OBJECTIVE: This study explores patient motivations and preferences for sharing medical data with researchers using the iAgree platform. We examine how study characteristics, including data type requested and data-sharing arrangements, influence consent decisions, and assess the role of demographic factors, privacy concerns, and perceived benefits in shaping data-sharing behavior. MATERIALS AND METHODS: We conducted a mixed-methods study with 527 US adults (≥18 years) recruited via advisory boards, social media, clinics, and newsletters. Participants completed 3 of 4 simulated studies on iAgree, each varying by data elements requested and data-sharing scope. Participants provided consent and data-sharing decisions and completed a post-simulation survey capturing demographics, data-sharing motivations, privacy concerns, and patient activation. We used logistic regressions to examine associations between demographics, privacy concerns, and patient activation and: (1) consent status and (2) willingness to share particular data elements. Finally, we applied thematic analysis to open-ended responses. RESULTS: Consent status did not significantly vary by data type or study design. However, participants citing altruism, personal benefit, and patient solidarity were more likely to share data. Higher privacy concerns were linked to lower willingness to share family health and mental health information. Participants with higher patient activation were also less likely to share data. DISCUSSION: Demographic factors were not significantly associated with consent or willingness to share data, countering common assumptions about disparities in sharing preferences. CONCLUSION: Altruism and perceived benefit drive willingness to share health data, while privacy concerns and patient activation may reduce it, emphasizing the need for patient-centered, transparent consent models.

Patient-controlled data-sharing systems are increasingly promoted as a way to empower patients with greater autonomy over their health data. Yet it remains unclear how different stakeholders, especially patients and health system leaders, perceive the benefits and challenges of enabling granular control over the sharing of de-identified medical data for research. To address this gap, we developed a high-fidelity prototype of a patient-controlled, web-based consent platform and conducted a two-phase mixed-methods study: semi-structured interviews with 16 health system leaders and a survey with 523 patient participants. While both groups appreciated the potential of such a platform to enhance transparency and autonomy, their views diverged in meaningful ways. Leaders viewed transparency and granular control through the lens of informed consent and institutional ethics, whereas patients interpreted these factors as safeguards against potential risks and uncertainties. Our findings underscore critical tensions such as individual control and research integrity. We offer design implications for building trustworthy, context-aware systems that support flexible granularity, provide ongoing benefit‑centered transparency, and adapt to diverse literacy and privacy needs.

OBJECTIVES: Traditional methods for medical device post-market surveillance often fail to accurately account for operator learning effects, leading to biased assessments of device safety. These methods struggle with non-linearity, complex learning curves, and time-varying covariates, such as physician experience. To address these limitations, we sought to develop a machine learning (ML) framework to detect and adjust for operator learning effects. MATERIALS AND METHODS: A gradient-boosted decision tree ML method was used to analyze synthetic datasets that replicate the complexity of clinical scenarios involving high-risk medical devices. We designed this process to detect learning effects using a risk-adjusted cumulative sum method, quantify the excess adverse event rate attributable to operator inexperience, and adjust for these alongside patient factors in evaluating device safety signals. To maintain integrity, we employed blinding between data generation and analysis teams. Synthetic data used underlying distributions and patient feature correlations based on clinical data from the Department of Veterans Affairs between 2005 and 2012. We generated 2494 synthetic datasets with widely varying characteristics including number of patient features, operators and institutions, and the operator learning form. Each dataset contained a hypothetical study device, Device B, and a reference device, Device A. We evaluated accuracy in identifying learning effects and identifying and estimating the strength of the device safety signal. Our approach also evaluated different clinically relevant thresholds for safety signal detection. RESULTS: Our framework accurately identified the presence or absence of learning effects in 93.6% of datasets and correctly determined device safety signals in 93.4% of cases. The estimated device odds ratios' 95% confidence intervals were accurately aligned with the specified ratios in 94.7% of datasets. In contrast, a comparative model excluding operator learning effects significantly underperformed in detecting device signals and in accuracy. Notably, our framework achieved 100% specificity for clinically relevant safety signal thresholds, although sensitivity varied with the threshold applied. DISCUSSION: A machine learning framework, tailored for the complexities of post-market device evaluation, may provide superior performance compared to standard parametric techniques when operator learning is present. CONCLUSION: Demonstrating the capacity of ML to overcome complex evaluative challenges, our framework addresses the limitations of traditional statistical methods in current post-market surveillance processes. By offering a reliable means to detect and adjust for learning effects, it may significantly improve medical device safety evaluation.

OBJECTIVES: To introduce quantum computing technologies as a tool for biomedical research and highlight future applications within healthcare, focusing on its capabilities, benefits, and limitations. TARGET AUDIENCE: Investigators seeking to explore quantum computing and create quantum-based applications for healthcare and biomedical research. SCOPE: Quantum computing requires specialized hardware, known as quantum processing units, that use quantum bits (qubits) instead of classical bits to perform computations. This article will cover (1) proposed applications where quantum computing offers advantages to classical computing in biomedicine; (2) an introduction to how quantum computers operate, tailored for biomedical researchers; (3) recent progress that has expanded access to quantum computing; and (4) challenges, opportunities, and proposed solutions to integrate quantum computing in biomedical applications.

OBJECTIVE: Blockchain has emerged as a potential data-sharing structure in healthcare because of its decentralization, immutability, and traceability. However, its use in the biomedical domain is yet to be investigated comprehensively, especially from the aspects of implementation and evaluation, by existing blockchain literature reviews. To address this, our review assesses blockchain applications implemented in practice and evaluated with quantitative metrics. MATERIALS AND METHODS: This systematic review adapts the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework to review biomedical blockchain papers published by August 2023 from 3 databases. Blockchain application, implementation, and evaluation metrics were collected and summarized. RESULTS: Following screening, 11 articles were included in this review. Articles spanned a range of biomedical applications including COVID-19 medical data sharing, decentralized internet of things (IoT) data storage, clinical trial management, biomedical certificate storage, electronic health record (EHR) data sharing, and distributed predictive model generation. Only one article demonstrated blockchain deployment at a medical facility. DISCUSSION: Ethereum was the most common blockchain platform. All but one implementation was developed with private network permissions. Also, 8 articles contained storage speed metrics and 6 contained query speed metrics. However, inconsistencies in presented metrics and the small number of articles included limit technological comparisons with each other. CONCLUSION: While blockchain demonstrates feasibility for adoption in healthcare, it is not as popular as currently existing technologies for biomedical data management. Addressing implementation and evaluation factors will better showcase blockchain's practical benefits, enabling blockchain to have a significant impact on the health sector.

OBJECTIVE: We aimed to develop a distributed, immutable, and highly available cross-cloud blockchain system to facilitate federated data analysis activities among multiple institutions. MATERIALS AND METHODS: We preprocessed 9166 COVID-19 Structured Query Language (SQL) code, summary statistics, and user activity logs, from the GitHub repository of the Reliable Response Data Discovery for COVID-19 (R2D2) Consortium. The repository collected local summary statistics from participating institutions and aggregated the global result to a COVID-19-related clinical query, previously posted by clinicians on a website. We developed both on-chain and off-chain components to store/query these activity logs and their associated queries/results on a blockchain for immutability, transparency, and high availability of research communication. We measured run-time efficiency of contract deployment, network transactions, and confirmed the accuracy of recorded logs compared to a centralized baseline solution. RESULTS: The smart contract deployment took 4.5 s on an average. The time to record an activity log on blockchain was slightly over 2 s, versus 5-9 s for baseline. For querying, each query took on an average less than 0.4 s on blockchain, versus around 2.1 s for baseline. DISCUSSION: The low deployment, recording, and querying times confirm the feasibility of our cross-cloud, blockchain-based federated data analysis system. We have yet to evaluate the system on a larger network with multiple nodes per cloud, to consider how to accommodate a surge in activities, and to investigate methods to lower querying time as the blockchain grows. CONCLUSION: Blockchain technology can be used to support federated data analysis among multiple institutions.

OBJECTIVE: This qualitative study aimed to understand patient and researcher perspectives regarding consent and data-sharing preferences for research and a patient-centered system to manage consent and data-sharing preferences. MATERIALS AND METHODS: We conducted focus groups with patient and researcher participants recruited from three academic health centers via snowball sampling. Discussions focused on perspectives on the use of electronic health record (EHR) data for research. Themes were identified through consensus coding, starting from an exploratory framework. RESULTS: We held two focus groups with patients (n = 12 patients) and two with researchers (n = 8 researchers). We identified two patient themes (1-2), one theme common to patients and researchers (3), and two researcher themes (4-5). Themes included (1) motivations for sharing EHR data, (2) perspectives on the importance of data-sharing transparency, (3) individual control of personal EHR data sharing, (4) how EHR data benefits research, and (5) challenges researchers face using EHR data. DISCUSSION: Patients expressed a tension between the benefits of their data being used in studies to benefit themselves/others and avoiding risk by limiting data access. Patients resolved this tension by acknowledging they would often share their data but wanted greater transparency on its use. Researchers expressed concern about incorporating bias into datasets if patients opted out. CONCLUSIONS: A research consent and data-sharing platform must consider two competing goals: empowering patients to have more control over their data and maintaining the integrity of secondary data sources. Health systems and researchers should increase trust-building efforts with patients to engender trust in data access and use.

In this editorial, the first 4 Editors-in-Chief of the Journal of the American Medical Informatics Association (JAMIA) reflect on its history and future.The origins and characterization of each Editor's era represent the "lived experience" of each Editor rather than a comparison of common metrics over time.We also qualitatively assess JAMIA's progress in meeting its original vision and goals, and posit considerations for its future.Table 1 summarizes key JAMIA-related events.

OBJECTIVE: Sepsis has a high rate of 30-day unplanned readmissions. Predictive modeling has been suggested as a tool to identify high-risk patients. However, existing sepsis readmission models have low predictive value and most predictive factors in such models are not actionable. MATERIALS AND METHODS: Data from patients enrolled in the AllofUs Research Program cohort from 35 hospitals were used to develop a multicenter validated sepsis-related unplanned readmission model that incorporates clinical and social determinants of health (SDH) to predict 30-day unplanned readmissions. Sepsis cases were identified using concepts represented in the Observational Medical Outcomes Partnership. The dataset included over 60 clinical/laboratory features and over 100 SDH features. RESULTS: Incorporation of SDH factors into our model of clinical and demographic features improves model area under the receiver operating characteristic curve (AUC) significantly (from 0.75 to 0.80; P < .001). Model-agnostic interpretability techniques revealed demographics, economic stability, and delay in getting medical care as important SDH predictive features of unplanned hospital readmissions. DISCUSSION: This work represents one of the largest studies of sepsis readmissions using objective clinical data to date (8935 septic index encounters). SDH are important to determine which sepsis patients are more likely to have an unplanned 30-day readmission. The AllofUS dataset provides granular data from a diverse set of individuals, making this model potentially more generalizable than prior models. CONCLUSION: Use of SDH improves predictive performance of a model to identify which sepsis patients are at high risk of an unplanned 30-day readmission.

Concerns regarding inappropriate leakage of sensitive personal information as well as unauthorized data use are increasing with the growth of genomic data repositories. Therefore, privacy and security of genomic data have become increasingly important and need to be studied. With many proposed protection techniques, their applicability in support of biomedical research should be well understood. For this purpose, we have organized a community effort in the past 8 years through the integrating data for analysis, anonymization and sharing consortium to address this practical challenge. In this article, we summarize our experience from these competitions, report lessons learned from the events in 2020/2021 as examples, and discuss potential future research directions in this emerging field.

After 25 years of service to the American Medical Informatics Association (AMIA), Ms Karen Greenwood, the Executive Vice President and Chief Operating Officer, is leaving the organization. In this perspective, we reflect on her accomplishments and her effect on the organization and the field of informatics nationally and globally. We also express our appreciation and gratitude for Ms Greenwood's role at AMIA.

Broad health data sharing raises myriad ethical issues related to data protection and privacy. These issues are of particular relevance to Native Americans, who reserve distinct individual and collective rights to control data about their communities. We sought to gather input from tribal community leaders on how best to understand health data privacy and sharing preferences in this population. We conducted a workshop with 14 tribal leaders connected to the Strong Heart Study to codesign a research study to assess preferences concerning health data privacy for biomedical research. Workshop participants provided specific recommendations regarding who should be consulted, what questions should be posed, and what methods should be used, underscoring the importance of relationship-building between researchers and tribal communities. Biomedical researchers and informaticians who collect and analyze health information from Native communities have a unique responsibility to safeguard these data in ways that align to the preferences of specific communities.

OBJECTIVE: The Genomic Medicine Working Group of the National Advisory Council for Human Genome Research virtually hosted its 13th genomic medicine meeting titled "Developing a Clinical Genomic Informatics Research Agenda". The meeting's goal was to articulate a research strategy to develop Genomics-based Clinical Informatics Tools and Resources (GCIT) to improve the detection, treatment, and reporting of genetic disorders in clinical settings. MATERIALS AND METHODS: Experts from government agencies, the private sector, and academia in genomic medicine and clinical informatics were invited to address the meeting's goals. Invitees were also asked to complete a survey to assess important considerations needed to develop a genomic-based clinical informatics research strategy. RESULTS: Outcomes from the meeting included identifying short-term research needs, such as designing and implementing standards-based interfaces between laboratory information systems and electronic health records, as well as long-term projects, such as identifying and addressing barriers related to the establishment and implementation of genomic data exchange systems that, in turn, the research community could help address. DISCUSSION: Discussions centered on identifying gaps and barriers that impede the use of GCIT in genomic medicine. Emergent themes from the meeting included developing an implementation science framework, defining a value proposition for all stakeholders, fostering engagement with patients and partners to develop applications under patient control, promoting the use of relevant clinical workflows in research, and lowering related barriers to regulatory processes. Another key theme was recognizing pervasive biases in data and information systems, algorithms, access, value, and knowledge repositories and identifying ways to resolve them.

The Cox proportional hazards model is a popular semi-parametric model for survival analysis. In this paper, we aim at developing a federated algorithm for the Cox proportional hazards model over vertically partitioned data (i.e., data from the same patient are stored at different institutions). We propose a novel algorithm, namely VERTICOX, to obtain the global model parameters in a distributed fashion based on the Alternating Direction Method of Multipliers (ADMM) framework. The proposed model computes intermediary statistics and exchanges them to calculate the global model without collecting individual patient-level data. We demonstrate that our algorithm achieves equivalent accuracy for the estimation of model parameters and statistics to that of its centralized realization. The proposed algorithm converges linearly under the ADMM framework. Its computational complexity and communication costs are polynomially and linearly associated with the number of subjects, respectively. Experimental results show that VERTICOX can achieve accurate model parameter estimation to support federated survival analysis over vertically distributed data by saving bandwidth and avoiding exchange of information about individual patients. The source code for VERTICOX is available at: https://github.com/daiwenrui/VERTICOX.

OBJECTIVE: This study aims at reviewing novel coronavirus disease (COVID-19) datasets extracted from PubMed Central articles, thus providing quantitative analysis to answer questions related to dataset contents, accessibility and citations. METHODS: We downloaded COVID-19-related full-text articles published until 31 May 2020 from PubMed Central. Dataset URL links mentioned in full-text articles were extracted, and each dataset was manually reviewed to provide information on 10 variables: (1) type of the dataset, (2) geographic region where the data were collected, (3) whether the dataset was immediately downloadable, (4) format of the dataset files, (5) where the dataset was hosted, (6) whether the dataset was updated regularly, (7) the type of license used, (8) whether the metadata were explicitly provided, (9) whether there was a PubMed Central paper describing the dataset and (10) the number of times the dataset was cited by PubMed Central articles. Descriptive statistics about these seven variables were reported for all extracted datasets. RESULTS: We found that 28.5% of 12 324 COVID-19 full-text articles in PubMed Central provided at least one dataset link. In total, 128 unique dataset links were mentioned in 12 324 COVID-19 full text articles in PubMed Central. Further analysis showed that epidemiological datasets accounted for the largest portion (53.9%) in the dataset collection, and most datasets (84.4%) were available for immediate download. GitHub was the most popular repository for hosting COVID-19 datasets. CSV, XLSX and JSON were the most popular data formats. Additionally, citation patterns of COVID-19 datasets varied depending on specific datasets. CONCLUSION: PubMed Central articles are an important source of COVID-19 datasets, but there is significant heterogeneity in the way these datasets are mentioned, shared, updated and cited.

OBJECTIVE: To utilize, in an individual and institutional privacy-preserving manner, electronic health record (EHR) data from 202 hospitals by analyzing answers to COVID-19-related questions and posting these answers online. MATERIALS AND METHODS: We developed a distributed, federated network of 12 health systems that harmonized their EHRs and submitted aggregate answers to consortia questions posted at https://www.covid19questions.org. Our consortium developed processes and implemented distributed algorithms to produce answers to a variety of questions. We were able to generate counts, descriptive statistics, and build a multivariate, iterative regression model without centralizing individual-level data. RESULTS: Our public website contains answers to various clinical questions, a web form for users to ask questions in natural language, and a list of items that are currently pending responses. The results show, for example, that patients who were taking angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers, within the year before admission, had lower unadjusted in-hospital mortality rates. We also showed that, when adjusted for, age, sex, and ethnicity were not significantly associated with mortality. We demonstrated that it is possible to answer questions about COVID-19 using EHR data from systems that have different policies and must follow various regulations, without moving data out of their health systems. DISCUSSION AND CONCLUSIONS: We present an alternative or a complement to centralized COVID-19 registries of EHR data. We can use multivariate distributed logistic regression on observations recorded in the process of care to generate results without transferring individual-level data outside the health systems.

Our goal is to summarize the collective experience of 15 organizations in dealing with uncoordinated efforts that result in unnecessary delays in understanding, predicting, preparing for, containing, and mitigating the COVID-19 pandemic in the US. Response efforts involve the collection and analysis of data corresponding to healthcare organizations, public health departments, socioeconomic indicators, as well as additional signals collected directly from individuals and communities. We focused on electronic health record (EHR) data, since EHRs can be leveraged and scaled to improve clinical care, research, and to inform public health decision-making. We outline the current challenges in the data ecosystem and the technology infrastructure that are relevant to COVID-19, as witnessed in our 15 institutions. The infrastructure includes registries and clinical data networks to support population-level analyses. We propose a specific set of strategic next steps to increase interoperability, overall organization, and efficiencies.

OBJECTIVE: Survival analysis is the cornerstone of many healthcare applications in which the "survival" probability (eg, time free from a certain disease, time to death) of a group of patients is computed to guide clinical decisions. It is widely used in biomedical research and healthcare applications. However, frequent sharing of exact survival curves may reveal information about the individual patients, as an adversary may infer the presence of a person of interest as a participant of a study or of a particular group. Therefore, it is imperative to develop methods to protect patient privacy in survival analysis. MATERIALS AND METHODS: We develop a framework based on the formal model of differential privacy, which provides provable privacy protection against a knowledgeable adversary. We show the performance of privacy-protecting solutions for the widely used Kaplan-Meier nonparametric survival model. RESULTS: We empirically evaluated the usefulness of our privacy-protecting framework and the reduced privacy risk for a popular epidemiology dataset and a synthetic dataset. Results show that our methods significantly reduce the privacy risk when compared with their nonprivate counterparts, while retaining the utility of the survival curves. DISCUSSION: The proposed framework demonstrates the feasibility of conducting privacy-protecting survival analyses. We discuss future research directions to further enhance the usefulness of our proposed solutions in biomedical research applications. CONCLUSION: The results suggest that our proposed privacy-protection methods provide strong privacy protections while preserving the usefulness of survival analyses.

Large observational data networks that leverage routine clinical practice data in electronic health records (EHRs) are critical resources for research on coronavirus disease 2019 (COVID-19). Data normalization is a key challenge for the secondary use of EHRs for COVID-19 research across institutions. In this study, we addressed the challenge of automating the normalization of COVID-19 diagnostic tests, which are critical data elements, but for which controlled terminology terms were published after clinical implementation. We developed a simple but effective rule-based tool called COVID-19 TestNorm to automatically normalize local COVID-19 testing names to standard LOINC (Logical Observation Identifiers Names and Codes) codes. COVID-19 TestNorm was developed and evaluated using 568 test names collected from 8 healthcare systems. Our results show that it could achieve an accuracy of 97.4% on an independent test set. COVID-19 TestNorm is available as an open-source package for developers and as an online Web application for end users (https://clamp.uth.edu/covid/loinc.php). We believe that it will be a useful tool to support secondary use of EHRs for research on COVID-19.

Our primary objective is to provide the clinical informatics community with an introductory tutorial on calibration measurements and calibration models for predictive models using existing R packages and custom implemented code in R on real and simulated data. Clinical predictive model performance is commonly published based on discrimination measures, but use of models for individualized predictions requires adequate model calibration. This tutorial is intended for clinical researchers who want to evaluate predictive models in terms of their applicability to a particular population. It is also for informaticians and for software engineers who want to understand the role that calibration plays in the evaluation of a clinical predictive model, and to provide them with a solid starting point to consider incorporating calibration evaluation and calibration models in their work. Covered topics include (1) an introduction to the importance of calibration in the clinical setting, (2) an illustration of the distinct roles that discrimination and calibration play in the assessment of clinical predictive models, (3) a tutorial and demonstration of selected calibration measurements, (4) a tutorial and demonstration of selected calibration models, and (5) a brief discussion of limitations of these methods and practical suggestions on how to use them in practice.

OBJECTIVE: Predicting patient outcomes using healthcare/genomics data is an increasingly popular/important area. However, some diseases are rare and require data from multiple institutions to construct generalizable models. To address institutional data protection policies, many distributed methods keep the data locally but rely on a central server for coordination, which introduces risks such as a single point of failure. We focus on providing an alternative based on a decentralized approach. We introduce the idea using blockchain technology for this purpose, with a brief description of its own potential advantages/disadvantages. MATERIALS AND METHODS: We explain how our proposed EXpectation Propagation LOgistic REgRession on Permissioned blockCHAIN (ExplorerChain) can achieve the same results when compared to a distributed model that uses a central server on 3 healthcare/genomic datasets, and what trade-offs need to be considered when using centralized/decentralized methods. We explain how the use of blockchain technology can help decrease some of the problems encountered in decentralized methods. RESULTS: We showed that the discrimination power of ExplorerChain can be statistically similar to its counterpart central server-based algorithm. While ExplorerChain inherited some benefits of blockchain, it had a small increased running time. DISCUSSION: ExplorerChain has the same prerequisites as a distributed model with a centralized server for coordination. In a manner similar to secure multi-party computation strategies, it assumes that participating institutions are honest, but "curious." CONCLUSION: When evaluated on relatively small datasets, results suggest that ExplorerChain, which combines artificial intelligence and blockchain technologies, performs as well as a central server-based method, and may avoid some risks at the cost of efficiency.

BACKGROUND: Privacy-related concerns can prevent equitable participation in health research by US Indigenous communities. However, studies focused on these communities' views regarding health data privacy, including systematic reviews, are lacking. METHODS: We conducted a systematic literature review analyzing empirical, US-based studies involving American Indian/Alaska Native (AI/AN) and Native Hawaiian or other Pacific Islander (NHPI) perspectives on health data privacy, which we define as the practice of maintaining the security and confidentiality of an individual's personal health records and/or biological samples (including data derived from biological specimens, such as personal genetic information), as well as the secure and approved use of those data. RESULTS: Twenty-one studies involving 3234 AI/AN and NHPI participants were eligible for review. The results of this review suggest that concerns about the privacy of health data are both prevalent and complex in AI/AN and NHPI communities. Many respondents raised concerns about the potential for misuse of their health data, including discrimination or stigma, confidentiality breaches, and undesirable or unknown uses of biological specimens. CONCLUSIONS: Participants cited a variety of individual and community-level concerns about the privacy of their health data, and indicated that these deter their willingness to participate in health research. Future investigations should explore in more depth which health data privacy concerns are most salient to specific AI/AN and NHPI communities, and identify the practices that will make the collection and use of health data more trustworthy and transparent for participants.

Scientific collaborations benefit from sharing information and data from distributed sources, but protecting privacy is a major concern. Researchers, funders, and the public in general are getting increasingly worried about the potential leakage of private data. Advanced security methods have been developed to protect the storage and computation of sensitive data in a distributed setting. However, they do not protect against information leakage from the outcomes of data analyses. To address this aspect, studies on differential privacy (a state-of-the-art privacy protection framework) demonstrated encouraging results, but most of them do not apply to distributed scenarios. Combining security and privacy methodologies is a natural way to tackle the problem, but naive solutions may lead to poor analytical performance. In this paper, we introduce a novel strategy that combines differential privacy methods and homomorphic encryption techniques to achieve the best of both worlds. Using logistic regression (a popular model in biomedicine), we demonstrated the practicability of building secure and privacy-preserving models with high efficiency (less than 3 min) and good accuracy [<;1% of difference in the area under the receiver operating characteristic curve (AUC) against the global model] using a few real-world datasets.

Genetic ancestry is a critical co-factor to study phenotype-genotype associations using cohorts of human subjects. Most publicly available molecular datasets are, however, missing this information or only share self-reported race and ethnicity, representing a limitation to identify and repurpose datasets to investigate the contribution of ancestry to diseases and traits. We propose an analytical framework to enrich the metadata from publicly available cohorts with genetic ancestry information and a resulting diversity score at continental resolution, calculated directly from the data. We illustrate this framework using The Cancer Genome Atlas datasets searched through the DataMed Data Discovery Index. Data repositories and contributors can use this framework to provide genetic diversity measurements for controlled access datasets, minimizing the work involved in requesting a dataset that may ultimately prove inadequate for a researcher's purpose. With the increasing global scale of human genetics research, studies on disease risk and susceptibility would benefit greatly from the adequate estimation and sharing of genetic diversity in publicly available datasets following a framework such as the one presented.

OBJECTIVE: Decentralized privacy-preserving predictive modeling enables multiple institutions to learn a more generalizable model on healthcare or genomic data by sharing the partially trained models instead of patient-level data, while avoiding risks such as single point of control. State-of-the-art blockchain-based methods remove the "server" role but can be less accurate than models that rely on a server. Therefore, we aim at developing a general model sharing framework to preserve predictive correctness, mitigate the risks of a centralized architecture, and compute the models in a fair way. MATERIALS AND METHODS: We propose a framework that includes both server and "client" roles to preserve correctness. We adopt a blockchain network to obtain the benefits of decentralization, by alternating the roles for each site to ensure computational fairness. Also, we developed GloreChain (Grid Binary LOgistic REgression on Permissioned BlockChain) as a concrete example, and compared it to a centralized algorithm on 3 healthcare or genomic datasets to evaluate predictive correctness, number of learning iterations and execution time. RESULTS: GloreChain performs exactly the same as the centralized method in terms of correctness and number of iterations. It inherits the advantages of blockchain, at the cost of increased time to reach a consensus model. DISCUSSION: Our framework is general or flexible and can also address intrinsic challenges of blockchain networks. Further investigations will focus on higher-dimensional datasets, additional use cases, privacy-preserving quality concerns, and ethical, legal, and social implications. CONCLUSIONS: Our framework provides a promising potential for institutions to learn a predictive model based on healthcare or genomic data in a privacy-preserving and decentralized way.

OBJECTIVES: To introduce healthcare or biomedical blockchain applications and their underlying blockchain platforms, compare popular blockchain platforms using a systematic review method, and provide a reference for selection of a suitable blockchain platform given requirements and technical features that are common in healthcare and biomedical research applications. TARGET AUDIENCE: Healthcare or clinical informatics researchers and software engineers who would like to learn about the important technical features of different blockchain platforms to design and implement blockchain-based health informatics applications. SCOPE: Covered topics include (1) a brief introduction to healthcare or biomedical blockchain applications and the benefits to adopt blockchain; (2) a description of key features of underlying blockchain platforms in healthcare applications; (3) development of a method for systematic review of technology, based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement, to investigate blockchain platforms for healthcare and medicine applications; (4) a review of 21 healthcare-related technical features of 10 popular blockchain platforms; and (5) a discussion of findings and limitations of the review.

Big data and data science investigations hold great promise for making efficient use of data generated in the course of daily life: from social media transactions, news, and a variety of apps used by a large portion of the world’s population, including data generated for health care and life sciences research. Data science brings new insights when large-scale datasets are brought together to characterize and address complex problems. The past decade has seen a plethora of federal and private investments in biomedical data science collection, organization, and analysis, including the National Institutes of Health’s Big Data to Knowledge program, the Patient-Centered Outcomes Research Institute’s PCORnet, and investments from various industries. The work is maturing and interesting, and exciting results are emerging. Biomedical data science offers new and powerful tools to better understand health and disease through insights gleaned from data. Linking data science advances with knowledge representation and clinical information understanding, which have been traditional topics in the biomedical informatics field since its early days, has the potential to accelerate data-driven discovery. Biomedical informatics has also been addressing data-driven discovery. However, until this decade, examples where big data were available for this type of pursuit were limited. Biomedical informatics has thus evolved and overlaps significantly with biomedical data science, the subfield of data science that is concerned with discoveries using primarily clinical and other health-relevant data. All data science investigations must address important and interesting questions that are relevant to the areas they are applied to, have access to comprehensible datasets, and devise and apply methods robust enough to cope with complex unstructured observations.

OBJECTIVE: Finding relevant datasets is important for promoting data reuse in the biomedical domain, but it is challenging given the volume and complexity of biomedical data. Here we describe the development of an open source biomedical data discovery system called DataMed, with the goal of promoting the building of additional data indexes in the biomedical domain. MATERIALS AND METHODS: DataMed, which can efficiently index and search diverse types of biomedical datasets across repositories, is developed through the National Institutes of Health-funded biomedical and healthCAre Data Discovery Index Ecosystem (bioCADDIE) consortium. It consists of 2 main components: (1) a data ingestion pipeline that collects and transforms original metadata information to a unified metadata model, called DatA Tag Suite (DATS), and (2) a search engine that finds relevant datasets based on user-entered queries. In addition to describing its architecture and techniques, we evaluated individual components within DataMed, including the accuracy of the ingestion pipeline, the prevalence of the DATS model across repositories, and the overall performance of the dataset retrieval engine. RESULTS AND CONCLUSION: Our manual review shows that the ingestion pipeline could achieve an accuracy of 90% and core elements of DATS had varied frequency across repositories. On a manually curated benchmark dataset, the DataMed search engine achieved an inferred average precision of 0.2033 and a precision at 10 (P@10, the number of relevant results in the top 10 search results) of 0.6022, by implementing advanced natural language processing and terminology services. Currently, we have made the DataMed system publically available as an open source package for the biomedical community.

OBJECTIVE: To present user needs and usability evaluations of DataMed, a Data Discovery Index (DDI) that allows searching for biomedical data from multiple sources. MATERIALS AND METHODS: We conducted 2 phases of user studies. Phase 1 was a user needs analysis conducted before the development of DataMed, consisting of interviews with researchers. Phase 2 involved iterative usability evaluations of DataMed prototypes. We analyzed data qualitatively to document researchers' information and user interface needs. RESULTS: Biomedical researchers' information needs in data discovery are complex, multidimensional, and shaped by their context, domain knowledge, and technical experience. User needs analyses validate the need for a DDI, while usability evaluations of DataMed show that even though aggregating metadata into a common search engine and applying traditional information retrieval tools are promising first steps, there remain challenges for DataMed due to incomplete metadata and the complexity of data discovery. DISCUSSION: Biomedical data poses distinct problems for search when compared to websites or publications. Making data available is not enough to facilitate biomedical data discovery: new retrieval techniques and user interfaces are necessary for dataset exploration. Consistent, complete, and high-quality metadata are vital to enable this process. CONCLUSION: While available data and researchers' information needs are complex and heterogeneous, a successful DDI must meet those needs and fit into the processes of biomedical researchers. Research directions include formalizing researchers' information needs, standardizing overviews of data to facilitate relevance judgments, implementing user interfaces for concept-based searching, and developing evaluation methods for open-ended discovery systems such as DDIs.

JAMIA has documented the evolution of biomedical informatics through its dissemination of original research and applications, brief communications and case studies, thought-provoking perspectives, and insightful reviews. The number and diversity of data-driven models have increased substantially in the past few years. From the first developments in machine and statistical learning that were applied to health sciences decades ago, our field has flourished to include biomedical data science as one of its important components, which is possible only because of other informatics work that allows data to be standardized, integrated, and used in various learning models. This issue is focused on biomedical data science and illustrates a broad range of techniques and application areas in this field; articles submitted in response to a specific request for papers are featured in an editorial by Brennan et al. (p. 2). In addition to the articles covered in the editorial, this issue highlights tools and applications of data science in a variety of domains, all of which use clinical text as a source of data: Trivedi (p. 81) presents an interactive tool for processing clinical text, Luo (p. 93) uses convolutional neural networks to classify relations in clinical notes, and Bejan (p. 61) introduces an approach to find homelessness and adverse childhood experiences described in clinical narratives. Additionally, nonclinical text is increasing in importance for health care and public health. Xie (p. 72) uses recurrent neural networks to find e-cigarette adverse events in social media posts, while Vigo (p. 88) describes a method to collect seasonal allergy symptoms for the British population. New types of structured data and new ways to integrate them are also continuously being produced: Doostparasti (p. 99) describes a novel approach for integrating -omics data to enhance phenotype classification performance, and Yu (p. 54) introduces a phenotyping algorithm that does not depend on expert-labeled observations. The articles listed above are only a few examples of the scope of informatics activities covered in JAMIA. Starting with this January issue, readers will be able to easily group articles into themes based on technologies used or application areas. This grouping is made possible by JAMIA’s change in frequency and format (to monthly online), which will allow for more frequent indexing. Readers will be able to compare approaches and discover solutions that are best suited to their problems. Stay tuned for additional data science articles in future monthly issues, as well as articles focused on clinical informatics systems (including clinical decision support), clinical research systems, translational bioinformatics, global public health informatics, and many other subfields of informatics that help us, through information technology, understand and address human health and disease.

This issue of JAMIA focuses on two of the most important and complex sources of data in informatics: electronic health records (EHRs) and patient-reported data. The overwhelming and relatively rapid adoption of EHRs in the US in the past decade has motivated numerous investigations related to the accuracy, completeness, and standardization of clinical data, as well as a wide spectrum of applications targeting patient safety, clinical decision support, and optimized workflows. It is interesting to reflect on how early reports of EHR interventions (such as computerized provider order entry, pioneered in homegrown systems now replaced by commercial products) have paved the way for a wide variety of applications that make use of EHRs not only for point-of-care decision support, but also for quality improvement, population health management, and research. Additionally, evolving technology that resulted in user-friendly interfaces has allowed the development of apps and other patient-facing systems that, together with EHRs, help portray a more complete picture of individual health and disease progression than was possible just a few years ago.

