Electronic health records (EHRs) facilitate storing, organizing, and sharing personal health information. The academic literature suggests that multimedia information (MM; image, audio, and video files) should be incorporated into EHRs. We examined the acceptability of MM-enabled EHRs for Ontario-based software vendors and physicians, using a qualitative analysis of primary and acute care EHR vendor websites, and a survey of physician perceptions regarding MM features in EHRs. Primary care EHR vendors provided more product-specific information than acute care vendors; however, neither group emphasized MM features in their EHRs. Physicians had slightly positive perceptions of image and video features, but not of audio features. None of the external factors studied predicted physicians‘ intention to use MM. Our findings suggest that neither vendors nor physicians are enthusiastic about implementing or using MM in EHRs, despite acknowledging potential benefits. Further research is needed to explore how to incorporate MM into EHR systems.

Abstract The 2009 American Recovery and Reinvestment Act has earmarked 27 billion dollars to promote the adoption of Health Information Technologies (HIT) in the US, and to gain access to these funds, providers must document “Meaningful Use” during the care process. While individual HIT use according to lean measures, including meaningful use, is prevalent in the IS literature, few studies have incorporated rich measures to account for the task, the technology, and the user in a team context. This dissertation conceptualizes Team Deep Structure Use of Computerized Provider Order Entry (CPOE) as an IT- enabled coordination mechanism, and Relational Coordination as the inherent ability of clinical teams to coordinate care spontaneously using informal, relationship based mechanisms. IT-enabled and Relational Coordination mechanisms are each evaluated across five maximally different patient conditions to simultaneously examine their impact on our outcome measure, Patient Satisfaction with the clinical care team. The extant literature has established a deep understanding of IT adoption shortly after implementation, yet the literature is silent on the antecedents of IT use according to rich measures well after the shake down phase, a period in which the majority of organizations operate. We incorporate the Adaptive Structuration Theory (AST) constructs of Faithfulness of Appropriation, and Consensus on Appropriation as the focal antecedents of Deep Structure Use of the clinical system by team members. To our knowledge, no prior research has linked these two AST constructs to clinical outcomes through the incorporation of a rich use mediator such as Deep Structure Use of a Health IT. To test our model, we relied on survey responses from 555 physicians, nurses and mid-levels which had cared for 261 patients across five patient conditions, ranging from vaginal birth, to organ transplant, as well as pneumonia, knee/hip replacement and cardiovascular surgery. Our results confirm that the Adaptive Structuration constructs of Faithfulness of Appropriation and Consensus on Appropriation, generate positive and statistically significant path coefficients predicting Team Deep Structure Use of CPOE. We also report differential effects on Patient Satisfaction with the care team resulting from technology use. Results range from a significant positive path coefficient (.285) associated with higher Team Deep Structure Use on combined Pneumonia and Organ Transplant teams, to a significant negative path coefficient (-.174) on cardiovascular surgery teams. As expected, Pneumonia, Organ Transplant and Cardiovascular Surgery teams all reported positive effects on Patient Satisfaction with the care team as a result of higher Relational Coordination scores. For teams caring for patient conditions consistently associated with a shorter length of stay, including vaginal birth and knee/hip replacement, higher reported use of IT- enabled, or Relational Coordination mechanisms, did not result in a significant increase in Patient Satisfaction. This dissertation contributes to the growing Health IT literature, and has practical implications for clinicians, hospital administrators and Health IT professionals. This dissertation is the first to operationalize a rich measure of use of an HIT by clinical teams, and to simultaneously measure the impact of IT enabled and Relational Coordination mechanisms on Patient Satisfaction. Secondly, through the introduction of Adaptive Structuration constructs, our model establishes a methodology for predicting rich, nuanced use in teams well after the initial shake down phase associated with recent HIT implementation. Through the juxtaposition of the impact of IT-enabled and Relational Coordination mechanisms across patient conditions, practitioners can design interventions and adjust the level of resources applied to process improvement accordingly.

