Anticipated Impacts on Veteran's Healthcare: Decision support, surveillance, and quality care initiatives have incredible potential to increase the health of veterans. The rich clinical data contained in the electronic health record enables development of these applications; however, much of that data is locked in free-text reports and remains inaccessible to computerized applications. Natural language processing (NLP) can be applied to provide structured output from free-text input. The VA has been at the forefront of developing NLP techniques within the VHA. Background: A new generation of tools is needed to realize the vision of effective use and analysis of free-text clinical records for both improving care quality and informing clinical research. Clinical natural language processing tools have been developed for classifying, extracting, and summarizing information from narrative clinical texts, in support of diverse tasks including the identification of patients with nosocomial infections, phenotype-genotype correlation, and selection of patient cohorts for research studies. The VA has been a leader in developing NLP capability, with the hope of implementing NLP in system-facing, patient-facing, and team-facing applications being developed in hi2 HMP and iEHR. Many NLP tools and components are already being developed and hosted on Veteran's Informatics and Computing Infrastructure (VINCI) (Project 1). Unfortunately, a large gap between development and clinical/research practice limits applicability-most applications require customization of NLP tools to specific domains, NLP tasks often require the combination of multiple (often incompatible) tools, and there is a need to integrate external displays to aid interpretation. Advances in human-computer interaction and information visualization suggest approaches that can help overcome many of these limitations. A growing body of experience with interactive visualizations of clinical data provides multiple models of how abstracted information from clinical records might best be displayed to users, while extensive experience in fields such as end-user programming and user-centered design provides guidance on how some of the shortcomings in clinical NLP application might be addressed. Objective: We envision a flexible, user-centered development environment that will provide VA clinicians and medical researchers with the ability to customize and configure NLP and visualization components with little or no NLP-specific expertise or custom software development. As shown in Figure 1, we will build upon a solid foundation of successful VA NLP components (Project 1) developed largely by members of this CREATE team to realize this goal, applying user-centered design and evaluation techniques to address the gap between existing NLP technology and the needs of end users (providers and researchers in projects 3 and 4). The resulting development environment will allow non-NLP experts to customize, investigate, and apply NLP to clinical tasks and to develop visual interfaces that integrate and display EHR data. Methods: Consistent with the entire INFORMED CREATE, our approach is a user-centered approach in which the data needs of users are considered in the context of their clinical workflow. Users for this project are researchers and providers in Projects 3 and 4 seeking to develop applications for extracting, analyzing, and displaying structured and unstructured data from the EHR. The development environment will be a general-purpose toolkit for researchers developing diverse types of displays for many different domains.
We envision a flexible, user-centered development environment that will provide VA clinicians and medical researchers with the ability to customize and configure NLP and visualization components with little or no NLP-specific expertise or custom software development. We will build upon a solid foundation of successful VA NLP components (Project 1) developed largely by members of this CREATE team to realize this goal, applying user-centered design and evaluation techniques to address the gap between existing NLP technology and the needs of end users (providers and researchers in projects 3 and 4). The resulting development environment will allow non-NLP experts to customize, investigate, and apply NLP to clinical tasks and to develop visual interfaces that integrate and display EHR data.