Biomedical informatics has evolved so rapidly in the past two decades that the emergence of subspecialization areas such as clinical research informatics was natural. The scope of clinical research informatics is also broad, and it will not be surprising if in a couple of years it is further subspecialized. In this issue of JAMIA, we bring some of the best work in clinical research informatics, as well as related articles on a technology that is behind many clinical applications: natural language processing (NLP). Clinical research informatics applications typically involve informatics approaches to collect, process, analyze, and display health care and biomedical data for research. Adams (p. 295) describes a system for collecting mechanical ventilator waveform data for research, Brokamp (p. 309) proposes an approach for geocoding and characterizing community and environmental exposures, and Hripcsak (p. 289) proposes a way to conduct high-fidelity phenotyping from electronic health records (EHRs). An analysis by Boland (p. 275) uncovers exposures that are responsible for birth season–disease effects, and a method proposed by Chen (p. 345) targets bias reduction in association studies that use EHR data. As clinical text is still rich in narrative text, NLP is frequently used to preprocess data for research. Ramanathan (p. 321) uses hierarchical attention networks to extract information from cancer pathology reports, and Sohn (p. 353) studies variations in clinical documentation due to issues in NLP system portability. The need for resource-consuming customization of NLP systems to increase portability is a known challenge. To address this challenge, Soysal (p. 331) presents a toolkit for efficiently building customized clinical NLP pipelines. However, not all data are ready for research, either because of their quality or because of privacy or other concerns that prevent their public disclosure. Shang (p. 248) proposes a framework for evaluating data suitability for observational studies, while Smith (p. 224) assesses the quality of administrative data for research. Weng (p. 239) evaluates the representativeness of eligible patients for clinical trials on type 2 diabetes. Additionally, data need to be standardized, and the various standards need to be reconciled. Utilization of an ontology to unify two standards is demonstrated by Campbell (p. 259). Once data are ready for dissemination, they can be reused in new exploratory or confirmatory analyses: Xu (p. 300) describes an open source discovery index for finding biomedical datasets, and Johnson (p. 337) reports on the results of a user needs analysis and usability assessment for a data search engine. This issue of JAMIA also presents articles related to participant privacy and data security in the context of research applications. O’Keefe (p. 315) utilizes a checklist approach to assess privacy risks in publications about population health, and Harle (p. 360) reports on patient preferences regarding an e-consent application for research involving EHRs. However, many times it is not practical or necessary to operate on real records: Walonoski (p. 230) describes a tool for generating synthetic EHRs. Finally, Peisert (p. 267) describes a health science secure network design for big data analysis. Based on the research published in this issue, we predict that clinical research informatics will continue to be a growing subspecialty of biomedical informatics. We hope to see an increasing trend toward systems that can be embedded in patient- and clinician-facing systems for real-time utilization. Stay tuned to JAMIA’s next issues, which will be focused on patient-centered systems and clinical decision support systems.

Never before has the adoption of networked communication devices been so high across healthcare provider and patient communities. It is thus no surprise that JAMIA receives an increasing number of outstanding submissions in this area, which is the focus of this issue of the journal. Patient portals and other direct-to-consumer informatics applications have become increasingly available in the last decade. Many of these applications can be considered health interventions and thus deserve to be evaluated in light of their costs and benefits. In this issue of JAMIA, Wolff (p. 408) provides an environmental scan of shared access to a patient portal, Grossman (p. 370) lists the main recommendations from early adopters of acute care patient portals, Dumitrascu (p. 447) associates portal use and hospital outcomes, Giardina (p. 440) evaluates patient perceptions of receiving test results via portals, and Lee (p. 413) discusses the need to modernize current recommendations for electronic communication between patients and clinicians. In addition to patient portals, the consumers of healthcare services are now also actively enhancing data collected by healthcare providers, rating services and their providers, and using text messaging as a means of communication with providers and researchers. Cantor (p. 419) proposes open data to measure social determinants of health. Daskivich (p. 401) describes how online physician ratings fail to predict actual performance on measures of quality, value, and peer-review. Moon (p. 423) provides a perspective on ethical issues in texting for HIV care in Mozambique, and Giguere (p. 393) reports on how concerned participants are about privacy and security when using text messaging services to report on product adherence in a rectal microbicide trial. Informatics interventions in public health, point of care, and educational settings are also featured in this issue of JAMIA. McClung (p. 435) describes the use of mobile applications in mass vaccination campaigns, and Bekemeier (p. 428) discusses the generation of standard public health services data and evidence for decision making. Van der Veen (p. 385) reports on the association between workarounds and medication administration errors in bar-code-assisted medication administration in hospitals. Fernando (p. 380) lists lessons learned from piloting mHealth informatics practice curriculum into a medical elective. It is gratifying to see the evolution of healthcare into a more inclusive, participatory partnership among providers, patients, and their caregivers. It is equally exciting to see that informatics has played a critical role in enabling new forms of communication and promoting easier access to information that can change the course of an individual’s health history. In upcoming issues, JAMIA will continue to bring to our readers the best scholarly work in health informatics.

In recent years, we have witnessed a rise in the use of informatics within healthcare settings. The development of applications such as clinical decision support (CDS) systems ensure that clinicians and allied health personnel are armed with the knowledge acquired and curated by the health sciences community. The impact of these applications in practice are directly measured by the foundational informatics methods and tools employed within the applications. In this issue of JAMIA, we present articles focused on clinical informatics and interventions that directly affect patient care. The use of CDS systems in healthcare is becoming more prominent; hence the importance of CDS systems’ design, functionality, and accuracy. Miller (p. 585) reviews functional requirements and design of CDS systems. A systematic review by Varghese (p. 593) analyzes the recent literature on CDS systems with regards to patient outcomes in the inpatient setting, while Austrian (p. 523) shows the impact of a CDS system that monitors sepsis and related mortality and length of stay in emergency care. Relevant to the current opioid crisis, an article by Finley (p. 515) reports on the feasibility of a decision support system for prescription monitoring in a military healthcare setting. Dagliati (p. 538) presents a dashboard system to support diabetes care, while Li (p. 548) and Hodge (p. 603) study electronic problem lists. Horsky (p. 465) shows how improved design will result in higher accuracy for complex medication reconciliation processes. Effective CDS systems also rely on updated knowledge bases. Lacson (p. 507) discusses the curation and contents of evidence libraries, and Van Allen (p. 458) contrasts interactive and static reports for cancer genomics. Tamblyn (p. 482) describes how patient safety can be improved with a computer-assisted tool for medication reconciliation that integrates population-based community data. Brown (p. 568) uses machine learning techniques for automated neuro-MRI protocol selection, and Wu (p. 530) describes a general-purpose system to extract data from clinical notes. As the name implies, decision support systems assist clinicians in making decisions, but sometimes automated systems fail to consider important contextual information or simply make mistakes of their own. Nanji (p. 476) describes the characteristics of medication-related CDS overrides. Wright (p. 496) analyzes CDS system malfunctions, and Stone (p. 564) presents two cases of unintended adverse consequences of a CDS system. Tolley (p. 575) reviews the factors that contribute to medication errors when clinicians use computerized order entry in pediatrics, and Ni (p. 555) describes a system for real-time medication administration error detection in a neonatal ICU. Finally, there is a lot of discussion in our community on the costs and benefits of EHRs. Wright (p. 572) shows that the transition to a new electronic health record system does not affect hospital bond ratings. JAMIA has been featuring the best work in clinical informatics since its inception. Much has changed in the past two-and-a-half decades, and the spectrum of topics we publish has expanded to reflect the expansion of biomedical informatics. As always, we continue to select the most original work and to value both foundations and applications of informatics that lead to better patient care, disease prevention, and scientific discoveries.

The adoption of electronic health records (EHRs) in the U.S. was greatly accelerated by the HITECH “meaningful use” (MU) regulations, which require that all healthcare institutions either implement these systems or pay penalties. Accordingly, in the past decade, JAMIA has received many more manuscripts focused on the experiences from selecting, customizing, implementing, and evaluating the use of EHRs in various clinical settings, and on exchanging information contained in the EHRs across healthcare institutions. This issue of JAMIA focuses on the latest experiences of using and evaluating EHRs and a health information exchange (HIE). Understanding where EHRs can be improved is an important factor in their continued adoption. Goss (p. 661) proposes a value set for adverse reaction documentation, and Kannampallil (p. 739) describes the association between issuing medication orders for the wrong patient and the number of open charts in a clinician’s monitor. Wright (p. 709) explains how free-text electronic prescriptions can result in communication failure, and Percha (p. 679) proposes an expansion of the radiology lexicon using contextual patterns contained in radiology reports. The importance of customizing EHR systems to clinical workflows in different settings has been extensively documented in the biomedical informatics literature. Veinot (p. 746) describes a process to model clinical information interactions in primary care, and Ramelson (p. 715) reports on an enhanced referral management system. Krousel-Wood (p. 618) compares healthcare provider perceptions on transitioning from a small EHR system into a comprehensive commercial system. Price-Haywood (p. 702) analyzes dose effects of communication between patients and the care team via secure portal messaging, and Reading (p. 759) reports on the converging and diverging needs among patients and providers who are using patient-generated health data. In addition to their role of assisting clinicians in documenting their activities and using the information to provide care, EHR systems have an important role for healthcare quality, management, and biomedical research. Cho (p. 730) reports on how specific eMeasurements can be automatically populated from EHR systems. Holman (p. 694) describes how MU can result in both benefits and burdens for family physicians, Holmgren (p. 654) assesses the relationship between specific EHR systems and MU performance. Additionally, Casucci (p. 670) uses Medicaid data to study effects of chronic disease combinations on 30-day hospital readmissions, an important healthcare quality measure. Fraser (p. 627) discusses barriers to the success of an electronic pharmacovigilance system, and Baron (p. 645) proposes an approach for imputing multi-analyte values in longitudinal clinical data for use in machine learning systems. In an era where healthcare data integration becomes the norm, several HIE approaches are being pursued across counties, states, and nations. Motulsky (p. 722) analyzes usage and accuracy of medication data from HIE in Quebec, Canada, Schmit (p. 635) describes how differences in state laws can adversely impact or facilitate this type of exchange, and Klapman (p. 686) reports on emergency care clinicians’ experiences of HIE across five countries. As this JAMIA issue illustrates, we live in an exciting time in the evolution of EHR systems and HIE. Never before has adoption been so high with an understanding of the multiple aspects of their use and usefulness approached from so many different perspectives.

This issue of JAMIA is focused on informatics applications to enhance patient safety. This is one of the most important, yet underemphasized, aspects of the informatics curriculum across the country. Although media attention occasionally concentrates on what can go wrong when information systems are employed in practice, there is also much to say on what might go wrong if information systems were not employed. Additionally, a proper amount of standardization of clinical practices can elevate sub-optimal care to an acceptable level, often reducing cost and patient suffering as a result. Decision support for medication prescribing and dispensing has always been one of the most direct ways for information systems to promote patient safety. Whalen (p. 849) reports on lessons learned in a pediatrics hospital from a transition to a new electronic health record (EHR) system, and Walsh (p. 911) studies the accuracy of the medication list in the EHR and its implications for care, research, and improvement. Cheng (p. 873) shows how using drug knowledgebase information to distinguish between commonly confused drugs can prevent errors, and Vajravelu (p. 780) proposes a new algorithm to analyze multiple pharmacologic exposures using EHR data. Additionally, Samwald (p. 895) shares the experience of implementing pharmacogenomics decision support across seven European countries. The domains in which information systems can improve patient safety are numerous. Waters (p. 901) studies current use, interest, and perceived usability of clinical pathways for primary care, while Sittig (p. 915) describes the levels of adherence to recommended EHR safety practices across eight healthcare organizations. EHR-based intervention and reports on several safety topics are also presented in this issue of JAMIA: Ray (p. 863) uses statistical anomaly detection models for decision support system malfunctions, Chen (p. 790) analyzes interaction patterns of trauma providers that are associated with increased patients’ lengths of stay in the hospital, while Vahdat (p. 827) reports on a simulation study of the effects of EHR implementation on timeliness of care in a dermatology clinic. Berger (p. 833) integrates physical abuse measures into a pediatric clinical decision support system, while Meyer (p. 841) evaluates a mobile application to improve clinical laboratory test ordering. The applications and algorithms described in this issue of JAMIA would be hard to implement without standardization of terminologies, ontologies, and foundational research in natural language processing and information retrieval. Examples of advances in these areas are also featured: Cuzzola (p. 819) links UMLS to DBpedia to promote knowledge discovery, Wang (p. 809) describes efforts involving RxNorm that are leading to a normalized clinical drug knowledge base in China, Vreeman (p. 886) presents a unified terminology for radiology procedures (the “LOINC RSNA Radiology Playbook”), and Blosnich (p. 907) shows how it is possible to use EHR-based clinician text notes to validate transgender-related ICD codes. Additional articles describe approaches that enable a variety of information systems: Mei (p. 800) describes an interactive medical word sense disambiguation method, Kilicoglu (p. 856) reports on the results of automatic recognition of self-acknowledged limitations in the clinical research literature, and Baladron (p. 774) proposes a tool for filtering PubMed search results by sample size. JAMIA continues to publish a combination of application and foundational articles that allows our readers to stay abreast with the best developments in the field. Patient safety is an important topic that has been in the informatics portfolio since its start and for which novel solutions continue to emerge. New topics are also continuously enlarging the informatics portfolio. Stay tuned for the August issue that highlights articles on informatics applications focused on patients, their family and friends, and the expanding scope of our field.

It is hard to believe that, just a decade ago, the technology of smart phones and the patient-centered healthcare movement were just starting to take off, and many of their current instantiations were not spelled out in visions of a connected healthcare system. Accordingly, informatics has evolved and expanded so fast in the past ten years that it is not a surprise that the volume, quality, and interest in scholarly articles describing the many facets of including providers and patients, as well as patients’ family and friends, in healthcare has increased so much that we are able to dedicate a full issue of JAMIA to research and applications in these areas. Kishore (p. 931) describes a model of the effects of online informational and emotional support on self-care behavior of HIV patients, Cheung (p. 955) evaluates a recommender app for measuring longitudinal user engagement in apps for depression and anxiety treatment, and Utrankar (p. 976) reports on technology use and preferences in supporting clinical practice guideline awareness and adherence in individuals with sickle cell disease. Additionally, Taylor (p. 989) describes the role of family and friends in helping older adults manage personal health information, while Sharko (p. 1008) describes the unique privacy needs of adolescent patients and the resulting complexity of the decision-making process. Different methods have been borrowed from various disciplines over time to fill the needs of provider-, patient- and other caregiver-facing applications. Bautista (p. 1018) reports on a psychometric evaluation of a scale to measure nurses’ use of smartphones for work purposes, Reese (p. 1026) uses card sorting methods to elicit expert knowledge in an ICU setting, and Pandolfe (p. 1047) proposes an architecture for a medication reconciliation application that aims at increasing patient activation and education. While the intent of information systems is always to improve care and promote health, positive and negative consequences have been reported in the literature. In this issue of JAMIA, Nouri (p. 1089) systematically reviews criteria for assessing the quality of mHealth apps, Veinot (p. 1080) discusses how informatics interventions can worsen inequality, Meyerhoefer (p. 1054) reports on provider and patient satisfaction with the integration of ambulatory and hospital EHR systems, and Plante (p. 1074) reveals trends in user ratings and reviews of a blood pressure-measuring smartphone app. Increased data sharing of clinical data, partly due to the popularity of patient-facing applications and a realization that faster biomedical discoveries may happen with the use of “big data,” also brings important issues related to ethics and how information is relayed to users. Stahl (p. 1102) discusses the role of ethics in data governance of a large neuro-ICT project, Tao (p. 1036) discusses the effects of graphical formats of self-monitoring test results for consumers, Karpefors (p. 1069) proposes a visual summary of the incidence, significance, and temporal aspects of adverse events in clinical trials, and Wright (p. 1064) describes the development and evaluation of a user interface for reviewing clinical microbiology results. The increased availability of data for studies has yet to be paired with increased transparency of these data and analytical methods to allow easy reproduction of results. Coiera (p. 963) discusses whether health informatics suffers from replication issues. Particularly in the area of predictive models, it is important that data and methods be available to assess reproducibility and generalizability to other data sets. Reps (p. 969) proposes a standardized framework to generate and evaluate patient-level prediction models using observational healthcare data. This issue of JAMIA also includes several other articles in the area of predictive modeling: Abbas (p. 1000) describes a machine learning approach for early detection of autism, Goldstein (p. 924) proposes a model to predict ambulatory no-shows across different specialties and clinics, and Kalpathy-Cramer (p. 945) reports on a distributed deep learning networks for medical imaging data. As the articles in this issue of JAMIA exemplify, the discipline of informatics has expanded into exciting new territory in which the intersection is no longer just of biomedical sciences and computer science, but rather of a variety of other domains involved in understanding how information systems can help improve healthcare, disease prevention and promote healthier behaviors. JAMIA has kept aligned with this expanded scope and will continue to bring the best scholarly work in informatics to our readers.

This issue of JAMIA focuses on various uses of electronic health records (EHRs) to provide better care for individuals and assist caregivers within and across institutions. EHRs have been used for clinical decision support (CDSS) for a long time, but their wide adoption in the United States was achieved only in the past decade. Several informatics studies based on EHR data have followed since then. In this issue of JAMIA, we present articles related to the use of CDSS to improve patient care. Grundmeier (p. 1160) evaluates a predictive model for site infections, Sperl-Hillen (p. 1137) reports on a randomized trial of a CDSS to reduce cardiovascular risk, Cheung (p. 1202) shows how the risk for a specific cardiovascular condition was reduced in a major academic medical center upon implementation of a CDSS, Slight (p. 1183) shows how a CDSS can promote safety in the form of aversion of adverse events related to medications, Ramirez (p. 1167) describes a CDSS for diabetes medication dose adjustments, and finally Gamble (p. 1240) systematically reviews availability and clinical drug information coverage in machine-readable format, without which the implementation of CDSS is not practical. EHRs can be utilized to improve patient care in other ways as well. Mummadi (p. 1228) reports on the effect of displaying price information to clinicians using a computerized order entry system, while Laranjo (p. 1248) systematically reviews conversational agents in healthcare. An increasing demand towards transparency of EHR usage can also explain the increasing adoption of OpenNotes described by Fossa (p. 1153). While widely adopted and playing important roles in supporting decision making and safety in healthcare, it is important to note that EHRs are not a panacea for a highly fragmented and somewhat cost-ineffective healthcare system: Krumholtz (p. 1218) provides a perspective on the promise of EHRs to benefit patients, Ratwani (p. 1197) assesses usability and safety of EHR systems, and Gilmore-Bykovski (p. 1206) addresses the lack of structure in clinical documentation of cognitive and behavior dysfunctions. The association of intra-hospital sharing of EHR information and health system organizational structure is reported by Holmgren (p. 1147). This issue of JAMIA also includes studies on the sharing of health information across healthcare institutions. The benefits and gaps of health information exchange (HIE) across institutions are described by Menachemi (p. 1259) and Everson (p. 1114), respectively. Vest (p. 1189) reports that adoption of intra-hospital HIE is negatively associated with inter-hospital HIE. Beskow (p. 1122) describes patients’ perspectives on the use of EHRs that results in their being contacted by researchers. It is important to understand that an EHR is not the only source of health data that can impact a patient’s care: this issue of JAMIA provides examples of the impact of online support groups, wearable technology, mobile-based pediatric clinical decision support, and telehealth as reported by Friedman (p. 1130), Bumham (p. 1221), McCulloh (p. 1175), and Gurbeta (p. 1213), respectively. The articles we present here are prime examples of the diversity of our field and the many practical ways in which informatics can help promote health and mitigate disease. At the same time, it is important that we experiment with other novel applications that will be mainstream in the near future. Stay tuned for the October issue in which we focus on clinical and translational research informatics.

Pattern recognition using machine learning methods is an area that exploded in recent years, given the increasing amount of available data. JAMIA has published an increased number of articles in this area in the past few years. In this issue, Zhang (p. 1351) reports on a data resource for sleep research, Feller (p. 1366) proposes a visual analytics approach for pattern recognition in personal health records, while Xiao (p. 1419) reports on a systematic review of deep learning models applied to electronic health record (EHR) data. Machine learning models to detect pulmonary nodules in CT scans are described by Grutzemacher (p. 1301), while models to predict adverse drug events are reported by Davazdahema (p. 1311) and Mower (p. 1339). Albers (p. 1392) provides a broad perspective on mechanistic machine learning using physiologic knowledge, while Pencina (p. 1273) simulates models to predict incremental value of biomarkers. Developing new approaches to facilitate automation of clinical research is another area in which informatics has evolved considerably in the past few years. In particular, biomedical natural language processing and other methods to structure narrative text and voice recordings have motivated informatics research. Sarker (p. 1274) reports on systems for medication-related text classification, Zamjahn (p. 1284) proposes a method to streamline the evaluation of video recordings, and Parr (p. 1292) describes the automated mapping of laboratory tests to standardized codes in EHRs. Algorithms and tools for processing and linking EHRs are reported by Hoopes (p. 1322) and Klann (p. 1331). Sinnott (p. 1359) proposes a method to improve the power of genetic association studies, while Aronson (p. 1375) describes how the eMERGE consortium established data flows that are particularly relevant to genomic medicine. Some new modalities of health data are not yet integrated into EHRs though are increasingly being used in research. Streaming data from sensors, for example, in the form of continuous heart rate, activity and location tracking is becoming more common. Donevant (p. 1407) reviews the literature on mHealth studies, Speier (p. 1351) evaluates utility of activity-related data, while Goldenholtz (p. 1402) proposes a way to utilize location data without compromising privacy. Regardless of their area of sub-specialization, informaticians worldwide now have a wealth of opportunities ahead that were not available to their predecessors. The rapid accumulation of data and knowledge due to new techniques and approaches brings opportunities as well as challenges related to the protection of privacy, inadequacy of computational resources to store, process and integrate large amounts of multi-modality data, antiquated regulatory frameworks, and a relatively low number of trained professionals. However, recognition of the value informatics brings to clinical practice and research has been facilitated by general understanding of the value of data science and artificial intelligence that is now pervasive in our daily lives. These are exciting times and we have a unique opportunity to make a difference towards better health for all.

This issue of JAMIA focuses on the use of patient- and provider-facing applications for medication reconciliation and related topics. It also presents original research on intended and non-intended consequences of clinical decision support systems, algorithms and tools for “phenotyping” from electronic health records (EHRs), as well as tools and approaches to promote clinical research involving EHRs, the biomedical literature, and patient-provider communications. Yin (p. 1444) describes how online portal communications between breast cancer patients and physicians were used to determine medication discontinuation, Cronin (p. 1470) describes patient and clinician views on a patient-reported outcomes portal, and Yang (p. 1516) focuses on discontinuation of new electronic prescriptions. Maryen (p. 1488) and Prey (p. 1460) report on patient-facing applications that use tablets and a web site, respectively, for medication reconciliation. Clinical decision support systems’ impact on clinical documentation and outcomes is also reported in this issue of the journal. Powers (p. 1556) systematically reviews the literature on the efficacy and unintended consequences of hard-stop alerts in EHRs, Wright (p. 1552) describes three cases of decision support system malfunction, Zhang (p. 1547) discusses how to develop and maintain decision support systems using clinical knowledge and machine learning, while Singh (p. 1481) uses machine learning for psychiatric patient triaging. Lacson (p. 1507) studies human factors leading to diagnostic errors in radiology, and Orestein (p. 1501) describes the influence of simulation on EHR use patterns among pediatric residents. The detail in documenting care via EHRs is highly variable across clinicians and institutions. Adelman (p. 1534) describes the use of EHRs to report inpatient stroke quality of care, while Rutkowski (p. 1524) shows that the number of diagnoses codes in inpatient discharge notes is associated with counts and rates of birth defects. Identification and validation of case definitions for medical conditions is systematically reviewed by McBrien (p. 1567), and reports on the portability of a phenotyping algorithm across institutions and EHR systems by Pacheco (p. 1540). While phenotypes are critical for research, two other modalities of data serve as fundamental companions: environmental exposures, illustrated in a framework for development and validation of prenatal exposures by Boland (p. 1432) and genetics, illustrated by Zhou’s (p. 1452) research on identifying symptom candidate genes via network embedding. Finally, AMIA’s 2018 code of professional conduct and ethics (p. 1579) is published in this issue. Never before has our specialty been so involved in complex privacy protections for clinical and research data. The code contains critical information on various aspects of our profession.

We have been highlighting the expanded quantity and quality of articles published in JAMIA for the past few years. They reflect how informatics has grown from a relatively small and lesser-known healthcare and biomedical science specialty, when the journal started 25 years ago, into a well-recognized discipline with distinct foundations and applications that are relevant to various domain areas. The 2018 closing issue of JAMIA exemplifies the breadth and depth of informatics: it presents applications in global and public health (p. 1608, p. 1586), healthcare (p. 1600, p. 1634), and behavioral science (p. 1675), and it describes foundational work in vocabulary mapping (p. 1618), privacy protection of patient records (p. 1593), statistical methods for longitudinal data (p. 1669), and a study on the integrity of clinical information in diagnostic imaging orders (p. 1651). Articles in this issue also review how mobile health applications can be leveraged for citizen science (p. 1685), discuss factors that are important for patient portal engagement (p. 1626), and show how deep neural networks can be used to provide expert-level sleep scoring (p. 1643). Finally, AMIA’s list of core competencies (to be achieved as a result of health informatics education) is presented. JAMIA’s articles represent the best work in our field, and it is no surprise that they have been featured in the lay press, as well as in popular “Year in Review” panels at AMIA conferences. In 2018, we provided our readers with an assortment of established and emerging research topics, as well as authoritative reviews and perspectives from informaticians around the world. Reporting on the growth of our discipline while affording new authors the opportunity to feature their best work together with established senior professionals in our field is an important function of JAMIA. We wish the entire JAMIA family (readers, authors, reviewers, editorial and production teams) happy holidays, and congratulate everyone for the impressive achievements in 2018. We look forward to 2019, in which a new and highly qualified editorial team will guide AMIA’s flagship publication to new heights.

In my first editorial, I stated my goals to implement changes that would disseminate JAMIA to a broader audience, expand its contents, and optimize its management.1 In my final editorial 8 years later, I share with you my journey toward accomplishing these goals. In 1994, Bill Stead organized a group of senior American Medical Informatics Association (AMIA) members to found the Journal of the American Medical Informatics Association (JAMIA).2 Bill, the founding editor-in-chief from 1994 to 2003, is one of the pioneers of the biomedical informatics field and a recognized leader in the academic medical center community. As its first editor, Bill set JAMIA’s original vision and mission; he stepped down after a decade of service and at a stage when the journal was a recognized asset to AMIA and the informatics community in general.3 Randy Miller (editor-in-chief, 2004–2010) succeeded him, further advancing JAMIA’s mission and solidifying its status as AMIA’s flagship publication. Eight years ago, when Randy passed me the JAMIA torch, he gallantly wrote “All’s well that ends well for JAMIA editors,”4 referring to the outcome of the search for his successor. I learned from Randy to be attentive to every single detail. (A hallmark of an editor-in-chief, in addition to setting up the vision and strategy for the journal, is the search for perfection. I found the best role models for this in Bill and Randy.) It was an honor to be selected for the role. I had been an associate editor for a few years, but I had not planned to be the editor when I first started serving as a reviewer. (When I was in graduate school, Ted Shortliffe taught me how to review a biomedical informatics article, and Mark Musen taught me how to write one. Bob Greenes ensured I did so while I was junior faculty at the Decision Systems Group, Brigham and Women’s Hospital, Harvard Medical School. I am lucky to have had their guidance and support for so many years.) I was ecstatic and somewhat surprised to have been selected, particularly because English is not my native language and I still had much to learn about editorial processes. I immediately accepted, having little time to reflect on what it truly meant to steer AMIA’s flagship publication, and how critical this role was for so many readers, authors, reviewers, and JAMIA’s editorial team. If I had thought too much about it, it might have been overwhelming, but being somewhat naïve turned out to be an asset: I did not think at any single moment that I would not be able to do the job; I just did not know how much of my time it would consume, which innovations I would bring forward, and which barriers I would need to overcome. As with similar professional or personal challenges, this was one to be attacked head-on, with confidence, a knowledge-seeking attitude, humility, and pride. In a “trial by fire,” I learned how to deal with extraordinary situations that took a lot of unexpected time, such as response to plagiarism, accusations of delivering biased or uninformed reviews, discovery of hidden conflicts of interest, authorship disputes, retractions, corrections, attempts to influence editorial decisions, and threats of lawsuits and retaliations. Fortunately, there are many sources of knowledge and support for many of these items, and the publisher’s and AMIA’s staff were always ready to help, so these temporary problems were overcome quickly. I learned to be efficient with time so my daytime job would not suffer from my dedication to JAMIA, and continued to improve my own writing for clarity, grammar, and style. I had the invaluable help from a technical editor: Dr Michele Day has provided insightful requests for clarification, suggested word replacements, and noticed lack of flow from paragraph to paragraph for most of the 60-plus “highlight” pieces and editorials. She taught me how to write better English (which may have resulted in better writing in Portuguese, too, but the hypothesis remains untested). I started the online-only special issues of JAMIA, and later helped the journal “go green” at the same time we transitioned from a bimonthly to a monthly publication. Another innovation I introduced was the JAMIA Journal Club. The rationale was simple and timely. When I became JAMIA’s editor-in-chief, I had recently started a new biomedical informatics program at the University of California, San Diego, after spending many years as a faculty member in Boston’s Harvard–Massachusetts Institute of Technology system. Given the small size of our new program, I missed meeting with various colleagues in journal clubs and seminars. Additionally, I thought JAMIA could benefit from live presentations by authors of outstanding papers, and the virtual journal clubs would provide an open forum to discuss the latest informatics innovations, especially for informaticians who hold positions in institutions without training programs or academic informatics groups. With a live (and recorded) journal club, JAMIA could also be known to a wider audience that could “listen to” instead of read an article. For these reasons, we started the monthly JAMIA Journal Club in 2012. The JAMIA Journal Club has been accomplishing its goals and is still ongoing because of the work of the student editorial board,5 which was another JAMIA innovation later replicated by other journals. However, this one I did not invent: I encouraged it to continue because it was a brilliant idea. Trainees could witness the review process as reviewers under the supervision of an associate editor, and understand the statistics and trends for the journal, thus cultivating a new generation of editors. I thank all readers and authors of JAMIA, the AMIA staff, and publishers. I am especially thankful to the associate editors who served as student editorial board organizers, our current associate editors for their input in the directions of the journal and selection of peers for the editorial board, and all associate editors that have rotated in the position the past 8 years, including associate guest editors of special issues (there were 44 in total). They brought new themes to JAMIA, as well as new authors and perspectives that enriched our field. Their contributions helped JAMIA continue to stand out at a time when a plethora of new informatics-related dissemination venues emerged and there was great concern about the sustainability of traditional scientific journals.6 I will not name everyone here, as I am afraid of making a critical omission, but I would like to ask that our community keeps recognizing their efforts. The editorial team is a secret sauce in running the journal: it is composed of AMIA members voted by peers as a result of a process that has improved over many years. The vote by the incumbent associate editorial team recognizes informaticians for being outstanding experts in their respective areas, as well as for their ability to review manuscripts fairly, insightfully, constructively, and in a timely manner; our authors and readers deserve no less. The editorial team helps ensure that our service to the scientific community is completed with utmost integrity and that it is inclusive, impactful, and impeccable. A key function of the editor-in-chief is to organize the team to achieve this goal. I trust that we were very effective at that, as can be shown by conventional and nonconventional measures of journal success; our team processed over 10 000 articles in the past 8 years, and we lowered the average and median times to first decision to under 30 days. Our articles have been read by millions of people worldwide, and we have received submissions from over 90 countries. We had millions of downloads and views, and the skyrocketing number of citations reflects the dissemination of informatics across many other disciplines. We achieved all this because we “stood on the shoulders of giants,” who made the journal an invaluable asset to AMIA and the informatics community at large, and because we kept improving on this legacy. At the end of my second term, one thing is certain: time flies, whether one is having fun or not. In this case, I had lots of fun, and with the same blend of sadness, happiness, accomplishment, and anxiety I felt when I left my oldest son for the first time in daycare or when my youngest son departed for college, and I am passing the JAMIA torch to the new editor-in-chief, Sue Bakken. It is reassuring to know that the journal will be in great hands, as she is exceedingly qualified and will take JAMIA to new heights. I thank you all for the unique opportunity to serve as your editor-in-chief for 8 productive and enjoyable years. Looking back, it was a lot of work, a lot of rewards, but, most importantly, a lot learned from people with so many different backgrounds, aspirations, beliefs, and goals.

Motivation: We introduce PRINCESS, a privacy-preserving international collaboration framework for analyzing rare disease genetic data that are distributed across different continents. PRINCESS leverages Software Guard Extensions (SGX) and hardware for trustworthy computation. Unlike a traditional international collaboration model, where individual-level patient DNA are physically centralized at a single site, PRINCESS performs a secure and distributed computation over encrypted data, fulfilling institutional policies and regulations for protected health information. Results: To demonstrate PRINCESS' performance and feasibility, we conducted a family-based allelic association study for Kawasaki Disease, with data hosted in three different continents. The experimental results show that PRINCESS provides secure and accurate analyses much faster than alternative solutions, such as homomorphic encryption and garbled circuits (over 40 000× faster). Availability and Implementation: https://github.com/achenfengb/PRINCESS_opensource. Contact: [email protected]. Supplementary information: Supplementary data are available at Bioinformatics online.

MOTIVATION: Inappropriate disclosure of human genomes may put the privacy of study subjects and of their family members at risk. Existing privacy-preserving mechanisms for Genome-Wide Association Studies (GWAS) mainly focus on protecting individual information in case-control studies. Protecting privacy in family-based studies is more difficult. The transmission disequilibrium test (TDT) is a powerful family-based association test employed in many rare disease studies. It gathers information about families (most frequently involving parents, affected children and their siblings). It is important to develop privacy-preserving approaches to disclose TDT statistics with a guarantee that the risk of family 're-identification' stays below a pre-specified risk threshold. 'Re-identification' in this context means that an attacker can infer that the presence of a family in a study. METHODS: In the context of protecting family-level privacy, we developed and evaluated a suite of differentially private (DP) mechanisms for TDT. They include Laplace mechanisms based on the TDT test statistic, P-values, projected P-values and exponential mechanisms based on the TDT test statistic and the shortest Hamming distance (SHD) score. RESULTS: Using simulation studies with a small cohort and a large one, we showed that that the exponential mechanism based on the SHD score preserves the highest utility and privacy among all proposed DP methods. We provide a guideline on applying our DP TDT in a real dataset in analyzing Kawasaki disease with 187 families and 906 SNPs. There are some limitations, including: (1) the performance of our implementation is slow for real-time results generation and (2) handling missing data is still challenging. AVAILABILITY AND IMPLEMENTATION: The software dpTDT is available in https://github.com/mwgrassgreen/dpTDT. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Background: Implementation of patient preferences for use of electronic health records for research has been traditionally limited to identifiable data. Tiered e-consent for use of de-identified data has traditionally been deemed unnecessary or impractical for implementation in clinical settings. Methods: We developed a web-based tiered informed consent tool called informed consent for clinical data and bio-sample use for research (iCONCUR) that honors granular patient preferences for use of electronic health record data in research. We piloted this tool in 4 outpatient clinics of an academic medical center. Results: Of patients offered access to iCONCUR, 394 agreed to participate in this study, among whom 126 patients accessed the website to modify their records according to data category and data recipient. The majority consented to share most of their data and specimens with researchers. Willingness to share was greater among participants from an Human Immunodeficiency Virus (HIV) clinic than those from internal medicine clinics. The number of items declined was higher for for-profit institution recipients. Overall, participants were most willing to share demographics and body measurements and least willing to share family history and financial data. Participants indicated that having granular choices for data sharing was appropriate, and that they liked being informed about who was using their data for what purposes, as well as about outcomes of the research. Conclusion: This study suggests that a tiered electronic informed consent system is a workable solution that respects patient preferences, increases satisfaction, and does not significantly affect participation in research.

OBJECTIVES: To introduce blockchain technologies, including their benefits, pitfalls, and the latest applications, to the biomedical and health care domains. TARGET AUDIENCE: Biomedical and health care informatics researchers who would like to learn about blockchain technologies and their applications in the biomedical/health care domains. SCOPE: The covered topics include: (1) introduction to the famous Bitcoin crypto-currency and the underlying blockchain technology; (2) features of blockchain; (3) review of alternative blockchain technologies; (4) emerging nonfinancial distributed ledger technologies and applications; (5) benefits of blockchain for biomedical/health care applications when compared to traditional distributed databases; (6) overview of the latest biomedical/health care applications of blockchain technologies; and (7) discussion of the potential challenges and proposed solutions of adopting blockchain technologies in biomedical/health care domains.