<p><bold>Background:<bold> Hospital adoption of health information technology (HIT) systems is promoted as essential to decreasing medical error and their associated 44,000 annual deaths and $17 billion in healthcare costs (Institute of Medicine, 2001; Kohn, Corrigan, & Donaldson, 1999). Leading national healthcare groups, such as the Institute of Medicine, Agency for Healthcare Research and Quality, Institute for Healthcare Improvement, and the Leap Frog Group continue to advocate for increased use of HIT (AHRQ, 2010; Beidler, 2010; Institute of Medicine, 2001; Page, 2003; The Leapfrog Group, 2009), such as provider order entry and electronic health record systems, as a way to improve healthcare quality in hospitals. Even under intense pressure to adopt HIT, however, a mere 2% of US hospitals report having a comprehensive electronic health record system. Further, more than 50% of US hospitals have only rudimentary HIT systems (Jha et al., 2009). With the ARRA HITECH Act of 2009, the pressure on hospitals to quickly adopt HIT and achieve meaningful use is mounting.</p><p>While a large body of literature exists about HIT implementation, the majority is anecdotal case reports. The remaining studies investigated attitudes about HIT or the impact of HIT on patient care processes and outcomes. Thus, best strategies for implementing HIT in hospitals remain unknown. Study design choices, such as the use of self report data, retrospective data collection methods, subjects from single care units or single healthcare professions further limit our understanding HIT implementation in complex hospital care settings.</p><p><bold>Methods:<bold> This prospective, longitutdinal case study used a novel approach, sensemaking, to understanding how project teams may work to implement HIT in an academic medical center. Sensemaking, defined as the social process of establishing the meaning of events and experiences (Weick, 1995), is associated with learning and problemsolving in research studies of healthcare and nonhealthcare settings. Through direct observation and document review I observed project team social interaction and activities over the course of the 18 month preimplementation phase of an HIT implementation project in a single tertiary care hopsital.</p><p><bold>Conclusions:<bold> In this study, I described team actions and activities that enhanced clinician team member sensemaking including: frequent, collective interaction with HIT and focusing team members' attention on specific aspects of HIT function. Further, study findings demonstrated that team members' perceptions of HIT and care processes varied across healthcare professions, management levels, and departments. Supportive social interaction from team leaders and members encouraged team member participation and resulted in members' voicing observations, perceptions and attitudes about the HIT and hospital care processes. Sensemaking of HIT teams not only resulted in identification of needed HIT design changes, but also revealed assumptions and information which may prove critical to successful HIT implementation in hospital care environments. Based on study findings, I suggested strategies for selecting and preparing HIT team members as well as for HIT team activities. This study advanced our understanding of how project teams function and bring about change in complex hospital care environments by not only identifying HIT implementation issues within but also describing the link between team member social interaction and implementation actions.</p> / Dissertation

This paper considers the adoption of Health Information Technology (HIT) by physician clinics with ten or fewer physicians. The paper considers the theoretical economics literature on technology adoption for a new technology and has a place in the empirical tests of these models. The two major hypotheses tested in the paper are that the probability of adopting HIT increases with the number of physicians working at the clinic and if the clinic is part of a chain of clinics, and that it also increases with increased competition at the market level measured by the number of clinics per 10,000 residents in a county. To test these hypotheses, the paper first estimates a baseline logit model followed by three hazard rate models. In each case, clinic size is found to have positive though not significant effect on the probability of adoption (in the logit model) or to decrease the predicted time to adoption for the clinic (in the hazard rate models), being in a chain of clinics is found to have a strong positive and significant on the probability of adoption, and increased competition is found to have a positive though not significant effect on the probability of adoption.

The implementation of planned organizational change is ultimately a communication-related phenomenon, and as such, it is imperative that organizational communication scholars examine the interactions surrounding EHR implementation and understand how users (e.g. healthcare practitioners) utilize, evaluate, and deliberate this new technological innovation. Previous research on planned organizational change has called for researchers to adopt a more dynamic perspective that emphasizes the active agency of organizational members throughout implementation processes and focuses on informal implementers and change reinvention (work-arounds) as individuals actively reinterpret and personalize their work roles during implementation socialization. This dissertation seeks to fill this gap in research by demonstrating how communication between doctors, nurses, and other health professionals affects the adoption, maintenance, alternation, modification, or rejection of EHR systems within health care organizations. To delve into these inquiries and examine the intersecting domains of medical informatics and organizational communication research, this dissertation proceeds in the following manner: First, a literature review, capitalizing on Laurie Lewis’s work in planned organizational change and social constructionist views of technology use in organizations, outlines the assumptions that undergird this research. Next, this dissertation builds a model that predicts the communicative and structural antecedents of the study outcome variables, which include 1) organizational resistance to EHR implementation, 2) employees’ perception of EHR implementation success, 3) levels of change reinvention—or work-arounds—due to change initiatives and activities, and 4) employees’ perceptions of the quality of the organizational communication surrounding the change. Hypotheses guiding the model specification are provided and are followed by a description of the empirical methods and procedures that were utilized to explore the variable relationships. Results of the SEM model suggest that work-arounds could play a mediating role governing the relationship between informal social influence and the outcome variables in the study. In addition, one-way ANOVAs and multiple regression analyses reveal that physicians are the most resistant to EHR implementation and perceived change communication quality positively predicts perceived EHR implementation success and perceived relative advantage of EHR and negatively predicts employee resistance. A discussion of the expected and unexpected results is offered in addition to study limitation and future directions. / text

