Scholars and practitioners in public health and social welfare increasingly recognize the need for preventative interventions that address root causes rather than respond to emergent crisis. However, they face significant challenges in designing tools and demonstrating success for these initiatives. We characterize these crucial, but difficult to develop and scale solutions, using Dan Heath’s term “upstream work”. We then explore design solutions to support upstream work through a multi-phase co-design process to assist fire departments developing alternate EMS response programs to reduce 911 call volume. We contribute to literature on designing to support data practices in community organizations and further delineate the key challenge of these programs as upstream initiatives: demonstrating success to stakeholders. We then present our co-designed prototype, a data dashboard to make the promising work of preventative programs visible for different stakeholder audiences. Finally we reflect on good practices for designing to support community based upstream initiatives.

Despite extensive efforts to create fairer machine learning (ML) datasets, there remains a limited understanding of the practical aspects of dataset curation. Drawing from interviews with 30 ML dataset curators, we present a comprehensive taxonomy of the challenges and trade-offs encountered throughout the dataset curation lifecycle. Our findings underscore overarching issues within the broader fairness landscape that impact data curation. We conclude with recommendations aimed at fostering systemic changes to better facilitate fair dataset curation practices.

This article describes the successful collaboration “in the wild” between Clinical Documentation Integrity Specialists (CDIS) and an Artificial Intelligence (AI)-embedded software to conduct knowledge work. CDIS review patient charts in near real-time to improve clinicians’ documentation, with the goal to make medical documentation more accurate, consistent and complete. CDIS collaborate with an AI-embedded “Computer Assisted Coding” (CAC) system that scans records from the Electronic Healthcare Record and auto-suggests codes based on natural language processing. CDIS find the CAC's suggestions are often inaccurate—often humorously so. Still, they find the CAC to be a useful helper, like Robin is to Batman. This human-AI collaboration is contingent on several factors: the flexible integration of the AI into the workflow similar to the notion of unremarkable AI; supporting the CDIS’ sensemaking; the CDIS’ knowledge about the CAC being predictably unreliable, an experience by the CDIS of the AI's value; humans remaining in control; and ability to experiment with the AI, which spurs reflection and learning for these knowledge workers.

During a global pandemic such as COVID-19, laypeople bear a large burden of responsibility for assessing risks associated with COVID-19 and taking action to manage risks in their everyday lives, yet epidemic-related information is characterized by uncertainty and ambiguity. People perceive risks based on partial, changing information. We draw on crisis informatics research to examine the multiple types of risk people perceive in relation to the COVID-19 pandemic, the information sources that inform perceptions of COVID-19 risks, and the challenges that people have in getting the information they need to understand risks, using qualitative interviews with individuals across the United States. Participants describe multiple pandemic-related threats, including illness, secondary health conditions, economic, socio-behavioral, and institutional risks. We further uncover how people draw on multiple information sources from technological infrastructures, people, and spaces to inform the types of their risk perceptions, uncovering deep challenges to acquiring needed risk information.

CSCW has previously investigated the "information work" that patients do in relation to their health, illnesses, and treatments. Our study contributes to this ongoing stream of work by investigating the information work that chronically ill patients do in relation to their physical bodily experiences. We use an interpretivist approach and qualitative methods (semi-structured interviews and observations) to examine how patients in a midwestern U.S. city carry out information work related to their bodily experiences of Chronic Kidney Disease (CKD). Our findings highlight four themes related to such work: bodily experiences of CKD; making meaning of bodily experiences through experiential learning; communicating bodily experiences; and attuning daily life around bodily experiences. In so doing, we show how patients continuously engage in information work in relation to their bodily experiences. For example, we found participants experienced physical sensations such as pain, cramping, and extreme fatigue, which elicited complex labor, yet most patients received little education from health care professionals about what bodily sensations they might have and what these physical experiences might mean. While patients and clinicians worked collaboratively to adjust treatment based on bodily information, such collaborative adjustment often did not occur until patients learned to interpret and communicate bodily experiences through a difficult and physically painful "trial and error" learning process. Our empirical insights extend on the literature of information work that chronically ill patients do and inform design of information technologies and care processes for patients with chronic disease.

