OBJECTIVES: This article describes the challenges faced by the National Library of Medicine with the rise of artificial intelligence (AI) and access to human knowledge through large language models (LLMs). BACKGROUND AND SIGNIFICANCE: The rise of AI as a tool for the acceleration and falsification of science is impacting every aspect of the transformation of data to information, knowledge, and wisdom through the scientific processes. APPROACH: This perspective discusses the philosophical foundations, threats, and opportunities of the AI revolution with a proposal for restructuring the mission of the National Library of Medicine (NLM), part of the National Institutes of Health, with a central role as the guardian of the integrity of scientific knowledge in an era of AI-driven science. RESULTS: The NLM can rise to new challenges posed by AI by working from its foundations in theories of Information Science and embracing new roles. Three paths for the NLM are proposed: (1) Become an Authentication Authority For Data, Information, and Knowledge through Systems of Scientific Provenance; (2) Become An Observatory of the State of Human Health Science supporting living systematic reviews; and (3) Become A hub for Culturally Appropriate Bespoke Translation, Transformation, and Summarization for different users (patients, the public, as well as scientists and clinicians) using AI technologies. DISCUSSION: Adapting the NLM to the challenges of the Internet revolution by developing worldwide-web-accessible resources allowed the NLM to rise to new heights. Bold moves are needed to adapt the Library to the AI revolution but offer similar prospects of more significant impacts on the advancement of science and human health.

OBJECTIVES: This study aims to summarize the usage of large language models (LLMs) in the process of creating a scientific review by looking at the methodological papers that describe the use of LLMs in review automation and the review papers that mention they were made with the support of LLMs. MATERIALS AND METHODS: The search was conducted in June 2024 in PubMed, Scopus, Dimensions, and Google Scholar by human reviewers. Screening and extraction process took place in Covidence with the help of LLM add-on based on the OpenAI GPT-4o model. ChatGPT and Scite.ai were used in cleaning the data, generating the code for figures, and drafting the manuscript. RESULTS: Of the 3788 articles retrieved, 172 studies were deemed eligible for the final review. ChatGPT and GPT-based LLM emerged as the most dominant architecture for review automation (n = 126, 73.2%). A significant number of review automation projects were found, but only a limited number of papers (n = 26, 15.1%) were actual reviews that acknowledged LLM usage. Most citations focused on the automation of a particular stage of review, such as Searching for publications (n = 60, 34.9%) and Data extraction (n = 54, 31.4%). When comparing the pooled performance of GPT-based and BERT-based models, the former was better in data extraction with a mean precision of 83.0% (SD = 10.4) and a recall of 86.0% (SD = 9.8). DISCUSSION AND CONCLUSION: Our LLM-assisted systematic review revealed a significant number of research projects related to review automation using LLMs. Despite limitations, such as lower accuracy of extraction for numeric data, we anticipate that LLMs will soon change the way scientific reviews are conducted.

INTRODUCTION: This study aimed to identify barriers and facilitators to the implementation of family cancer history (FCH) collection tools in clinical practices and community settings by assessing clinicians' perceptions of implementing a chatbot interface to collect FCH information and provide personalized results to patients and providers. OBJECTIVES: By identifying design and implementation features that facilitate tool adoption and integration into clinical workflows, this study can inform future FCH tool development and adoption in healthcare settings. MATERIALS AND METHODS: Quantitative data were collected using survey to evaluate the implementation outcomes of acceptability, adoption, appropriateness, feasibility, and sustainability of the chatbot tool for collecting FCH. Semistructured interviews were conducted to gather qualitative data on respondents' experiences using the tool and recommendations for enhancements. RESULTS: We completed data collection with 19 providers (n = 9, 47%), clinical staff (n = 5, 26%), administrators (n = 4, 21%), and other staff (n = 1, 5%) affiliated with the NCI Community Oncology Research Program. FCH was systematically collected using a wide range of tools at sites, with information being inserted into the patient's medical record. Participants found the chatbot tool to be highly acceptable, with the tool aligning with existing workflows, and were open to adopting the tool into their practice. DISCUSSION AND CONCLUSIONS: We further the evidence base about the appropriateness of scripted chatbots to support FCH collection. Although the tool had strong support, the varying clinical workflows across clinic sites necessitate that future FCH tool development accommodates customizable implementation strategies. Implementation support is necessary to overcome technical and logistical barriers to enhance the uptake of FCH tools in clinical practices and community settings.