JAMIA has been a premier venue for publication of scholarly work on clinical decision support systems since its inception. With an initial emphasis on knowledge-based systems, i.e., encoding of clinicians’ knowledge into rules that would trigger alerts and reminders, JAMIA evolved into a phase of an increasing number of articles reporting on the use of EHR “big data” to build and validate predictive models that recognize patterns in large amounts of data to derive actionable recommendations for clinicians. Interestingly, this is happening at the same time that health information technology (HIT) is still evolving, data quality in EHRs is being improved, and health information exchange (HIE) continues to be evaluated for cost effectiveness. Key recommendations for HIT optimization are proposed by Chresswell (p. 186). Additionally, Wright (p. 192) advocates for more testing of EHR systems, Alexander (p. 69) focuses on IT in nursing homes, and Kharrazi (p. 2) promotes an agenda for population health informatics. HIT use by office-based physicians and healthcare reform programs is described in a study by Heisey-Grove (p. 133). Improving the quality of EHRs is addressed by Van der Bij (p. 84) and Jamieson (p. 126). The latter describes a randomized trial on the quality of admission notes from EHRs. A systematic review of EHR usability is reported by Ellsworth (p. 222), Yadav (p. 143) compares EHRs with paper records in the documentation of physical exams, Denny (p. 165) describes hypertension “phenotyping” from EHRs, and Das (p. 24) proposes how to generate discharge recommendations. HIT is now pervasive in healthcare and this issue of JAMIA features several articles on the use of HIT for predictive modeling: Goldstein reviews risk prediction (p. 202) and describes the challenges in predicting mortality over time horizons (p. 180), Gillame-Bert (p. 48) learns temporal rules that predict instability in patients undergoing continuous monitoring, and Lennon (p. 148) shows predictive value in particular combinations of pathology markers for pancreatic cysts. Manaktala (p. 91) proposes a clinical decision support system for sepsis mortality, and Haslam (p. 13) discovers disease relationships from clinical trial data. HIE systems help clinicians share information from a particular patient with each other. However, the cost effectiveness of various types of HIEs remains hard to measure. Downing (p. 116) reports on the policies of 11 health systems, Dixon (p. 99) describes the characteristics of veterans who enroll in HIE, Vest (p. 39) describes the organizational capacity and utility of clinical event notification, and Slovis (p. 30) studies the rate of duplicative CT exams. Pharmacy informatics is an important sub-field of specialization. Nelson (p. 197) describes the interaction between pharmacists and the EHR, and White (p. 175) reports on outcomes of a computer-based system for Vitamin D prescriptions. Medication reconciliation is systematically reviewed by Marien (p. 231). Also related to pharmacy are reports from Eschmann (p. 62), who describes a system to predict hyperkalemia from drug interactions, Heringa (p. 55), who shows how clustering related drug interaction alerts helps reduce the number of alerts, and Manzi (p. 77), who describes a clinical pharmacogenomics service. Other types of computer-based interaction with providers and patients are reported in this issue of the journal: McGrath (p. 213) systematically reviews computer-aided instruction in oral health, Ratanawongsa (p. 109) studies computer use and literacy in safety net outpatient communications, and Kelly (p. 156) reports how families stay engaged in pediatric care through a patient portal. HIT evolved rapidly in the past decade and is likely to continue to evolve at this pace until its many challenges are overcome. Our responsibility as professionals is to help drive HIT, clinical decision support, and predictive analytics to the next level, and to educate the next generation so that they can fill the many knowledge gaps that still exist.

Health information technology (HIT) has changed the way clinicians work, and brought some advantages and disadvantages to clinical practice. In this special focus issue of JAMIA, our guest associate editors (p. 244) introduce three articles on the effects of HIT for patient safety (p. 246, p. 261, p. 268). Additional articles report on algorithms to measure patient safety (p. 310), and an interactive tool for cross-over analyses of EHRs for patient safety (p. 323). Prescription of medications is an area in which the effects of HIT for patient safety are well studied. Articles reporting on the analyses of medication errors (p. 316), computerized prescriber order entry (CPOE) system-related patient safety reports (p. 316), variation in high priority drug-drug interaction alerts across institutions (p. 331), clinician response to electronic health record (EHR) prompts (p. 275), and alert override analysis (p. 409) are directly connected to topics in this special focus issue. This issue also includes articles on a tool for automated screening for medication errors (p. 281), an approach for automated identification of antibiotic overdoses and adverse events (p. 295), and the impact of CPOE on the length of stay and mortality rate in an academic medical center (p. 303), and an intensive care unit (p. 413). Completing the set of articles focused on HIT and patient safety are two systematic reviews—one on automation bias (p. 423) and another on types and causes of prescribing errors generated from CPOE systems (p. 432), a brief communication on changes in the quality of care due to Meaningful Use implementation (p. 394), approaches to leverage EHRs for failure mode and effects analysis on a cardiology unit (p. 288), and approaches to leverage EHRs to identify complex atrial fibrillation patients for targeted intervention (p. 339). This issue also covers broad informatics topics that indirectly affect patient safety—the reuse of clinical and population health data to gain new insights into health outcomes. Clinical decision support systems rely on evidence generated from large amounts of data, hence data sharing is a pre-requisite for the successful development of such systems. We present an experience of opening government data to the public (p. 345), a study on patient preferences towards clinical data sharing (p. 380), a case report on improving the discoverability of ‘omics data sets (p. 388), and a predictive model for heart failure (p. 361). Interfacing applications to EHR systems (p. 398) and providing decision support in acute care (p. 441) are also presented, including the use of social media for translating evidence into practice (p. 403). A systematic review of context-sensitive decision support (p. 460) helps readers understand the general value of clinical decision support systems. Finally, HIT influence in healthcare is dependent on patient and provider engagement. Online cancer communities for social support (p. 451) and maternal and newborn mHealth interventions (p. 352) are reported in this issue of JAMIA. The mHealth interventions are deployed in Guatemala, reminding us that generalizable informatics ideas from all over the world need to be disseminated in JAMIA. Related to the globalization of informatics and HIT is the article on lessons learned and benchmarks for HIT in 30 countries (p. 371). Direct and indirect informatics interventions are changing the way we study health and disease, and provide healthcare. JAMIA is proud to feature the full spectrum of topics and serve as the dissemination vehicle for our professional specialty worldwide.

This online issue of JAMIA contains original articles about usability and utilization of electronic health records (EHRs) (p. e35, p. e55, p. e191, p. e207, p. e157, p. e87), including explorations of utilization patterns to identify collaborative care teams (p. e111) and to monitor resident supervision (p. e2), as well as “phenotyping” approaches that use natural language processing and machine learning techniques (p. e121, p. e79, p. e143). We also report that mobile devices are increasingly being used to help coordinate care in teams (p. e69, p. e178). Utilization of EHRs for administrative purposes is certainly not new, as billing systems were primary motivators for their implementation several decades ago. Derivation of quality metrics for congestive heart failure management (p. e40) and for hospitalization event notification and readmission reduction (p. e150) are some applications described in this issue of JAMIA. There is increasing awareness in the health care community about the power of using EHR data for decision support. We present several use cases for predictive analytics based on EHRs, such as predicting neutropenia risk in cancer patients (p. e129) and forecasting high costs due to coding issues in pressure ulcer reporting (p. e95). At a higher level, the impact of health information exchange on emergency medicine care is also reported (p. e103). Information technology is powerful but may have unintended consequences. Focusing on patient-provider communication is thus important to assess the impact on care. EHRs affect communication between health care providers and patients (p. e18), and we report on a qualitative study involving both groups to improve the quality of visit summary notes (p. e61). Additionally, patient portals are increasingly being used to communicate results and to gather information directly from patients, hence articles that report on their usability and utilization are also featured (p. e173, p. e9, p. e28, p. e47, p. e166). Examples of public health applications from which we derive lessons that can be helpful for population health management (p. e194, p. e136) are also presented in this online issue of JAMIA. Biomedical informatics encompasses the full spectrum, from theory to application, and the articles in this issue were selected from a large number of high-quality submissions. JAMIA is proud to continue to serve as the premier venue for dissemination of this work.

Every issue of JAMIA presents articles related to biomedical data science: from algorithms that discover and validate data-driven patterns, to big data indexing strategies, to predictive models that use novel approaches to analyze new or reused data. In this issue of JAMIA, several articles relied on mining of large data sets: Chen (p. 472) focuses on electronic health records for prediction of clinical order patterns, while Ghassemi (p. 488) uses a time series for prediction of vasopressor effects. Ritchie (p. 577) uses a new approach that utilizes genomic interactions for prediction of clinical outcomes in ovarian cancer. When mining individual level data, it is critical that the privacy of individuals be protected. O’Keefe (p. 544) proposes an online data center approach to prevent re-identification, while Dernoncourt (p. 596) uses recurrent neural networks to de-identify patient notes. Data analysis related to medications is also a popular topic in JAMIA. Patel (p. 614) uses the biomedical literature to suggest drug repositioning based on drug-drug interactions, Noor (p. 556) proposes a drug-drug interaction discovery approach using semantic web technologies, and Shah (p. 565) suggests potentially synergistic drug combinations using data from electronic health records (EHRs) and gene expression measurements. Also related to utilization of EHRs for research, Elemento (p. 513) reports on a cancer precision medicine knowledge-base for interpretation of clinical-grade mutations, and Garcelon (p. 607) describes an approach to improve full text searches on family history. Banerjee (p. 550) presents a heart failure dashboard designed to reduce readmissions, and Delon (p. 588) uses data from a national health insurance information system to implement epidemiological surveillance of malaria. Clinical decision support systems (CDSS) for emergency care are reviewed by Bennett (p. 655), and the cost-benefit of CDSS for cardiovascular disease prevention is reviewed by Jacob (p. 669). Increasingly, novel applications that target patients as their primary users are published in JAMIA. Hui (p. 619) reviews mobile apps for self-management in asthma, while Zikmund-Fisher (p. 520) shows how graphics can help patients recognize urgent deviations in laboratory results. With the vast amount of health information available online, it is timely that Allam (p. 481) proposes an instrument to assess the quality of web page contents. Additionally, McClellan (p. 496) shows how data from social media can be used to monitor mental health discussions, and Lambooij (p. 529) reports on the use of personal health records in the Netherlands. Finally, electronic systems are now disseminated into many tasks that directly or indirectly relate to healthcare and biomedical research. Khodyakov (p. 537) reports on an online system to help large groups of stakeholders in prioritizing research topics, Herndon (p. 503) uses a stakeholder-engaged approach to develop and validate clinical quality measures, and Pagliari (p. 633) reviews human resource information systems in healthcare. JAMIA has always been a venue for dissemination of biomedical data science, and is now issuing a call for papers on biomedical data science that emphasize the development of a data commons in which multiple digital objects can be found, accessed, and reused through interoperation in reproducible workflows and digital “recipes.” This special focus issue will solidify the overdue partnership with data repositories—JAMIA articles will be accompanied by data utilized, as well as by software deposited in code repositories or containers that may encapsulate these types of digital objects. We look forward to advancing healthcare and biomedical research through your submissions for this and future issues.

The enormous growth of informatics in the past decade has resulted in a high demand for trained professionals. Different types of training programs have been developed to serve the full spectrum of our field, from theoretical foundations to practical applications. For example, clinical research data management requires certain professional competencies (p. 737), while clinical data management involves an overlapping, but not identical, set of skills (p. 832). The goal of informatics education is to train a cadre of professionals who can design, develop, implement, and evaluate information-based interventions that impact health sciences and clinical care. These interventions target a variety of users and utilize diverse approaches and technologies. Among them, electronic health records (EHRs) continue to be a critical source of research in informatics, with articles that describe how to identify gravely ill Social Security disability applicants from EHRs (p. 709), compare vital signs recorded on paper and electronically (p. 717), review digital interventions to improve cardiovascular health (p. 867), and analyze policies that promote safe and usable EHRs (p. 769), among many others. Identifying adverse drug events (ADEs) from electronic health records is an important area of investigation in informatics. Authors report on the decline in in-hospital ADEs resulting from meaningful use of information technology (p. 729), show how Twitter posts can serve as input for deep-learning pharmacovigilance models to label ADEs (p. 813), and describe a system to detect ADEs from nursing notes and laboratory test results (p. 697). Related to medication safety, different studies compare knowledge bases used for detection of drug-drug interactions (p. 806), analyze medication order voiding in computerized provider order entry systems (p. 762), and describe a pharmacogenomics information resource for pharmacists (p. 822). Any analysis based on electronic health data must rely on harmonized data and reproducible processes. Examples of such processes are a phenotyping algorithm for the Elixhauser Comorbidity Index (p. 845), a method for embedding nursing interventions into the World Health Organization classification (p. 722), and a system to automatically classify eligibility criteria in clinical trials (p. 781). Related to data organization are also a new implementation of MetaMap (p. 841) and a method to detect missing relations and concepts in SNOMED CT (p. 788). Genomic information is increasingly being utilized in clinical care. We report on the decision-support needs of primary care pediatricians (p. 851) and a strategy for mitigating privacy risks in genomic data queries (p. 799). Decision-support tools are important to fill information gaps and to continuously educate health care providers and patients on various topics, eg, the information needs of generalists and specialists (p. 754), communication of patient health information to guide health information technology design (p. 680), shared decision-making using personal health record technology (p. 857), and systems to decrease unnecessary vitamin D testing (p. 776) and to monitor mood (p. 746). As exemplified in this issue of the journal and documented in JAMIA for over 20 years, informatics is a diverse and fascinating field that not only supports and augments the work of health scientists and clinicians, but also changes the way they approach problems. It is thus our responsibility to ensure that our information interventions result in maximal societal benefit and are improved or replaced by better products generated by a new generation of informatics professionals.

The evolution of biomedical informatics, reflected by the articles JAMIA has published in the past few years, indicates a steep increase in articles describing direct interactions of patients, study participants, and caregivers with electronic health systems. Examples of patient interactions with apps for diabetes management are reviewed in two separate articles (p. 1014, p. 1024). Another review addresses health-related behavior interventions (p. 1002), and a research article compares the information needs of people working with patients who have special needs (p. 933). A lot more attention is given now than was just a few years ago to the utilization and effectiveness of patient portals (p. 903, p. 927), the role of patient-reported outcomes (p. 964, p. 897), and the quality of electronic patient-clinician interactions (p. 942). Equally exciting is the movement toward personalized medicine that takes into account genetic and other factors in prevention and treatment strategies (p. 921, p. 950). Health care provider interactions with electronic systems are certainly important as well. Reports on clinician interactions with information technology have always been featured in JAMIA. For example, in this issue of the journal, authors report on outcomes of interventions related to lab order systems (p. 958), discharge instructions (p. 975), and antimicrobial management (p. 981). Administrators also use information technology to measure and plan resources, with examples focusing on telemedicine (p. 969, p. 891), electronic health record system configuration (p. 992), and value-based care (p. 1036). Finally, JAMIA has always been the premier place to report on original biomedical informatics research and applications. Much work happens in informatics units operating “behind the scenes” to make systems work better, eg, by adapting natural language processing tools for diverse health care settings (p. 986), building a domain model for clinical protocol–driven research (p. 882), addressing major barriers to the implementation of pragmatic clinical trials (p. 996), or developing better decision support for adverse events due to medications (p. 913). The journal thus covers a wide variety of topics and users, from those focusing on the foundations of biomedical informatics to those designing, implementing, and evaluating their various uses and impact in health sciences all over the world. Reporting on the use of electronic media, artificial intelligence, and Internet-based communications is part of JAMIA’s mission, but sometimes it is important to reflect on our broader vision. The fact that digital artifacts are now part of most people’s day-to-day life is amazing, as people can get more information and engage directly with multiple systems that can provide data and/or knowledge related to their specific conditions. However, as devices and systems get more sophisticated, they may widen the digital divide, separating a portion of the world’s population who do not have easy access to the Internet, due to cost, literacy, or political barriers, from another portion of the population for whom nothing of importance seems to happen without Internet access. Our obligation as informatics professionals is to understand this divide and mitigate it. If we do not do so, we incur the risk of being so fixated in technology that we forget the reasons why we embraced the profession in the first place: we care about the health of individuals first and foremost, and aim to provide information to them and to their caregivers in a way that is most impactful for their health and well-being. Our goal is to deliver impactful information. We are now reaching an inflexion point in biomedical informatics in which nondigital media, human counseling, and in-person communications can enhance the technology we have been working so hard to design, implement, evaluate, and disseminate, instead of the other way around. This points to a state of maturity, a balance between technology and people, that characterizes many established fields. It is very gratifying to be able to document this evolution in the pages of JAMIA. In this issue of the journal, we present some articles addressing the digital divide directly or indirectly, and we expect to see much more on this topic presented in the near future.

This is a historical issue of JAMIA: It is the last time the journal will be printed and mailed to our subscribers, some of whom have grown accustomed to receiving it in print every 2 months for the past 23 years. After this issue, all future issues will only be made available online, thus reflecting our professional society’s evolution to a “green” approach to academic publishing. The online format will also allow us to increase the frequency of journal issues, which will now be published monthly. With the removal of print media, the partnership with digital object repositories and the increased visibility of metrics related to access and utilization will help us take the first steps toward having an “executable paper,” in other words, an article of record that includes digital objects that can be easily found and whose work can be more easily reproduced, adapted, and/or enhanced by interested readers. The inaugural online-only issue in January will present various articles describing different components on which an executable paper relies: interoperable, shareable digital objects containing data, as well as tools (software, workflows) and processes that are particularly important for biomedical data sciences. The executable paper will evolve from these components, many of which are already being described in JAMIA articles. This issue of JAMIA includes articles on assessing data quality, a top priority in informatics, since little value is derived from poor-quality data. A longitudinal analysis of electronic health record (EHR) data in a large pediatric network (p. 1072), automatic identification of implausible values (p. 1080), a call for formalization of drug indications (p. 1169), a study on biases introduced by filtering out EHRs with incomplete data (p. 1134), and a review on evidence appraisal (p. 1192) describe critical issues in evaluating the quality of data and of the derived knowledge. A tutorial on blockchain (p.1211) explains how this technology enables immutable ledgers that allow verification of data provenance and transformations. Disseminating data-driven approaches to produce generalizable knowledge is also critical to JAMIA. Many articles in this issue address highly important topics in biomedical informatics in general, and biomedical data science in particular. For example, we include articles that are prime examples of statistical or machine learning methods in support of health care or biomedical science: predictive models for lung cancer (p. 1046), acute (p. 1052) and chronic (p.1111) kidney injury, and asthma exacerbation (p. 1116); algorithms to identify nonmedical opioid use (p. 1204); and a hybrid system to identify clinical trial reports (p. 1165). Informatics is not limited to the generation of new knowledge from data via learning techniques; it is also concerned with how this information is effectively used in practice. Clinical implications ofinformatics systems, such as the effect of health information exchange on recognizing medication discrepancies (p. 1095), automated laboratory results notification (p. 1173), optimization of drug alerts (p. 1149), clinical sequencing of exome reports (p.1184), participatory design of decision support tools for nurses (p. 1102), eligibility criteria extraction from clinical trials (p. 1062), and the origins of the divide in advanced EHR adoption, are featured in this issue. The use of Facebook to combat Zika (p. 1155) and of information technology to support patient engagement (p.1088), and the effectiveness of bilingual patient portals (p. 1160) are also presented. JAMIA is continuously evolving, and the next issue will mark the beginning of a new phase, in which we will see innovation not only in content, but in format and scope. We will continue to publish early access articles online as soon as they are ready, but the faster indexing into monthly journal issues will allow our authors to refer to the final versions of their articles sooner, and allow the editorial team to have more flexibility in publishing materials and grouping the articles. We expect the online format to increase the number of articles that integrate various digital objects, which may also require new forms of peer review and attribution that have not yet been widely utilized. However, authors will continue to have a venue to disseminate their best work – regardless of their ability to pay author fees – and readers will continue to benefit from a rigorous peer-review process that selects the most valuable submissions for publication.

The Global Alliance for Genomics and Health (GA4GH) created the Beacon Project as a means of testing the willingness of data holders to share genetic data in the simplest technical context-a query for the presence of a specified nucleotide at a given position within a chromosome. Each participating site (or "beacon") is responsible for assuring that genomic data are exposed through the Beacon service only with the permission of the individual to whom the data pertains and in accordance with the GA4GH policy and standards.While recognizing the inference risks associated with large-scale data aggregation, and the fact that some beacons contain sensitive phenotypic associations that increase privacy risk, the GA4GH adjudged the risk of re-identification based on the binary yes/no allele-presence query responses as acceptable. However, recent work demonstrated that, given a beacon with specific characteristics (including relatively small sample size and an adversary who possesses an individual's whole genome sequence), the individual's membership in a beacon can be inferred through repeated queries for variants present in the individual's genome.In this paper, we propose three practical strategies for reducing re-identification risks in beacons. The first two strategies manipulate the beacon such that the presence of rare alleles is obscured; the third strategy budgets the number of accesses per user for each individual genome. Using a beacon containing data from the 1000 Genomes Project, we demonstrate that the proposed strategies can effectively reduce re-identification risk in beacon-like datasets.

OBJECTIVE: To develop an accurate logistic regression (LR) algorithm to support federated data analysis of vertically partitioned distributed data sets. MATERIAL AND METHODS: We propose a novel technique that solves the binary LR problem by dual optimization to obtain a global solution for vertically partitioned data. We evaluated this new method, VERTIcal Grid lOgistic regression (VERTIGO), in artificial and real-world medical classification problems in terms of the area under the receiver operating characteristic curve, calibration, and computational complexity. We assumed that the institutions could "align" patient records (through patient identifiers or hashed "privacy-protecting" identifiers), and also that they both had access to the values for the dependent variable in the LR model (eg, that if the model predicts death, both institutions would have the same information about death). RESULTS: The solution derived by VERTIGO has the same estimated parameters as the solution derived by applying classical LR. The same is true for discrimination and calibration over both simulated and real data sets. In addition, the computational cost of VERTIGO is not prohibitive in practice. DISCUSSION: There is a technical challenge in scaling up federated LR for vertically partitioned data. When the number of patients m is large, our algorithm has to invert a large Hessian matrix. This is an expensive operation of time complexity O(m(3)) that may require large amounts of memory for storage and exchange of information. The algorithm may also not work well when the number of observations in each class is highly imbalanced. CONCLUSION: The proposed VERTIGO algorithm can generate accurate global models to support federated data analysis of vertically partitioned data.

This issue of JAMIA focuses on the many “meaningful uses” of mobile health (mHealth) and health information technology in healthcare. An editorial by Tang et al. introduces nine articles (see pages 5 , 12 , 19 , 29 , 38 , 48 , 60 , 74 , 80 , 88 , 94 , 105 , 110 , 119 , 129 , 137 , 149 ) that compose the special focus issue on “Interactive Systems for Patient-Centered Care to Enhance Patient Engagement.” Other mHealth-related articles focus on the effect of tablet computers on hospitalized patients’ knowledge (see page 159 ), smartphone-based diagnosis of preeclampsia in resource-limited settings (see page 166 ), use of mobile technologies to enhance immunization (see page 207 ), mHealth adoption by healthcare professionals (see page 212 ), and the effect of mobile technologies in interventions for stress and anxiety (see page 221 ). Additionally, a review on whether electronic games help improve knowledge and self management in young people with chronic conditions (see page 230 ) is presented in this issue of the journal. Various technological platforms are currently used by patients and providers. However, the type of platform is only one factor in the quality of patient engagement, since the best healthcare system in the world will not be complete without an effective means to communicate health information to patients and their caregivers. Articles in this issue show how the use of infographics can mitigate health illiteracy (see page 174 ), and the importance of measuring a patient’s reading level (see page 202 ). Patient-provider electronic messages do not always match readability levels appropriate for a specific patient. Patient engagement is as critical as having access to a healthcare system where decisions are based on evidence. Data can have a transformative role in healthcare and research. Electronic systems for data management are impactful in various contexts, such as support of multi-center networks operating in resource-limited settings (see page 184 ), automated assessment of bias in clinical trials (see page 193 ), and health information exchange. As the articles in this issue of JAMIA illustrate, biomedical informatics continues to be an exciting field that is ripe for innovations that have a direct impact on human health. JAMIA is proud to disseminate innovations that stimulate researchers, clinicians, and patients to think of better ways to promote health and ease the burden of disease.

Earlier this year, a graduate student asked me how data scientists and informaticians could be viewed as “research parasites.” I have been reflecting on this question as a biomedical informatician passionate about data sharing and open science. While some may fear that using someone else’s data will disprove earlier published results or steal credit and productivity, many of us view the real benefits of open science as reproducibility of results and exploration and unveiling of new truths. Open science is no longer just the dream of visionaries and the privilege for a few, it is a reality demanded by the public and government agencies. The biomedical informatics field has exploded the past few decades because of access to a wealth of data from genomic sequencing, research on gene regulation, reinvention of artificial intelligence and machine learning, and clinical data research networks using electronic health records, to name a few. The boundaries of biology, medicine, and behavioral science are blurring, and being tied together by quantitation and computation. We, as biomedical informaticians and data scientists, should be proud of our accomplishments in biomedical informatics and grateful to the pioneers in our field – the risk takers who departed from traditional careers to forge a new discipline. Our field is now rightfully recognized as critical for the success of important initiatives to advance human health as evidenced by the prominence of informatics in the president’s precision medicine initiative. We should not be deterred by misperceptions of a few colleagues and should continue to share data and ideas, and collaborate with colleagues from many disciplines without any reservations. Our field is a prime example of how embracing diverse backgrounds, experiences, and cultures help us move forward faster. For instance, we are playing a major role in transforming healthcare and public health by developing new tools to track diseases and monitor the success of interventions, including those enabled by social media. Partnerships among a new generation of biomedical informatics scientists and biomedical and behavioral researchers will bring new ideas and ideals. By understanding and embracing the power of open science, together we will continue to thrive towards a common goal: to accelerate health science discoveries through information and computation.

Health-related data from individuals are at the “start” and “end” of a continuous clinical decision support loop: these data drive the development of models embedded in decision support systems whose implementation results in outcomes data for those individuals. The outcomes data are further used to improve and develop new predictive models and corresponding decision support systems. In this issue of JAMIA we present brief communications and research articles in areas that support the clinical decision support loop: data collection and quality assessment, analysis, predictive modeling in decision support systems, and evaluation of clinical outcomes. The quality of data collected in the process of care is highly variable and needs to be continuously monitored and improved. Moscow (see page e108 ) shows that this is particularly true for antibiotic prescriptions. Davhle (see page e99 ) evaluates the utilization of RxNorm in ambulatory care prescriptions, and Zhou (see page e79 ) examines food entries in a large allergy repository. Hsu (see page e152 ) proposes a data-driven approach for quality assessment of radiologic interpretations, while J. Bates (see page e113 ) classifies radiology reports for falls in an HIV study cohort. Cohen (see page e146 ) describes the barriers and benefits of meaningful use care coordination criteria among primary care providers. A systematic review by Hodgson (see page e169 ) describes risks and benefits of speech recognition for clinical documentation. Data integration is helpful in reducing uncertainty and improving phenotype characterization, and Scheurwegs (see page e11 ) and Denny (see page e20 ) show the advantages of combining structured and unstructured sources for this purpose. Patrick (see page e42 ) discusses opportunities and challenges in using personal health data for research. Data analyses result in predictive models that are employed in different health sciences domains, for example, predictive models embedded in electronic clinical decision support systems. Heatherly (see page e131 ) reports on a multi-institutional evaluation of clinical profile anonymization. Toerper (see page e49 ) reports on a prospective evaluation of a catheterization laboratory inpatient forecast tool, Barnes (see page e2 ) presents a model that predicts inpatient lengths of stay in real time, and Apley (see page e71 ) describes an approach to validating sampling for logistic regression models. Different modalities of alerts and reminders are available through texting, email, and patient or clinician portals; however, patient use of these modalities is variable. McIvers (see page e88 ) shows that text messages may not be effective in improving rates of hepatitis B vaccinations, while Tesfalul (see page e142 ) studies the potential impact of mobile technologies on specialty care coordination in a resource-constrained setting. Otte-Trojel (see page e162 ) does a systematic review of the literature on patient portal development, while Lazard (see page e157 ) emphasizes the need for design simplicity in these portals. Xu (see page e34 ) describes a visualization system to navigate educational materials, and Lyles (see page e28 ) reports on improvements in adherence to medical refills via EHRs across racial and ethnic groups. Chow (see page e58 ) identifies patient and provider predictors of patient receipt of therapies recommended by a clinical decision support system, and Turley (see page e118 ) proposes an information model to assess concordance between patient advance care directives and actual delivery of care at the end of life. Several articles in this issue of JAMIA also focus on clinician adherence to clinical decision support. Bauer (see page e125 ) shows that clinicians’ responses to alerts and reminders can be predicted by their familiarity with the topic and experience with the system, and D. Bates (see page e93 ) studies factors associated with provider response to clinical decision support system warnings. Some alerts and reminders can be disruptive, and hence Hoffman (see page e138 ) proposes a novel measure for their evaluation. JAMIA is proud to cover the full spectrum of these areas, from foundational to applied informatics, and to provide our readers with carefully selected, outstanding reports of research studies, perspectives, and reviews from all over the world.

This issue of JAMIA starts with an editorial by Patricia Flatley Brennan, the new director of the National Library of Medicine (NLM), one of the institutes that composes the National Institutes of Health. Brennan describes the road ahead for an institution that has embraced and promoted informatics, intra- and extramurally, for several decades (p. 858 ). Before data science became fashionable and artificial intelligence (AI) silently awoke from hibernation enabling various AI applications that directly affect millions of Internet users around the world, NLM was already promoting and supporting informatics research in these areas. Among other initiatives, NLM built information systems to democratize access to the biomedical literature, structured biomedical concepts into a unified medical language system, organized high performance computing projects, and built capacity in informatics through its training programs. Clinical informatics is an area in which this investment has been very visible, and is the emphasis of most of the articles presented in this issue of JAMIA. This issue of JAMIA features clinical informatics research and applications that span the spectrum of point-of-care applications to support clinicians and patients in tasks such as identifying high risk heart failure patients (p. 872 ), detecting colorectal cancer in primary care (p. 879 ), supporting generic drug prescribing (p. 892 ), understanding the tradeoffs of digital information sharing and privacy when engaging families in an intensive care unit setting (p. 995 ), and reminding patients to review doctor’s notes (p. 951 ). A complex infrastructure to enable the delivery of user-facing applications is necessary. Various components of this infrastructure are analyzed in articles presented in this issue, including information flow, standardization, and knowledge management tools that help inform the development of applications. Analyses of consistency of problem lists among physicians (p. 859 ), clinical code co-occurrence (p. 866 ), alert overrides (p. 924 ), medication errors (p. 942 ), mobile technology usage (p. 979 ), health insurance status identification from electronic health records (EHR) (p. 984 ), data validation in EHR system transitions (p. 991 ), information flows in clinical decision support (p. 1001 ), approaches to EHR “phenotyping” (p. 1007 ), and information technology effects on patient outcomes (p. 1016 ) are presented in this issue. Standards for interfacing with EHR systems (p. 899 ), and for representing common data elements (p. 956 ) are also described. These efforts help explain and improve the bases upon which user-facing applications are built. Related to research, this issue of the journal also describes informatics projects presenting new approaches for data interchange (p. 909 ), project management (p. 916 ), metadata expansion for geospatial location (p. 934 ), and pharmacovigilance (p. 968 ). Outstanding contributions from authors worldwide, insightful suggestions from reviewers and careful selection of manuscripts by the editorial team make JAMIA a great generalist informatics journal. Credit is also due to previous editors, associate editors and editorial board members who made the journal the premier venue for biomedical/health informatics scholarly publication. This JAMIA issue, focused on clinical informatics, is a tribute to former editors Bill Stead and Randy Miller, as well as to two associate editors who have worked relentlessly for this journal for over 20 years: Betsy Humphreys and Patti Brennan. We celebrate their multiple achievements in these two decades of intense informatics growth and look forward to their continued contributions.

The field of biomedical and health informatics is growing at a fast pace, both in breadth and depth. The evolution and dissemination of informatics is evident from the articles that have been published in JAMIA the past few years. In this issue of JAMIA , not only do a large number of articles report on studies that rely on machine learning and data science to build predictive models in a variety of biomedical domains, but there are also a high number of articles describing patient-facing applications in addition to articles about electronic health records (EHRs), and clinical decision support, which used to be the most frequently appearing topics in the journal just a few years ago. Additionally, contributions from authors outside the US are on the rise, expanding the knowledge base and affording our readers with learning experiences that are generalizable to many settings. Caron (p. 1159 ) describes a predictive model relating enteric disease episodes and sexually transmitted diseases, and Osborne (p. 1077 ) describes how natural language processing and machine learning can increase the number of reportable cases of cancer. Zhong (p. 1060 ) uses EHRs for surveillance of childhood diabetes, Denny (p. 1046 ) reports on a catalog of phenotyping algorithms to enhance study reproducibility, and Agarwal (p. 1166 ) shows that learning of phenotypes is possible even when some labels are noisy—an important finding as the quality of EHR data is not perfect. Singer (p. 1107 ) reports on a study comparing problem lists and billing data in chronic diseases, Madden (p. 1143 ) quantifies missing clinical and behavioral data in EHRs, and Sáez (p. 1085 ) shows the results of probabilistic temporal and multi-site quality control in a mortality registry. Clinical decision support, clinical information modeling, and physician-driven tools are also not devoid of quality issues, as described by Wright (p. 1068 ), Moreno-Conde (p. 1127 ), and Mazur (p. 1113 ), respectively. Tools focused on patient information are described in articles by Payton (p. 1121 ), Hill (p. 1136 ), Wolff (p. 1150 ), and Donaldson (p. 1174 ). Patil (p. 1096 ) reports differences in patient preferences regarding EHR storage, as well as access and sharing, reminding us of the importance of preserving privacy while generating new knowledge. Hripcsak (p. 1040 ) explains how temporal relations in clinical data can be preserved without compromising patient privacy. Finally, the digital divide still needs to be addressed: McCloud (p. 1053 ) explains that barriers to the urban poor seeking health information extend beyond access. The issue also features outstanding brief communications by Fong (p. 1180 ), Sharif (p. 1185 ), Milberg (p. 1190 ), Shea (p. 1195 ), and Peisert (p. 1199 ) that cover the high diversity of our topics. As a generalist journal in biomedical and health informatics, JAMIA continues to evolve to remain aligned with the growth of our field while remaining highly selective with its featured articles. JAMIA is proud to conclude this year with articles from traditional and new informatics. Readers should stay tuned for a wealth of new knowledge in 2017.

Biomedical Informatics is a growing interdisciplinary field in which research topics and citation trends have been evolving rapidly in recent years. To analyze these data in a fast, reproducible manner, automation of certain processes is needed. JAMIA is a "generalist" journal for biomedical informatics. Its articles reflect the wide range of topics in informatics. In this study, we retrieved Medical Subject Headings (MeSH) terms and citations of JAMIA articles published between 2009 and 2014. We use tensors (i.e., multidimensional arrays) to represent the interaction among topics, time and citations, and applied tensor decomposition to automate the analysis. The trends represented by tensors were then carefully interpreted and the results were compared with previous findings based on manual topic analysis. A list of most cited JAMIA articles, their topics, and publication trends over recent years is presented. The analyses confirmed previous studies and showed that, from 2012 to 2014, the number of articles related to MeSH terms Methods, Organization & Administration, and Algorithms increased significantly both in number of publications and citations. Citation trends varied widely by topic, with Natural Language Processing having a large number of citations in particular years, and Medical Record Systems, Computerized remaining a very popular topic in all years.

