Mark L. Braunstein

Understanding variations in pediatric asthma care processes in the emergency department using visual analytics

Health care delivery processes consist of complex activity sequences spanning organizational, spatial, and temporal boundaries. Care is human-directed so these processes can have wide variations in cost, quality, and outcome making systemic care process analysis, conformance testing, and improvement challenging. We designed and developed an interactive visual analytic process exploration and discovery tool and used it to explore clinical data from 5784 pediatric asthma emergency department patients.

A visual analytics approach to understanding care process variation and conformance

With greater pressures of providing high-quality care at lower cost due to a changing financial and policy environment, the ability to understand variations in care delivery and associated outcomes and act upon this understanding is of critical importance. Building on prior work in visualizing healthcare event sequences and in collaboration with our clinical partner, we describe our process in developing a multiple, coordinated visualization system that helps identify and analyze care processes and their conformance to existing care guidelines. We demonstrate our system using data of 5,784 pediatric emergency department visits over a 13-month period for which asthma was the primary diagnosis.

Supervised embedding of textual predictors with applications in clinical diagnostics for pediatric cardiology

OBJECTIVE Electronic health records possess critical predictive information for machine-learning-based diagnostic aids. However, many traditional machine learning methods fail to simultaneously integrate textual data into the prediction process because of its high dimensionality. In this paper, we present a supervised method using Laplacian Eigenmaps to enable existing machine learning methods to estimate both low-dimensional representations of textual data and accurate predictors based on these low-dimensional representations at the same time. MATERIALS AND METHODS We present a supervised Laplacian Eigenmap method to enhance predictive models by embedding textual predictors into a low-dimensional latent space, which preserves the local similarities among textual data in high-dimensional space. The proposed implementation performs alternating optimization using gradient descent. For the evaluation, we applied our method to over 2000 patient records from a large single-center pediatric cardiology practice to predict if patients were diagnosed with cardiac disease. In our experiments, we consider relatively short textual descriptions because of data availability. We compared our method with latent semantic indexing, latent Dirichlet allocation, and local Fisher discriminant analysis. The results were assessed using four metrics: the area under the receiver operating characteristic curve (AUC), Matthews correlation coefficient (MCC), specificity, and sensitivity. RESULTS AND DISCUSSION The results indicate that supervised Laplacian Eigenmaps was the highest performing method in our study, achieving 0.782 and 0.374 for AUC and MCC, respectively. Supervised Laplacian Eigenmaps showed an increase of 8.16% in AUC and 20.6% in MCC over the baseline that excluded textual data and a 2.69% and 5.35% increase in AUC and MCC, respectively, over unsupervised Laplacian Eigenmaps. CONCLUSIONS As a solution, we present a supervised Laplacian Eigenmap method to embed textual predictors into a low-dimensional Euclidean space. This method allows many existing machine learning predictors to effectively and efficiently capture the potential of textual predictors, especially those based on short texts.

Enterprise Transformation Through Mobile ICT: a Framework and Case Study in Healthcare

This study examines the transformative value and impact of mobile information and communication technology (ICT) in enterprises. Based on a comprehensive field study of mobile ICT adoption, implementation, and use in enterprises across multiple industries, a multi-stage framework for mobile ICT-enabled transformation was developed. The study revealed that mobile ICT enterprise value consists of increased convenience, higher levels of efficiency, new core competences, and creation of competitive advantage. The organizational impact is realized in existing and newly created data and processes, business models and strategies, and entire markets and industries. Four distinct stages of transformation are identified and described, namely mobilization, enhancement, reshapement, and redefinition. The framework is applied in the context of healthcare. The study contributes to the theoretical understanding of enterprise transformation and IT-enabled change and provides important guidelines for management of emerging technologies.