OBJECTIVES: Stressful life events, such as going through divorce, can have an important impact on human health. However, there are challenges in capturing these events in electronic health records (EHR). We conducted a scoping review aimed to answer 2 major questions: how stressful life events are documented in EHR and how they are utilized in research and clinical care. MATERIALS AND METHODS: Three online databases (EBSCOhost platform, PubMed, and Scopus) were searched to identify papers that included information on stressful life events in EHR; paper titles and abstracts were reviewed for relevance by 2 independent reviewers. RESULTS: Five hundred fifty-seven unique papers were retrieved, and of these 70 were eligible for data extraction. Most articles (n = 36, 51.4%) were focused on the statistical association between one or several stressful life events and health outcomes, followed by clinical utility (n = 15, 21.4%), extraction of events from free-text notes (n = 12, 17.1%), discussing privacy and other issues of storing life events (n = 5, 7.1%), and new EHR features related to life events (n = 4, 5.7%). The most frequently mentioned stressful life events in the publications were child abuse/neglect, arrest/legal issues, and divorce/relationship breakup. Almost half of the papers (n = 7, 46.7%) that analyzed clinical utility of stressful events were focused on decision support systems for child abuse, while others (n = 7, 46.7%) were discussing interventions related to social determinants of health in general. DISCUSSION AND CONCLUSIONS: Few citations are available on the prevalence and use of stressful life events in EHR reflecting challenges in screening and storing of stressful life events.

OBJECTIVE: The annual American College of Medical Informatics (ACMI) symposium focused discussion on the national public health information systems (PHIS) infrastructure to support public health goals. The objective of this article is to present the strengths, weaknesses, threats, and opportunities (SWOT) identified by public health and informatics leaders in attendance. MATERIALS AND METHODS: The Symposium provided a venue for experts in biomedical informatics and public health to brainstorm, identify, and discuss top PHIS challenges. Two conceptual frameworks, SWOT and the Informatics Stack, guided discussion and were used to organize factors and themes identified through a qualitative approach. RESULTS: A total of 57 unique factors related to the current PHIS were identified, including 9 strengths, 22 weaknesses, 14 opportunities, and 14 threats, which were consolidated into 22 themes according to the Stack. Most themes (68%) clustered at the top of the Stack. Three overarching opportunities were especially prominent: (1) addressing the needs for sustainable funding, (2) leveraging existing infrastructure and processes for information exchange and system development that meets public health goals, and (3) preparing the public health workforce to benefit from available resources. DISCUSSION: The PHIS is unarguably overdue for a strategically designed, technology-enabled, information infrastructure for delivering day-to-day essential public health services and to respond effectively to public health emergencies. CONCLUSION: Most of the themes identified concerned context, people, and processes rather than technical elements. We recommend that public health leadership consider the possible actions and leverage informatics expertise as we collectively prepare for the future.

The COVID-19 pandemic exposed multiple weaknesses in the nation's public health system. Therefore, the American College of Medical Informatics selected "Rebuilding the Nation's Public Health Informatics Infrastructure" as the theme for its annual symposium. Experts in biomedical informatics and public health discussed strategies to strengthen the US public health information infrastructure through policy, education, research, and development. This article summarizes policy recommendations for the biomedical informatics community postpandemic. First, the nation must perceive the health data infrastructure to be a matter of national security. The nation must further invest significantly more in its health data infrastructure. Investments should include the education and training of the public health workforce as informaticians in this domain are currently limited. Finally, investments should strengthen and expand health data utilities that increasingly play a critical role in exchanging information across public health and healthcare organizations.