UNLABELLED: New models of healthcare delivery such as accountable care organizations and patient-centered medical homes seek to improve quality, access, and cost. They rely on a robust, secure technology infrastructure provided by health information exchanges (HIEs) and distributed research networks and the willingness of patients to share their data. There are few large, in-depth studies of US consumers' views on privacy, security, and consent in electronic data sharing for healthcare and research together. OBJECTIVE: This paper addresses this gap, reporting on a survey which asks about California consumers' views of data sharing for healthcare and research together. MATERIALS AND METHODS: The survey conducted was a representative, random-digit dial telephone survey of 800 Californians, performed in Spanish and English. RESULTS: There is a great deal of concern that HIEs will worsen privacy (40.3%) and security (42.5%). Consumers are in favor of electronic data sharing but elements of transparency are important: individual control, who has access, and the purpose for use of data. Respondents were more likely to agree to share deidentified information for research than to share identified information for healthcare (76.2% vs 57.3%, p < .001). DISCUSSION: While consumers show willingness to share health information electronically, they value individual control and privacy. Responsiveness to these needs, rather than mere reliance on Health Insurance Portability and Accountability Act (HIPAA), may improve support of data networks. CONCLUSION: Responsiveness to the public's concerns regarding their health information is a pre-requisite for patient-centeredness. This is one of the first in-depth studies of attitudes about electronic data sharing that compares attitudes of the same individual towards healthcare and research.

OBJECTIVE: The Cox proportional hazards model is a widely used method for analyzing survival data. To achieve sufficient statistical power in a survival analysis, it usually requires a large amount of data. Data sharing across institutions could be a potential workaround for providing this added power. METHODS AND MATERIALS: The authors develop a web service for distributed Cox model learning (WebDISCO), which focuses on the proof-of-concept and algorithm development for federated survival analysis. The sensitive patient-level data can be processed locally and only the less-sensitive intermediate statistics are exchanged to build a global Cox model. Mathematical derivation shows that the proposed distributed algorithm is identical to the centralized Cox model. RESULTS: The authors evaluated the proposed framework at the University of California, San Diego (UCSD), Emory, and Duke. The experimental results show that both distributed and centralized models result in near-identical model coefficients with differences in the range [Formula: see text] to [Formula: see text]. The results confirm the mathematical derivation and show that the implementation of the distributed model can achieve the same results as the centralized implementation. LIMITATION: The proposed method serves as a proof of concept, in which a publicly available dataset was used to evaluate the performance. The authors do not intend to suggest that this method can resolve policy and engineering issues related to the federated use of institutional data, but they should serve as evidence of the technical feasibility of the proposed approach.Conclusions WebDISCO (Web-based Distributed Cox Regression Model; https://webdisco.ucsd-dbmi.org:8443/cox/) provides a proof-of-concept web service that implements a distributed algorithm to conduct distributed survival analysis without sharing patient level data.

BACKGROUND: Centralized and federated models for sharing data in research networks currently exist. To build multivariate data analysis for centralized networks, transfer of patient-level data to a central computation resource is necessary. The authors implemented distributed multivariate models for federated networks in which patient-level data is kept at each site and data exchange policies are managed in a study-centric manner. OBJECTIVE: The objective was to implement infrastructure that supports the functionality of some existing research networks (e.g., cohort discovery, workflow management, and estimation of multivariate analytic models on centralized data) while adding additional important new features, such as algorithms for distributed iterative multivariate models, a graphical interface for multivariate model specification, synchronous and asynchronous response to network queries, investigator-initiated studies, and study-based control of staff, protocols, and data sharing policies. MATERIALS AND METHODS: Based on the requirements gathered from statisticians, administrators, and investigators from multiple institutions, the authors developed infrastructure and tools to support multisite comparative effectiveness studies using web services for multivariate statistical estimation in the SCANNER federated network. RESULTS: The authors implemented massively parallel (map-reduce) computation methods and a new policy management system to enable each study initiated by network participants to define the ways in which data may be processed, managed, queried, and shared. The authors illustrated the use of these systems among institutions with highly different policies and operating under different state laws. DISCUSSION AND CONCLUSION: Federated research networks need not limit distributed query functionality to count queries, cohort discovery, or independently estimated analytic models. Multivariate analyses can be efficiently and securely conducted without patient-level data transport, allowing institutions with strict local data storage requirements to participate in sophisticated analyses based on federated research networks.

We are thrilled to release the first issue of JAMIA produced by our new publisher, Oxford University Press. JAMIA has a new cover and overall article layout that we hope you enjoy browsing through. In addition to these improvements, we continue to strive for innovations that will support JAMIA’s mission as the premiere vehicle for dissemination of informatics ideas worldwide. The number of submissions we receive increases every year leading us to select only the very best articles on informatics research and applications in areas as diverse as clinical informatics, translational bioinformatics, imaging informatics, public health informatics, and more. We take this opportunity to thank Kevin Johnson, Vanderbilt University, who has been a valuable member of our associate editor team. We will miss his insightful reviews and innovative ideas, as well as his contributions in training the next generation of editors. Kevin was responsible for the student editorial board (SEB) from 2004–2010. As an example of Kevin's contributions, Michael Chiang, Oregon Health Sciences University, a former student member, is now an associate editor of JAMIA and directs the SEB. We are pleased to announce that Mark Frisse, Vanderbilt University, and Dina Demner-Fushman, National Library of Medicine, accepted our invitation to serve as new JAMIA associate editors, joining the eight other members of our team. In addition, Berry de Bruijn, PhD, National Research Council, Canada, Farah Magrabi, PhD, Maquerie University, Australia, and Adam Wright, PhD, Partners Healthcare, are joining our editorial board in 2015. This issue of the journal features a variety of informatics subspecialties and settings written by authors from more than 16 countries representing a high number of academic and non-academic institutions. As evidenced by articles in this issue, the healthcare environment is changing due to advances in the informatics field. Data are driving the development of learning healthcare systems and informing patients, as described by Friedman (2977) and Valdez (2826), respectively. Related work by Rosen (2606) describes how sensor data are changing the way we measure team work, while Lussier (2491) explains how a transition to ICD-10-CM may have an important effect in patient safety. Apps for mobile devices and other technologies are increasingly being used to deliver health-related applications, such as depression screening (2840), medication reminders by use of synthetic speech (2820), and text messaging to improve adherence to medications (2845). JAMIA is now issuing a call for papers in this important area of mobile health monitoring and social networking. Clinical research informatics, imaging informatics and translational bioinformatics applications are also on the rise. Articles by Swerts, Choquet, and Decaestecker present software for pooled analyses across biobanks (2577), strategies to determine minimum common elements for rare disease data sets (2794), and methods for pathology image registration, respectively. Tang and Kim present translational bioinformatics methods for protecting the privacy of shared DNA data (2794) and for integrating multi-omics data (2481). Text mining and biomedical natural language processing (NLP) continue to have a strong presence in JAMIA. Jung (2902) evaluates out-of-the-box software for text mining, Liou compares semi-automated and NLP tools for generating summary statements in clinical notes, and Pradhan evaluates the recognition of disorders in clinical text. New applications for venous thromboembolism detection, clinical trial prescreening, and drug repurposing signals are presented by Rochefort (2768), Ni (2887), and Xu (2649), respectively. Another research area that frequently appears in JAMIA is the usage of Electronic Health Record (EHR) systems. Lenert (2871) investigates the effects of EHR systems on trainees’ communication skills, Burke (2726) shows that EHRs improve clinical note quality, Coleman (2822) describes temporal and other factors that affect physician’s electronic prescriptions, and Wells (3055) discusses organizational strategies for promoting use of personal health records. Data collection from EHRs also assists administrators in the implementation of value-based strategies to constrain costs (2511). In a systematic review, Hosltiege (2886) shows that clinical decision support systems hold promise in improving antibiotic prescribing. Finally, a systematic review by Laranjo (2841) shows that social networking sites may be influential in health behavior change, thus motivating future research on the combination of EHRs, PHRs, and social media for health improvement. In closing, we hope you enjoy this new era of JAMIA. Our editorial and production staff remain dedicated to expanding your knowledge base by selecting high quality, innovative articles. As always, we thrive on and welcome your feedback to improve our processes and product. We look forward to hearing from you.

President Obama’s initiative on Precision Medicine is very exciting to the biomedical informatics community as informatics will be key to its success. Informatics will be a critical component to integrate the large, scientific ecosystem of individualized interventions based on genotypes, data gathered by mHealth devices and advanced imaging devices, electronic health records (EHRs) that follow phenotypes over time, and more. The Precision Medicine ecosystem will bring together molecular biologists, clinicians, engineers, computer scientists, and biomedical informaticians to address an important challenge: how to manage and extract knowledge from data that are highly diverse (Variety), quickly accumulating (Velocity), and large-sized (Volume), referring to the three “Vs” currently used to define Big Data. Accordingly, biomedical informatics sub-specialties will need to coalesce to help deliver “the right intervention to the right person at the right time.” This issue of JAMIA includes articles on innovative and impactful research that accelerate us towards addressing this challenge. One featured topic is the visualization of healthcare data to assist clinicians and administrators in decision making. An editorial by Caban and Gotz (003556) introduces articles describing visualizations used to summarize patient records (002945), improve patient safety (002963), prognosticate kidney disease (002927) and explore its co-morbidities (02936). Innovative displays for understanding variations in asthma care (002937) and heart failure (002960) are also featured. Articles on interactive network visualization (002965) and a review of current EHR visualization techniques by West et al (002955) complete the special focus on visualization. Another main topic in this issue is clinical informatics. EHR adoption in children’s hospitals is studied by Nakamura (003245), in rural settings by Whitacre (003251), in primary care clinics by O’Malley (003284), and in safety net settings by Kern (003296). A systematic review by Fritz (003131) lists criteria for successful EHR implementation in low-resource settings. Applications based on the EHR are the focus of articles by Plaisant (003334), who reports on a novel interface design for medication reconciliation, Klann (003040), who describes how continuity of care documents can be used for research, and Vest (002760), who describes health information exchange strategies to reduce hospital readmission. The Veterans Administration’s experience with opening clinical notes to patients is reported by Nazi (003144). EHR-based clinical decision support involving information technology is on the rise: Evans (002816) and Simpao (02538) describe alert systems to detect physiologic deterioration in hospitalized patients and to prevent adverse events resulting from drug-drug interactions, respectively. Amster (002865) discusses National Quality Forum-specified eMeasures, and Sittig (002988) presents a perspective on IT-based patient safety goals. Value-based strategies are also on the rise with EHRs continuing to play a significant role in documentation, and ideally also in facilitating data-driven decisions. The high adoption of EHR systems should not, however, overshadow some issues that deserve open discussion by our community. For example, Koppel (003923) describes potential implications of low diversification of EHR vendors in the US healthcare systems, including effects in patient care and costs, while Turer describes needed ICD-10 crosswalks and reimbursement mappings. Howley (002686) and Riskin (003065) examine the financial impact of EHRs and the value of information technology in healthcare, respectively. As our field evolves, we will continue to face increasing challenges in determining the exact cost-effectiveness of IT interventions in a frail healthcare system, while at the same time trying to develop the necessary ecosystem to enable Precision Medicine. Innovations in biomedical sciences, clinical services and policy will need to rely on highly integrated information systems. There has never been a more exciting time for our profession.

This special focus issue contains prime examples of how our field is changing—technology and processes increasingly involve patient interaction with systems such as patient portals and health promotion apps, and joint decision-making with health care providers. A report from AMIA’s policy meeting (see page e2) discusses the role our professional society can play in ensuring that systems—for care, collaboration, and communication—become more patient centered than they are today. This issue of JAMIA covers a large number of patient-centered topics. A brief communication by Gelb (see page e39) discusses how to direct the public to evidence-based content online. Studies by Suominen (see page e48), Mistry (see page e177), and Shade (see page e104) report on how information technology interventions can improve the capture of patient information in nursing handoffs, enhance medication adherence, and increase patient engagement in HIV care, respectively. A systematic review by Kung (see page e194) reports on randomized controlled trials of Internet-based interventions to reduce caregiver stress. The use of crowdsourcing for dietary self-monitoring is presented by Turner-McGrievy (see page e112). Weng (see page e141) describes a system that helps match patients to clinical trials. This type of “phenotyping” also described by Lin (see page e151) is for the purpose of identifying episodes of drug toxicity. As with any other intervention, information technology also carries some risks: Campos-Castillo (see page e130) discusses electronic health record (EHR) implications for patient disclosure, and Sunyaev (see page e28) presents an interesting account of privacy policies employed in mobile health apps. Clinician-centered applications also have obvious implications in patient care. Dhavle (see page e7) presents a perspective on the “perfect” electronic prescription, Dhiman (see page e13) reviews studies on clinical decision support systems (CDDS), Caraballo (see page e21) provides an illustration of a cardiology CDDS, and Popejoy (see page e93) proposes a method to quantify care coordination. Although EHRs are the most common source of data, other data types such as images (see page e81) and claims data (see page e34) are also used to improve patient care and public health. Despite the potential enhancement of patient care with the above-described applications, gaps exist that need to be filled. Documentation of family history is a challenging problem that is addressed by Chen (see page e67), while Duncan (see page e120) reports on the important problem of identity resolution. Finally, although clinical informatics has become central to the practice of medicine today, medical students’ awareness of and interest in this area is still highly variable, as reported by Banerjee (see page e42). In summary, this special online focus issue presents multiple examples of how informatics is changing the way patients, caregivers, and health care providers are using electronic systems to enhance their experience. Informatics is clearly essential for the contemporary practice of medicine. Data-driven medicine requires a large amount of data. In the next journal issue, we will focus on the importance of standardized data.

The role of a scientific journal is to disseminate generalizable knowledge that can be applied to new settings and advance science. This is an exciting landmark for the current editorial team: after about 4.5 years of service we are celebrating the acceptance of our 1,000th article. Through hundreds of articles, JAMIA has documented the rise of new subspecialties of informatics, the launch of large targeted funding initiatives, and the many ways in which patients participate in their own care and in health-related research. Informatics is growing faster than we are able to train new professionals, and hence everyone is facing increasing demands and new responsibilities in their jobs. I remain impressed by the dedication of associate editors, editorial board members, and reviewers. Their contributions are critical to ensure a quality product, so I urge leaders in our field to give high, proper credit to reviewer contributions when assessments for appointments, promotions, and overall stature in the field are requested, especially since reviewers’ roles may be significantly expanded in the near future. Scientific publishing is changing in many positive ways, and informatics is playing a significant role in this evolution. Promotion of resource sharing to advance discoveries of current and future generations of scientists is an important component of JAMIA’s mission. Reproducibility of science is now facilitated through the development of repositories that are able to index, preserve, and disseminate the critical components of analytic processes. Journals publish the main algorithms and findings as well as discuss implications of studies, while repositories index and provide the ingredients (data and software) and full “recipes” (e.g., analytical workflows) to ensure reproducibility of results. The combination of technology and policy facilitates access to data and software environments, allowing team science to expand well beyond a single institution or a small group of collaborators. Combined with incentives for sharing and appreciation for team science, these advances have the potential to accelerate discoveries in ways that would not have been dreamed possible just a few years ago. It is thus important that journal publishers, editors, reviewers, authors, and owners take a minute to reflect on their roles and responsibilities in bringing novel, reproducible, and sustainable knowledge to our readers. Articles as we know them today will not disappear, since the ever-increasing volume of data and journals calls for a process in which the most relevant materials are selected, edited, and disseminated to specific audiences. However, many informatics journal articles should be accompanied by data, software, and/or systems that will help reproduce the science and facilitate reuse for new analyses. Identifying reviewers with the expertise and time to verify and reproduce results is very challenging, as it goes well beyond their current call of duty. On the other hand, there is no community more skilled than our own to serve this role. JAMIA plans to continue to be at the forefront of innovative informatics publishing. Every year we publish many articles that report on new software systems, data analyses, and new approaches to organize and evaluate information to generate new knowledge. JAMIA promotes reproducibility and reuse of these new algorithms, software, data, workflows, and knowledge in addition to evaluation of systems and dissemination of new perspectives. Certain software may be proprietary and some data may not be shared publicly, but it is important that working versions be available for reviewers to assess quality and to verify results. A variety of data and software repositories exist and JAMIA does not prescribe the use of a particular one, particularly given that some data need to be placed under special access controls to protect patient and institutional privacy. However, we have been increasingly requesting proper deposition into repositories, as authors’ web sites lack sustainability given fluctuations in funding, changes of institution, etc. We are also giving more attention to the citation of such resources in the manuscripts we review than we had in the past, with the anticipation that citations to data and software may become as important as citations to articles. This opens a new set of opportunities to study and develop bibliometric equivalents that can be used to assess academic progression and assist in funding decisions. Here again the informatics community is well suited to address this challenge. A critically important aspect of reproducibility is the use of standards, as it is only feasible to perform analyses when data are represented in a way that makes them comparable. This issue of the journal addresses a fundamental aspect of reproducibility, which is its dependence on the use of biomedical and healthcare data standards. Guest editors Richesson and Chute (see page 492) describe how each of the selected articles fills an important gap in our field. We hope you will enjoy this issue of JAMIA and join us in the pursuit of ever increasing reproducibility and reuse of resources unveiled in this journal.

A “one-size-fits-all” application no longer works to meet user needs and expectations as multiple factors dictate adoption by an individual user, including costs, efficacy, and convenience, among many others. In this issue of JAMIA, we focus on tailoring informatics interventions to patients as well as clinicians. Many of these interventions are based on the information available in the electronic health record (EHR). In Brief Communications, Dalal (see page 905) reports on the importance of following up on nonurgent results in the EHR, while Li (see page 896) describes a large experiment in “phenotype profiling” conducted in Taiwan. Heintzman (see page 909) discusses how EHRs are still missing important data related to health insurance, while Dixon (see page 917) comments on informatics implications to significantly expand care of Veterans and other populations outside the main health systems that have historically cared for them. In Perspectives, Hripcsak (see page 921) advocates for more informatics support for social and behavioral domains and measures, and Chung (see page 914) describes the need for incorporating patient-reported outcomes into the EHR. The Research and Applications articles cover health IT needs in medical homes (see page 815), clinical decision support for management of asthma (see page 773), tailoring of alerts (see page 881) and risk assessment (see page 872) in kidney disease, EHR- (see page 755) and personally controlled information systems (see pages 748, 805), readability of discharge instructions (see page 857), patient portal utilization (see page 888), and patient preferences towards data sharing (see page 821). Sheikh (see page 849) shows how the aims of healthcare reform may leverage health IT, while Ancker (see page 864) measures the association between EHR use and healthcare quality. Hripcsak (see page 794) proposes how to parameterize time in EHR studies. The important sub-specialty of Pharmacy Informatics is represented by articles related to return on investment for drug–drug interaction alerting systems (see page 764), cost–effectiveness of an electronic medication management system (see page 784), management of duplicate medication alerts (see page 831), and user perceptions of an electronic network for prescriptions (see page 838). Reviews in the areas of test results display (see page 900) and clinical information modeling (see page 925) complete this issue of the journal. We hope to continue to provide our readers with the best scholarly work in informatics through the rigorous selection of manuscripts such as the ones featured here. We take this opportunity to thank Chuck Friedman for his many years of service to the journal and to thank Sue Bakken for accepting the role of associate editor. Sue, whose work was published in the first issue of JAMIA in 1994 and has served on the editorial board for many years, is joining a team of outstanding associate editors. Johanna Westbrook has kindly accepted to serve as assistant editor for the journal. I also want to thank Michele Day for editing the Highlights section of JAMIA and many guest associate editors who, during the past 4 years, have helped bring new topics and perspectives to the journal.

Electronic health records (EHRs) have been increasingly adopted in the United States because of governmental incentives and the realization that healthcare should not lag behind other industries in deriving knowledge from “big data.” There has been much discussion about the use of big data to discover patterns for targeted therapy and disease prevention. However, much less is said about the readiness of EHR data for such data mining initiatives. The existence of data is often equated with the existence of good data, ie, high-quality, standardized data that can be used in sophisticated data analyses to reveal patterns that previously escaped observation. Although the difficulties of preparing data for such analyses are well known to informaticians, biostatisticians, and computer scientists specializing in machine learning, not all stakeholders (ie, administrators, clinicians, researchers, and patients) appreciate the challenges of using data from different health systems for meaningful analyses. It is not uncommon for such stakeholders to assume that, if health systems utilize the same software versions of the same EHR system, then the data will be immediately comparable. Because bringing data from different health systems together is so difficult, due to privacy concerns and institutional policies, many believe that the investment in bringing the data “together” (in a federated or in a centralized way) is sufficient to allow immediate analyses. This is not so. The inconvenient truth is that much needs to be done to EHR data before they can be used for analyses and decision-making – a topic that has been a focus of the informatics community for many years. The first differences that becomes apparent when data from different health systems are brought together relate to the overall format of those data. Data format in and of itself can be relatively easy to fix, but addressing health systems’ heterogeneous utilization of ontologies and terminology standards and differences in semantics (eg, the various definitions of “uncontrolled diabetes”) requires intervention from data modeling experts as well as biomedical or behavioral domain experts. While the last issue of JAMIA focused on data standards, this issue focuses on the transformation of narrative text into structured data using natural language processing techniques. This fundamental preprocessing step towards data mining extracts concepts from clinical notes and the body of biomedical research literature (see pages 938 to 1020 for articles in our Special Focus on natural language processing). Once data are harmonized and structured, as well as processed to prevent re-identification (see page 1029, see page 1072), they can potentially be used for various types of data mining, such as malpractice mitigation (see page 1020). More examples of data mining initiatives are included in this issue: associating month of birth (as a proxy for seasonal maternal-infant exposures) with certain diseases (see page 1042), and risk stratification for acute kidney disease (see page 1054). The current status and future direction of EHR systems should be discussed further. EHR systems are currently used for public health monitoring, eg, tuberculosis contact investigation (see page 1089) and have also been used extensively for quality improvement and clinical decision support (see page 1081), eg, to decrease the rate of adverse events via e-prescribing (see page 1094). However, these types of activities can only be effective across EHR systems if those systems are truly interoperable (see page 1099) and used properly. For this to happen, EHR systems need to be designed with end users in mind (see page 1102). Our informatics community is working relentlessly to prepare EHR data for data mining and is uniquely positioned to evaluate EHR data's quality and usefulness in a variety of applications. JAMIA will continue to publish “gold nuggets” found in the course of data mining as well as negative results that significantly contribute to the body of informatics knowledge.

Every two years since 2012 we have been issuing analyses of JAMIA publications. As a general informatics journal, JAMIA’s publication trends provide an indication of how the informatics field is evolving in terms of topics, outreach, and overall impact. This year’s analysis (see page 1153 ) shows how we automated some laborious portions of the analysis, promoted new topics and reconstructed our list of most highly cited articles since 2009. The analyses show that the landscape of informatics is changing fast. Cross-institutional informatics activities such as health information exchange (HIE) and clinical data research networks are on the rise. In this issue, we present brief communications describing center projects involving biomedical big data (see pages 1115 – 1152 ) and articles on HIE (see pages 1169 , 1183 ). We also present novel contributions in distributed analytics and web services (see pages 1187 , 1212 , 1271 ) as well as phenotyping algorithms (see pages 1220 , 1251 ). These algorithms can be used for population health management and research, and have received increasing attention from our readers since the rapid uptake of electronic health records (EHRs) triggered by the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009. No longer confined to a few institutions, EHRs are increasingly utilized for applications such as drug management (see pages 1196 , 1243 , 1261 ) and will soon be integrated with clinical genomics applications (see pages 1173 , 1231 ). Usually connected to clinical decision support systems, several informatics applications, such as the ones presented here, have the potential to improve healthcare as well as clinical and translational research if presented in a user-friendly manner (see page 1179 ). These applications rely on data that can be compared across systems. Standardization of codes and utilization of proper ontologies is thus critical to their success. Articles on laboratory code standardization (see page 1205 ) and on integration of ontologies for rare diseases and radiological diagnoses (see page 1164 ) are illustrative of work in this area. JAMIA values a diverse representation of topics and has increasingly promoted a diverse representation of institutions and countries. We rely on a large team of editors, editorial board members, reviewers, and editorial staff to handle submissions from all over the world. The current editorial team will be celebrating five years of service with JAMIA in December. Over the years, we have processed over 6,500 manuscripts and had the opportunity to work with new and experienced authors, associate editors, editorial board members and reviewers. Several decisions are made on a daily basis, involving a very large team. Figure 1 illustrates our workflow and our target turnaround times. Peer-review process workflow and target duration of review activities. We live in an era in which the voluntary reviewer system and the heightened competitiveness of the academic and industrial worlds provide few incentives for reviewer participation. Timely and thorough review of meritorious manuscripts is one of our main goals, but it sometimes presents challenges. Publication and peer-review models are changing fast and finding the one that best suits the interests of our stakeholders while also keeping realistic expectations of volunteer reviewers ability to provide timely contributions is not an easy task. We recommend that academic leaders credit the scholarly work involved in thoroughly reviewing a manuscript when considering appointments and promotions. We receive periodic feedback from AMIA leaders and from our publisher. We encourage readers and authors to communicate with the JAMIA editorial office (in an identified or de-identified way) to provide feedback on current processes and to share ideas for improvements.

OBJECTIVE: To propose a new approach to privacy preserving data selection, which helps the data users access human genomic datasets efficiently without undermining patients' privacy. METHODS: Our idea is to let each data owner publish a set of differentially-private pilot data, on which a data user can test-run arbitrary association-test algorithms, including those not known to the data owner a priori. We developed a suite of new techniques, including a pilot-data generation approach that leverages the linkage disequilibrium in the human genome to preserve both the utility of the data and the privacy of the patients, and a utility evaluation method that helps the user assess the value of the real data from its pilot version with high confidence. RESULTS: We evaluated our approach on real human genomic data using four popular association tests. Our study shows that the proposed approach can help data users make the right choices in most cases. CONCLUSIONS: Even though the pilot data cannot be directly used for scientific discovery, it provides a useful indication of which datasets are more likely to be useful to data users, who can therefore approach the appropriate data owners to gain access to the data.

SUMMARY: MAGI is a web service for fast MicroRNA-Seq data analysis in a graphics processing unit (GPU) infrastructure. Using just a browser, users have access to results as web reports in just a few hours->600% end-to-end performance improvement over state of the art. MAGI's salient features are (i) transfer of large input files in native FASTA with Qualities (FASTQ) format through drag-and-drop operations, (ii) rapid prediction of microRNA target genes leveraging parallel computing with GPU devices, (iii) all-in-one analytics with novel feature extraction, statistical test for differential expression and diagnostic plot generation for quality control and (iv) interactive visualization and exploration of results in web reports that are readily available for publication. AVAILABILITY AND IMPLEMENTATION: MAGI relies on the Node.js JavaScript framework, along with NVIDIA CUDA C, PHP: Hypertext Preprocessor (PHP), Perl and R. It is freely available at http://magi.ucsd.edu.

The database of genotypes and phenotypes (dbGaP) developed by the National Center for Biotechnology Information (NCBI) is a resource that contains information on various genome-wide association studies (GWAS) and is currently available via NCBI's dbGaP Entrez interface. The database is an important resource, providing GWAS data that can be used for new exploratory research or cross-study validation by authorized users. However, finding studies relevant to a particular phenotype of interest is challenging, as phenotype information is presented in a non-standardized way. To address this issue, we developed PhenDisco (phenotype discoverer), a new information retrieval system for dbGaP. PhenDisco consists of two main components: (1) text processing tools that standardize phenotype variables and study metadata, and (2) information retrieval tools that support queries from users and return ranked results. In a preliminary comparison involving 18 search scenarios, PhenDisco showed promising performance for both unranked and ranked search comparisons with dbGaP's search engine Entrez. The system can be accessed at http://pfindr.net.

There is currently limited information on best practices for the development of governance requirements for distributed research networks (DRNs), an emerging model that promotes clinical data reuse and improves timeliness of comparative effectiveness research. Much of the existing information is based on a single type of stakeholder such as researchers or administrators. This paper reports on a triangulated approach to developing DRN data governance requirements based on a combination of policy analysis with experts, interviews with institutional leaders, and patient focus groups. This approach is illustrated with an example from the Scalable National Network for Effectiveness Research, which resulted in 91 requirements. These requirements were analyzed against the Fair Information Practice Principles (FIPPs) and Health Insurance Portability and Accountability Act (HIPAA) protected versus non-protected health information. The requirements addressed all FIPPs, showing how a DRN's technical infrastructure is able to fulfill HIPAA regulations, protect privacy, and provide a trustworthy platform for research.

BACKGROUND AND OBJECTIVE: Short-read sequencing is becoming the standard of practice for the study of structural variants associated with disease. However, with the growth of sequence data largely surpassing reasonable storage capability, the biomedical community is challenged with the management, transfer, archiving, and storage of sequence data. METHODS: We developed Hierarchical mUlti-reference Genome cOmpression (HUGO), a novel compression algorithm for aligned reads in the sorted Sequence Alignment/Map (SAM) format. We first aligned short reads against a reference genome and stored exactly mapped reads for compression. For the inexact mapped or unmapped reads, we realigned them against different reference genomes using an adaptive scheme by gradually shortening the read length. Regarding the base quality value, we offer lossy and lossless compression mechanisms. The lossy compression mechanism for the base quality values uses k-means clustering, where a user can adjust the balance between decompression quality and compression rate. The lossless compression can be produced by setting k (the number of clusters) to the number of different quality values. RESULTS: The proposed method produced a compression ratio in the range 0.5-0.65, which corresponds to 35-50% storage savings based on experimental datasets. The proposed approach achieved 15% more storage savings over CRAM and comparable compression ratio with Samcomp (CRAM and Samcomp are two of the state-of-the-art genome compression algorithms). The software is freely available at https://sourceforge.net/projects/hierachicaldnac/with a General Public License (GPL) license. LIMITATION: Our method requires having different reference genomes and prolongs the execution time for additional alignments. CONCLUSIONS: The proposed multi-reference-based compression algorithm for aligned reads outperforms existing single-reference based algorithms.

This is a very special year for JAMIA, which was founded 20 years ago by Dr William Stead, Editor-in-chief from 1994 to 2002, and subsequently led by Dr Randolph Miller, Editor-in-chief from 2003 to 2010. Following the steps of these remarkable editors, we strive to produce a balanced journal whose articles represent the best work in our field. Consistent with JAMIA's mission of publishing prime informatics articles after rigorous peer-review, and also consistent with JAMIA's vision of disseminating biomedical and health informatics work, our goal is to continue to expand our readership so this work reaches out to a broad audience in the USA and abroad. Our readers come from different backgrounds and represent important driving forces in informatics, healthcare, and biomedical science. Regardless of whether they are academic-, government-, or industry-based, research- or application-focused, skilled in technical or non-technical aspects of informatics, decision makers, creators, implementers, or evaluators, our readers are seeking information that will transform the way they go about their daily businesses. JAMIA's responsibility is to provide them with high quality information. The collective influence of our readers in shaping the future of healthcare through changes in technical, sociological, or policy-related issues is enormous, particularly at a time when relevant changes in healthcare policy, in public engagement, and in the global economy are all happening simultaneously and are highly dependent on technology. We are fortunate to live in a time when it is possible to find out what people want, develop technology that can be delivered to fit their needs, and implement policy that can help deliver this technology when and where it is most needed. As we are living through these changes now, it may be hard for us to notice them, but 20 years from now we will look back and understand how much was changed and what a critical role informatics had in advancing healthcare and biomedical science. This important period of unprecedented growth in informatics is reflected in JAMIA articles that not only document scientific innovations, but also drive discussion on what kind of innovation is needed and how it should be deployed in real settings. We are often asked what kind of articles are within scope of the journal. JAMIA is a generalist journal in biomedical informatics: we value innovations in informatics that constitute (A) novel methods or approaches to solve difficult problems, (B) original applications that result in new knowledge, (C) reviews that synthetize the literature and provide insightful discussion of timely topics (eg, systematic reviews when the literature is mature enough, scoping reviews for less mature literature, tutorials on ‘must-know’ topics of high significance to our readers), (D) generalizable case reports describing how an institution or group of investigators successfully addressed a significant problem, (E) brief communications on highly innovative projects, (F) insightful perspectives that spark discussion on critical topics, (G) invited editorials that help contextualize journal issues that have a special focus, or (H) correspondence that documents controversy in areas of high interest and relevance. In this issue of JAMIA, we present a large array of article categories. In Perspectives, Gaynor (See page 2) starts with a provocative opinion on how common carrier and neutrality principles that are applicable to the telecommunication industry should apply to the nationwide health information network. Huser (See page 8) advocates for electronic health record (EHR) donation for research, and Rudin (See page 13) presents a vision for care coordination using EHRs. In Brief Communications, Kang (See page 17) describes the impact of social media in cardiovascular care, Gupta (See page 23) assesses the accuracy of clinician-reported data in clinical decision support, Khor (See page 27) describes a smoking detection system that uses a small training corpus, and Kim (See page 31) evaluates a novel system that enhances phenotype-based searches in dbGaP (See page 31). In Case Reports, Garrido (See page 181) studies publicly reported e-Measures of healthcare quality, and Hurdle (See page 185) describes a single center's experience in implementing infrastructure for hosting research data. In Reviews, Abraham (See page 154) and Topaz (See page 163) systematically report on the literature related to hand-off tools and the Omaha system, respectively, while Schoenbill (See page 171) discusses ethical, logistical and technological considerations when handling genetic data in EHRs. Correspondence from Gospodarevskaya (See page 190) calls for a framework for categorizing economic evaluations of health information systems, a topic that has experienced a significant rise in JAMIA publications in the past few years. Finally, a diverse set of Research and Applications articles address important topics such as organization and quality of electronic health data (See pages 49, 64, 82, 90, 97, 111), users' perspectives (See pages 37, 56, 73, 117, 146), policy (See page 111), and various topics in data mining and clinical decision support (See pages 105, 132, 139, 146). It is currently not possible to publish all materials that some segments of the JAMIA readership could be interested in. The editorial team has to prioritize high quality manuscripts in terms of their novelty, applicability and impact of findings, as well as relevance to our diverse and broad readership, in addition to the manuscript's scientific merit. Together with AMIA leaders, we will seek models to accommodate meritorious submissions that currently cannot fit into the journal's issues due to the overwhelming increase in the number of articles that have been submitted to journal lately, as well as to current constraints on number of pages we can publish per year. In other words, JAMIA itself will need to innovate in order to keep up with changing times. I believe that the role of a scientific journal such as JAMIA is to promote change, document it, and disseminate it to the broadest possible audience. I stand in the shoulders of those who made JAMIA the great journal it is today, and I am honored to be leading it together with an outstanding editorial team during a most exciting time for our field.

