AI-enabled decision-support systems aim to help medical providers rapidly make decisions with limited information during medical emergencies. A critical challenge in developing these systems is supporting providers in interpreting the system output to make optimal treatment decisions. In this study, we designed and evaluated an AI-enabled decision-support system to aid providers in treating patients with traumatic injuries. We first conducted user research with physicians to identify and design information types and AI outputs for a decision-support display. We then conducted an online experiment with 35 medical providers from six health systems to evaluate two human-AI interaction strategies: (1) AI information synthesis and (2) AI information and recommendations. We found that providers were more likely to make correct decisions when AI information and recommendations were provided compared to receiving no AI support. We also identified two socio-technical barriers to providing AI recommendations during time-critical medical events: (1) an accuracy-time trade-off in providing recommendations and (2) polarizing perceptions of recommendations between providers. We discuss three implications for developing AI-enabled decision support used in time-critical events, contributing to the limited research on human-AI interaction in this context.

Hemorrhage, or severe blood loss due to injury, is a leading cause of preventable deaths after injury. This study uses and extends activity theory to understand the dynamics of team-based hemorrhage control during trauma resuscitation and to explore potential computerized mechanisms to support this time- and safety-critical process. We reviewed videos of 25 resuscitation cases and analyzed hemorrhage control activities using nine activity theory prompts, including a new prompt-speech intention-a critical but underexplored dimension of teamwork in prior activity theory analyses. Through this process, we identified the most common delay-causing activities and developed routine and non-routine activity models for each. A comparison of these models showed that variations from the routine models emerged due to changes in the division of labor, instruments, community, and speech intentions. We contribute to research on designing socio-technical systems by (1) identifying needs and opportunities for computerized support that address delays in complex medical teamwork and (2) examining how an intervention changes an activity model. We also show how adding detailed speech data aids in identifying contradictions between elements in an activity model.

Process data constructed from event logs provides valuable insights into procedural dynamics over time. The confidential information in process data, together with the data's intricate nature, makes the datasets not sharable and challenging to collect. Consequently, research is limited using process data and analytics in the process mining domain. In this study, we introduced a synthetic process data generation task to address the limitation of sharable process data. We introduced a generative adversarial network, called ProcessGAN, to generate process data with activity sequences and corresponding timestamps. ProcessGAN consists of a transformer-based network as the generator, and a time-aware self-attention network as the discriminator. It can generate privacy-preserving process data from random noise. ProcessGAN considers the duration of the process and time intervals between activities to generate realistic activity sequences with timestamps. We evaluated ProcessGAN on five real-world datasets, two that are public and three collected in medical domains that are private. To evaluate the synthetic data, in addition to statistical metrics, we trained a supervised model to score the synthetic processes. We also used process mining to discover workflows for synthetic medical processes and had domain experts evaluate the clinical applicability of the synthetic workflows. ProcessGAN outperformed the existing generative models in generating complex processes with valid parallel pathways. The synthetic process data generated by ProcessGAN better represented the long-range dependencies between activities, a feature relevant to complicated medical and other processes. The timestamps generated by the ProcessGAN model showed similar distributions with the authentic timestamps. In addition, we trained a transformer-based network to generate synthetic contexts (e.g., patient demographics) that were associated with the synthetic processes. The synthetic contexts generated by our model outperformed the baseline models, with the distributions similar to the authentic contexts. We conclude that ProcessGAN can generate sharable synthetic process data indistinguishable from authentic data. Our source code is available in https://github.com/raaachli/ProcessGAN.

Automatically integrating data within interactive clinical checklists allows for enhanced dynamic displays, while also providing information needed for checklist adaptation to the context of the medical event. In this mixed-methods study, we used user-centered design sessions with clinicians to design a checklist interface that automatically captures and displays dynamic patient data. We compared the manual and automatic checklist versions during video-guided simulation sessions, evaluating the effects of automatic capture on clinicians' interactions with dynamic data and their situation awareness. Despite clinicians' concerns that automatic data capture would affect situation awareness, we found no significant difference in awareness scores. Participants preferred the automatic version, highlighting its improved accuracy and completeness. From our findings, we propose a framework for capturing dynamic data and designing dynamic data interfaces within interactive checklists. We conclude by discussing barriers and design opportunities for supporting awareness of data trends through checklists.

Almost half of the preventable deaths in emergency care can be associated with a medical delay. Understanding how clinicians experience delays can lead to improved alert designs to increase delay awareness and mitigation. In this paper, we present the findings from an iterative user-centered design process involving 48 clinicians to develop a prototype alert system for supporting delay awareness in complex medical teamwork such as trauma resuscitation. We used semi-structured interviews and card-sorting workshops to identify the most common delays and elicit design requirements for the prototype alert system. We then conducted a survey to refine the alert designs, followed by near-live, video-guided simulations to investigate clinicians' reactions to the alerts. We contribute to CSCW by designing a prototype alert system to support delay awareness in time-critical, complex teamwork and identifying four mechanisms through which teams mitigate delays.