Patient — Physician collaboration on FHIR (Fast Healthcare Interoperability Resources)

With electronic health record adoption now far advanced the focus has largely shifted to interoperability - the ability to meaningfully share data among these systems and to collect and share data with devices, sensors, apps and tools increasingly available to patients. At the same time new forms of reimbursement encourage greater collaboration among providers and their patients. These twin imperatives have led to the rapid development and adoption of FHIR (Fast Healthcare Interoperability Resources), a more facile approach to representing and sharing health data based on familiar, standard technologies widely used by other industries and a potential universal health app platform. This paper explains FHIR and the impact it will potentially have on patient-physician collaboration.

Data and Analytics Challenges for a Learning Healthcare System

The Institute of Medicine (IOM) describes U.S. healthcare as inefficient, often ineffective and, for far too many patients, unsafe [Institute of Medicine 2001]. It calls for a “learning health system designed to generate and apply the best evidence for the collaborative healthcare choices of each patient and provider; to drive the process of discovery as a natural outgrowth of patient care; and to ensure innovation, quality, safety, and value in health care” [Smith et al. 2013]. A learning health system would thus continously improve based on the long recognized, but so far largely unmet, goal to use wide and deep “big data” from generally deployed electronic record systems with “the potential to transform medical practice by using information generated every day to improve the quality and efficiency of care” [Murdoch and Detsky 2013]. Progress toward a learning health system has long been impeded by three major challenges: low adoption of electronic records, the lack of interoperability among clinical information systems, and effective platforms to analyze and visualize large-scale digital health data. After decades of scant progress, recent federal programs have spurred electronic health record (EHR) adoption to levels over 90% for hospitals and approaching 70% for community-based providers.1 Advanced imaging technologies and genomics are an emerging part of medical practice and contribute additional, important, but very large, new data. Simultaneously, there is rapid growth in nontraditional

Opinion & Special Articles: Professionalism in neurology Maintaining patient rapport in a world of EMR

Health information technology is playing a critical role in fostering more efficient and effective health care systems by improving how information is recorded, organized, and exchanged through the use of electronic medical records (EMR). EMR are defined, according to the International Organization for Standardization, as “repositories of patient data in digital form, stored and exchanged securely, and accessible by multiple authorized users. They contain retrospective, concurrent, and prospective information, and their primary purpose is to support continuing, efficient, and quality integrated healthcare.”1

Health Big Data and Analytics

As the quote from the IOM conveys, our ability to generate scientific knowledge has outstripped the ability of healthcare providers to absorb, integrate and use that knowledge in daily patient care. The ability of information systems to analyze huge amounts of clinical data and provide useful insights and guidance is a major focus for research and commercial development now that the substantial majority of healthcare providers are using electronic records that have the potential to provide the data for that analysis. This feedback loop is a key part of the IOM’s vision of a learning healthcare system, as indicated by the title of the book from which the quote is taken. In this section, we’ll explore early results from these exciting and potentially transformational analytic technologies.

A Brief History and Overview of Health Informatics

This first chapter provides a brief history of some, but certainly not all, of the key subdomains within the health informatics field and further explains the potential significance of the FHIR standard that will occupy much of the rest of the book. To do this, the chapter begins with a discussion of early electronic records and clinical decision support tools and then shifts gears to introduce the concept of health information exchange. Later, we discuss interoperability challenges that date back decades and the various ways that existing technologies have been used, sometimes with limited success, to simplify and coordinate the sharing of information among providers. The chapter ends with the premise that widespread adoption of modern web technologies (and FHIR in particular) is transforming health informatics. To help illustrate this the chapter ends with a demonstration FHIR app developed by a team of Georgia Tech students did using these emerging technologies to help predict the onset of a life threatening condition in ICU patients.

The US Healthcare System

This chapter briefly describes the US healthcare system and some of the most important of its many problems. This is a complex topic that I cannot adequately cover in a short, introductory book, so I have provided a number of suggested supplemental readings. For a very complete and detailed discussion of the topics raised here (and others) I suggest the Institute of Medicine publication The Healthcare Imperative: Lowering Costs and Improving Outcomes: Workshop Series Summary which is available for purchase or free download.