Highlights. We studied these concepts with 13 radiologists and clinicians who assessed the VLM concepts as valuable, yet articulated many design considerations. Reflecting on our findings, we discuss implications for integrating VLM capabilities in radiology, and for healthcare AI more generally.
doi.org/10.1145/3613904.3642013
Integration of artificial intelligence (AI) into clinical decision support systems (CDSS) poses a socio-technological challenge that is impacted by usability, trust, and human-computer interaction (HCI). AI-CDSS interventions have shown limited benefit in clinical outcomes, which may be due to insufficient understanding of how health-care providers interact with AI systems. Large language models (LLMs) have the potential to enhance AI-CDSS, but haven't been studied in either simulated or real-world clinical scenarios. We present findings from a randomized controlled trial deploying AI-CDSS for the management of upper gastrointestinal bleeding (UGIB) with and without an LLM interface within realistic clinical simulations for physician and medical student participants. We find evidence that LLM augmentation improves ease-of-use, that LLM-generated responses with citations improve trust, and HCI varies based on clinical expertise. Qualitative themes from interviews suggest the perception of LLM-augmented AI-CDSS as a team-member used to confirm initial clinical intuitions and help evaluate borderline decisions.
doi.org/10.1145/3613904.3642024