A method has been developed to track infectious diseases by using data mining of active Twitter accounts and its efficacy was demonstrated during the West African Ebola outbreak of 2014. Using a meme based n-gram semantic usage model to search the Twitter database for indications of illness, flight and death from the spread of Ebola in Africa, principally from Guinea, Sierra Leone and Liberia. Memes of interest relate disease to location and severity and are coupled to the density of Tweets and re-Tweets. The meme spreads through the community of social users in a fashion similar to nonlinear wave propagation- like a shock wave, visualized as a spike in Tweet activity. The spreading was modeled as a system isomorphic to a modified SIR (Susceptible, Infected, Removed disease model) system of three coupled nonlinear differential equations using Twitter variables. The nonlinear terms in this model lead to feedback mechanisms that result in unusual behavior that does not always reduce the spread of the disease. The resulting geographic Tweet densities are coupled to geographic maps of the region. These maps have specific threat levels that are ported to a mobile application (app) and can be used by travelers to assess the relative safety of the region they will be in.

Health information technology is continuously changing and becoming more complex and susceptible to errors. It is both an essential and disruptive innovation that requires proper management of risks arising from its use. To properly manage these risks, there is a need to, first, determine how healthcare organizations in Canada are addressing the issue of errors arising from the use of health information technology (i.e., technology-induced errors). The purpose of this thesis is to determine the level of technology-induced error awareness in Canadian healthcare organizations, to identify processes and procedures at these organizations aimed at addressing, managing, and preventing technology-induced errors, as well as to identify factors that contribute to technology-induced errors. The study finds that, based on the currently available literature, information about these errors in healthcare is not complete. This prevents the development and application of effective health information technology risk management solutions. The research from the semi-structured interviews finds that the definition of technology-induced errors is not consistent among the study participants. The research from the semi-structured interviews also finds a lack of consensus on factors that cause technology-induced errors as well as a lack of reporting mechanisms available that are specifically aimed at reporting technology-induced errors in healthcare. This confirms that there is a lack of technology-induced error awareness among Canadian healthcare organizations, which prevents the ability to properly address, manage, and prevent these errors. / Graduate / 0769 / 0723 / paulebw@uvic.ca

A rise in chronic conditions has put a strain on our healthcare system. Treatment for chronic conditions spans time, agencies, and providers, making coordination a complex problem. Information systems such as electronic health records should be helping with the challenge of coordination, but research shows that often they do not. This thesis aims to alleviate this problem by examining the design of health information technology with an emphasis on social and organizational processes. The focus of this thesis is on the implications of continuous care over time: the shift from a single provider to team-based services, the emergence of patients and families as primary caregivers in the home, and the diffusion of data-driven decision making. I investigated these trends to understand the role of data in coordinating long-term care, and inform the design of information systems. I studied behavioral and mental health services for children, which are coordinated across clinical, home, and special education settings. I found coordination that was unstructured, unpredictable, and adaptive. I developed a conceptual framework, collaborative reflection, to describe my observations and distinguish my findings from the processes of time-critical and protocol-based care. I also found ways in which coordination was not data-driven, due to a lack of support and tools. Collaborative reflection thus illustrates how long-term coordination works when it is data-driven, informing a discussion of what is needed for coordination to be data-driven. Based on the process of collaborative reflection, and using participatory design, I developed Lilypad—a tablet-based information system for data-driven coordination. I conducted a five-month deployment study of Lilypad in the field, to examine its social impact. This study validated designing for collaborative reflection to improve the use of data in coordination. The contributions of this thesis are: a description of unstructured and informal workflow that drives long-term coordination in health services; the theoretical construct of collaborative reflection to inform the design of systems that improve coordination; a field deployment validation, demonstrating how designing for collaborative reflection improves coordination and avoids common unintended consequences of health information technology.