OBJECTIVE: Opioid-related overdose (OD) deaths continue to increase. Take-home naloxone (THN), after treatment for an OD in an emergency department (ED), is a recommended but under-utilized practice. To promote THN prescription, we developed a noninterruptive decision support intervention that combined a detailed OD documentation template with a reminder to use the template that is automatically inserted into a provider's note by decision rules. We studied the impact of the combined intervention on THN prescribing in a longitudinal observational study. METHODS: ED encounters involving an OD were reviewed before and after implementation of the reminder embedded in the physicians' note to use an advanced OD documentation template for changes in: (1) use of the template and (2) prescription of THN. Chi square tests and interrupted time series analyses were used to assess the impact. Usability and satisfaction were measured using the System Usability Scale (SUS) and the Net Promoter Score. RESULTS: In 736 OD cases defined by International Classification of Disease version 10 diagnosis codes (247 prereminder and 489 postreminder), the documentation template was used in 0.0% and 21.3%, respectively (P < .0001). The sensitivity and specificity of the reminder for OD cases were 95.9% and 99.8%, respectively. Use of the documentation template led to twice the rate of prescribing of THN (25.7% vs 50.0%, P < .001). Of 19 providers responding to the survey, 74% of SUS responses were in the good-to-excellent range and 53% of providers were Net Promoters. CONCLUSIONS: A noninterruptive decision support intervention was associated with higher THN prescribing in a pre-post study across a multiinstitution health system.

COVID-19 vaccination uptake has been suboptimal, even in high-risk populations. New approaches are needed to bring vaccination data to the groups leading outreach efforts. This article describes work to make state-level vaccination data more accessible by extending the Bulk Fast Healthcare Interoperability Resource (FHIR) standard to better support the repeated retrieval of vaccination data for coordinated outreach efforts. We also describe a corresponding low-foot-print software for population outreach that automates repeated checks of state-level immunization data and prioritizes outreach by social determinants of health. Together this software offers an integrated approach to addressing vaccination gaps. Several extensions to the Bulk FHIR protocol were needed to support bulk query of immunization records. These are described in detail. The results of a pilot study, using the outreach tool to target a population of 1500 patients are also described. The results confirmed the limitations of current patient-by-patient approach for querying state immunizations systems for population data and the feasibility of a Bulk FHIR approach.

Prior authorization (PA) may be a necessary evil within the healthcare system, contributing to physician burnout and delaying necessary care, but also allowing payers to prevent wasting resources on redundant, expensive, and/or ineffective care. PA has become an "informatics issue" with the rise of automated methods for PA review, championed in the Health Level 7 International's (HL7's) DaVinci Project. DaVinci proposes using rule-based methods to automate PA, a time-tested strategy with known limitations. This article proposes an alternative that may be more human-centric, using artificial intelligence (AI) methods for the computation of authorization decisions. We believe that by combining modern approaches for accessing and exchanging existing electronic health data with AI methods tailored to reflect the judgments of expert panels that include patient representatives, and refined with "few shot" learning approaches to prevent bias, we could create a just and efficient process that serves the interests of society as a whole. Efficient simulation of human appropriateness assessments from existing data using AI methods could eliminate burdens and bottlenecks while preserving PA's benefits as a tool to limit inappropriate care.

Research Patient Data Repositories (RPDRs) have become essential infrastructure for traditional Clinical and Translational Science Award (CTSA) programs and increasingly for a wide range of research consortia and learning health system networks.1–5 Almost every institution with a CTSA or Clinical Translational Research (CTR) program (found in states with lower amounts of National Institutes of Health funding) hosts an RPDR for the benefit of affiliated researchers. These repositories aim to enable healthcare research based upon the patient populations they serve. Within the institution, RPDRs are valuable for a range of research activities. They are used to identify patients for clinical trial recruitment using privacy-preserving methods to search and extract specific cohorts of trial-eligible patients.6 They aid in developing and validating computable phenotypes that are increasingly important for accurately identifying patient cohorts in a reproducible fashion.7 RPDRs provide de-identified patient data for population health research and support a growing body of artificial intelligence to predict patient outcomes.8 Further, clinical studies can often be simulated using data from an RPDR.9 Beyond the institution, aggregates of de-identified datasets from multiple institutions linked with privacy-preserving hash codes provide an unprecedented opportunity to conduct population health research, perform comparative effectiveness analyses and apply artificial intelligence methods over large and diverse populations.10 The data contained within the RPDR vary across institutions, based on institutional strengths and weaknesses; the papers published in this issue reflect that variability (see Table 1). Data are commonly acquired from local electronic health records (EHRs) and other clinical information systems that capture information during clinical care. Data consist of diagnoses, problem lists, procedures, prescribed medications, laboratory exams, and many types of free-text reports. Overall, the benefits of the RPDR for accelerating translational research can be significant. For example, at Harvard, in 2006, between $94 and $136 million in annual research funding was linked to the use of data from the RPDR.11