Implementation of new health information systems such as Electronic Health Records (EHR) is expected to reap many benefits. However, the transition from one information system to another is often associated with inefficiency, ineffectiveness, and patient safety hazards. These negative consequences are difficult to predict and avoid before system transitions take place. The changed physical form of information remains an unexamined facet of healthcare system transitions. Using ethnographic methods in two clinical sites, we discovered a recurrent set of problems that emerged due to physical disconnections between information and practice predicated on implementation of new information systems. "Physical misalignments" are instances where workers cannot bring information sources to hand in the precise time and place in which they are needed. We identify three types of physical misalignments, then discuss how physical misalignments can be proactively identified and corrected before, during, and after implementation of new health information systems.

Crisis informatics has examined how institutions and individuals seek, communicate, and curate information in response to crises. The public's communication and perception of risks on social media remain understudied. In this study, we report a qualitative analysis of public perceptions of risks and risk management measures on Reddit during the Zika crisis, an emerging epidemic associated with high uncertainty regarding pathology, epidemiology, and broad consequences. We found two types of perceived risks: ones directly caused by the Zika virus, and ones potentially introduced by authorities' risk management measures. Risk perceptions unfolded along multiple dimensions beyond the imminent and personal level. Reddit users discussed in a speculative way to foresee various risks in the long run or at larger geographical scales. We discuss the multidimensionality and speculative nature of risk perception on social media, and derive implications for crisis informatics research and public health research and practice.

CSCW research has investigated how people at workplace and organizational settings gain knowledge required for work, but less is known regarding how "organizational outsiders" obtain knowledge about organizations and organizational landscapes that provide a service. Gaining knowledge about service landscapes is particularly difficult because they are often complex, non-transparent, and fragmented. We address this gap through an interview study of 19 U.S. parents regarding how they navigated health services for their young children, and how they gained competence in navigation practices. We describe a similar process all participants went through including four inherently iterative stages: seeking and integrating knowledge, decision-making, encountering breakdowns, and repairing and reflecting. We further elucidate what constitutes navigational competence, or the creation of resources about how to navigate, for our participants. We discuss how our study could advance understanding of navigation practices, and the importance for HCI design to support these complex yet essential navigation practices and the accumulation of navigational competence.

Recently, diseases like H1N1 influenza, Ebola, and Zika virus have created severe crises, requiring public resources and personal behavior adaptation. Crisis Informatics literature examines interconnections of people, organizations, and IT during crisis events. However, how people use technology to cope with disease crises (outbreaks, epidemics, and pandemics) remains understudied. We investigate how individuals used social media in response to the outbreak of Zika, focusing on travel-related decisions. We found that extreme uncertainty and ambiguity characterized the Zika virus crisis. To cope, people turned to social media for information gathering and social learning geared towards personal risk assessment and modifying decisions when dealing with partial and conflicting information about Zika. In particular, individuals sought local information and used socially informed logical reasoning to deduce the risk at a specific locale. We conclude with implications for designing information systems to support individual risk assessment and decision-making when faced with uncertainty and ambiguity during public health crises.

We report a study of peer support in online health communities for pregnancy care along three gestational stages (trimesters) to investigate how pregnant women seek and receive peer support during different stages of pregnancy. Using Babycenter.com as our research setting, we found that pregnant women sought peer support due to constrained access to healthcare providers, dissatisfaction with healthcare services/medical advice, limited offline social support, and unavailability of information in other venues. While the particular topics of concern typifying each trimester were distinct, pregnant women consistently sought advice, informal and formal knowledge, reassurance, and emotional support from peers during each stage of pregnancy. BabyCenter.com peers provided support by leveraging their own experiential knowledge and passing along clinical expertise acquired during the course of their own healthcare. We discuss design implications for health services and IT systems that meet pregnant women's temporal and multi-faceted needs during prenatal care.

