OBJECTIVE: Geriatric syndromes such as functional disability and lack of social support are often not encoded in electronic health records (EHRs), thus obscuring the identification of vulnerable older adults in need of additional medical and social services. In this study, we automatically identify vulnerable older adult patients with geriatric syndrome based on clinical notes extracted from an EHR system, and demonstrate how contextual information can improve the process. MATERIALS AND METHODS: We propose a novel end-to-end neural architecture to identify sentences that contain geriatric syndromes. Our model learns a representation of the sentence and augments it with contextual information: surrounding sentences, the entire clinical document, and the diagnosis codes associated with the document. We trained our system on annotated notes from 85 patients, tuned the model on another 50 patients, and evaluated its performance on the rest, 50 patients. RESULTS: Contextual information improved classification, with the most effective context coming from the surrounding sentences. At sentence level, our best performing model achieved a micro-F1 of 0.605, significantly outperforming context-free baselines. At patient level, our best model achieved a micro-F1 of 0.843. DISCUSSION: Our solution can be used to expand the identification of vulnerable older adults with geriatric syndromes. Since functional and social factors are often not captured by diagnosis codes in EHRs, the automatic identification of the geriatric syndrome can reduce disparities by ensuring consistent care across the older adult population. CONCLUSION: EHR free-text can be used to identify vulnerable older adults with a range of geriatric syndromes.

OBJECTIVE: The Johns Hopkins Center for Population Health IT hosted a 1-day symposium sponsored by the National Library of Medicine to help develop a national research and development (R&D) agenda for the emerging field of population health informatics (PopHI). MATERIAL AND METHODS: The symposium provided a venue for national experts to brainstorm, identify, discuss, and prioritize the top challenges and opportunities in the PopHI field, as well as R&D areas to address these. RESULTS: This manuscript summarizes the findings of the PopHI symposium. The symposium participants' recommendations have been categorized into 13 overarching themes, including policy alignment, data governance, sustainability and incentives, and standards/interoperability. DISCUSSION: The proposed consensus-based national agenda for PopHI consisted of 18 priority recommendations grouped into 4 broad goals: (1) Developing a standardized collaborative framework and infrastructure, (2) Advancing technical tools and methods, (3) Developing a scientific evidence and knowledge base, and (4) Developing an appropriate framework for policy, privacy, and sustainability. There was a substantial amount of agreement between all the participants on the challenges and opportunities for PopHI as well as on the actions that needed to be taken to address these. CONCLUSION: PopHI is a rapidly growing field that has emerged to address the population dimension of the Triple Aim. The proposed PopHI R&D agenda is comprehensive and timely, but should be considered only a starting-point, given that ongoing developments in health policy, population health management, and informatics are very dynamic, suggesting that the agenda will require constant monitoring and updating.

In the September/October 2006 issues of JAMIA, Campbell et al.'s article “Types of Unintended Consequences Related to Computerized Provider Order Entry”1 lays out an innovative and comprehensive framework for categorizing the things that can go wrong when CPOE systems are implemented. We commend the authors for helping to move forward our collective understanding of this important area. As CPOE and other components of health information technology (HIT) logarithmically diffuse across the U.S. health care system, it is clear they will eventually become the standard all-encompassing platform for the delivery of medical care. As has been the case for all previous medical and non-medical technologies, HIT dissemination carries with it both positive and negative consequences. All nine types of “unintended consequences” outlined by Campbell et al. in their article should be of concern to health informaticists and others in involved in health care. We would like to suggest to the authors and your readers that one of the many unintended consequences they identified lurks as the most serious of all. Within Campbell et al.'s so-called “Type-7” category (new kinds of errors) is arguably the ultimate of unintended consequences: what we term “e-iatrogenesis.” We define e-iatrogenesis as patient harm caused at least in part by the application of health information technology. Our team has coined this new term as part of an ongoing Commonwealth Fund/Robert Wood Johnson Foundation supported project to develop new frameworks for measuring and promoting ambulatory care quality and safety within the HIT context. Using different labels, most scholarly discussion of actual or potential (i.e., “near-miss”) e-iatrogenic events have surrounded CPOE ordering errors, particularly relating to drugs or diagnostic tests.2–4 But we believe this is only the tip of the iceberg. An e-iatrogenic event can be associated with just about any aspect of a comprehensive HIT system and it may involve errors of commission or omission. These unintended adverse events may fall into technical, human-machine interface or organizational domains. Some e-iatrogenic events will represent the electronic version of “traditional” errors, such as a patient receiving the wrong drug dosage due to a human click-error. But other HIT precipitated or enabled errors may have no exact analog in the non-electronic context. For example, a clinical decision support system (CDSS) embedded within an electronic health record might contribute to a clinician's incorrect diagnosis or treatment plan; this could represent either a “type-one” or “two” error (e.g., making a diagnosis that was not present or missing one that was). Furthermore, while the focus of our discussion is human errors and technical design flaws, some e-iatrogenic events will not be due to errors, per se. Just as a properly used traditional medical intervention (e.g., a drug or procedure) may cause patient harm, it is inevitable that a well-designed HIT module used appropriately may also contribute to an undesirable outcome. As part of our ongoing development effort (known as the “e-indicator” quality measurement project) we are working with a consortium of five advanced HIT enabled integrated delivery systems, and we have also interviewed numerous early HIT adopters across the nation. Universally, we are hearing reports that e-iatrogenesis, and the broader area of unintended consequences, is of concern at all of these top-notch organizations. What will happen as HIT is rolled out at organizations further down the diffusion curve? By coining the term e-iatrogenesis and writing this letter our intent is to draw attention to this critical issue. We believe that both the health informatics and patient safety/quality communities must urgently collaborate to addresses this emerging problem. It is understandable that developers and vendors continue to promote the many ways that HIT can improve quality and reduce safety hazards. But we also believe it is imperative that all parties acknowledge that even with all its promise, CPOE and other types of health information technology represent a new 21st century vector for medical care-system induced harm—something we must all work to understand, measure, and mitigate.