The ability to analyze human specimens is the pillar of modern-day translational research. To enhance the research availability of relevant clinical specimens, we developed the Living BioBank (LBB) solution, which allows for just-in-time capture and delivery of phenotyped surplus laboratory medicine specimens. The LBB is a system-of-systems integrating research feasibility databases in i2b2, a real-time clinical data warehouse, and an informatics system for institutional research services management (SPARC). LBB delivers deidentified clinical data and laboratory specimens. We further present an extension to our solution, the Living µBiome Bank, that allows the user to request and receive phenotyped specimen microbiome data. We discuss the details of the implementation of the LBB system and the necessary regulatory oversight for this solution. The conducted institutional focus group of translational investigators indicates an overall positive sentiment towards potential scientific results generated with the use of LBB. Reference implementation of LBB is available at https://LivingBioBank.musc.edu.

OBJECTIVE: Integrated, real-time data are crucial to evaluate translational efforts to accelerate innovation into care. Too often, however, needed data are fragmented in disparate systems. The South Carolina Clinical & Translational Research Institute at the Medical University of South Carolina (MUSC) developed and implemented a universal study identifier-the Research Master Identifier (RMID)-for tracking research studies across disparate systems and a data warehouse-inspired model-the Research Integrated Network of Systems (RINS)-for integrating data from those systems. MATERIALS AND METHODS: In 2017, MUSC began requiring the use of RMIDs in informatics systems that support human subject studies. We developed a web-based tool to create RMIDs and application programming interfaces to synchronize research records and visualize linkages to protocols across systems. Selected data from these disparate systems were extracted and merged nightly into an enterprise data mart, and performance dashboards were created to monitor key translational processes. RESULTS: Within 4 years, 5513 RMIDs were created. Among these were 726 (13%) bridged systems needed to evaluate research study performance, and 982 (18%) linked to the electronic health records, enabling patient-level reporting. DISCUSSION: Barriers posed by data fragmentation to assessment of program impact have largely been eliminated at MUSC through the requirement for an RMID, its distribution via RINS to disparate systems, and mapping of system-level data to a single integrated data mart. CONCLUSION: By applying data warehousing principles to federate data at the "study" level, the RINS project reduced data fragmentation and promoted research systems integration.

Public health faces unprecedented challenges in its efforts to control COVID-19 through a national vaccination campaign. Addressing these challenges will require fundamental changes to public health data systems. For example, of the core data systems for immunization campaigns is the immunization information system (IIS); however, IISs were designed for tracking the vaccinated, not finding the patients who are high risk and need to be vaccinated. Health systems have this data in their electronic health records (EHR) systems and often have a greater capacity for outreach. Clearly, a partnership is needed. However, successful collaborations will require public health to change from its historical hierarchical information supply chain model to an ecosystem model with a peer-to-peer exchange with population health providers. Examples of the types of informatics innovations necessary to support such an ecosystem include a national patient identifier, population-level data exchange for immunization data, and computable electronic quality measures. Rather than think of these components individually, a comprehensive approach to rapidly adaptable tools for collaboration is needed.

OBJECTIVE: The rapidly evolving COVID-19 pandemic has created a need for timely data from the healthcare systems for research. To meet this need, several large new data consortia have been developed that require frequent updating and sharing of electronic health record (EHR) data in different common data models (CDMs) to create multi-institutional databases for research. Traditionally, each CDM has had a custom pipeline for extract, transform, and load operations for production and incremental updates of data feeds to the networks from raw EHR data. However, the demands of COVID-19 research for timely data are far higher, and the requirements for updating faster than previous collaborative research using national data networks have increased. New approaches need to be developed to address these demands. METHODS: In this article, we describe the use of the Fast Healthcare Interoperability Resource (FHIR) data model as a canonical data model and the automated transformation of clinical data to the Patient-Centered Outcomes Research Network (PCORnet) and Observational Medical Outcomes Partnership (OMOP) CDMs for data sharing and research collaboration on COVID-19. RESULTS: FHIR data resources could be transformed to operational PCORnet and OMOP CDMs with minimal production delays through a combination of real-time and postprocessing steps, leveraging the FHIR data subscription feature. CONCLUSIONS: The approach leverages evolving standards for the availability of EHR data developed to facilitate data exchange under the 21st Century Cures Act and could greatly enhance the availability of standardized datasets for research.