Two influential reports on data and computation from advisory committees to the NIH leadership resulted in important initiatives: (1) the report from the Working Group on Biomedical Computing for the Advisory Committee to the Director (ACD) of the National Institutes of Health (NIH) in 1999i led to the Biomedical Information Science and Technology Initiative (BISTI), and (2) the more recent report from the Working Group on Data and Informatics for the ACD in 2012ii led to the Big Data to Knowledge (BD2K) initiative.iii Several AMIA members participated in this working group. Both reports recommended strong support for data science and computation in biomedical sciences and healthcare. The BD2K initiative was launched at NIH in 2013 through the development of several focused workshops, calls for proposals for centers of excellence, for a data discovery index, for training programs, and through the creation of the new position of Associate Director of Data Sciences, reporting directly to the NIH director. According to Francis Collins, the charge is to “lead an NIH-wide priority initiative to take better advantage of the exponential growth of biomedical research datasets, which is an area of critical importance to biomedical research. The era of ‘Big Data’ has arrived, and it is vital that the NIH play a major role in coordinating access to and analysis of many different data types that make up this revolution in biological information.” The first NIH Director of Data Sciences is a member of AMIA and an elected fellow of the American College of Medical Informatics. Phillip Bourne, former Editor-in-Chief of PLoS Computational Biology, developer of the Protein Data Bank, and Professor of Pharmacology at the University California San Diego, introduces this special Big Data focus issue of JAMIA with an editorial describing his vision of a ‘Digital Enterprise’. Bourne's vision reflects a new reality in which informatics has moved from the periphery of the healthcare and biomedical research enterprise to the center of action. JAMIA has been documenting solutions to data acquisition, management, and knowledge generation, and will be a premier venue for reporting on BD2K and related activities. In this first special issue of JAMIA on Big Data, we present creative solutions to challenges in data acquisition, organization, and analysis, with a particular emphasis on electronic health record data. Technical and policy infrastructure for data acquisition, efficient storage, and management. Articles by LeDuc et al (See page 195), White et al (See page 379),iv and Sahoo et al (See page 263) focus on technical infrastructure for research data, and articles by Bloomrosen et al (See page 204) and Akagu et al (See page 374) focus on secondary use of healthcare data, from a sociotechnical perspective, which includes privacy concerns. Goldwater et al (See page 280) describes how the acquisition of electronic health data can be feasible in institutions that compose the U.S. federal safety net. EHRs are often distributed, so techniques for record linkage are important to integrate the data – Kum et al (See page 212) and Rajasekaran et al (See page 252) describe algorithms for this task. Additionally, new data modalities are increasingly augmenting the EHR, and some of these data can challenge current organizational structures and storage capabilities. Tenenbaum et al (See page 200)iv discusses standards for ‘omics’ data, and Li et al (See page 363) describes a novel algorithm for genomic data compression. Data processing and organization. EHR phenotyping, which was the focus of JAMIA's December 2013 issue, refers to data processing for accurate characterization of disease status and health conditions using data collected in the process of care. A review by Shivade et al (See page 221) provides the context for EHR phenotyping, and articles by Tate et al (See page 292) and Melton et al (See page 299) describe algorithms for EHR phenotyping based on structured data and resources for structured data derivation from clinical notes. Rosenman et al (See page 345) focuses on database queries on hospitalizations for acute congestive hearth failure, while Dentler et al (See page 285) and Perotte et al (See page 231) focus on quality measures and diagnosis code assignment on EHRs. Knowledge generation. The articles by Iyer et al (See page 353), Friedman et al (See page 308), and Liu et al (See page 245) focus on detecting drug-related adverse events based on EHRs and related sources. Huston et al (See page 238) uses data mining techniques to suggest drug repurposing. Data mining techniques for predicting hospital readmissions, early detection of neonatal sepsis, outcome of septic patients, and changes in hypertension control are presented in articles by He et al (See page 272), Mani et al (See page 326), Tagkopouls et al (See page 315), and Sun et al (See page 337), respectively. Big Data is a big deal in biomedical research and healthcare. I hope our readers enjoy this special issue and continue to submit the products of Big Data initiatives such as BD2K for dissemination through JAMIA. In the highly diverse biomedical informatics community, professionals with expertise in library science, statistics, management, computer science, software engineering, natural language processing, and implementation science focus on biomedical and healthcare data. Our community is uniquely positioned to translate these data into actionable knowledge to promote health and advance science.

The utilization of Clinical Decision Support (CDS) systems is increasing with implementation of Electronic Health Record (EHR) systems across the USA. One of the first and most studied CDS applications has been computerized provider order entry (CPOE), particularly CPOE related to medications. Many EHR systems check drug dosage, allergies, and drug-drug interactions as clinicians enter their orders, and issue alerts as needed. However, alert overriding in CPOE systems is common practice. Nanji et al (see page 487) reports that over 50% of alerts are overridden at a particular academic health center and proposes how to decrease unintended CPOE effects on workflows and communications (see page 481). Related to workflows, Coiera (see page 414) offers a model to study workspaces in which critical communications among clinicians occur. Many articles in this issue focus on medication management, including three systematic reviews that address barriers and facilitators for CPOE implementation (see page 535), quality of mobile applications to support medication self-management (see page 542), and identification of adverse drug events (see page 547). Tamblyn et al (see page 391) reports on the large gap between medication records of community-based pharmacies and reports of medications from emergency department records. Dixon et al (see page 517) provides an informatics approach to narrow this type of gap by combining information from multiple sources, including patient-reported data. Miller et al (see page 564) describes an innovative CPOE tool supporting clinical research and quality improvement, while Woods et al (see page 569) reports on a decrease in atypical medication orders with the introduction of alerts. Other articles in this issue address different types of CDS systems. Bellows et al (see page 432) reports on a CDS system that individualizes clinical practice guidelines. A test result notification system is evaluated in a randomized controlled trial by Dalal et al (see page 473), and recommended practices for those involved in CDS systems are described by Wright et al (see page 464). Collins et al (see page 438) describes functional specifications for maintaining continuity of care across different types of healthcare providers. Genetic and environment information are being collected at a rapid pace and will likely be more present in EHRs in the near future. Goldspiel et al (see page 522) provides solutions for integrating pharmacogenetic information and CDS into the EHR. Martin-Sanchez et al (see page 386) makes the case for increased inclusion of environmental factors in biomedical research information systems. Wall et al (see page 399) presents a new search tool for extraction of disease-associated genes from the literature. A balance of structured and unstructured data in the EHR will also likely be achieved in the near future. Structured data in EHRs are important to facilitate analysis and research. However, inclusion of unstructured data in the EHR is also important to ensure that the some aspects of healthcare provision are well accommodated. JAMIA is the premier venue for discussion on the appropriate balance of structured and unstructured data in EHRs. Morrison et al (see page 492) reports on an exploratory national evaluation of benefits and risks of increased structuring and coding of the EHR, while articles by Solti et al (see page 406), McDonald et al (see page 423), and Dligach et al (see page 448) report on elegant NLP strategies to structure information in narrative clinical text. Examples of data analytics that rely on structured data are also provided in this issue of JAMIA. Galvez et al (see page 529) presents a visual analytical tool for quality improvement, and Tabak et al (see page 455) and Hauskrecht et al (see page 501), respectively, describe a machine learning tool and new algorithm to enable clinical predictive modeling. Finally, in order for EHR and CDS systems to have a true impact, we need to train users and developers. To increase utilization of EHRs and build capacity in health information technology (HIT), Landman et al (see page 558) reports on the use of a simulation center to increase EHR usability, while Mohan et al (see page 509) evaluates the US-based ONC HIT curriculum. We hope our readers will enjoy this issue of the journal and will stay tuned for the next one, where we will present how EHR-derived data will be connected through clinical data research networks that are beginning to build a national infrastructure for patient-centered outcomes research.

This special focus issue on research data networks starts with a collection of articles describing a large initiative in the US that will use electronic health record data for patient-centered outcomes research in a privacy-preserving manner. This type of research will include observational and interventional studies. PCORnet, funded by the Patient-Centered Outcomes Research Institute (PCORI), leverages investments of several agencies such as the NIH, AHRQ, and FDA, as well as institutional support from healthcare systems to build a ‘network-of-networks' aimed at helping researchers answer questions that matter most to patients and their caregivers. An editorial by leaders of the NIH (See page 576) embodies the excitement that has permeated the biomedical science, health services research, and informatics communities around the big challenge of connecting highly diverse systems into a national network. Taken together, the 11 clinical data research networks (CDRNs) and the 18 patient-powered research networks (PPRNs) will have the potential to analyze de-identified data on over 100 million unique individuals located in all US states and territories. The PCORnet program is introduced by the PCORI leadership and associated coordinating center members (See page 578), and each of the 11 CDRNs based on health systems is described in a brief communication (See pages 587, 591, 596, 602, 607, 612, 615, 621, 627, 633, 637). The participating health systems will contribute their infrastructure, expertise, and processes to run pragmatic trials and establish cohorts that can be followed over time for comparative effectiveness research. PPRNs are another critical component of PCORnet. An article written by a consortium of PPRN and PCORI leaders (See page 583) summarizes the main characteristics of networks that originate from patient groups. The diversity of PPRNs is impressive: from networks that focus on rare diseases to ones that focus on several genetic disorders to those that focus on common conditions. These networks represent a bottom-up approach to organizing patient-reported outcomes, establishing a mechanism so that patients can actively participate in research, and posing questions that matter to patients. Several other articles address different aspects of data quality and organization (See pages 642, 692, 720, 758); health information networks (See pages 714, 650, 671, 730, 699); and personal health records, portals, and patient engagement activities (See pages 657, 664, 679, 687, 725, 737, 742, 751, 707). They help describe the context for the development and implementation of related initiatives. Several informatics leaders are directly involved in PCORnet, and the whole informatics community can participate in this effort. The collective knowledge of informatics experts has the potential to put into action many of the best practices in data modeling, privacy technology, distributed computing, software and social engineering, and data sharing policies that JAMIA has been publishing for a long time. These best practices result from investments from many agencies that were able to recognize the value of big data well before the concept became a national priority. Large initiatives such as PCORnet also pose some challenges: the ability to develop and combine different solutions that are suited for different health systems will be highly dependent on striking the right balance between flexibility and homogeneity. A one-size-fits-all approach is not likely to generalize beyond a few medical centers, and accommodation of every possible type of data model and technology would delay progress. The challenge is not just technological: many different data models, network software, analysis tools, and policies exist and every institutional prefers to keep using the ones already implemented. However, these models and tools do not cover some new types of data and new sharing models, and hence there are important gaps that need to be filled. We are facing important challenges in achieving true interoperability and developing a path that accommodates institutional preferences without losing sight of the common goal of using data to develop knowledge that helps promote health. The informatics community can be an important partner to patients, clinicians, administrators, and researchers in health services, behavioral, and biomedical science. Due to HITECH, we now finally have EHRs implemented in most healthcare settings in the US. It is important to synergize efforts to avoid duplication and waste of resources as we develop new ways to use these data for research in a way that respects patient privacy.

The increasing adoption of Electronic Health Records (EHR) systems in the USA has resulted in many new studies. In this issue of JAMIA, Krist (See page 764) reports on needed functionality to better support primary care, while Feeley (See page 772) describes how cancer care can benefit from information technology. However, it is well known that the quality of EHR notes is highly variable. Burke (See page 910) proposes an instrument to assess the quality of EHR clinical notes that will help make comparisons across providers and inform the development of systems. We are just starting to quantify and fully understand the limitations of current EHR systems. Hanauer (See page 925) compares associations found in structured diagnoses in clinical datasets and associations found in Medline. The study reveals that there is not much concordance among these sources, which calls for further investigation in this area, especially since electronic surveillance for public health or quality improvement often relies on EHR data. Examples are found in articles by Wang (See page 938) and de Bruin (See page 942), which report on the use of electronic sources for tuberculosis and nosocomial infection surveillance, respectively. There are advantages and disadvantages in the use of narrative text in the EHR. The expressiveness of narrative text will probably not ever be achieved by structured data. However, the paucity of structured data requires that information in narrative text be extracted, but this extraction may lead to significant problems when machines are used to interpret clinical text. Additionally, data in structured fields are not always standardized. Standards are necessary to ensure that data are computable. In this issue of JAMIA, Savaris (See page 917) describes a standardized storage model based on DICOM, Liou (See page 792) describes mappings of laboratory terminologies to LOINC, and Campbell (See page 885) analyzes the adequacy of SNOMED CT to represent histopathology findings. Natural language processing (NLP) applications for structuring clinical text have always been an important subject area for JAMIA. “Phenotyping” (i.e., extracting phenotypes or clinical events of interest from clinical records) is fast developing into one of the most common informatics interventions to prepare data for use in a variety of applications, and most applications rely on NLP. Solti (See page 776) describes an algorithm to identify adverse events and medical errors in a neonatal ICU setting; Marafino (See page 871) reports on the use of support vector machines for classification of diagnoses and procedures; and Abhyankar (See page 801) identifies dialysis patients using narrative text from ICU settings. Several other phenotyping applications help identify patients with influenza (See page 815), epilepsy (See page 866), asthma (See page 876), and on antidepressant medication (See page 785). Additional articles related to medication extraction (See page 858), lymphoma classification from pathology reports (See page 824), classification of radiology reports (See page 893), clinical text classification (See page 850), and assisted annotation of clinical text (See page 833) show the breadth of NLP applications based on EHRs. Methods for named entity recognition (See page 808), word sense disambiguation (See page 842) and for extraction of patterns from online medical forums (See page 902) also appear in this issue. As our field matures, it is gratifying to see how health professionals are increasingly aware of the importance of informatics, and how it impacts healthcare, biomedical research, and public health. With this recognition comes also the responsibility of learning from past lessons and moving ahead in the most cost effective way. Healthcare is changing fast and informatics innovations are critical to improve patient care, accelerate biomedical research, and promote public health.

The patient-centered movement started not long ago and has been accompanied by an increasing number of publications describing different aspects of patient-focused healthcare and research. For example, a study in this issue of the journal reports on characteristics of individuals who tweet weight loss attempts (see page 1032), and another provides a guide to conducting treatment fidelity of mHealth-based systems (see page 959). Patients use the Internet not only as a means to seek help independently, but also to provide information to and about healthcare systems. For example, patients amend electronic health records (EHRs) (see page 992) and rate physicians online (see page 1098). This connected health behavior has raised concerns that a large “digital divide” could promote or exacerbate health disparities. However, there is evidence that patients may be trying to bridge the health access gap by seeking health information online (see page 1113). Systems to promote access to sophisticated healthcare interventions such as bone marrow transplantation have been successfully implemented in developing countries (see page 1125). Improving public insurance application processes through information technology (see page 1045) showed the importance of connectedness. The value of information provided by librarians in healthcare settings (see page 1118) is yet another indication that concerns expressed in the past about technology-induced gaps may be less relevant today. Systems that help provide personalized care (see page 1069) and patient education (see page 1026) can also be considered part of the patient-centered movement. However, we still have some work to do. For example, patient-centered longitudinal care plan systems are envisioned for the future, but solutions to bridge vision and reality are needed (see page 1082). EHR systems must also evolve to be more patient-centered. They have been increasingly adopted world-wide, but the major inflection point in the USA resulted from the American Recovery and Reinvestment Act of 2009 that incentivized their meaningful use. The regional extension center program from the Office of the National Coordinator for Health IT supports and monitors meaningful use of EHR systems. Important area-level implementation differences (see page 976), adoption rates and variability among providers (see page 1001) and among hospitals (see page 984) were noted. But adoption is just the beginning: decision makers representing the government, vendors and health systems must still address EHR system interoperability (see page 1060), patient safety concerns (see page 1053), and the fact that not all technology specifications are ready for national standardization. The informatics community has been addressing the problem of structuring the EHR so that it can serve as the basis for clinical decision support. Examples date back several decades when the problem-oriented medical record was developed (see page 963). Several other examples exist, such as EHR data modeling initiatives in an integrated health delivery system (see page 1076), clinical decision support for antibiotic prescribing in upper respiratory infections (see page 1091), temporal trend monitoring in HbA1C testing (see page 1038), electronic systems for test result management (see page 1104), for decreasing the use of haloperidol in high risk patients (see page 1109), and for placement of HIV positive patients in antiretroviral therapy programs (see page 1009). Finally, the clinical research informatics community has developed data repositories (see page 1136) and data quality assessment resources (see page 1129) to facilitate healthcare quality improvement and outcomes research. Genomic data has received a lot of attention lately. However, in most academic health systems there is still a clear separation between basic and translational research involving genome data and clinical research. Computational environments for genomic research (see page 969) typically have different specifications than those required of clinical systems. Strategies that use specific genome information to inform care for a particular patient (see page 1015) will require solutions that introduce both technical and policy innovation into existing systems, or most likely the introduction of completely new systems designed to address the challenges of integrating large amounts of data from molecular, individual, and population levels.

The field of biomedical informatics has experienced an enormous expansion in the past few years. As I have edited JAMIA since 2011, I have had the unique opportunity of following this evolution in terms of the volume, diversity, and impact of manuscripts, partially measured by the number of article downloads and citations.1,2 The remarkable work of previous editors Bill Stead and Randy Miller, of current and past associate editors, of editorial board members, and of volunteer reviewers has ensured JAMIA's position as the premier journal for scholarly biomedical informatics publications. Over the past 4 years, the number of annual submissions and published articles has more than doubled, and the journal has continued to accrue a large number of citations (figure 1). Increase in the number of JAMIA submissions, publications, and citations since 2010. The contributions of our authors are key to making the journal an exciting vehicle for documenting the scientific progress of our multi- and inter-disciplinary profession. A summary of the most cited articles from 2011–2013, as listed in the Web of Science (http://apps.webofknowledge.com), is shown in table 1. These original articles or educational reviews represent some of the best work in JAMIA, and clearly reflect a variety of informatics subspecialties. Among the excellent collection of articles, some deserve special mention for their role in expanding the scope of the journal in new directions. For example, articles by Archer et al3 and Sarkar et al4 described patient-centered systems—even before this movement became widespread—and paved the way for many other outstanding articles. Frisse et al5 were among the first to describe an economic evaluation of health information exchange, and their article was followed by other highly cited articles addressing the economic impact of health IT. Influential phenotyping articles by Kho et al,6 Carroll et al,7 and Newton et al8 reported on outcomes of the eMERGE project. A popular article by Hripcsak and Albers9 provided an excellent guide to ‘next generation’ phenotyping. McGraw10 reported on public trust and the privacy of electronic health records (EHRs), leading the way to a number of articles related to the secondary use of EHRs and associated strategies and technologies to protect patient privacy. Most cited JAMIA articles 2011–2013 Most cited JAMIA articles 2011–2013 The expansion of topics and authors in JAMIA reflects the evolution of informatics in general as well as AMIA initiatives such as the Joint Summits in Translational Science and the annual AMIA Policy Meetings. In the past 4 years, many articles have been published in special focus issues managed by regular and guest associate editors, covering translational bioinformatics,11,12 clinical research informatics,13 privacy,14 imaging,15 EHR phenotyping,16 natural language processing,17 and big data.18,19 These articles were well received by our readers. Special focus issues on visualization and standards that are scheduled for publication in early 2015 will add to this important collection. Large funding initiatives in the USA such as NIH's National Centers for Biomedical Computing led to seminal articles describing specific areas.20–27 These brief communications were highly cited because they made software or data resources available to the public. As our field has evolved, open-source software and data sharing have become critical and JAMIA is an important venue for their dissemination. For example, Natter et al28 described open-source software for disease registries. Overhage et al29 described a data model for comparative effectiveness research that has been adopted by several institutions around the country. We expect to receive future submissions reporting on the NIH Big Data to Knowledge initiative,30 which embodies the spirit of informatics and is a landmark for biomedical and behavioral research in general. JAMIA's mission is not limited to disseminating scholarly work: it also plays an important role in education. An open access monthly journal club features editor's choice articles in webinars presented live by the article authors. Through direct interaction with the authors or by later viewing of recorded sessions, informatics trainees and a broad community of non-informaticians are exposed to the best work in our field (http://healthsciences.ucsd.edu/som/medicine/divisions/dbmi/education/journal-club/Pages/default.aspx). Over 25 000 views from users around the globe attest to the value of this resource. Another important outreach and learning opportunity was provided to informatics trainees who were selected through a rigorous selection process to serve on the student editorial board. This innovative activity, initiated well before I started editing the journal, continues to provide an opportunity for trainees to be guided through the process of conducting a balanced, fair, and constructive review of articles originating from both new and seasoned authors. Additionally, by working directly with the editorial office, 17 guest editors have had the opportunity to understand how the entire editorial process works, from the formulation of calls for papers to triage, peer-review, author revision, and finally journal publication. Editing a journal is a perfect example of a team activity in which each player is essential. Editorial board members and volunteer peer reviewers represent a broad spectrum of expertise areas, institutions, and countries and provide support for all our decisions. Their insightful, constructive, and timely reviews are deeply appreciated. Experienced associate editors coordinate the review of manuscripts in their areas of expertise, contribute new ideas to optimize our processes, and help shape the vision for the journal. The editorial office staff, publisher production team, and AMIA leaders are also critical in making this complex organization run smoothly. They work tirelessly behind the scenes to bring the best possible product to our readers. The large JAMIA team has a clear, common goal in mind: to enrich our reader's knowledge of informatics. We strive to publish articles that will bring new contributions to the informatics community or will help disseminate informatics to new, untapped audiences. Through a diverse combination of original research and application reports, reviews, perspectives, brief communications, and case reports, we are able to disseminate exceptionally innovative and practical ideas describing how informatics in healthcare and biomedical and behavioral research can contribute to the overall goals of providing better healthcare, improving outcomes, and unraveling disease processes and population trends. I remain committed to continuously improving the journal's processes and contents to meet the changing needs of a growing audience. I have been entrusted with an important role in informatics, and thank everyone who made this possible (a complete list of names is not provided, as it is so long that it would violate JAMIA's word limits for an editorial). I manage an all-star team and look forward to serving as JAMIA's editor-in-chief for a second term.

This article describes the patient-centered Scalable National Network for Effectiveness Research (pSCANNER), which is part of the recently formed PCORnet, a national network composed of learning healthcare systems and patient-powered research networks funded by the Patient Centered Outcomes Research Institute (PCORI). It is designed to be a stakeholder-governed federated network that uses a distributed architecture to integrate data from three existing networks covering over 21 million patients in all 50 states: (1) VA Informatics and Computing Infrastructure (VINCI), with data from Veteran Health Administration's 151 inpatient and 909 ambulatory care and community-based outpatient clinics; (2) the University of California Research exchange (UC-ReX) network, with data from UC Davis, Irvine, Los Angeles, San Francisco, and San Diego; and (3) SCANNER, a consortium of UCSD, Tennessee VA, and three federally qualified health systems in the Los Angeles area supplemented with claims and health information exchange data, led by the University of Southern California. Initial use cases will focus on three conditions: (1) congestive heart failure; (2) Kawasaki disease; (3) obesity. Stakeholders, such as patients, clinicians, and health service researchers, will be engaged to prioritize research questions to be answered through the network. We will use a privacy-preserving distributed computation model with synchronous and asynchronous modes. The distributed system will be based on a common data model that allows the construction and evaluation of distributed multivariate models for a variety of statistical analyses.

Previous reference-based compression on DNA sequences do not fully exploit the intrinsic statistics by merely concerning the approximate matches. In this paper, an adaptive difference distribution-based coding framework is proposed by the fragments of nucleotides with a hierarchical tree structure. To keep the distribution of difference sequence from the reference and target sequences concentrated, the sub-fragment size and matching offset for predicting are flexible to the stepped size structure. The matching with approximate repeats in reference will be imposed with the Hamming-like weighted distance measure function in a local region closed to the current fragment, such that the accuracy of matching and the overhead of describing matching offset can be balanced. A well-designed coding scheme will make compact both the difference sequence and the additional parameters, e.g. sub-fragment size and matching offset. Experimental results show that the proposed scheme achieves 150% compression improvement in comparison with the best reference-based compressor GReEn.

In this paper, we develop a novel genome compression framework based on distributed source coding (DSC)[3], which is specially tailored to the need of miniaturized devices. At the encoder side, subsequences with adaptive code length can be compressed flexibly through either low complexity DSC based syndrome coding or hash coding with the decision determined by the existence of variations between source and reference known from the decoder feedback. Moreover, to tackle the variations between source and reference at the decoder, we carefully designed a factor graph based low-density parity-check (LDPC) decoder, which automatically detects insertion, deletion and substitution.

Recent studies recognize that health is influenced broadly by a multitude of factors of different types, including medical, genetic, environmental, social and behavioral factors. Developing successful health interventions therefore requires taking into account all these factors as well as the interactions between them. However, intervention designers have traditionally had access only to a very limited subset of health data (typically medical record data). Other health data, such as environmental or physical activity data, although already collected and stored, have been very difficult to access, since they are maintained by different providers and isolated in their own proprietary silos. This prevents physicians and intervention designers from acquiring a true overview of all factors influencing a condition and acting towards its prevention or cure. To solve this problem, we propose DELPHI: a platform allowing the integration of disparate health data into a single Whole Health Information Model (WHIM), providing a 360-degree view of an individual's health. DELPHI supports the integration of data and thus enables the design of applications and services that utilize the WHIM to offer the next generation of health services. In this paper, we describe DELPHI's architecture, outline the technical challenges encountered and describe an asthma management use case that will be enabled by DELPHI.

UNLABELLED: WebGLORE is a free web service that enables privacy-preserving construction of a global logistic regression model from distributed datasets that are sensitive. It only transfers aggregated local statistics (from participants) through Hypertext Transfer Protocol Secure to a trusted server, where the global model is synthesized. WebGLORE seamlessly integrates AJAX, JAVA Applet/Servlet and PHP technologies to provide an easy-to-use web service for biomedical researchers to break down policy barriers during information exchange. AVAILABILITY AND IMPLEMENTATION: http://dbmi-engine.ucsd.edu/webglore3/. WebGLORE can be used under the terms of GNU general public license as published by the Free Software Foundation.

One of the most exciting research areas in Translational Bioinformatics1,2 is related to the redefinition of fundamental notions of what constitutes a ‘disease.’ Nosology, the systematic classification of diseases, dates back to Carl Linnaeus, with the Genera Morborum3 Today, the improvement in our abilities to make molecular measurements related to health and disease has largely driven the revolution towards personalized medicine. For example, in diseases like non-small cell lung cancer or breast cancer, standard-of-care is now including sequencing of genes such as EGFR or quantitating panels of RNA such as those included in Oncotype DX, respectively, to drive therapeutic decisions for new subtypes of patients. While experts, including those at the National Research Council, are seeing the potential of scaling beyond these early case examples towards redefining our entire nosology,4 it is in the field of cancer where personalized or precision medicine has had best traction. It is no coincidence that many contributions to this special issue of JAMIA focus on cancer. Personalized medicine, also known as precision medicine, has often been equated with the use of molecular measurements to characterize disease. The special feature in this issue of JAMIA challenges this limited view.

OBJECTIVES: We present SHARE, a new system for statistical health information release with differential privacy. We present two case studies that evaluate the software on real medical datasets and demonstrate the feasibility and utility of applying the differential privacy framework on biomedical data. MATERIALS AND METHODS: SHARE releases statistical information in electronic health records with differential privacy, a strong privacy framework for statistical data release. It includes a number of state-of-the-art methods for releasing multidimensional histograms and longitudinal patterns. We performed a variety of experiments on two real datasets, the surveillance, epidemiology and end results (SEER) breast cancer dataset and the Emory electronic medical record (EeMR) dataset, to demonstrate the feasibility and utility of SHARE. RESULTS: Experimental results indicate that SHARE can deal with heterogeneous data present in medical data, and that the released statistics are useful. The Kullback-Leibler divergence between the released multidimensional histograms and the original data distribution is below 0.5 and 0.01 for seven-dimensional and three-dimensional data cubes generated from the SEER dataset, respectively. The relative error for longitudinal pattern queries on the EeMR dataset varies between 0 and 0.3. While the results are promising, they also suggest that challenges remain in applying statistical data release using the differential privacy framework for higher dimensional data. CONCLUSIONS: SHARE is one of the first systems to provide a mechanism for custodians to release differentially private aggregate statistics for a variety of use cases in the medical domain. This proof-of-concept system is intended to be applied to large-scale medical data warehouses.

OBJECTIVE: Privacy-preserving data publishing addresses the problem of disclosing sensitive data when mining for useful information. Among existing privacy models, ε-differential privacy provides one of the strongest privacy guarantees and makes no assumptions about an adversary's background knowledge. All existing solutions that ensure ε-differential privacy handle the problem of disclosing relational and set-valued data in a privacy-preserving manner separately. In this paper, we propose an algorithm that considers both relational and set-valued data in differentially private disclosure of healthcare data. METHODS: The proposed approach makes a simple yet fundamental switch in differentially private algorithm design: instead of listing all possible records (ie, a contingency table) for noise addition, records are generalized before noise addition. The algorithm first generalizes the raw data in a probabilistic way, and then adds noise to guarantee ε-differential privacy. RESULTS: We showed that the disclosed data could be used effectively to build a decision tree induction classifier. Experimental results demonstrated that the proposed algorithm is scalable and performs better than existing solutions for classification analysis. LIMITATION: The resulting utility may degrade when the output domain size is very large, making it potentially inappropriate to generate synthetic data for large health databases. CONCLUSIONS: Unlike existing techniques, the proposed algorithm allows the disclosure of health data containing both relational and set-valued data in a differentially private manner, and can retain essential information for discriminative analysis.

The bioethics advisory committee to the President has recently issued a report emphasizing the importance of protecting health information, particularly the data about an individual's genome.1 The report does not prescribe how to balance the need for sharing information to accelerate discoveries with the potential risk of privacy breach. However, it does mention the potential benefits of data sharing and calls for the development of solutions that minimize the risk of privacy breach. Similarly, a recent report from the NIH's ‘Workshop on Establishing a Central Resource of Data from Genome Sequencing Projects’2 recommends that ‘sequence/phenotype/exposure data sets (be) deposited in one or several central databases.’ Studies on human genomes require good characterization of individual phenotypes, and some of these data may be retrieved from electronic health records. Lessons learned from over a decade of research in privacy technology can help guide solutions to the problem of combining phenotype and genome data in a way that preserves confidentiality. This issue of JAMIA, in addition to several articles we have been publishing in the past few years on technology and policy,3–12 explains regulatory constraints, presents a collection of the latest research results on privacy technology, and displays diverse perspectives on the topic of reusing clinical data for research, healthcare quality improvement, and public health. I am grateful to guest associate editors Malin, O'Keefe, and El Eman for organizing a call for papers and the subsequent reviews for submissions on privacy technology and policy for this special focus issue. These articles represent a variety of subtopics and approaches, ranging from differential privacy (a technical framework that guides safe data disclosure by quantifying the risk of privacy breach to an individual) to discussions on legal aspects of protecting privacy. Malin and his co-guest editors discuss particular articles in their comprehensive editorial.13 This issue of the journal also contains reports on how different institutions have been approaching the utilization of clinical and genomic data for research and public health. Hripcsak (see page 117) discusses how EHRs need to evolve to fulfill current need for comprehensively phenotyping individuals, and Marsolo (see page 122) describes lessons learned when integrating a commercial EHR with research systems. In a provocative article, Witten and Tibshirani (see page 125) express concern about the traditional publishing models, particularly with regards to their inability to ensure that experiments are reproducible. The authors also discuss how the pre-publication review model may be antiquated in an era in which readers have an opportunity to comment on any published article. Farley (see page 128) proposes a platform for biomedical knowledge computing, and Cusack (see page 134) reports on AMIA recommendations for data capture and documentation. Data sharing requires an environment in which the professionals who handle the data adhere to the highest ethical standards and implement systematic processes that (a) measure data quality, (b) respect to consumer preferences, (c) successfully identify research cohorts, and (d) are scalable. The articles by Weiskopf and Weng (see page 144), Ancker (see page 152), Ge (see page 157), Hurdle (see page 164), and Natter (see page 172), address each of these issues, respectively. Cumin (see page 180) provides an excellent example of data sharing, describing two available datasets for anesthetic records. Avillach (see page 184) describes a European experience for harmonization of the process involved in the identification of medical events in healthcare databases. Jones (see page 193) reuses administrative claims data to supplement a state disease registry. Sharing the knowledge obtained from the analyses of the shared data is an equally important endeavor: Kawamoto (see page 199) discusses a potential framework for knowledge sharing in the context of clinical decision support. It is exciting to start the new year with an issue of JAMIA that will certainly generate a lot of discussion and lead to a potential re-examination of several current practices and paradigms. My goal is to continue to promote this discussion throughout the year and to keep an open mind to adapting the journal to new trends in scholarly publishing. New publishing models may serve not only our diverse informatics community, but also the scientific and lay communities at large, extending our journal beyond its current boundaries. Finally, as I complete 2 years of service to JAMIA, I remain indebted to an outstanding editorial team, authors, reviewers, and readers who continue to provide valuable feedback so we can further improve the journal. The author is partially funded by NIH grant U54HL108460. None. Commissioned; not peer reviewed.

Just 10 years ago ‘informatics’ was a relatively unknown concept to the healthcare and biomedical science communities. Over the past years, the field expanded significantly and informatics became not only well known, but also an integral part of daily business in these communities. Accordingly, JAMIA has been continuously expanding the informatics pyramid of published work, from the foundational work upon which important applications are built (such as new algorithms and innovative approaches), all the way to illustrative experiences of system implementations that have generalizable lessons and randomized trials of informatics interventions. Practitioners suggest that JAMIA should publish a larger number of applied articles that describe the design, implementation, or evaluation of systems. Academics suggest that JAMIA should publish a larger quantity of foundational research reports. Defining the boundary can be challenging; what some consider applied work may be considered somewhat theoretical by others. Whether there is a dividing line between theory and practice of informatics and where this line would sit does not really matter (hence our section ‘Research and applications’). What matters is that our current and future readers get value for the time they spend reading the articles or abstracts, that informatics becomes even better known outside its own community, and that the dissemination of information afforded by this journal translates into developments that positively impact healthcare and accelerate biomedical science. In JAMIA, articles from any tier of the pyramid are judged by their quality and innovation, and not by the theoretical or applied nature of the work. This issue of the journal reinforces the point that the journal welcomes diverse kinds of work; we focus on applications of health information technology (IT) and clinical decision support (CDS), while also integrating articles that support some foundations for these areas. This issue starts with Payne discussing health IT and economics (see page 212), Strasberg discussing some of the challenges of making targeted information available at the point of care (see page 218), and Friedman providing a perspective on what informatics is and is not (see page 224). In our section ‘Focus on health information technology’, Vest discusses changes in the electronic health record (EHR) market given health IT certification and meaningful use (see page 227), Harle describes the characteristics of hospitals that successfully responded to meaningful use requirements (see page 233), and Hernández-Ávila evaluates the process of designing and implementing an EHR system for a public health system (see page 238). Once deployed, there are marked differences in EHR use and satisfaction, as well as patient outcomes. Keenan and Hoonakker describe the challenges from the perspective of nurses and other clinicians (see pages 245, 252). Hilligoss describes the impact of EHR on admission handoffs in an emergency department (see page 260), Tundia describes effects on outpatient preventive care (see page 268), and Czaja reports on factors that determine healthcare consumers' use of e-health information sources (see page 277). The underlying infrastructure to achieve high utilization of EHR resources is critically important: Malin describes a practical approach to link medical records scattered in different locations (see page 285). Zunner describes a semi-automated approach to map laboratory concepts into LOINC (see page 293), and Sánchez-de-Madariaga proposes a markup language to facilitate data extraction from EHRs (see page 298). EHR systems can help clinicians do a better job. In our section ‘Focus on clinical decision support’, Adelman reports on a clinical trial evaluating computerized provider order entry (see page 305), and Carroll describes results of a clinical trial of an informatics intervention to screen for maternal depression (see page 311). Mainous reports the impact of CDS on antibiotic prescribing (see page 317), and Torsvik compares visualization techniques for laboratory results (see page 325). Modeling temporal relationships and extracting information from narrative text have always been a challenge to CDS. Hanauer proposes an approach for modeling these temporal relationships (see page 332). Natural language processing approaches are used by Carrell and Quinn to support sharing of de-identified clinical text and to determine cancer types in clinical records, respectively (see pages 342, 349). Jindal utilizes natural language processing to resolve coreferences in biomedical text (see page 356). Wliczynski evaluates the robustness of MEDLINE clinical queries, which help users retrieve answers to their clinical questions from the biomedical literature (see page 363). In ‘Brief communications’, de Bruin describes a CDS for surveillance of ICU-acquired infections (see page 369), and Ríos-Bedoya reports on the effects of using text messaging to screen for alcohol use among adolescents (see page 373). In ‘Case reports’, Martin shares generalizable lessons learned while customizing a commercial EHR CDS module (see page 377), and Sánchez-Mendiola reports on the implementation of biomedical informatics education for medical students (see page 381). Finally, Kawamoto systematically reviews the literature on CDS for genetically-guided personalized medicine (see page 388). As this issue illustrates, the field of biomedical informatics is diverse, complex, and exciting. We strive to make JAMIA work for a diverse, highly demanding readership. Your suggestions and feedback are important so that the premier informatics journal continues to work for you.