The Live Well Springfield (LWS) movement is a collaborative effort of partner organizations in Springfield, Massachusetts. The project promotes healthy living by increasing knowledge and awareness of food and physical activity. A key LWS strategy was the creation of a website to function as an information hub. In addition to local event and health information, the website features 16 narratives depicting residents practicing healthy lifestyle choices, designed to encourage community engagement. To date, there has been no evaluation of the website’s reach and effect. A mixed methods approach, surveys and focus group discussions, was designed to collect data from people who live, work, or attend school in Springfield. Focus group participants were recruited in person at Springfield Community College, via recruitment posters (distributed at STCC), and through email requests from a previously compiled list of residents willing to be contacted. A website evaluation survey was developed using eHealth research constructs and the Expectation-Confirmation Model (ECM). This survey measured users’ perceived quality and satisfaction with the website. The survey was accessible via the livewellspringfield.org homepage, the LWS Facebook page, and emailed directly to potential respondents. The validated eHealth Literacy Scale (eHEALS) was incorporated into the survey and focus group sessions to assess self-reported skills for using eHealth resources. Each hour-long focus group (n=5 and n=6, respectively) was video/audio recorded and fully transcribed. Focus group transcripts were analyzed to thematically organize responses to narratives and fact-based health messages and assess the appeal, relevance, effectiveness, perceived purpose, and appropriateness. Survey data was analyzed to produce frequencies, descriptive statistics, and correlations. A mean eHEALS score of 4.22 of 5.00 (SD=0.83) was calculated from 36 responses, suggesting this sample felt very knowledgeable and confident using eHealth resources. Health Literacy Advisor (HLA) software was used to analyze an aggregate of all narratives, resulting in a Fry-based reading grade level of 8.4. On a five-point Likert scale, mean satisfaction with the website was 4.71 (SD=0.53), and mean likelihood to return was 4.76 (SD=0.51). Content analysis of focus group transcripts resulted in 184 responses coded for one or more themes. The largest proportion of responses (40.2%) related to effectiveness. One third of these effectiveness-related responses were negative toward the fact-based examples. Although the narratives were greatly preferred in both groups, all respondents made comments or agreed with suggestions to have both affective narratives and strictly fact-based health messages accessible, regardless of initial preferences. Results and interpretations will be reported to LWS partners to inform potential revisions of the website revisions and contribute to ongoing activities of the LWS initiative.

Health information networks continue to expand under the Affordable Care Act yet little research has been done to query and analyze multiple patient populations in parallel. Differences between hospitals relating to patient demographics, treatment approaches, disease prevalences, and medical coding practices all pose significant challenges for multi-site analysis and interpretation. Furthermore, numerous methodological issues arise when attempting to analyze disease association in heterogeneous health care settings. These issues will only continue to increase as greater numbers of hospitals are linked. To address these challenges, I developed the Shared Health Research Informatics Network (SHRINE), a distributed query and analysis system used by more than 60 health institutions for a wide range of disease studies. SHRINE was used to conduct one of the largest comorbidity studies in Autism Spectrum Disorders. SHRINE has enabled population scale studies in diabetes, rheumatology, public health, and pathology. Using Natural Language Processing, we de-identify physician notes and query pathology reports to locate human tissues for high-throughput biological validation. Samples and evidence obtained using these methods supported novel discoveries in human metabolism and paripartum cardiomyopathy, respectively. Each hospital in the SHRINE network hosts a local peer database that cannot be overridden by any federal agency. SHRINE can search both coded clinical concepts and de-identified physician notes to obtain very large cohort sizes for analysis. SHRINE intelligently clusters phenotypic concepts to minimize differences in health care settings. I then analyzed a statewide sample of all Massachusetts acute care hospitals and found diagnoses codes useful for predicting Acute Myocardial Infarction (AMI). The AMI association methods selected 96 clinical concepts. Manual review of PubMed citations supported the automated associations. AMI associations were most often discovered in the circulatory system and were most strongly linked to background diabetic retinopathy, diabetes with renal manifestations, and hypertension with complications. AMI risks were strongly associated with chronic kidney failure, liver diseases, chronic airway obstruction, hemodialysis procedures, and medical device complications. Learning the AMI associated risk factors improved disease predictions for patients in Massachusetts acute care hospitals.