OBJECTIVES: We describe our approach in using health information technology to provide a continuum of services during the coronavirus disease 2019 (COVID-19) pandemic. COVID-19 challenges and needs required health systems to rapidly redesign the delivery of care. MATERIALS AND METHODS: Our health system deployed 4 COVID-19 telehealth programs and 4 biomedical informatics innovations to screen and care for COVID-19 patients. Using programmatic and electronic health record data, we describe the implementation and initial utilization. RESULTS: Through collaboration across multidisciplinary teams and strategic planning, 4 telehealth program initiatives have been deployed in response to COVID-19: virtual urgent care screening, remote patient monitoring for COVID-19-positive patients, continuous virtual monitoring to reduce workforce risk and utilization of personal protective equipment, and the transition of outpatient care to telehealth. Biomedical informatics was integral to our institutional response in supporting clinical care through new and reconfigured technologies. Through linking the telehealth systems and the electronic health record, we have the ability to monitor and track patients through a continuum of COVID-19 services. DISCUSSION: COVID-19 has facilitated the rapid expansion and utilization of telehealth and health informatics services. We anticipate that patients and providers will view enhanced telehealth services as an essential aspect of the healthcare system. Continuation of telehealth payment models at the federal and private levels will be a key factor in whether this new uptake is sustained. CONCLUSIONS: There are substantial benefits in utilizing telehealth during the COVID-19, including the ability to rapidly scale the number of patients being screened and providing continuity of care.

The novel coronavirus disease 2019 infection poses serious challenges to the healthcare system that are being addressed through the creation of new unique and advanced systems of care with disjointed care processes (eg, telehealth screening, drive-through specimen collection, remote testing, telehealth management). However, our current regulations on the flows of information for clinical care and research are antiquated and often conflict at the state and federal levels. We discuss proposed changes to privacy regulations such as the Health Insurance Portability and Accountability Act designed to let health information seamlessly and frictionlessly flow among the health entities that need to collaborate on treatment of patients and, also, allow it to flow to researchers trying to understand how to limit its impacts.

Public health needs up-to-date information for surveillance and response. As healthcare application programming interfaces become widely available, a novel data gathering mechanism could provide public health with critical information in a timely fashion to respond to a fast-moving epidemic. In this article, we extrapolate from our experiences using a Fast Healthcare Interoperability Resource-based architecture for infectious disease surveillance for sexually transmitted diseases to its application to gather case information for an outbreak. One of the challenges with a fast-moving outbreak is to accurately assess its demand on healthcare resources, since information specific to comorbidities is often not available. These comorbidities are often associated with poor prognosis and higher resource utilization. If the comorbidity data and other clinical information were readily available to public health workers, they could better address community disruption and manage healthcare resources. The use of FHIR resources available through application programming and filtered through tools such as described herein will give public health the flexibility needed to investigate rapidly emerging disease while protecting patient privacy.

OBJECTIVE: In an effort to improve the efficiency of computer algorithms applied to screening for coronavirus disease 2019 (COVID-19) testing, we used natural language processing and artificial intelligence-based methods with unstructured patient data collected through telehealth visits. MATERIALS AND METHODS: After segmenting and parsing documents, we conducted analysis of overrepresented words in patient symptoms. We then developed a word embedding-based convolutional neural network for predicting COVID-19 test results based on patients' self-reported symptoms. RESULTS: Text analytics revealed that concepts such as smell and taste were more prevalent than expected in patients testing positive. As a result, screening algorithms were adapted to include these symptoms. The deep learning model yielded an area under the receiver-operating characteristic curve of 0.729 for predicting positive results and was subsequently applied to prioritize testing appointment scheduling. CONCLUSIONS: Informatics tools such as natural language processing and artificial intelligence methods can have significant clinical impacts when applied to data streams early in the development of clinical systems for outbreak response.