Conspiracy theories have gained much academic and media attention recently, due to their large impact on public events. Crisis informatics researchers have examined conspiracy theories as a type of rumor. However, little is known about how conspiracy theories are produced and developed on social media. We present a qualitative study of conspiracy theorizing on Reddit during a public health crisis--the Zika virus outbreak. Using a mixed-methods approach including content analysis and discourse analysis, we identified types of conspiracy theories that appeared on Reddit in response to the Zika crisis, the conditions under which Zika conspiracy theories emerge, and the particular discursive strategies through which Zika conspiracy theories developed in online forums. Our analysis shows that conspiracy talk emerged as people attempted to make sense of a public health crisis, reflecting their emergent information needs and their pervasive distrust in formal sources of Zika information. Practical implications for social computing researchers, health practitioners, and policymakers are discussed.

Contemporary decisions about the management of populations, public services, security, and the environment are increasingly made through knowledge gleaned from "big data" and its attendant infrastructures and algorithms. Though often described as "raw," this data is produced by techniques of measurement that are imbued with judgments and values that dictate what is counted and what is not, what is considered the best unit of measurement, and how different things are grouped together and "made" into a measureable entity. In this paper, we analyze these politics of measurement and how they relate to action through two case studies involving high stake public health measurements where experts intentionally leverage measurement to change definitions of harm and health. That is, they use measurement for activism. The case studies offer a framework for thinking about of how the politics of measurement are present in user interfaces. It is usually assumed that the human element has been scrubbed from the database and that significant political and subjective interventions come from the analysis or use of data after the fact. Instead, we argue that human-computer interactions start before the data reaches the computer because various measurement interfaces are the invisible premise of data and databases, and these measurements are political.

Electronic Medical Records promise to simultaneously enhance coordination and provide transparency and accountability in work process. As such, EMR are purported to benefit both hospitals and patients. In this paper we use grounded empirical data to explore how this promise plays out in the everyday tasks of healthcare providers. Building on the small body of CSCW literature that suggests that the accounting functions of EMR are impinging on the ability of medical personnel to coordinate work, we draw on the theoretical lens of new institutionalism to outline how certain institutional logics around safety and accountability are shaping the experience of EMR systems in situ. We suggest that the institutional logics that currently characterize U.S. healthcare are embedded in the EMR design itself, structuring how institutional values such as 'safety' are achieved and evaluated. Using over one year of ethnographic research in an obstetrical unit, we find that the institutional logics of 'safety' embedded in the EMR create negative organizational outcomes, effectively undermining coordination and necessitating inaccurate accounts of work. We provide design implications to address these issues in the current institutional environment and envision how systems might be designed to promote alternate logics of safety that are social, dynamic, and cast humans as expert agents in the system.

In this paper, we describe a practice that is common across multiple heterogeneous contexts but enacted differently depending on the unique constellation of resources and demands present in each local context. Using the case of informal documentation practices in two departments of a single hospital, Emergency and Labor & Delivery, we describe how clinicians in each department develop contextualized informal documentation practices after deployment of a new EMR system. We describe three underlying functions of informal documentation that are inherent to the practice of medical personnel: "memory work," abstraction work," and "future work." We then find that the newly deployed EMR technology does not support these kinds of work. We argue that hospital documentation work systems should be designed with an eye to such universal work practices, while keeping in mind that the effectiveness of informal documentation practices is rooted in its adaptive and flexible deployment in heterogeneous work settings.

CSCW has long been concerned with how work is coordinated. A rich body of literature examines the mechanisms underlying cooperative work and the articulation of discrete tasks into meaningful sequences of action. However, there is less treatment of how workers balance multiple streams of work at once. In hospitals, the introduction of Health Information Technologies coupled with increased requirements for documentation means that workers must simultaneously care for and integrate two work trajectories: that related to the patient and that related to the medical record. Using data from an ethnographic study of labor & delivery nurses in a mid-size hospital, I describe the situated, embodied, and effortful work of coordinating multiple streams of action into a single coherent performance of work, a process I refer to as choreography, and present a number of choreography practices. I then describe implications of this perspective for CSCW.