The rapid accumulation of large amounts of data from Internet searches, social network sites, and self-monitoring devices, in addition to the widespread adoption of electronic health records (EHR) in the past few years have increased opportunities as well as challenges in effectively integrating and using these data. JAMIA has recently issued requests for papers (RFPs) for two special issues: one is dedicated to EHR Phenotype Extraction, and another one is dedicated to Big Data. The first RFP addresses the problem that, even though they are electronic, the data in EHRs are not always easy to structure and to harmonise across different institutions. Scalable and reliable approaches to extract meaningful phenotypes that can be integrated with genetic and environmental data are urgently needed, and we expect to publish the best work in this area in this special issue. The second RFP relates to ‘meaningful use’ of big health-related data. We will feature the most innovative approaches for efficient storage, pre-processing, analysis, and sharing of data from ‘omics’, EHRs, imaging, and Internet sources of information. In this issue of JAMIA, we also present innovative work that is related to these topics. These articles will intrigue and motivate readers to apply existing algorithms and tools to new problems, as well as to develop new solutions to derive knowledge from data originating from very heterogeneous sources. White (see page 404) uses data originated from a log of web searches, Harpaz (see page 413) and Xu (see page 420) use data from EHRs, and Avillach (see page 446) uses the literature to detect and validate adverse drug events. El Eman (see page 453) proposes a method to combine distributed sources to detect rare adverse drug events while preserving privacy, which is a topic also covered by Mohammed (see page 462). Articles from Phansalkar (see page 489) and Duke (see page 494) cover drug-drug interaction alert systems, while Olsho (see page 470) and Galanter (see page 477) report on the impact of computerised provider order entry (CPOE) systems in the rate of medication errors. Also relating to medications, Fung (see page 482) and McDonald (see page 499) describe systems to extract information from narrative text in drug labels and to automate the medication regimen complexity index, respectively. The Internet has been used as a vehicle to collect information, but also to deliver interventions. Samwald (see page 409) describes a prototype system for patients to carry their pharmacogenomics information in a way that is easily interpretable by clinicians. Patient empowerment is also addressed by Turner-McGrievy (see page 513), who compares traditional versus mobile app self-monitoring of physical activity and dietary intake. Osborn (see page 519) describes patient experiences using web portals and secure messaging for diabetes management, and Tang (see page 526) reports on the impact of engaging diabetic patients in online disease management. Mathieu (see page 568) reviews the quality of Internet-base randomised controlled trials, and Silverstein (see page 535) describes an innovative web-based stereoscopic visualisation system that helps clinicians interactively collaborate at various sites. Today, EHRs constitute a major source of data for biosurveillance. Maslove (see page 427) uses data from EHRs to monitor acquired infections using social network techniques, and Cheng (see page 435) uses these type of data to detect outbreaks using real-time structural models. Surveillance techniques are used by Mahajan (see page 441) to identify cases of Hepatitis B and by Ong (see page 506) to monitor health information system failures. Clinical data warehouses derived from EHRs are the primary sources of this information. Hruby (see page 563) reports on a centralised research data repository for outcomes research, while Tao (see page 554) describes the experience of using the Clinical Element Model to represent concepts for secondary use of EHR data. Maslove (see page 544) describes a method for discretization of continuous features in clinical data to facilitate the use of machine learning applications, among other purposes. Finally, a review by Dixon (see page 577) summarises the literature on communication of data from public health systems to clinicians via EHRs, and a review from Edworthy (see page 584) covers the topic of best practices for the design of medical audible alarm systems. Healthcare and biomedical research are collecting data at a fast pace. Added to the relatively new sources of data originating from Internet use, this constitutes an ideal time to propose disruptive approaches for extracting knowledge from data. We expect to see a lot of highly innovative, and possibly unconventional, submissions to our special focus as well as to our regular issues in the upcoming months.

Recent initiatives from the National Institutes of Health (NIH; http://bd2k.nih.gov) and other federal agencies emphasize that health-relevant big data present unique challenges. Professionals who specialize in handling these challenges, that is, data scientists, are in very high demand in all disciplines. Although the name ‘data scientist’ could imply a significant difference between those who handle data versus those who handle information (ie, processed data), when biomedical, healthcare, and health behavioral data are concerned, there is no distinction: biomedical informatics is biomedical data science. It is great to see that agencies now recognize the value of our discipline and rightfully place it at the top of their priority list. An increasing number of academic institutions are following suit. There are many reasons why the awareness of biomedical informatics as a scientific discipline has raised in the past few years. Biomedical and behavioral research have been deeply impacted by technologies that enable rapid and relatively inexpensive data collection. New collection sources that generate big data are becoming more and more important and are beginning to merge with more traditional biomedical data sources …

The increasing adoption of electronic health records (EHRs) and the corresponding interest in using these data for quality improvement and research have made it clear that the interpretation of narrative text contained in the records is a critical step. The biomedical literature is another important information source that can benefit from approaches requiring structuring of data contained in narrative text. For the first time, we dedicate an entire issue of JAMIA to biomedical natural language processing (NLP), a topic that has been among the most cited in this journal for the past few years. We start with a description of a contest to select the best performing algorithms for detection of temporal relationships in clinical documents (see page 806), followed by a general review of significance and brief description of commonly used methods to address this task (see page 814). Top performing approaches are featured in seven articles from five different countries—Canada (see page 843), China (see page 849), France (see page 820), Serbia (see page 859), and the US (see page 828, 836, 867).

OBJECTIVE: Competing tools are available online to assess the risk of developing certain conditions of interest, such as cardiovascular disease. While predictive models have been developed and validated on data from cohort studies, little attention has been paid to ensure the reliability of such predictions for individuals, which is critical for care decisions. The goal was to develop a patient-driven adaptive prediction technique to improve personalized risk estimation for clinical decision support. MATERIAL AND METHODS: A data-driven approach was proposed that utilizes individualized confidence intervals (CIs) to select the most 'appropriate' model from a pool of candidates to assess the individual patient's clinical condition. The method does not require access to the training dataset. This approach was compared with other strategies: the BEST model (the ideal model, which can only be achieved by access to data or knowledge of which population is most similar to the individual), CROSS model, and RANDOM model selection. RESULTS: When evaluated on clinical datasets, the approach significantly outperformed the CROSS model selection strategy in terms of discrimination (p<1e-14) and calibration (p<0.006). The method outperformed the RANDOM model selection strategy in terms of discrimination (p<1e-12), but the improvement did not achieve significance for calibration (p=0.1375). LIMITATIONS: The CI may not always offer enough information to rank the reliability of predictions, and this evaluation was done using aggregation. If a particular individual is very different from those represented in a training set of existing models, the CI may be somewhat misleading. CONCLUSION: This approach has the potential to offer more reliable predictions than those offered by other heuristics for disease risk estimation of individual patients.

OBJECTIVE: Predictive models that generate individualized estimates for medically relevant outcomes are playing increasing roles in clinical care and translational research. However, current methods for calibrating these estimates lose valuable information. Our goal is to develop a new calibration method to conserve as much information as possible, and would compare favorably to existing methods in terms of important performance measures: discrimination and calibration. MATERIAL AND METHODS: We propose an adaptive technique that utilizes individualized confidence intervals (CIs) to calibrate predictions. We evaluate this new method, adaptive calibration of predictions (ACP), in artificial and real-world medical classification problems, in terms of areas under the ROC curves, the Hosmer-Lemeshow goodness-of-fit test, mean squared error, and computational complexity. RESULTS: ACP compared favorably to other calibration methods such as binning, Platt scaling, and isotonic regression. In several experiments, binning, isotonic regression, and Platt scaling failed to improve the calibration of a logistic regression model, whereas ACP consistently improved the calibration while maintaining the same discrimination or even improving it in some experiments. In addition, the ACP algorithm is not computationally expensive. LIMITATIONS: The calculation of CIs for individual predictions may be cumbersome for certain predictive models. ACP is not completely parameter-free: the length of the CI employed may affect its results. CONCLUSIONS: ACP can generate estimates that may be more suitable for individualized predictions than estimates that are calibrated using existing methods. Further studies are necessary to explore the limitations of ACP.

Adoption studies of social media use by clinicians were systematically reviewed, up to July 26th, 2011, to determine the extent of adoption and highlight trends in institutional responses. This search led to 370 articles, of which 50 were selected for review, including 15 adoption surveys. The definition of social media is evolving rapidly; the authors define it broadly to include social networks and group-curated reference sites such as Wikipedia. Facebook accounts are very common among health science students (64-96%) and less so for professional clinicians (13-47%). Adoption rates have increased sharply in the past 4 years. Wikipedia is widely used as a reference tool. Attempts at incorporating social media into clinical training have met with mixed success. Posting of unprofessional content and breaches of patient confidentiality, especially by students, are not uncommon and have prompted calls for social media guidelines.

There is much debate on which types of computer-based systems have the most impact in healthcare delivery and patient outcomes. Safety surveillance systems, which have been around for several years, are probably at the top of the list. These provider-oriented clinical decision support systems allow healthcare providers to monitor the safety of medications and other interventions that are critical to prevent poor outcomes. However, another rapidly growing type of system related to personal health records (PHR) is likely to be a contender for the top position within the next few years. These ‘consumer’-oriented systems currently have a primary focus on providing information to patients, but soon will follow the evolution of provider-oriented systems to expand into consumer-oriented decision support systems. In this issue of JAMIA, we cover safety surveillance systems and patient-centric systems, which nicely complement articles covering the same topics that were published in our extraordinary online issue last December. A perspective by Coiera et al (see page 2) reviews the main developments in safety surveillance in the USA and abroad, highlighting opportune areas that remain uncharted. For example, data from different sources can be used in surveillance systems. Magrabi (see page 45) describes how US Food and Drug Administration reports can be used to develop a classification system focused on safety, and Overhage (see page 54) validates a data model for safety surveillance research using, among other sources, data from electronic health record (EHR) systems. Li (see page 6) reviews the literature on patient safety implications of frequent interruptions in the course of clinician care. Proper evaluation is essential for studying the efficacy of decision support systems as it relates to safety. Ancker (see page 61) describes a framework for evaluating the effect of health information technology on both quality and safety. In a systematic review, Augestad (see page 13) reports the rate of adherence of informatics randomized controlled trials to CONSORT guidelines. Specifically relating to pharmacotherapy, McKibbon (see page 22) reviews the effectiveness of several medication management systems, and Forster (see page 31) reviews the impact of different adverse drug event detection systems. In terms of research, Du's tutorial (see page 39) reviews methods for comparing count data, which are frequently used in safety systems, and explains the pitfalls of commonly used methods based on ordinary least squares. Eppenga (see page 66) compares the accuracies of different pharmacotherapy clinical decision support systems, and Rodriguez-Gonzalez (see page 72) describes medication administration errors resulting from automated prescription and dispensing systems. Tatonetti (see page 79) describes a new algorithm to identify drug–drug interactions. Special populations, such as the elderly, require different drug dosing. Griffey (see page 86) reports the results of a real-time decision support system evaluation for this population. Other decision support systems and tools are included in this issue of the journal: Niland (see page 111) briefly describes a system to grade adverse events based on laboratory values, Shiffman (see page 94) describes a tool for assisting in knowledge capture for computer-based practice guideline systems, and Chiu (see page 102) reports on how a detailed pedigree information system can be used to support genetic studies. Perceptions are key to the adoption of these decision support systems. Usability studies can uncover the reasons why healthcare providers do not exhibit uniform levels of adoption. Goddard (see page 121) provides a systematic review of automation biases related to human computer interactions that extend beyond the healthcare field to fields such as the aviation industry. The internet has transformed the way people access information and how they connect to each other or to institutions that hold their data. Informatics solutions that engage patients to play a major role in their own care are long overdue. As reported by Beard et al (see page 116), attempts have been made to make EHRs directly accessible to patients. However, the healthcare sector has historically lagged behind in terms of opening up information to its ‘consumers’, and adoption has been slow, except for certain types of patient portals and PHR systems. Patient portals are now commonly found in many institutions that have electronic records as major EHR vendors provide this module. Nielsen (see page 128) describes factors related to the usage of a portal for multiple sclerosis in which users frequently communicate with their providers about medications and side effects. In addition to provider-based patient portals, several initiatives to develop PHR controlled by the patients have been created in the past decade. Li (see page 134) proposes a standard for travelers' records that could be used across the globe. Marquard (see page 137) presents a case-based evaluation of a PHR system. As evidenced by this issue's articles, the healthcare industry is changing rapidly. The increasing adoption of electronic records and decision support systems will likely make healthcare safer and more cost effective, as well as increasingly engage healthcare providers and consumers as major agents of change. JAMIA is proud to feature excellent articles on every subspecialty of informatics, and will continue to publish supplementary online issues. Visit us at http://www.jamia.org to see the special online-only issues and continue to sign up for our free JAMIA Journal Club, featuring presentations by the authors of Editor's Choice papers. This highly interactive event offers everyone an opportunity to ask questions and learn more about specific domain areas within informatics, directly from the people who are making changes in the way we provide care and perform research.

This is a particularly exciting issue of JAMIA . Not only do we display exceptional work spanning informatics research that integrates data from different biological levels (from molecules to tissues to individuals), but we also show how this research is greatly enhanced by clever integration of knowledge from publicly shared resources (from nucleotide sequences to gene and protein networks to data from the biomedical literature). The articles in this issue cover a broad range of approaches developed in different institutions spread over five countries and 12 US states, and are prime examples of the importance of a quantitative approach to health sciences that requires computational analysis of massive amounts of data that are now being generated at an accelerated pace. Upon recognizing the importance of providing biomedical and behavioral researchers with algorithms, tools, and computational facilities that accelerate scientific discoveries, the NIH sponsored the creation of several National Centers for Biomedical Computing (NCBCs) 8 years ago. An editorial by Berg ( see page 151 ) discusses the impact of these centers, which are described in eight brief communications ( see pages 166 – 206 ). These NCBCs embody the very nature of biomedical informatics: a field dedicated to the improvement of human health through the development of new algorithms and software tools for data capture, analysis, and knowledge dissemination, resulting from the combined efforts of researchers from fields seemingly as diverse as computer science, engineering, physics, statistics, library sciences, and biomedical and behavioral sciences, to name a few. While each of the centers focuses on different informatics aspects, they all share the goal of enabling collaborative ‘big science’ through the dissemination of innovative algorithms, tools, and services to the scientific community. An article by …

This issue of the journal focuses on an important but often underpublished area of biomedical informatics: the cost-effectiveness of informatics interventions in healthcare. The adoption of electronic health records (EHR) across the USA has been accelerated in response to legislation, but there is still much uncertainty regarding costs as well as short and long-term effects, given the many different ways in which systems are implemented and the high diversity of institutions served. A systematic review by O'Reilly (see page 423) covers economic evaluations of medication management systems, and the author also describes the cost-effectiveness of a clinical decision support system (CDSS) for diabetes in another article (see page 341). Frisse (see page 328) reports on the financial impact of EHR in an emergency department, and Subramanian (see page 439) analyzes the financial impact of a CDSS for renal dose adjustments. The financial aspects are not sufficient to assess the full impact of information systems and informatics interventions. Several articles relate to the impact of systems in the quality of care. Connelly (see page 334) describes how patients with congestive heart failure benefit from EHR when they visit emergency departments. Kennebeck (see page 443) reports on how EHR impact workflows in a pediatric emergency department, and Lanham (see page 382) associates communication patterns among healthcare providers with EHR use in an ambulatory setting. Zandieh (see page 401) describes the impact of transitions between EHR systems in the same setting, and Herwehe (see page 448) reports on the impact of an EHR system for public health information exchange in HIV/AIDS. CDSS have the potential to improve the quality of care via EHR. However, not all CDSS are created equal, and their utilization is highly variable. Predicting the usage of CDSS embedded in EHR is not an easy task, but McCoy (see page 346) proposes a framework to study CDSS alerts and responses. A review by Yen (see page 413) discusses some of the key methodologies that are used to study the usability of health information systems, and Lindblom (see page 407) describes how the usability of a CDSS is related to user characteristics such as computer anxiety. Unertl (see page 392) discusses patterns of use and impact on workflows from the viewpoint of health information exchange, and Grossmann (see page 353) quantifies the diversity of experiences of physician practices and pharmacies in transmitting and processing electronic prescriptions. Also related to electronic prescriptions and barriers, Thomas (see page 375) describes prescribers' expectations for the electronic prescription of controlled substances, Stenner (see page 368) describes a text message system designed to assist patients with medication management, and Appari (see page 360) reports on a national study of US hospitals to determine associations between information technology and the quality of medication administration. From all the work mentioned above, it is clear that informatics and health information technology are in different stages of development in different settings, and in different countries. Perspectives from Canadian experts and from the AMIA Policy Committee on the future of health information systems in the USA are provided by Zimlichman (see page 453) and McGowan, (see page 460) respectively. Bates and Edmonds (see page 495) describe what AMIA has been doing to prepare for this future. To prepare for the present, and for a near future in which we can realize the full potential of all the biomedical data that are currently being collected, we will need to train a new generation of informaticians who will help push the limits of the present technology into uncharted territory. The boundaries between clinical informatics, bioinformatics, and consumer health informatics will no longer matter, as these skilled professionals will develop integrative approaches to create a seamless flow of data to actionable knowledge. Training in biomedical informatics thus involves analyses of several types of information. Processing information found in the biomedical literature, for example, continues to be an important area of research, and the combination of skills from biomedical librarians and computer scientists helps the development of faster, targeted retrieval algorithms. The papers by Van de Glind (see page 468) and Petrova (see page 479) describe filter strategies for article retrieval, and Goodwin (see page 473) reports on an approach to predict biomedical document access based on past use. The need for informatics professionals is currently so high that, in addition to programmes for long-term research training, such as those sponsored by the National Library of Medicine, new modalities of short programmes are constantly being proposed, creating a need for continuous evaluation. Tian (see page 489) describes one method to evaluate web-based instructional modules using Markov chain models. Disseminating informatics beyond our community and bringing new talent to our field is certainly something we care much about in JAMIA. Our editorial team has now completed reviews of more than 1500 manuscripts since our start in January 2011, and our sample size is almost large enough to enable the exploration of new trends in the field. I already notice an increased effort on the part of our authors to make their articles openly accessible to the world. I trust that this effort is reflective of the authors' desire to fulfil a societal responsibility as scientists and engineers—openly share work that has an immediate impact in the lives of others.

This issue of JAMIA focuses on the opportunities and challenges for implementing and evaluating practical systems to handle health information for a variety of end users, including patients. Early on, electronic systems usually consisted of simple electronic health records (EHRs) with the purpose of providing clinicians with efficient access to patient information during the encounter. As our field evolved, selecting, implementing, and evaluating these systems became very complex, as EHR systems have increasingly added (a) EHR-based clinical data warehouses to support quality improvement and research, (b) clinical decision support (CDS) features, and (c) programming interfaces to support health information exchange (HIE) across different systems. As the informatics community takes important steps toward developing systematic frameworks to evaluate the cost-effectiveness of these different system features, JAMIA presents studies and perspectives that document the current state of development in health information systems. Two randomized controlled trials present systematic evaluations of CDS. Tamblyn (See page 635) compares the effectiveness of a new generation of computer-based drug alerts, and Wright (See page 555) compares the completeness of problem lists when CDS is utilized. Parsons (See page 604) validates the use of EHR for clinician performance monitoring, Adler-Milstein (See page 537) describes organizational complements to the EHR that improve physician performance, and Fleurant (See page 541) reports on EHR implementation challenges. Related to patient use, Slack (See page 545) and Rosenbloom (See page 549) describe patient interactions with EHRs related to past history documentation and influenza prevention, respectively. Wagner (See page 626) studies the relationship between hypertension control and use of personal health records in a randomized trial, and Yu (See page 514) systematically reviews web tools for management of diabetes and cardiovascular disease. Several authors study the impact of EHR, CDS, and HIE systems on clinician behavior. Hains (See page 506) systematically reviews the impact of Picture Archiving and Communication Systems on ICU clinicians. Ronquillo (See page 570) focuses on genetic testing behavior by physicians, and Love (See page 610) studies the association of provider's perception of healthcare quality with EHR quality. Articles by Eastabrooks (See page 575), Dowding (See page 615), Abramson (See page 644), and Mandl (See page 649) describe EHR impact in clinical care. Crotty (See page 621) describes the impact of an interactive web tool in a residency program. Related to CDS, Koch (See page 583) describes information needs of ICU nurses, and Hripcsak (See page 529) proposes a visualization tool for operating ranges of a classification system. Related to HIE, Lenert (See page 498) provides a perspective on the successes and failures of HIE initiatives, and Schank (See page 562) reports on a statewide survey of healthcare provider beliefs about the value of HIE. This issue also describes early experiences with novel aspects of EHR and CDS systems: Seto (See page 503) provides a perspective on multimedia EHRs, Mandl (See page 597) describes system architecture for interoperable EHR applications, and Okoniewska (See page 674) describes experience with an in-hospital positioning system. Feldman (See page 591) reports on the association between key EHR findings and high-risk diagnoses, and Zheng (See page 660) describes an approach to structure clinical narratives. Dalal (See page 523) describes an email notification system for laboratory tests pending at discharge, and Jung (See page 533) describes the execution of medical logic modules expressed in ArdenML. Kalb (See page 668) reports on internet-mediated survey response by parents of children with autism, and Rajput (See page 655) evaluates a mobile-based system for disease surveillance. The articles in this issue represent the diverse uses and evaluations of a variety of systems. Heated debate has stirred the informatics community about the importance of developing homegrown versus commercial systems, the need for institutional- and user-customization, the lack of universally accepted evaluation methods and benchmarks, as well as the relative paucity of interoperable solutions. This issue of JAMIA sheds light on some of these topics, but much remains to be done, and we will not accomplish this alone. The informatics community must increase its outreach, widely disseminate its knowledge, and promote higher involvement of other academics, industry and government representatives, as well as the public. Without a concerted, inclusive effort to address the massive challenge of effectively implementing and evaluating computer-based health information systems, we are at risk of wasting our resources in limited, fragmented, non-sustainable initiatives.

Healthcare and biomedical research are part of a global electronic ecosystem in which social media, mobile devices, and increasing public engagement have been playing game-changing roles. ‘Web 2.0’ has been used to describe a highly interactive virtual environment in which geographic, cultural, and language barriers are overcome by people's needs to form networks to communicate and collaborate due to common interests. Preventing or alleviating the burden of disease is a most powerful global common goal; hence it is no surprise that there are increasing reports describing informatics advances that enable ‘Health 2.0.’ For example, McCoy (see page 713) describes how her team successfully used crowd sourcing (ie, outsourcing of a particular task to a large community of people) to construct a knowledge base of associations between medication and problems, in contrast to more traditional mechanisms based on a small panel of experts. The latter model is described by Phansalkar (see page 735), who reports on the construction of a knowledge base of drug–drug interactions. von Muhlen (see page 777) briefly reviews how clinicians have been participating in social networks, and provides some recommendations on how to participate responsibly, so that both clinician and patient privacy are preserved. Public uptake of Web 2.0 is very high, and higher utilization of this resource is part of the proposed agenda for public health informatics. Patients have become increasingly engaged in promoting their own health and have been using different types of media to report outcomes and provide data that can help in their care. Connelly (see page 705) describes a mobile system to monitor nutrition for patients undergoing dialysis, Kass-Hout (see page 775) describes how self-reported fever may have higher value for syndromic surveillance than measured temperature recorded in electronic health records (EHRs), and Whitford (see page 744) evaluates the reliability, validity, and acceptability of patient-reported data for research related to maternal decisions on infant feeding. Using randomized controlled trials, Schnipper (see page 728) checks the value of personal health records for medication accuracy and safety, and Lau (see page 719) reports higher rates of influenza vaccination due to a personally controlled health management system. Vervloet (see page 696) systematically reviews the effectiveness of electronic reminders for patient adherence to chronic medications. To support personalized medicine involving genomics, Overby (see page 840) describes the construction of a rule-based system derived from pharmacogenomics knowledge resources. There is little question that the construction of intelligent systems will depend more and more on high quality data derived from patient and clinician sources, and that EHRs and the biomedical literature will continue to be major sources of information. Boussadi (see page 782) provides an excellent example of how a clinical data warehouse can help refine medication prescription alerts, and Wu (see page 758) presents a novel approach to build shared predictive models across institutions without the need to share individual patient data. This is important to preserve patient privacy in clinical data warehouses, a topic that is also addressed by Vinterbo (see page 750).i Weng (see page 684) describes the challenges of using EHRs for research. One important challenge is the high prevalence of narrative text in the EHR and other knowledge sources. Several articles in this issue focus on natural language processing and information retrieval. They display a variety of statistical and rule-based approaches to extract machine-readable information from narrative text in radiology reports (see pages 913, 792, 859), pathology reports (see page 833), clinical notes (see pages 809, 817, 824) and several knowledge resources, including the biomedical literature (see page 800). This issue contains a series of articles on the specific topic of coreference resolution (ie, identifying whether concepts are associated through an identity or equivalence relation) (see pages 786, 867, 875, 883, 888, 897, 906). It also features statistical learning techniques for medical literature retrieval (see page 851), as well as novel information retrieval methods that are relevant to translational bioinformatics (see page 765). Our goal is to have JAMIA continue to be the prime source of scholarly work in informatics. The journal covers all aspects of this fascinating and rapidly evolving discipline. To achieve this, the diverse editorial team is constantly seeking ways to improve processes so that we can continue to provide timely responses to authors (the median time from submission to first decision is below 30 days), and to diversify our portfolio so that it represents the multiple facets of informatics. Please enjoy the outstanding articles assembled for this issue of JAMIA, and take a moment to provide us with feedback that will help shape the journal's path into the future.

The role of biomedical informatics in science has never been so prominent, and we strive to make this journal a vehicle for wide dissemination of the best biomedical informatics work well beyond our own field. This issue of the journal is dedicated to translational science enabled by informatics. It combines the topics of AMIA's Joint Summits on Translational Science—Translational Bioinformatics and Clinical Research Informatics—and presents the best articles from the conference, as well as several other related articles that have been submitted to JAMIA through the regular journal process. As the readers can verify, the outstanding quality of the articles selected from the conference make them indistinguishable from the regular articles. It was particularly exciting to work with our guest editors Nigam Shah, Michael Kahn, and Chunhua Weng to edit this special issue. This was true not only because we learned much from each other's perspectives, but also because JAMIA was able make its contribution to the success of the Joint Summits, by motivating authors to submit their best work to the conference for potential publication in the journal. As director of a biomedical informatics academic unit within a medical school, I appreciate that trainees and junior faculty have the option to submit their best papers for publication in conference proceedings, but may hesitate given the unequal weights assigned by promotion committees to proceedings and journal publications. On the other hand, everyone in our community wants to have an opportunity to see live presentations and interact with the authors who are producing the best work in our field. By inviting the best conference papers to be submitted to this special issue of the journal, we hope we are contributing to both these goals. As the readers will notice, even though the articles in this issue are categorized as primarily representing translational bioinformatics or clinical research informatics, these boundaries are fuzzy. Shah and Tenembaum (see page e2) describe how informatics analysis of Big Data has already started to be translated to the bedside and to influence patient outcomes. They describe how translational bioinformatics has bridged the gap between research and clinical practice, in ways that resemble how basic science discoveries have been translated into healthcare. With a strong, but not exclusive, focus on whole genome sequencing analysis, translational bioinformatics articles are increasingly appearing in JAMIA, reflecting the growth of a critical sub-specialty of informatics that is attracting a large number of new trainees to our field. Two of our prior issues (July 2011 and March 2012) were focused on translational bioinformatics and were very well received by our readers. Another one of our notable prior issues (December 2011) was focused on clinical research informatics. In this issue, Kahn and Weng (see page e36) propose a conceptual model to highlight how several articles relate to the translational continuum of basic research to clinical trials. As Kahn and Weng report, clinical research informatics' presence in JAMIA has experienced exponential growth, reflecting the growth of the consortium of Clinical and Translational Science Awards, and the recognition that, without informatics, it is not possible to implement ‘Big Science.’ Enjoy this issue and spread the word: biomedical informatics has a key role in Big Science, and Big Data. JAMIA has a big role in disseminating the scholarly work of our community.

These are exciting times for a career in informatics: not only are there a multitude of advertised jobs, but employment opportunities have grown from a relatively small academic community to a large diverse community including government, commercial, and non-profit organizations. What was once considered an ‘exotic’ career for those in health sciences who were passionate about computation (or for those in quantitative sciences who were passionate about health sciences) has become an essential profession that is impacting healthcare and biomedical science in an unprecedented way. Therefore, trained professionals to fill informatics positions are now in high demand. In this issue of JAMIA, we honor some of our distinguished informatics colleagues who were elected to the American College of Medical Informatics in 2010 and 2011 (see page 920). These professionals have continued to contribute to the informatics field for over 10 years. We also report on the core competencies for graduate education in biomedical informatics (see page 931), which have expanded significantly in the past decade. This long-term, graduate education is training a new generation to become our future leaders. Simultaneously, short-term programs, such as undergraduate and certificate programs, are serving the need to rapidly fill gaps in the current work force. Informatics research and development have also expanded in the past decade. Foundational informatics research occurs in all sectors (academia, industry, and government), and applications are utilized across the board. In this issue, we present some clear examples of how practical applications of informatics, such as clinical decision support (see pages 942, 980, 995, 1003) and surveillance systems (see pages 988, 1011, 954, 1075, 1103, 939), are impacting quality of care in our healthcare system. Physician use of social media is reported in a brief communication (see page 960). We also report on how secondary use of clinical data for research can provide important insights for healthcare quality improvement (see page 965), and how telemonitoring technology improves glycemic control (see page 973). We include reports on improvements in electronic health record (EHR) systems (see pages 1043, 1050, 1032, 1019, 1025, 1089), strengths and weaknesses of electronic prescribing (see page 1059) and describe preparatory work for EHR system implementation in a developing country (see page 1039). The field of informatics has also expanded to include new sub-areas, which have been increasingly featured in JAMIA—translational bioinformatics (see pages 1095, 1066) and clinical research informatics (see page 1110), including a report on a how a biorepository information system was integrated with a commercial EHR system (see page 1115). Additionally, we report on a model of financial effects of health information exchange (see page 1082). As evidenced by this issue of JAMIA, those who are starting an informatics career, in addition to those currently in the field, have a large array of opportunities. We look forward to their future contributions to biomedical informatics and to publishing these contributions in JAMIA.

iDASH (integrating data for analysis, anonymization, and sharing) is the newest National Center for Biomedical Computing funded by the NIH. It focuses on algorithms and tools for sharing data in a privacy-preserving manner. Foundational privacy technology research performed within iDASH is coupled with innovative engineering for collaborative tool development and data-sharing capabilities in a private Health Insurance Portability and Accountability Act (HIPAA)-certified cloud. Driving Biological Projects, which span different biological levels (from molecules to individuals to populations) and focus on various health conditions, help guide research and development within this Center. Furthermore, training and dissemination efforts connect the Center with its stakeholders and educate data owners and data consumers on how to share and use clinical and biological data. Through these various mechanisms, iDASH implements its goal of providing biomedical and behavioral researchers with access to data, software, and a high-performance computing environment, thus enabling them to generate and test new hypotheses.

OBJECTIVE: Postoperative nausea and vomiting (PONV) is a frequent complication in patients undergoing ambulatory surgery, with an incidence of 20%-65%. A predictive model can be utilized for decision support and feedback for practitioner practice improvement. The goal of this study was to develop a better model to predict the patient's risk for PONV by incorporating both non-modifiable patient characteristics and modifiable practitioner-specific anesthetic practices. MATERIALS AND METHODS: Data on 2505 ambulatory surgery cases were prospectively collected at an academic center. Sixteen patient-related, surgical, and anesthetic predictors were used to develop a logistic regression model. The experimental model (EM) was compared against the original Apfel model (OAM), refitted Apfel model (RAM), simplified Apfel risk score (SARS), and refitted Sinclair model (RSM) by examining the discriminating power calculated using area under the curve (AUC) and by examining calibration curves. RESULTS: The EM contained 11 input variables. The AUC was 0.738 for the EM, 0.620 for the OAM, 0.629 for the RAM, 0.626 for the SARS, and 0.711 for the RSM. Pair-wise discrimination comparison of models showed statistically significant differences (p<0.05) in AUC between the EM and all other models, OAM and RSM, RAM and RSM, and SARS and RSM. DISCUSSION: All models except the OAM appeared to have good calibration for our institution's ambulatory surgery data. Ours is the first model to break down risk by anesthetic technique and incorporate risk reduction due to PONV prophylaxis. CONCLUSION: The EM showed statistically significant improved discrimination over existing models and good calibration. However, the EM should be validated at another institution.

OBJECTIVE: The classification of complex or rare patterns in clinical and genomic data requires the availability of a large, labeled patient set. While methods that operate on large, centralized data sources have been extensively used, little attention has been paid to understanding whether models such as binary logistic regression (LR) can be developed in a distributed manner, allowing researchers to share models without necessarily sharing patient data. MATERIAL AND METHODS: Instead of bringing data to a central repository for computation, we bring computation to the data. The Grid Binary LOgistic REgression (GLORE) model integrates decomposable partial elements or non-privacy sensitive prediction values to obtain model coefficients, the variance-covariance matrix, the goodness-of-fit test statistic, and the area under the receiver operating characteristic (ROC) curve. RESULTS: We conducted experiments on both simulated and clinically relevant data, and compared the computational costs of GLORE with those of a traditional LR model estimated using the combined data. We showed that our results are the same as those of LR to a 10(-15) precision. In addition, GLORE is computationally efficient. LIMITATION: In GLORE, the calculation of coefficient gradients must be synchronized at different sites, which involves some effort to ensure the integrity of communication. Ensuring that the predictors have the same format and meaning across the data sets is necessary. CONCLUSION: The results suggest that GLORE performs as well as LR and allows data to remain protected at their original sites.

BACKGROUND: Cross-platform analysis of gene express data requires multiple, intricate processes at different layers with various platforms. However, existing tools handle only a single platform and are not flexible enough to support custom changes, which arise from the new statistical methods, updated versions of reference data, and better platforms released every month or year. Current tools are so tightly coupled with reference information, such as reference genome, transcriptome database, and SNP, which are often erroneous or outdated, that the output results are incorrect and misleading. RESULTS: We developed AnyExpress, a software package that combines cross-platform gene expression data using a fast interval-matching algorithm. Supported platforms include next-generation-sequencing technology, microarray, SAGE, MPSS, and more. Users can define custom target transcriptome database references for probe/read mapping in any species, as well as criteria to remove undesirable probes/reads. AnyExpress offers scalable processing features such as binding, normalization, and summarization that are not present in existing software tools. As a case study, we applied AnyExpress to published Affymetrix microarray and Illumina NGS RNA-Seq data from human kidney and liver. The mean of within-platform correlation coefficient was 0.98 for within-platform samples in kidney and liver, respectively. The mean of cross-platform correlation coefficients was 0.73. These results confirmed those of the original and secondary studies. Applying filtering produced higher agreement between microarray and NGS, according to an agreement index calculated from differentially expressed genes. CONCLUSION: AnyExpress can combine cross-platform gene expression data, process data from both open- and closed-platforms, select a custom target reference, filter out undesirable probes or reads based on custom-defined biological features, and perform quantile-normalization with a large number of microarray samples. AnyExpress is fast, comprehensive, flexible, and freely available at http://anyexpress.sourceforge.net.

OBJECTIVE: To determine whether statistical and machine-learning methods, when applied to electronic health record (EHR) access data, could help identify suspicious (ie, potentially inappropriate) access to EHRs. METHODS: From EHR access logs and other organizational data collected over a 2-month period, the authors extracted 26 features likely to be useful in detecting suspicious accesses. Selected events were marked as either suspicious or appropriate by privacy officers, and served as the gold standard set for model evaluation. The authors trained logistic regression (LR) and support vector machine (SVM) models on 10-fold cross-validation sets of 1291 labeled events. The authors evaluated the sensitivity of final models on an external set of 58 events that were identified as truly inappropriate and investigated independently from this study using standard operating procedures. RESULTS: The area under the receiver operating characteristic curve of the models on the whole data set of 1291 events was 0.91 for LR, and 0.95 for SVM. The sensitivity of the baseline model on this set was 0.8. When the final models were evaluated on the set of 58 investigated events, all of which were determined as truly inappropriate, the sensitivity was 0 for the baseline method, 0.76 for LR, and 0.79 for SVM. LIMITATIONS: The LR and SVM models may not generalize because of interinstitutional differences in organizational structures, applications, and workflows. Nevertheless, our approach for constructing the models using statistical and machine-learning techniques can be generalized. An important limitation is the relatively small sample used for the training set due to the effort required for its construction. CONCLUSION: The results suggest that statistical and machine-learning methods can play an important role in helping privacy officers detect suspicious accesses to EHRs.

Biomedical informatics is a young, highly interdisciplinary field that is evolving quickly. It is important to know which published topics in generalist biomedical informatics journals elicit the most interest from the scientific community, and whether this interest changes over time, so that journals can better serve their readers. It is also important to understand whether free access to biomedical informatics articles impacts their citation rates in a significant way, so authors can make informed decisions about unlock fees, and journal owners and publishers understand the implications of open access. The topics and JAMIA articles from years 2009 and 2010 that have been most cited according to the Web of Science are described. To better understand the effects of free access in article dissemination, the number of citations per month after publication for articles published in 2009 versus 2010 was compared, since there was a significant change in free access to JAMIA articles between those years. Results suggest that there is a positive association between free access and citation rate for JAMIA articles.

The current focus and funding of health information technology forces the informatics research community to ask many questions. A key one is: Are we providing policy makers the evidence they need to make the best decisions? Corollary to this question are two more: Are we using the best methods? Are we looking for evidence in all the right places? Liu and Wyatt1 present their perspective on the critical role of randomized clinical trials (RCTs) in the assessment of clinical information systems. The authors briefly review the sources of skepticism that claim that RCT-based evaluation is not useful in evaluating health information systems due to ethical and technical grounds. Liu and Wyatt present counter-arguments to make a compelling case that clinical information systems can be very influential in determining clinical outcomes, so they should be subject to the same rigorous evaluation standards as other types of clinical interventions, such as medications and procedures. Regardless of design, justification of the match between purpose and design is paramount. JAMIA agrees that RCTs are the preferred standard for evaluation in general, but it is not appropriate in all circumstances. As Liu and Wyatt point out, RCTs, which are often expensive, are appropriate for the evaluation of systems that expose added risk to subjects or that are associated with high costs. However, the growing field of comparative effectiveness research helps our thinking beyond RCTs and is directly relevant, not only because informatics systems should be evaluated using those principles, but because informatics systems often provide the data for those studies. Thus, leading comparative effectiveness research researchers provide guidance on when non-RCT designs are appropriate.2 For example, “To evaluate real-world applicability… to study multiple treatment paradigms simultaneously…to understand current practices…where trials have not been or cannot be performed…when treatment adherence differs…when providers have different training…” Each of these items applies when evaluating informatics systems: such systems can embody multiple interventions (eg, think of the many decision support options employed); baseline studies prior to implementation help us to understand current practice; trials have rarely been performed, especially of commercial systems; adherence, in terms of adoption, varies greatly within an institution; and training for HIT systems is notoriously variegated. Thus, while many reports that are based on RCTs have appeared in the journal recently, other reports evaluate system features that may not expose patients, clinicians, or healthcare workers to added risk and are relatively inexpensive to implement. These studies are important for documenting the impact of a clinical information system in clinical outcomes and/or processes and for guiding subsequent system development. These reports may ultimately influence decision makers to implement certain system features at their sites. More examples of studies not based on RCT designs appear in this issue, such as quantitative evaluations of clinical decision support interventions for appropriate test ordering in primary care and for timely discontinuation of antibiotics after surgery. Arguably, these interventions did not impose added risk to patients and were relatively inexpensive when compared to the cost of implementing a whole clinical information system. In these circumstances, other study designs may be appropriate. Similarly, JAMIA publishes qualitative evaluations and reports of studies that employ methodologies designed to model behavioral and social systems and does not restrict its publications to quantitative evaluations or studies that utilize methodologies designed to model physical systems. Therefore, JAMIA ignores the perceived dichotomy between “soft” versus “hard” sciences. The field of health and biomedical informatics is diverse and each study is unique; thus, it is important to understand what is most appropriate for a particular investigation and to avoid a priori rule-in or rule-out of particular methodologies. Beyond correct methodologies, we must study at the right level of organization. JAMIA is encouraging the submission of articles in all areas addressed in AMIA's strategic realignment,3 especially those that have been relatively under-represented in the journal. At the opposite extreme of hospital-based informatics is public health informatics. Although computer systems have been used since 1938, when Illinois used IBM tabulation equipment for vital statistics,4 the self consciousness of public health informatics (PHI) as a field is relatively new as evidenced by the 2001 Spring AMIA meeting,5 the 2003 publication of the PHI textbook,4 and the funding for syndromic surveillance in the mid-2000s by the CDC Centers for Excellence in Public Health Informatics and the Department of Homeland Security. In addition to being relatively new, the scope of PHI is also broad. In the spirit of evidence-based inquiry, we reviewed JAMIA's publications in this area, mapping article titles to PHI topics and to the 10 Essential Public Health Services.6 Of 273 articles (of all sorts) published in 2008–2010, 51 (19%) were directly related to public health and addressed only four of the functions identified by the Essential Public Health Services (see next page). From CDC. Ten Essential Public Health Services. http://www.cdc.gov/nphpsp/essentialServices.html. From CDC. Ten Essential Public Health Services. http://www.cdc.gov/nphpsp/essentialServices.html. Consider some of the functions not on the list. “Mobilize community partnerships and action to identify and solve health problems” does seem to be an epiphenomenon of health information exchange activities, but we have not published research that addresses community mobilization head on. “Enforce laws and regulations that protect health and ensure safety” is perhaps represented by articles on privacy, but those articles mostly address the technology of deidentification. Meanwhile, education of public health professionals makes no appearance. Even the functions that do appear on the list are not well-represented. While “consumer health” is on the list, most of the articles deal with personal health records, and not the public health function of “inform, educate, and empower people about health issues.” Certainly more research exists about informatics and developing countries or rural domestic areas beyond a couple of articles. The same goes for each of the other areas. The final two functions—evaluating the “effectiveness, accessibility, and quality of personal and population-based health services” and “research for new insights and innovative solutions to health problems”—are part of every JAMIA submission, but return us to the question of research methods. The broad scope of PHI means that there should be a wide range of topics that JAMIA will publish, for example, health messaging; uses of mobile technologies in developing countries; decision support for public health practice; emergent behavior of social networks and their effect on public health. We plan to issue a call for papers soon, which will address these under-represented areas in the journal. With both appropriate methodology and proper level of analysis, the best evidence will be reproducible and generalizable to different settings. Therefore, JAMIA requests that authors thoroughly justify study design choices, submit data and code as appropriate for peer-review and for potential inclusion in online appendices, and discuss study limitations explicitly. We have broadened the scope of the journal to encompass all areas of biomedical and health informatics and look forward to receiving submissions which represent research, applications, reviews, and perspectives in all areas. Our collective contributions can have a large impact in healthcare in the USA and abroad. Commissioned; internally peer reviewed.

OBJECTIVES: To provide an overview and tutorial of natural language processing (NLP) and modern NLP-system design. TARGET AUDIENCE: This tutorial targets the medical informatics generalist who has limited acquaintance with the principles behind NLP and/or limited knowledge of the current state of the art. SCOPE: We describe the historical evolution of NLP, and summarize common NLP sub-problems in this extensive field. We then provide a synopsis of selected highlights of medical NLP efforts. After providing a brief description of common machine-learning approaches that are being used for diverse NLP sub-problems, we discuss how modern NLP architectures are designed, with a summary of the Apache Foundation's Unstructured Information Management Architecture. We finally consider possible future directions for NLP, and reflect on the possible impact of IBM Watson on the medical field.

This issue of JAMIA reports on a collection of informatics research and applications developed and implemented worldwide. I highlight below some of the latest findings in our field and discuss how they are interconnected. The increasing adoption of electronic health records (EHR) and secondary use of clinical data for research have fueled research in privacy technology. In this issue, Malin (page 3) uses simple techniques to show that the same level of privacy preservation attained with the ‘safe harbor’ rule from the US government can be achieved, even if select portions of the data that are not supposed to be disclosed under the safe harbor rule (eg, the ages of individuals above 89 years old) are made available. The article is important because it illustrates the rationale behind the safe harbor rule but proposes technological alternatives that may influence how new policies and regulations are formulated. More research in this burgeoning field is sorely needed to enable effective use of clinical data for research without compromising individual privacy. A secondary finding, related to privacy, in a study from Atkinson (page 24) shows that patients are reluctant to enter their personal information in a web application that helps locate appropriate clinical trials. The main focus of the study is the evaluation of user perceptions of a web-based system. The authors systematically study patient perceptions of two web-based systems and show that a counterbalanced design (ie, a design in which the order of presentation of two different systems are alternated) is important to avoid overestimation of preference toward the system that is presented last. On a related topic, Or (page 51) utilizes a model based on technology acceptance theory to determine the factors that influence the use of a home-care web-based self-management tool by patients with chronic cardiac disease. For this population, the main factors were the perceived usefulness and ease of use, social influence (‘peer pressure’), and the patients' healthcare knowledge. Clinician perceptions are important for acceptance of new clinical systems. Hincapie (page 60) describes how physicians in limited focus groups perceived a particular health information exchange (HIE) system. Physicians thought that HIE could improve care, but that it would not have much impact if data were not comprehensive, as was the case in their setting. This might explain why limited scope HIE initiatives have failed. Additionally, opponents of electronic health systems have raised concerns about the potential for EHRs to increase the length of visits and to decrease the clinician's face time with the patients and/or their families. Fiks (page 38) shows that pediatricians spend a large proportion of visit time interacting with the families while using the computer. They also show that, after adjusting for visit characteristics such as number and types of diagnoses, computer use is not associated with increased time per visit; nor is it associated with decreased face time. On the other hand, advocates for electronic health systems often assume that these systems improve efficiency and accuracy. Blaya (page 11) reports on negative results of a cluster randomized controlled trial that implemented an electronic communication system for laboratory results. Although the turnaround time to receive tuberculosis culture and drug-sensitivity results was significantly reduced for regional health centers, indicating that the system was effective at that level, the system did not benefit the end users directly, as it did not cover the ‘last mile’ of communication between the regional centers and local point-of-care users. Lessons learnt from this study should inform those who are designing new interventions, especially in situations where resources are very limited as in many global health informatics projects. On the topic of clinical documentation, Collins (page 45) reports that critical goals and actions discussed in ICU rounds are not adequately documented in clinical notes authored by attending physicians, nurses, respiratory therapists, and residents. The results indicate an urgent need to define a common information source for the care plan that is consistent across team members, since its absence may compromise patient safety. Clinical documentation is fraught with difficulties. For example, the use of certain abbreviations in clinical documents is unsafe. Myers (page 17) reports on the teaching effects of a system that required users to remove inappropriate abbreviations when writing clinical notes in an EHR system. Fewer inappropriate abbreviations in subsequent handwritten notes were written by subjects in this ‘hard stop’ group, but not for the group in which the corrections were done automatically by the EHR system. Further work is certainly needed to clarify whether this observation can generalize to other situations. Also related to patient safety, the article by Saverno (page 32) describes the heterogeneous implementation of decision-support tools to detect drug–drug interactions in different pharmacies. Results are concerning: the median sensitivity across 64 pharmacies in Arizona for some important drug interactions was only 0.85, ranging from 0.23 to 1. Interestingly, the variability was not related to the type of system being used (different vendors were represented) or to the type of pharmacy (such as hospital- or community-based), but rather related to the particular configuration at the site. More research on this topic is certainly warranted to understand how best to implement decision-support tools for pharmacists. Readers will find in this issue and in the JAMIA online archives a wealth of information that can be applicable to their professional activities. In addition to these Research and Application articles, this issue contains peer-reviewed Case Reports, a Brief Communication Correspondence, and Perspectives on important topics such as ethics and EHR vendors, unintended consequences of Health Information Technology (HIT) interoperability, and lessons learnt from the decade long HIT experience in the UK. Reviews that systematically synthesize important topics and primers on biomedical and health informatics research areas not previously highly represented in JAMIA (algorithms and methods, translational bioinformatics, clinical research and public health informatics) will be featured in future issues. JAMIA welcomes submissions in all article categories, as well as feedback from readers and authors. I hope this synopsis helps readers navigate and enjoy the journal.

It is with great honor that I write my first editorial as the new Editor-in-Chief of JAMIA, the flagship journal of our informatics association. Under the leadership of Randy Miller, JAMIA has become the number-one-ranked journal in biomedical and health informatics. To maintain this status and keep up with the growth of our field, I plan to implement changes that will disseminate JAMIA to a broader audience, expand its contents, and streamline its management. These changes will happen over the course of several months. I hope to hear and discuss new ideas with all our readers. In addition to discussing the journal, I expect to hear perspectives on our rapidly evolving profession. Promoting and documenting discussions that are shaping the history of our field is an important role for JAMIA. Biomedical and health informatics now encompass the full spectrum of translational bioinformatics, informatics for clinical research and practice, consumer health informatics, and public health. Reflecting these changes, JAMIA will expand its contents significantly and reach out to a broader readership, including not only the current basis of academicians but also the growing number of informatics practitioners and collaborators from other disciplines. The current team of distinguished associate editors has been augmented to cover the new research areas and the increasing number of submissions. Our outstanding editorial board has also been significantly expanded, and we are launching several initiatives to increase submissions in the new areas. JAMIA will quickly deliver knowledge to this broader audience using various media, with an emphasis on disseminating the most innovative developments in the field. The journal will bring material online as soon as it is accepted and will make it easily accessible in mobile platforms, in addition to traditional printed issues. We will work within cost constraints toward making JAMIA a monthly publication. Additionally, article categories and requirements have been simplified to make it easier for authors, reviewers, and editors to concentrate on content instead of format. JAMIA will continue to publish original articles in foundations and applications of biomedical and health informatics, but I expect to include more perspectives, reviews, and letters depicting important and oftentimes controversial subjects. All JAMIA articles will be made freely available at the JAMIA website and at PubMed Central 12 months after publication. This will help disseminate knowledge to new readers, increase the number of accessed articles, and ultimately increase the number of citations. Immediate free access will be available for Editor's Choice articles (check some outstanding ones in this issue), as well as articles selected by AMIA. In addition, authors may continue to use the unlock mechanism that requires payment of a moderate fee to have the article freely available immediately. This fee is lower than that required by many open access journals, and I encourage authors to use this opportunity, as it helps to disseminate knowledge earlier and also facilitates coverage by news media. Our field has matured and has been attracting an increasing number of professionals from different backgrounds. JAMIA has become its premier journal, and it is therefore fitting to have it reach out to a broader national and international readership. JAMIA is the scholarly journal in biomedical and health informatics with the highest impact factor in its category. We will continue to require rigorous peer reviews for all manuscripts but will employ an editorial management system that facilitates tracking of increased submissions and reviewer assignments, and streamlines the communication between authors, reviewers, and the editorial office. The system will be very familiar to those who submitted, reviewed, or edited manuscripts for the AMIA Annual Symposium or for journals such as the New England Journal of Medicine, Bioinformatics, IEEE Transactions, and many others. The transition to a new management system is never easy. I thank everyone in advance for the patience and constructive feedback as we debug the new system. The turnaround time to return decisions to authors is critical so that a journal can attract submissions that depict novel and timely ideas. It is dependent not only on the efficiency of the editorial office, but also on the expedience of peer reviews and author responses. Although the new system and management primarily enhance efficiency at the editorial office, we must count on authors and reviewers to help shorten the time from submission to publication so that we can bring relevant, up-to-date material to our readers in a timely fashion. Authors who rightfully demand fast and insightful reviews should respond quickly to requested changes and not forget the importance of timely reviews when they assume the role of reviewers themselves. With these changes, we improve on an already exceptional journal. I am indebted to my two outstanding predecessors, William Stead and Randolph Miller, the editorial staff, and all anonymous reviewers without whom JAMIA would not have achieved its current status. I am also indebted to a magnificent team of associate editors, editorial board members, and authors who provide constructive feedback on how to improve JAMIA and work tirelessly to improve it at every new issue. Finally, I acknowledge the help from several entities that made my appointment possible: the search committee for the JAMIA Editor-in-Chief position, led by Reed Gardner, for selecting me to serve in this role; the AMIA Publications Committee, led by Eneida Mendonça, for understanding the importance of open access as a key strategy for broader dissemination and increased impact; the AMIA staff, led by Ted Shortliffe and Karen Greenwood, for their help in operations; and the AMIA Board of Directors, led by Nancy Lorenzi, for supporting the independence of the Editor-in-Chief and editorial team in making key decisions for JAMIA. I owe special thanks to Randy Miller for the encouragement and the confidence deposited in me to lead his legacy forward. We are living in a most exceptional time in history, when informatics permeates all aspects of biomedical science, healthcare, and global health initiatives. With the help and support from our community, together we will take our journal to new and exciting directions. Please join us in this important journey.

Electronic health record systems are in different phases of implementation in the US and abroad. From the perspective of enhancing the quality of healthcare, one of the most attractive features of electronic health record systems is the ability to implement computer-based clinical decision support. However, utilization of two main subsets of electronic health record systems, electronic medical records systems (EMRs) and personal health records systems (PHRs), is still very heterogeneous across institutions and individuals. EMRs are not always perceived as advantageous from the standpoint of individual clinicians, with paper records still being the main form of documentation in several institutions. PHRs are still not well integrated into EMRs and their adoption by patients is not widespread. In this issue, we include articles that focus on how electronic health records are being used and how clinical decision support is making an impact on clinical care. Do describes the results of a pilot study on PHR usage and users' perceptions (see page 118). The results are very encouraging, showing that concerns about privacy are not an impediment for use, and that patients perceive benefits in using PHRs that are not tethered to a particular healthcare system. However, more research and further studies in this area are needed. From the clinician's perspective, Hripcsak reports on whether and how notes from an EMR are utilized by members of the clinical team (see page 112). Clinicians spent an average of 54 min/day authoring and 21 min/day viewing notes at one academic medical center. Yet, a significant number of these notes were never viewed. The authors speculate that oral communication at turn of shifts may obviate the role of certain written notes for immediate care. Nevertheless, these notes are still important for documentation, and Hripcsak's study should motivate more research in this area. Developing new ways of documenting oral communications and evaluating the balance between clinical usefulness and legal documentation in electronic health record systems will require multi-centric studies conducted by multidisciplinary teams of researchers, since ease of use and decreased time in documentation are very important factors in EMR adoption. Related to this topic, Rosenbloom discusses tradeoffs between structured and narrative text entry in EMRs (see page 181). Aziz briefly presents personal sensors that can augment health data collection in the home environment (see page 156). Also related to usability of EMRs are the articles by Sykes on clinician attitudes related to EMR implementation in an Emergency Department (see page 125), by Pearce on whether computers intrude on the relationship between physicians and patients (see page 138), and by Vest on factors that motivate health information exchange (see page 143). Use of an electronic health record system is the first step toward computer-based clinical decision support systems (CDSs). Different outcomes for CDSs are reported in this issue. Positive results in real clinical settings are reported by Were (see page 150 on improved CD4 monitoring for people living with HIV), Haynes (see page 164 on improved discontinuation of antibiotics after surgery), and Zlabek (see page 169 on improved patient safety markers after implementation of a commercial EMR). Florez-Arango also reports increased adherence to guidelines by healthcare workers using CDSs on mobile devices in a controlled experimental setting (see page 131). El-Kareh, on the other hand, describes and discusses a case in which such improvements could not be realized (see page 160). Furthermore, CDSs are not easy to design, implement, and maintain. Wright discusses the difficulties in maintaining knowledge bases for CDSs and how the governance structure in different institutions is organized to address this issue (see page 187). Moreover, Liu evaluated methods for evaluating clinical information systems (see page 173 and accompanying editorial by Lehmann on page 110). Lastly, trained professionals are needed to design, implement, and evaluate electronic health records systems and CDSs. Kampov-Polevoi surveyed training programs and confirmed that electronic health records and computer-based CDSs are core components of biomedical and health informatics curricula in the US (see page 195). We hope that this issue of JAMIA will communicate to clinicians, researchers, healthcare administrators, and many other decision makers the relevant, new, and actionable information related to usage and usefulness of electronic health record systems and CDSs.

This issue of JAMIA contains articles that explore several aspects of informatics. Related to public health and algorithms, El Emam (see page 212) proposes a simple protocol to protect the privacy of providers that disclose data for public health purposes. While we often think of technology that emphasizes patient privacy, an important motivation for institutions to share data relates to the ability to maintain the privacy of their providers. The authors describe the protocol in a way that is accessible to everyone, and a related online appendix describes the technical details. Also related to public health and disease surveillance, Que (see page 218) describes a new algorithm for rapid outbreak detection that first ranks regions according to risk, then uses these ranks to define clusters, in contrast with other techniques that search for predetermined cluster shapes around a high risk area. By publishing highly technical articles written in a manner that can be understood by all, we restate JAMIA's intention to bring important findings to our diverse readership. Since Clinical Decision Support Systems (CDSS) are becoming increasingly available, it is important to develop a systematic approach for their categorization. Wright (see page 232) provides a taxonomy for these systems. Three CDSSs are reported in this issue: Horwitz (see page 243) studies the effects of management protocols for asthma, Campion (see page 251) studies the impact of insulin therapy protocols for diabetes, and Curry (see page 267) describes the advantages of using CDSS in diagnostic imaging. Naun (see page 225) reports on the pros and cons of using poison control data for decision support related to pharmaceutical surveillance. This issue also features two reviews: Georgiou (see page 335) reports on a systematic review of computerized provider order entry (CPOE) systems for clinical imaging. Jaspers (see page 327) utilizes a “meta” systematic review approach to review the literature on the effects of CDSS on provider performance and patient outcomes. Clinical research informatics is an area of increasing activity in our field, and readers should expect to see an increasing number of articles addressing the main challenges and solutions in upcoming issues of JAMIA. In this issue, Segagni (see page 314) describes a statistical tool for research based on R that can be plugged into existing systems. Richesson (see page 341) discusses challenges related to use or appropriate standards in data collection forms. We must not forget, however, that informatics is as much concerned about algorithms and systems, as it is concerned about people, their information needs, and their social networks. Understanding barriers for EHR and CDSS adoption is important to devise strategies for their increased utilization. Rao (see page 271) provides evidence of specific barriers for EHR adoption in small physician practices, and Schnipper (see page 309) describes the difficulties in adopting a medication reconciliation tool. Chan (see page 276) demonstrates the importance of user-centered design for CPOE systems to be effective, and Munasinghe (see page 322) confirms this finding by showing increased CPOE usage after order sets were organized in a nested fashion that was modeled after physicians' practice patterns. Sub-optimal processes can also lead to errors in patient identification and increased nosocomial infections, and Dunn (see page 259) describes an analytical model for studying risks associated with in-patient transfers that suggests where improvements can be made. Related to networks of people, the web has created new paradigms for biomedical information search and communication. Zheng shows that social networking techniques can be used to improve the quality and efficiency of information retrieval from EHRs (see page 282). Nambisan and Weitzman approach the problem from the patient perspective. They report on information seeking patterns (see page 298) and safety of patient support websites (see page 292), respectively. On a related topic, Sarkar (see page 318) describes concerning disparities in patient usage of a diabetes online resource, indicating that the gap cannot be solely justified on the basis of differences in access to technology. More research in this area is clearly needed for informatics to achieve its fullest societal impact. Also related to this topic, Schnall (see page 305) describes the information needs, not of physicians or patients, but other agents who are critically important in healthcare—case managers. This issue of JAMIA thus provides a sample of the diversity of approaches and disciplines that make our field so fascinating. I hope our readers will enjoy reading these articles as much as I did.

If you are reading this editorial, you are probably someone who cares as much about free dissemination of scientific information as I do. Readers want to have free access to high quality scientific articles. Authors want to have their work viewed by as many people as possible. Hence, an open-access model is a great idea. Given the costs need to be covered to allow a journal to be sustainable, a hybrid model offers this particular advantage, as well as an important additional one, related to non-discrimination for author access, that I will describe later. First, I would like to clarify what “open-access” is, since there is evidence that both authors and readers are confused.1 There are primarily three models for scientific journal publishing1: the “reader-pays” model, utilized by traditional (ie, non-open-access) journals2; the “author-pays” model, utilized in most open-access journals; and hybrid models that combine the two. A frequent misconception is to therefore equate the concepts of “open-access” and “free”: articles in open-access journals are not really “free”, but rather paid for upfront by the authors. That is, in open-access journals, the cost is picked up in large part by the authors, and not by the subscribers. The reverse is true in traditional journals (though some journals have a large volume of advertisements that can defray costs). In hybrid models, the costs are distributed to both readers and authors. JAMIA utilizes a hybrid, “delayed open-access” model. In this model, all articles are made freely available after one year of publication, but authors have the option to remove the one-year embargo (and thus have their articles become “open-access” immediately) by paying a fee. For JAMIA, the “author-pays” fee is optional and lower than the fee charged by many popular open-access journals. Subscriptions cover the costs for all other articles. By offering two options, JAMIA allows authors who do not have resources to pay the publishing fees to still have their work appear in the journal. I strongly subscribe to the principle of publishing articles of merit, regardless of the author's ability to pay, and hence I view this model as the one that better suits the journal at present. Because of the prevailing confusion about open-access, many authors have told me they prefer to submit to open-access journals, as they want their work to be highly disseminated and highly cited. However, the hybrid model also allows this to happen: see for example the high representation of open-access articles in the list of “most-viewed” articles that we display in JAMIA's web site (http://www.jamia.org), which include many “editor's choice” articles that are open by default. Some authors are unaware that JAMIA offers this opportunity, and I hope this editorial helps clarify this option. Related to the issue of author access, the hybrid model prevents JAMIA from inadvertently creating a deep publishing divide that would happen if important work from authors who can't pay were excluded. This is currently the most appropriate model for fair publication by our highly diverse informatics community: while authors in some institutions can sponsor unlock fees, authors from less advantaged settings can still publish their best work in the journal and have it publicly accessible after 1 year. I hope this clarification allays the concerns of some readers and authors, and encourages authors to consider the immediate open-access alternative for articles accepted in JAMIA. I also take this opportunity to clarify another important point of confusion. JAMIA is changing not only in terms of layout and expediency, but also philosophy: the journal, like our professional society AMIA,2,3 is reinventing itself, placing a strong emphasis on knowledge dissemination and broadened scope,4 while keeping its scientific quality through a peer-reviewed system that placed it at the top of its category in terms of impact factor. We are reaching out to new scientific communities such as those represented in the Translational Bioinformatics and Clinical Research Informatics Summits (don't miss the upcoming issues highlighting some of the best articles from the conferences), public health informatics,5 as well as highly technical contributions from the computer science and engineering communities. JAMIA is looking for innovative contributions that represent what informatics practitioners and academics are doing today. These contributions are not limited to evaluation of systems or to health informatics topics, and will start to appear more frequently in every new issue. JAMIA seeks high quality submissions in all AMIA strategic areas,2 and considers diversity to be fundamental for it to remain at the center stage in the education and practice of current and future informatics leaders. We have simplified our article categories to encourage submissions by those who are not accustomed to think of JAMIA as a venue for publication. To expedite the process and ensure appropriate review, it is important for authors to categorize their articles properly: mature work that often contains thorough system evaluation or a detailed description of a novel application or adaptation of existing technologies and methods should be submitted in the category “Research and Applications,” while insights gained from the design and preliminary evaluation of systems, or short and generally applicable descriptions of new methods fit in the category, “Brief Communication”. Reports of implementations that offer insights that can guide other practitioners should be categorized as “Case Reports” (note that, in contrast to other medical specialty journals, in JAMIA, the “case” may be a system, software, or method). “Reviews” can report on what is available in the literature on topics of interest or constitute tutorials that help readers understand the importance of certain areas of study. The word limitations for each category should not be used as a factor for category selection since JAMIA allows online supplements without word limits to accompany the primary article. The main function of a scientific journal is to keep readers informed of the best work in the field. JAMIA is proudly pursuing this goal and continually evolving to reflect the dynamic nature of our field. I welcome your feedback and express once again my true appreciation for the tireless efforts of the editorial and management teams, anonymous reviewers, authors, and critics who helped us transition to a new editorial management system. We share the goal of improving the journal through every issue, and together we will continue to evolve models that allow the best informatics articles to be published in JAMIA. None. Not commissioned; internally peer reviewed.

This landmark issue of JAMIA is the first to introduce articles reflecting our extended scope in translational bioinformatics (TBI), as described in the editorial by Butte and Shah (see page 352). Just as medicine has evolved to rely on both molecular and clinical phenotyping, biomedical informatics has evolved to encompass the integration and analysis of information from different biological levels. Sarkar (see page 354) reviews the role of TBI and explains how it bridges biology and medicine through methods for information handling that significantly overlap those used in clinical informatics. Altman (see page 358) reviews the most notable TBI articles in 2010, some of which were published in JAMIA. Wei (see page 370), who was the recipient of the 2011 Marco Ramoni Awardi, presents a Bayesian approach to using data from genome-wide association studies for predictive modeling. The TBI articles cover a range of clinical and molecular approaches. Pathak (see page 376) reports on the challenges of and solutions for phenotype mapping across institutions participating in the Electronic Medical Records and Genomics (eMERGE) Network, while Xu (see page 387) reports how natural language processing significantly helped researchers conduct pharmacogenetic studies at an institution participating in the same network. Chen (see page 392) describes how using a new approach to integrate data from different biological levels can help in the investigation of genes responsible for prostate cancer progression. Foran (see page 403) describes software used to process information from tissue microarrays using grid technology, and Helmer (see page 416) describes a network framework designed to accommodate collaborative biomedical research. Related to this topic, advanced networking and high performance computing are reviewed by Locatis (see page 523). However, a network cannot be fully functional unless the information that it carries is semantically mapped across its nodes. Rector (see page 432) discusses the problems in using SNOMED in practical applications, and Nelson (see page 441) describes the challenges and solutions for standardizing medication names and formulations. This is critical for studies such as those conducted by Hasan (see page 449), in which the goal is to detect omissions in medication lists, and for studying the tradeoffs between manual versus automated record review for detection of adverse events, as described by Tinoco (see page 491). Related to recognition of concepts in electronic health records (EHRs), Savova (see page 459) describes a method to construct a resource for public sharing of gold standard annotations from clinical narratives, and Wilcox (see page 511) provides a method to estimate mismatches (ie, inclusion of information from the wrong patient) in EHRs. Related to information access in EHRs, Boxwala (see page 498) describes a method to monitor potentially inappropriate accesses, and Welter (see page 506) reports on difficulties in conducting content-based image retrieval. Archer (see page 515) reviews specific issues encountered in personal health records. All this fascinating work for structuring, analyzing, and disseminating data through a network is prerequisite to the end goal of improving care and preventing disease. Clinical decision support (CDS) is one way to move toward this goal. Carroll (see page 485) describes a specific CDS for screening used in pediatric clinics associated with a major academic center, and Seidling (see page 479) reports on factors associated with CDS. Related to CDS for quality improvement and public health, Carnevale (see page 466) describes a CDS to detect nosocomial infections, and Kirchhoff (see page 473) reports on the feasibility of utilizing machine learning methods to translate public health information into different languages. Finally, because research in our field is still primarily funded by federal agencies, it is important to understand some factors that are associated with funding. Boyack (see page 423) studies the positive association of funding and the impact of articles. This author also reports an increasing trend in co-funding by NIH institutes, reflecting the current collaborative nature of biomedical research. The articles from this special issue are prime examples of the diversity and richness of research and innovative uses of applications. However, we must recognize that the field of biomedical informatics could not have attained the achievements exemplified in these articles without the pioneering work of colleagues who dedicated their professional lives to it. This issue of JAMIA celebrates exceptional individuals who helped establish the foundations of biomedical informatics and who have touched our lives in many different ways. We remember them fondly, knowing that, by following their footsteps, we will keep their legacy alive for current and future generations of biomedical informatics professionals.

Meaningful use of electronic health records (EHRs) for patient care or for research requires data to be comparable. Many portions of EHRs continue to be unstructured, presenting significant challenges for biomedical informatics. This issue of the journal displays several solutions to this problem that are based on natural language processing (NLP) techniques. A high-level review by Nadkarni (see page 544) is intended to introduce the main components of NLP for the novice, and to briefly describe machine learning methods that are successfully being employed in the field. It includes a discussion on Watson, a contestant on ‘Jeopardy!,’ a popular question-and-answer TV show, and the ensuing speculations about its potential extensions to medical NLP. However, despite some notable examples of successful NLP applications in clinical care, progress in the field has been relatively slow. Chapman and colleagues (see page 540) discuss the need to steer current NLP research efforts so that new developments can be accelerated, and research products can become readily usable in healthcare applications. The authors advocate for a concerted, collaborative effort to develop open-source software components and infrastructure to share annotated data and tools. The guest editorial raises important questions about the ability of the NLP community to collaborate in addressing big challenges that go beyond the scope of what can be accomplished by a single group. There are positive signs that the NLP community is ready for this: it has been able to share data and tools in the organization of software competitions, which are related to specific medical NLP tasks. The community may have reached a point in which researchers can start building on each other's achievements. JAMIA has historically featured the best papers from NLP competitions (see pages 557–613), and takes this opportunity to encourage readers, researchers, and developers in all informatics subspecialties to think beyond competitions, to envision a future in which most effort is spent on collaborative initiatives that help assemble big teams to solve big problems, and to set a goal of developing new solutions that have a direct impact on healthcare. Several NLP and information retrieval articles are featured in this issue: Garla (see page 614) provides an excellent example on how existing open-source components of an NLP pipeline, which was developed at another institution, can be enhanced. The authors evaluate their new components for the problem of detecting hepatic decompensation from findings contained in radiology reports. F Liu (see page 625) quantifies the improved performance of clinical query systems after speech recognition systems are adapted for medical language. Lokker (see page 652) reports on the value of the Clinical Queries filters in PubMed for retrieval of relevant articles for practicing physicians. Hunter (see page 621) reports on a novel application of data-to-text generation in the context of ICU summaries, where clinicians receive information in a familiar format. Several other examples of how information from EHRs and the literature can be automatically processed for clinical decision support are featured in this JAMIA issue: Botsis (see page 631) compares the performance of several machine learning and rule-based classifiers for text-mining in the context of vaccine-adverse event reporting. Herasevich (see page 639) evaluates a rule-based system to detect sepsis in EHRs from an intensive-care unit. H Liu (see page 645) evaluates the adequacy of a proposed e-prescription standard, assessing whether it is ready for adoption given potential ambiguities and imperfect mapping to controlled terminologies. Other authors address the issue of structuring text from the biomedical literature: Huang (see page 660) describes a new approach for assigning MESH terms to documents based on nearest-neighbor documents. Shetty (see page 668) compares the performance of several filters for detection of drug-adverse events from information contained in PubMed. Once data are structured, the ability to exchange them for clinical care or research is greatly enhanced. Thus, without the wide adoption of structured EHRs, our field will not move forward. This issue also provides examples of realistic and timely accounts of critical issues in implementing EHRs: Banas (see page 721) describes the difficult road toward EHR implementation and the lessons learned that can help those who are starting their journey; Zheng describes problems with methods used to study clinical processes and workflows that need to be understood for successful EHR implementation; and Carroll (see page 717) provides a word of caution, showing that the ability of EHRs to accurately report critical information such as medication adherence is still limited. Finally, a perspective from Simborg (see page 675) addresses a way to regulate fraud in EHRs by calling on the Office of the National Coordinator for Healthcare Information Technology to regulate accountability. Furthermore, health information exchange (HIE) presents its own set of challenges. Kuperman (see page 678) provides a lucid account on HIE efforts of the past decade, explaining why some efforts failed, while others have (at least partially) succeeded, and describes where current efforts are focused. However, there is evidence for optimism: Gadd (see page 711) reports on clinicians' high perception of usability of an HIE that has been operational since 2004. Li (see page 683) describes the Dolphin project, initiated in 1998, which is a system to exchange and translate records from two prefectures in Japan and one in China that overcame important logistical, cultural, and language barriers. LaBorde (see page 698) describes a method for analyzing patient crossover between different institutions that can be used by decision-makers to understand the need for HIE and the expected volume of transactions. Biomedical informatics is evolving at a rapid pace, and by leveraging each other's efforts, we will achieve our collective goal of improving care and preventing disease through computation. It is exciting to see how far we have come as a scientific community and to document this progress within the pages of JAMIA.

This issue of JAMIA completes my first year as the Editor-in-Chief. The extended scope, improved workflow, and increase in editorial staff have allowed us to reduce the median review time to <30 days, even with a nearly 70% increase in original submissions. It is exciting to see an increasing number of authors with diverse backgrounds submitting from many different institutions in numerous countries, reinforcing our intent to reflect the best work of informatics without borders. This issue focuses on electronic health records (EHRs, including medical and personal health records (PHRs)) and Clinical Decision Support Systems (CDSS). The debate on what really constitutes meaningful use of information technology (IT) in healthcare has never been so intense, with informatics professionals playing a central role in designing, implementing, and evaluating relevant information systems. EHRs and CDSS are critical components of meaningful use. An editorial by Johnson (see page 730) elaborates on the current role of computer-based provider order entry systems and CDSS designed to reduce pharmacotherapy-related errors. Several articles on this topic appear in this issue (see pages 754–804). This issue features excellent reviews and tutorials that help readers place research and applications articles in context: Holroy-Leduc (see page 732) systematically reviews the international literature on the effects of EHRs in healthcare documentation, processes, and outcomes. Gooch (see page 738) reviews implementation challenges addressed by different strategies for modeling workflows, practice guidelines, and care pathways. Kern (see page 749) describes lessons learned in applying principles of community-based participatory research to evaluate health IT initiatives. The usability of different types of EHRs is still highly variable. Baker (see page 805) shows that there is little difference between adoption of EHR-based versus paper-based reminder for certain clinicians, concluding that inflexibility to change in workflows and practice probably extends beyond IT interventions for these clinicians. Zheng (see page 883) describes workaround strategies that clinicians developed to circumvent limitations of EHRs. Carayon (see page 812) describes significant improvement in nurses' perceptions of usefulness after 3 to 12 months of EHR implementation. Dennehy (see page 820) describes the results of applying a similar model in 30 safety net clinics in two different states. Weir (see page 827) provides a qualitative analysis of the importance of IT in a quality-improvement initiative for geriatrics education conducted at 33 primary care clinics in Utah. Also related to utilization, clinician access to mobile platforms for healthcare information is on the rise. An example is illustrated in the brief communication by Desai (see page 875) about a nephrology information resource. Different settings motivate the development of specialized EHRs: Bostrom (see page 835) compares a specialized application to manage bladder cancer with a conventional EHR in a randomized controlled study that includes comparison of human factors, time spent, and quality of documentation. Lenert (see page 842) reports on the feasibility of a wireless EHR system for deployment in disasters. Rudin (see page 853) describes the frequency and type of transitions of care that occur in the treatment of senior patients in order to inform the design of systems. Related to EHRs and decision support, Wright (see page 859) evaluates the performance of human-developed rules to infer problem lists from other items in the EHR, a prerequisite for the successful implementation of CDSS. Tools to facilitate and evaluate CDSS implementation are also described in this issue. Landis Lewis (see page 868) shows that even in resource-poor environments in which simple EHR systems are implemented, it can be beneficial to implement clinical practice guidelines. Ash (see page 879) describes a case in which a diverse, independent physician association successfully adopted a CDSS by identifying barriers and facilitators for its implementation. As this issue exemplifies, JAMIA is our community's prime source for scholarly work in biomedical informatics. I remain indebted to the authors, reviewers, and readers for their positive feedback and relentless energy dedicated to making JAMIA a better journal. During the next year I plan to publish focus issues on Clinical Research Informatics, Translational Bioinformatics, and Imaging Informatics. We will continue to emphasize clinical informatics and novel informatics approaches to healthcare and biomedical research. As always, feedback from the community is welcome. I look forward to continually featuring the best informatics work in JAMIA in 2012 (stay tuned for an extraordinary online issue in January!).

The Patient-Centered Outcomes Research Institute recently announced several funding opportunities for research focused on empowering patients to make informed decisions. The literature on patient-centered systems has been increasingly present in JAMIA , and we anticipate receiving several submissions on this topic in the upcoming year. Another topic of continued interest is patient safety: the recent IOM report on health IT and patient safety makes important recommendations regarding actions that federal agencies and the private sector should take to maximize the safety of electronic health record systems and other health IT software. It recognizes that, although one of the most impactful areas is medication safety, there are important gaps in the literature. JAMIA helps fill some of these gaps, featuring the outstanding work by the informatics community to address patient safety challenges. Finally, the ongoing discussion related to the new National Center for Advancing Translational Sciences (NCATS) at NIH illustrates the critical role translational research now plays in biomedical sciences, which would not be possible without the developments in clinical research informatics . Together with translational bioinformatics, this is …

Nearly a decade since the completion of the first draft of the human genome, the biomedical community is positioned to usher in a new era of scientific inquiry that links fundamental biological insights with clinical knowledge. Accordingly, holistic approaches are needed to develop and assess hypotheses that incorporate genotypic, phenotypic, and environmental knowledge. This perspective presents translational bioinformatics as a discipline that builds on the successes of bioinformatics and health informatics for the study of complex diseases. The early successes of translational bioinformatics are indicative of the potential to achieve the promise of the Human Genome Project for gaining deeper insights to the genetic underpinnings of disease and progress toward the development of a new generation of therapies.

BACKGROUND: The amount of data deposited in the Gene Expression Omnibus (GEO) has expanded significantly. It is important to ensure that these data are properly annotated with clinical data and descriptions of experimental conditions so that they can be useful for future analysis. This study assesses the adequacy of documented asthma markers in GEO. Three objective measures (coverage, consistency and association) were used for evaluation of annotations contained in 17 asthma studies. RESULTS: There were 918 asthma samples with 20,640 annotated markers. Of these markers, only 10,419 had documented values (50% coverage). In one study carefully examined for consistency, there were discrepancies in drug name usage, with brand name and generic name used in different sections to refer to the same drug. Annotated markers showed adequate association with other relevant variables (i.e. the use of medication only when its corresponding disease state was present). CONCLUSIONS: There is inadequate variable coverage within GEO and usage of terms lacks consistency. Association between relevant variables, however, was adequate.

BACKGROUND: This study describes a large-scale manual re-annotation of data samples in the Gene Expression Omnibus (GEO), using variables and values derived from the National Cancer Institute thesaurus. A framework is described for creating an annotation scheme for various diseases that is flexible, comprehensive, and scalable. The annotation structure is evaluated by measuring coverage and agreement between annotators. RESULTS: There were 12,500 samples annotated with approximately 30 variables, in each of six disease categories - breast cancer, colon cancer, inflammatory bowel disease (IBD), rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), and Type 1 diabetes mellitus (DM). The annotators provided excellent variable coverage, with known values for over 98% of three critical variables: disease state, tissue, and sample type. There was 89% strict inter-annotator agreement and 92% agreement when using semantic and partial similarity measures. CONCLUSION: We show that it is possible to perform manual re-annotation of a large repository in a reliable manner.

BACKGROUND: Large repositories of biomedical research data are most useful to translational researchers if their data can be aggregated for efficient queries and analyses. However, inconsistent or non-existent annotations describing important sample details such as name of tissue or cell line, histopathological type, and subject characteristics like demographics, treatment, and survival are seldom present in data repositories, making it difficult to aggregate data. RESULTS: We created a flexible software tool that allows efficient annotation of samples using a controlled vocabulary, and report on its use for the annotation of over 12,500 samples. CONCLUSION: While the amount of data is very large and seemingly poorly annotated, a lot of information is still within reach. Consistent tool-based re-annotation enables many new possibilities for large scale interpretation and analyses that would otherwise be impossible.

The American College of Medical Informatics is an honorary society established to recognize those who have made sustained contributions to the field. Its highest award, for lifetime achievement and contributions to the discipline now known more inclusively as biomedical informatics, is the Morris F Collen Award. Dr. Collen's own efforts as a pioneer in the field stand as the embodiment of creativity, intellectual rigor, perseverance, and personal integrity. At most once a year, the College gives its highest recognition to an individual whose attainments have, throughout a career, substantially advanced the science and art of biomedical informatics. In 2008, the College was proud to present the Collen Award to Robert A. Greenes, MD, PhD (Figure 1). ‘Bob’ Greenes, who like Dr. Collen himself was one of the pioneers in biomedical informatics, has had multiple achievements as a physician, computer scientist, researcher, educator, and eloquent spokesperson for the field. His career contributions make him most deserving of the recognition embodied in the Collen Award. Robert A. Greenes, M.D., PhD 2008 Collen Award Recipient. Bob was born on June 17, 1940 in Cleveland, OH. His parents owned a hardware store in the Polish neighborhood where he and his sister grew up. Upon graduation from Cleveland Heights High School, Bob did undergraduate work at the University of Michigan, where he majored in zoology and took a course in computing (Fig 2). He entered Harvard Medical School in 1962, where he looked for individuals doing work with computers in medicine and quickly sought ways to become engaged in clinical computing research. He later noted: I've had the wonderful luck and opportunity to begin my career in the field of biomedical informatics much at the same time that the field itself was beginning to take shape. In 1964 Octo Barnett1 came to Massachusetts General Hospital (MGH) to run the hospital portion of a project called ‘The Hospital Computer Project. I joined this project as soon as I learned about it. There was a lot of naïveté and optimism in those days. The idea was to implement a time-shared operating system on a PDP-1 computer, using teletype machines connected by 10-character-per-second modems. It was one of the first time-sharing systems in the country. Neil Pappalardo, Curt Marble, and I began to think about the possibility of developing medical applications using a new operating system and computer language that we would develop on a PDP-7 computer in the laboratory. Bob Greenes as undergraduate student, University of Michigan. This novel computer language evolved during the 1960s into what became known as the Massachusetts General Hospital Utility Multiprogramming System (MUMPS) and it formed the basis for many early applications at MGH. Closely involved with the development of MUMPS was Dr. Octo Barnett, who would later say of Bob Greenes: He was one of a brilliant group of young computer-niks who created a new language and a new vision of the potential role of computer technology in medical care. With the development of MUMPS, the Hospital Computer Project was awarded a grant from the NIH. Although the novel system was very limited in memory and speed by modern criteria, a series of remarkably successful applications emerged from the MGH group over the next few years.1–4 Some 40 years later, MUMPS (which has since come to be known as “M”) continues to be heavily utilized for both medical and nonmedical applications. Upon graduating magna cum laude from Harvard Medical School in 1966, Bob took what was an unusual path at the time. Instead of pursuing an internship and residency, he applied for a postdoctoral grant from the NIH to pursue his PhD in Applied Mathematics (with an emphasis on computer science) at Harvard. To our knowledge, he thus became the first physician in the country to obtain such a doctorate. This type of MD/PhD training later became a critical component of many biomedical informatics programs in the United States. Referring to his PhD training years, Bob has noted: This was a wonderful opportunity because the program was very flexible. I was able to take courses at both Harvard and MIT, including a course in decision and control taught by Howard Raiffa, one of the fathers and founders of the field of decision science. I decided to do a thesis on structured capture of progress notes by physicians' data entry in a hypertension clinic at Mass. General Hospital. To do this project, with the help of engineers we developed a touch-screen computer interface to a CRT display terminal. We did this by pasting aluminum strips on the screen at points where selections could be made and connecting those strips with wires to a capacitor circuit (Fig 3). Cathode-ray tube display terminal used at Massachusetts General Hospital in the late 60s. The touch screen input was based on capacitance sensitive strips underlying lines of text. Also involved in this early research was Edward H. Shortliffe, MD, PhD, President and CEO Designée for the American Medical Informatics Association and recipient of the 2006 Collen award: I was personally excited to hear that Bob Greenes was selected as this year's Morris Collen awardee, since Bob has been arguably the single most consistent influence on my professional career since my early days as a Harvard undergraduate when I went over to Octo Barnett's lab and was paired-up with Bob as his research assistant while he was doing his PhD in the late 1960s. The field that we now call biomedical informatics was unrecognized as a discipline at that time, and others have pointed out that Bob Greenes was a visionary in his appreciation of what was coming and the kind of professional commitments that were needed. For example, Milton Corn, MD, Director of Extramural Programs at the National Library of Medicine, has noted: I think there was some appreciation, some ability to see the future, in Bob, that made him realize long before almost anybody else that the role of the computer in improving health care was going to be enormous … and he took a chance. After completing his PhD, Bob briefly worked in the commercial world, serving as President of a young company, Automated Health Systems, Inc, that sold MUMPS-based clinical information systems into the hospital marketplace. Recognizing that he preferred the academic research environment, Bob then jointed the faculty at Stanford Medical School for a year. However, he determined that he needed to complete his clinical training to have the kind of impact that he desired, so he returned to Boston to pursue a residency in radiology at Massachusetts General Hospital. After completing his training, he joined the Harvard Radiology Faculty, based at the Brigham and Women's Hospital, where he combined his clinical work in radiology with a rapidly growing research and education program in clinical informatics and decision support. He benefited from the strong support of his departmental leadership, one of whom, Professor and Chairman Steven Seltzer, would later say of Bob: The way I really got to know Bob was that for almost 20 years we shared an office cluster. It showed me that Bob had many special characteristics above and beyond his great intellectual talent as an informatician and as an excellent diagnostic radiologist. It showed me that he was patient, that he was flexible, that he was willing to share, and that he was always a good colleague and friend. Upon completing his medical training in 1978, Bob founded the Decision Systems Group (DSG), a research and training unit within Brigham and Women's Hospital's (BWH) Department of Radiology, which currently has over 100 alumni, many of whom have taken lead positions in academia and industry around the world. Bob continued to develop applications for use in the clinic (Fig 4), 5 and to contribute to the evolution of MUMPS.6 He also worked on a system for measurement, calculation, reporting, and retrieval of obstetric ultrasound examination,7 and a graphical tool for clinical decision support.8 Bob Greenes in the Department of Radiology, Brigham and Women's Hospital, Harvard Medical School. In the late 70s and early 80s, the fields of medical informatics and health decision sciences were not yet as specialized. The Boston area had a large number of faculty whose research was dedicated to medical decision analysis. In this rich environment, Bob collaborated with several pioneers in health decision sciences, such as Barbara McNeil, Stephen Pauker, Milton Weinstein, and Colin Begg. Bob published results of his decision analysis research in Medical Decision Making, then in its third year of existence,8 as well as several other journals.9–12 He collaborated with his former mentor, Octo Barnett, and Rita Zielstorff in the first IAIMS grant to Harvard Medical School.13 At the DSG, Bob led several federally funded projects related to medical knowledge management and decision support. A key contribution of the DSG was its role in the development of the Unified Medical Language System (UMLS). Dr. Betsey Humphreys, Deputy Director of the National Library of Medicine, has noted: In my mind, Bob Greenes will always be associated with the early years of the UMLS project. He was at that time utterly consistent, and throughout the rest of his career as well, in saying that what we really needed were some good tools that would make the UMLS resources more useful and usable to people. So Bob was really ahead of the curve and brave to be working solely or heavily on the tool side when there was absolutely nothing like platform independence (in fact, in a period in which platform independence was a pipe-dream). The UMLS effort was precedent setting, not only for the results but for the way in which the coordinated project, with contracts from the NLM to several leading informatics research sites, brought people together across institutional boundaries and showed how collaboration could work at a distance and with crucial involvement by trainees as well as faculty members. Bob Greenes would later note: I've come to believe that informatics is a terrific vehicle for social engineering. If one has a good idea and can bring people together to test it, to explore it, to expand on it, and to use it, we've actually created something that was not there before. Bob supervised research on information retrieval by one of his first trainees, Dr. William Hersh.14 He also addressed practical problems at the Department of Radiology by developing practical computer-based solutions. For example, derivatives of pioneer work on a structured reporting system for ultrasound studies, developed by Dr. Doug Bell under Bob's supervision, are still in use at BWH.15,16 Another early trainee, Dr. Rick Shiffman, worked on the representation of clinical practice guidelines as decision tables.17 Bob allowed his trainees to explore their own ideas, but also brought them back firmly to well-grounded problems that could be tackled effectively and in the time available to them. As Dr. Shiffman (now an associate professor at Yale University) recalls: Among the first projects I proposed to Bob was creating an expert system to help with the diagnosis of dysmorphic children. At the time Bob himself was thinking about knowledge representation using algorithms. And he counter-offered a suggestion that I pursue a problem he was wrestling with. Now one of Bob's most enviable qualities is his ability to focus on a problem until it gets solved. Each month I'd offer a fresh expert systems proposal at our meetings and each month he'd patiently re-describe for me the issue with the CAT scan, and how he thought it'd make a good fellow project. Dr. Bill Hersh, Professor and Chair of the Department of Medical Informatics and Clinical Epidemiology at the Oregon Health and Sciences University, also has fond memories of Bob's mentoring skills: An important aspect of Bob's mentorship is that he taught me how to be a mentor … he showed me that it's not just a matter of teaching someone or critiquing their work, but really keeping an eye out for their career and the directions that they're headed. Another former trainee, Dr. Luke Sato (Chief Information Officer for the Harvard Risk Management Foundation), notes: I was just really struck with the sophistication, the innovation, and the out of the box thinking that was taking place in his laboratory. He really forced us to think non-traditionally and challenged us on our assumptions and forced us to think beyond what was really capable at the time. These themes are often recapitulated by other trainees, such as Dr. Steven Labkoff, now a Senior Director with Pfizer's Global Business Unit: One of the most compelling things about working in the DSG was best summed up in a poem by Edgar A Guest. The poem was, “It Couldn't Be Done”. The poem outlines stepping up to huge challenges. It summed up Bob's philosophy about our projects. He seldom if ever said “it can't be done”, or “you can't do that” in general. In point of fact, the lab was a place which encouraged me to push the envelope, test the limits, and where possible, break through. It was this kind of creative freedom that helped me to try ground breaking things. It is a philosophy I've kept with me throughout my career in the business world. Bob led a group of talented physician trainees and computer scientists in the exploration of several new technologies that later became mainstream in informatics, such as a web-based prototype for scientific journal publishing in the mid 90s (a prototype for the New England Journal of Medicine's online publication was developed at the DSG),18 and a web system for Partners Healthcare System (a merger of BWH and Massachusetts General Hospital, two of the academic medical centers affiliated with Harvard Medical School).19 Other contributions from this period include participation in the InterMed Collaboratory,20 a consortium of leading medical informatics centers from 1995 to the early 2000s involving medical informatics researchers from Columbia, Stanford, McGill, and Harvard Universities. As part of InterMed, the GuideLine Interchange format (GLIF),21 a model for representing and sharing guidelines in a computer-interpretable format, was created. Under Bob's leadership, this line of research, as well as many others, was continued at the DSG (Fig 5). 22 GELLO, an expression language for GLIF, later became an HL-7 standard.23 Bob and the DSG family and friends (2007) Front: Lucila Ohno-Machado, Carole Greenes, Bob Greenes, Octo Barnett. Middle: Rosa Figueroa, Qing Zeng, Hyeon-Eui Kim, Ronilda Lacson, Staal Vinterbo, Margarita Sordo. Back: James Signorovitch, Esther Shilcrat, Christopher Tsai, William Solomon, Pankaj Sarin, Sassikiran Kandula, Sergei Goryachev. In 2005, at the celebration of the 25th anniversary of the DSG, Bob became the first incumbent of the Distinguished Chair in Biomedical Informatics at BWH and also received a leadership award from AMIA. Over 200 colleagues, family, and friends from around the world came to the Harvard Faculty Club to celebrate this memorable event. Bob founded and was the director of the BIRT program for 15 years. The BIRT program is the program in biomedical informatics, with and postdoctoral trainees who research in biomedical informatics in the Boston This consortium of MIT, and Boston Bob worked with Professor from in the of a program in medical informatics at the of Health Sciences and of The program was in and has over who now several positions in the industry and and are the Department of and Computer at or the Department of Health Decision at Harvard. his role as training program director and director of one of the that the Bob had the opportunity to with several trainees, and from (Fig His in this role to Bob's leadership and continued to education in biomedical informatics. The of his role has been not only by his trainees and colleagues, but also by such as Dr. Director of the National Library of I think Bob more to to and on to a I think he a problem as an opportunity for and what a good to begin with. I can years he was very with in the had some programs that allowed to make and of almost and he got a lot out of he used in this teaching so that a could ahead and develop a problem by the and of program I think he has been an excellent to many people. training program faculty Octo Barnett, Howard and Bob Bob in many on and academic and he on the Biomedical Library of the National Library of which he as Chairman in as well as in other at NIH. A fellow of the American College of Medical Informatics since year in which the was he to positions of leadership in a number of professional including the American Medical Informatics the of and the American College of He was President of the American College of Medical Informatics in and for many years on the Radiology Information Systems He is also a at and where he not only his own research but also often is to that his knowledge of the field and his ability to the that from computer and hardware to both the clinical and medical In he was to in the of of the National of Bob Greenes has over 200 in the fields of medical informatics and is an associate of the Journal of Biomedical He is the of a Clinical Decision the published by (Fig Bob also two to the now Biomedical Computer in Health and and Bob Greenes: Clinical Decision The After that kept Bob and his Carole in Boston for most of their both of their professional were in their decision to their and positions at University in and Bob was by the opportunity to as of a new Department of Biomedical created at as part of their contribution to a new medical in created with the University of The University of College of in with University brought a new medical to the in the United and the only United a medical The University of College of was founded in in and has been the only School of in the It was established with strong support by and and with an early to biomedical informatics into the medical as part of a collaboration the two his new role in Bob has been an new with faculty and advanced He and his have the new program in a and he has his for what is in the The has out to be a very and … with the of collaboration and the to and things to health and One of his faculty Professor him in and helped to the program established before Bob was I to and to be the of the formed of biomedical informatics, one of my was to a One of the first that came to my was Bob his in informatics education and his as a in his field. The leadership has been remarkably visionary in the for an academic program in biomedical informatics and working to the The President PhD, has been that the has Bob Greenes to this leadership He came to from Harvard and has a our of biomedical informatics. more I think of this as an of the kind of impact he had throughout his he in and and Dr. Professor of Computer and Director of School of and Informatics (which is the to the new Department of Biomedical has the at which Bob has biomedical informatics and made it an part of many clinical research and academic in is of the kind of Bob in of able to things as well as with very Bob his novel and his to be a and his to be the first in our field to the important to which he to his time and His special contributions to medical decision and its have been as by Octo of the work of we in clinical decision have benefited by on Bob's It is and for his career and contributions to biomedical informatics to have been with the Morris Collen Award for

The AMIA Annual Symposium is the premier forum for live presentation of scientific studies in biomedical informatics. Since its inception in 1976 as the Symposium on Computer Applications in Medical Care (SCAMC), the AMIA Symposium has showcased leading-edge informatics studies and promoted the exchange of ideas within a diverse community. Recognizing the importance of the symposium, as well as the reality of current academic promotion systems, the Scientific Program Committee (SPC) of the 2008 AMIA Annual Symposium developed a strategy to recommend outstanding AMIA manuscript submissions for possible publication in JAMIA. Since MEDLINE® has for years indexed the articles that appear in the AMIA Symposium, JAMIA has traditionally treated those works as “already published.” Thus, in order for a paper published in the AMIA Proceedings to merit consideration as a JAMIA submission, JAMIA has required authors to add new methods, and produce substantially different results that lead to new insights. The intent of the current new initiative was to motivate authors to submit their best work and have it presented at the Annual Symposium, while also giving selected authors the opportunity to publish nonduplicative work in a scientific journal. The key enabling factor was early review of AMIA Symposium submissions that allowed us to invite authors of outstanding AMIA papers to convert them to JAMIA submissions directly, while substituting abstracts of less than 400 words to replace the selected manuscripts' original submissions in the AMIA Proceedings. In the first year of the initiative, we implemented a pilot strategy. The 2008 AMIA Scientific Program Committee nominated only a very small number of manuscripts for consideration in JAMIA. From twenty-two nominated manuscripts, the JAMIA Editorial Office selected nine for full review as potential JAMIA submissions. The SPC invited the authors to revise their manuscripts according to recommendations from AMIA reviewers, and to extend the text wherever necessary. The authors were free to follow the recommendation and submit to JAMIA, or to decide to leave their AMIA papers “as is” in the AMIA Proceedings and not submit to JAMIA. All of the papers appeared as regular presentations at the AMIA Meeting. Those papers that did not pass JAMIA peer review still appeared in the AMIA Proceedings as 400-word abstracts, and authors can still submit the corresponding full-length papers elsewhere if they choose to do so. We include in this issue a set of four AMIA submissions that passed full, rigorous JAMIA peer review after expansion as “selected AMIA manuscripts”. By coincidence, all four relate to information retrieval and natural language processing applications: Kilicoglu et al.1 demonstrated the feasibility of utilizing machine learning algorithms to automate the process of retrieving scientifically rigorous clinical research evidence from the literature, and discuss the potential for clinicians to benefit from this application. Lu et al.2 showed that document retrieval by a simple term-weighting approach (TF-IDF3) outperformed retrieval based on sentence-level co-occurrence in a given set of queries for MEDLINE. The authors advocate for the use of this approach in PubMed®. The two manuscripts related to natural language processing include: Uzuner et al.4 who extended the rule-based NegEx algorithm5 to cover alter-association assertions in an Extended NegEx (ENegEx) system and compared it to a Statistical Assertion Classifier (StAC). The authors showed that StAC models developed on a training set of discharge summaries outperform ENegEx when applied to a previously unseen (test) set of radiology reports. Xu et al.6 showed that a clustering-based method outperforms manual annotation in the task of building sense inventories of clinical abbreviations for randomly selected samples of hospital admission notes. Three other manuscripts originally submitted to the AMIA Symposium and later extended for JAMIA will appear in the next JAMIA issue7–9 and are currently available on JAMIA's Web site as publish-ahead-of-print “PrePrints.” The remaining two manuscripts are currently undergoing revision and their final JAMIA status will be determined at a later date. The SPC for the 2009 AMIA Annual Symposium has committed to continue and expand support for this initiative. We expect an even larger number of outstanding submissions to the AMIA Symposium by the deadline of March 13, 2009. The SPC will scale up the pilot strategy adopted in 2008 to promote a substantial increase in the number of manuscripts that are invited for potential publication in several biomedical informatics journals. Annual symposium attendees, biomedical informatics journal readers, and especially authors will benefit from the flexibility of this strategy, which constitutes a small but important step towards the goal of rewarding excellence in the field of biomedical informatics.

Monitoring vital signs and locations of certain classes of ambulatory patients can be useful in overcrowded emergency departments and at disaster scenes, both on-site and during transportation. To be useful, such monitoring needs to be portable and low cost, and have minimal adverse impact on emergency personnel, e.g., by not raising an excessive number of alarms. The SMART (Scalable Medical Alert Response Technology) system integrates wireless patient monitoring (ECG, SpO(2)), geo-positioning, signal processing, targeted alerting, and a wireless interface for caregivers. A prototype implementation of SMART was piloted in the waiting area of an emergency department and evaluated with 145 post-triage patients. System deployment aspects were also evaluated during a small-scale disaster-drill exercise.

UNLABELLED: A critical step in any SAGE, MPSS and SBS data analysis is tag-to-gene assignment. Current available tools are limited by a tag-by-tag annotation process and/or do not provide the dataset that is used to produce a complete tag-to-gene mapping. We developed ACTG, a web-based application that allows a large-scale tag-to-gene mapping using several reference datasets. ACTG can annotate SAGE (14 or 21 bp), MPSS (17 or 20 bp) and SBS (16 bp) data for both human and mouse organisms. AVAILABILITY: http://retina.med.harvard.edu/ACTG/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Single nucleotide polymorphisms (SNPs) are locations at which the genomic sequences of population members differ. Since these differences are known to follow patterns, disease association studies are facilitated by identifying SNPs that allow the unique identification of such patterns. This process, known as haplotype tagging, is formulated as a combinatorial optimization problem and analyzed in terms of complexity and approximation properties. RESULTS: It is shown that the tagging problem is NP-hard but approximable within 1 + ln((n2 - n)/2) for n haplotypes but not approximable within (1-epsilon) ln(n/2) for any epsilon > 0 unless NP subset DTIME(n(log log n)). A simple, very easily implementable algorithm that exhibits the above upper bound on solution quality is presented. This algorithm has running time O(np/2(2m-p+1)) < or = O(m(n2-n)/2) where p < or = min(n, m) for n haplotypes of size m. As we show that the approximation bound is asymptotically tight, the algorithm presented is optimal with respect to this asymptotic bound. CONCLUSION: The haplotype tagging problem is hard, but approachable with a fast, practical, and surprisingly simple algorithm that cannot be significantly improved upon on a single processor machine. Hence, significant improvement in computational efforts expended can only be expected if the computational effort is distributed and done in parallel.

OBJECTIVE: A variety of postmarketing surveillance strategies to monitor the safety of medical devices have been supported by the U.S. Food and Drug Administration, but there are few systems to automate surveillance. Our objective was to develop a system to perform real-time monitoring of safety data using a variety of process control techniques. DESIGN: The Web-based Data Extraction and Longitudinal Time Analysis (DELTA) system imports clinical data in real-time from an electronic database and generates alerts for potentially unsafe devices or procedures. The statistical techniques used are statistical process control (SPC), logistic regression (LR), and Bayesian updating statistics (BUS). MEASUREMENTS: We selected in-patient mortality following implantation of the Cypher drug-eluting coronary stent to evaluate our system. Data from the University of Michigan Consortium Bare-Metal Stent Study was used to calculate the event rate alerting boundaries. Data analysis was performed on local catheterization data from Brigham and Women's Hospital from July 1, 2003, shortly after the Cypher release, to December 31, 2004, including 2,270 cases with 27 observed deaths. RESULTS: The single-stratum SPC had alerts in months 4 and 10. The multistrata SPC had alerts in months 5, 10, and 18 in the moderate-risk stratum, and months 1, 4, 7, and 10 in the high-risk stratum. The only cumulative alerts were in the first month for the high-risk stratum of the multistrata SPC. The LR method showed no monthly or cumulative alerts. The BUS method showed an alert in the first month for the high-risk stratum. CONCLUSION: The system performed adequately within the Brigham and Women's Hospital Intranet environment based on the design goals. All three cumulative methods agreed that the overall observed event rates were not significantly higher for the new medical device than for a closely related medical device and were consistent with the observation that the initial concerns about this device dissipated as more data accumulated.

MOTIVATION: Interpretation of classification models derived from gene-expression data is usually not simple, yet it is an important aspect in the analytical process. We investigate the performance of small rule-based classifiers based on fuzzy logic in five datasets that are different in size, laboratory origin and biomedical domain. RESULTS: The classifiers resulted in rules that can be readily examined by biomedical researchers. The fuzzy-logic-based classifiers compare favorably with logistic regression in all datasets. AVAILABILITY: Prototype available upon request.

MOTIVATION: [corrected] The existence of several technologies for measuring gene expression makes the question of cross-technology agreement of measurements an important issue. Cross-platform utilization of data from different technologies has the potential to reduce the need to duplicate experiments but requires corresponding measurements to be comparable. METHODS: A comparison of mRNA measurements of 2895 sequence-matched genes in 56 cell lines from the standard panel of 60 cancer cell lines from the National Cancer Institute (NCI 60) was carried out by calculating correlation between matched measurements and calculating concordance between cluster from two high-throughput DNA microarray technologies, Stanford type cDNA microarrays and Affymetrix oligonucleotide microarrays. RESULTS: In general, corresponding measurements from the two platforms showed poor correlation. Clusters of genes and cell lines were discordant between the two technologies, suggesting that relative intra-technology relationships were not preserved. GC-content, sequence length, average signal intensity, and an estimator of cross-hybridization were found to be associated with the degree of correlation. This suggests gene-specific, or more correctly probe-specific, factors influencing measurements differently in the two platforms, implying a poor prognosis for a broad utilization of gene expression measurements across platforms.

Privacy protection is an important consideration when releasing medical databases to the research community. We show that while recent advances in anonymization algorithms provide increased levels of protection, it is still possible to calculate approximations to the original data set. In some cases, one can even uniquely reconstruct entries in a table before anonymization. In this paper, we demonstrate how knowledge of an anonymization algorithm based on ambiguating data cell entries can be used to undo the anonymization process. We investigate the effect of this algorithm and its reversal on data sets of varying sizes and distributions. It is shown that by using a computationally complex disambiguation process, information on individuals can be extracted from an anonymized data set.

Protecting individual data in disclosed databases is essential. Data anonymization strategies can produce table ambiguation by suppression of selected cells. Using table ambiguation, different degrees of anonymization can be achieved, depending on the number of individuals that a particular case must become indistinguishable from. This number defines the level of anonymization. Anonymization by cell suppression does not necessarily prevent inferences from being made from the disclosed data. Preventing inferences may be important to preserve confidentiality. We show that anonymized data sets can preserve descriptive characteristics of the data, but might also be used for making inferences on particular individuals, which is a feature that may not be desirable. The degradation of predictive performance is directly proportional to the degree of anonymity. As an example, we report the effect of anonymization on the predictive performance of a model constructed to estimate the probability of disease given clinical findings.

We present a prototype of a decision support system for anesthesia that applies set covering theory. The system is designed to generate dynamically configured check-lists for intra-operative problems. These lists have the potential to help anesthesiologists detect and manage problems in a timely manner. The items in the lists consist of major complications that should be considered for a particular case. A set covering algorithm that accommodates multiple problem sets was used to implement the prototype. A simulated case and the system behavior are presented. The ultimate goals of a system such as the one presented are to function as an intelligent alarm module for electronic monitors and to facilitate the task of correcting intra-operative problems.

OBJECTIVE: To allow exchange of clinical practice guidelines among institutions and computer-based applications. DESIGN: The GuideLine Interchange Format (GLIF) specification consists of GLIF model and the GLIF syntax. The GLIF model is an object-oriented representation that consists of a set of classes for guideline entities, attributes for those classes, and data types for the attribute values. The GLIF syntax specifies the format of the test file that contains the encoding. METHODS: Researchers from the InterMed Collaboratory at Columbia University, Harvard University (Brigham and Women's Hospital and Massachusetts General Hospital), and Stanford University analyzed four existing guideline systems to derive a set of requirements for guideline representation. The GLIF specification is a consensus representation developed through a brainstorming process. Four clinical guidelines were encoded in GLIF to assess its expressivity and to study the variability that occurs when two people from different sites encode the same guideline. RESULTS: The encoders reported that GLIF was adequately expressive. A comparison of the encodings revealed substantial variability. CONCLUSION: GLIF was sufficient to model the guidelines for the four conditions that were examined. GLIF needs improvement in standard representation of medical concepts, criterion logic, temporal information, and uncertainty.

This paper describes a medical application of modular neural networks (NNs) for temporal pattern recognition. In order to increase the reliability of prognostic indices for patients living with the acquired immunodeficiency syndrome (AIDS), survival prediction was performed in a system composed of modular NNS that classified cases according to death in a certain year of follow-up. The output of each NN module corresponded to the probability of survival in a given year. Inputs were the values of demographic, clinical and laboratory variables. The results of the modules were combined to produce survival curves for individuals. The NNs were trained by backpropagation and the results were evaluated in test sets of previously unseen cases. We showed that, for certain combinations of NN modules, the performance of the prognostic index, measured by the area under the receiver operating characteristic curve, was significantly improved (p 0.05). We also used calibration measurements to quantify the benefits of combining NN modules, and show why, when and how NNs should be combined for building prognostic models.