Adverse drug events (ADEs), defined as “any injuries resulting from medication use, including physical harm, mental harm, or loss of function,”1 are reported to account for approximately 30% of all adverse events,2 with results that can include repeated hospital admission and fatality. Information about causes of ADEs can be found in data that document concurrent use of multiple medications, drug interactions, and possible allergies such as “charts, laboratory [data], prescription data” and “administrative data.”3 However, much of the crucial information related to ADEs are detailed in free text narratives and are not easily accessible by computerized systems, requiring manual review and manual identification of this information. Natural language processing (NLP) holds potential for automatically extracting ADE-related information from narratives, to make it available for decision support systems that can alert clinicians to potential ADEs at the point of care. To assess and advance the state of the art in NLP for extraction of ADEs, the National NLP Clinical Challenges (n2c2) shared task in 2018 included a track on this topic.4 This track required the identification of potential ADE mentions, along with their link to the medication that caused them, and the administration details such as the dosage, route, and frequency information related to the medication causing the ADE. The systems that tackled extraction of ADEs and related concepts primarily utilized recurrent deep neural networks consisting of bidirectional long short-term memory units, achieving performances that reached 94% in F-measure. In linking ADEs to their causes, the systems were more diverse in their methods, utilizing a range of machine learning approaches including both deep learning and more traditional methods and achieving performances that reached 96% in F-measure. These results indicate that while they are not perfect, NLP systems can successfully extract ADE information from narratives with impressive accuracy. In this editorial, we highlight 4 systems. Others are summarized in Henry et al.4

As an interdisciplinary technologically driven field, the science of informatics is rapidly evolving. In this issue of Journal of the American Medical Informatics Association, we bring together a series of articles and commentaries that describe various aspects of the science of predictive modeling. These articles describe work to ensure that models are useful and valid on release and, perhaps more importantly, continue to be so as clinical processes and patient populations evolve over time. The upshot of the collection is to point out a new direction for informatics research and policy advocacy in the development of models for predictive analytics. Rather than focus on the mechanics of model building and validation, scientists should now be focused on how to document the model,1 when it is likely to yield benefits,2 what the model life cycle is,3 how to maintain models in a sustainable way,4 and even which types of health care offer the optimal predictive performance.5

If biomedical informatics is a science, we believe it best prospers by the careful application of the scientific method to testable hypotheses. From time to time this may require the use of simplified models to prove cause and effect or the lack thereof. Our experiment applied a controlled model of a clinical encounter and focused on the question of whether the use of an electronic health records system (EHR) on a computer is the cause of decrements in communications. As appropriate for a controlled experiment, cognitive load, except for use of the EHR, was balanced across both study arms. Our findings suggest that EHR use per se is not a problem but an advantage for residents , communications-wise.1 However, as Hauser and Zeng2 suggest, in practice the number and types of tasks that EHRs ask users to perform is far higher than typical with paper charts. These tasks are valuable but add to the cognitive load of the users; thus, in practice the cognitive load from using an EHR system might be far higher than in our controlled study. This does not invalidate the findings of our study or render it “simplistic.” Rather, it suggests that the next plausible hypothesis to test is whether higher user cognitive loads due to specific interface designs and/or additions of new tasks or interruptions are the cause of perceived problems with examination room communications. This hypothesis does have implications for strategies on how to address the issue of the perceived negative effects of computer usage in the examination room. If the problem is not the computer per se but the user’s cognitive load, then strategies such as LEVEL (Let the patient Look-on; Eye-contact; Value the computer; Explain actions; Log off) that focus on integrating the computer into the interview are not enough.3,4 They will be successful only to the degree that they slow care down and thus secondarily reduce cognitive load. Moreover, to answer Hauser and Zeng’s question about “why (would) a physician greets a patient more warmly, when walking into a room with a laptop,” the answer is, “Yes, if it takes less effort to come to understand a patient’s prior history and symptoms, it may be easier to remember to be social in complex environments.” There is an optimistic note in this—better designs for EHRs that reduce cognitive burden for providers may allow patients to have a more pleasant and person-centric experience. The science of informatics requires both carefully controlled experiments and real-world observational studies. To dismiss the structured experiment merely because it is structured is to dismiss an important part of the science of informatics.

BACKGROUND: The effects of electronic health records (EHRs) on doctor-patient communication are unclear. OBJECTIVE: To evaluate the effects of EHR use compared with paper chart use, on novice physicians' communication skills. DESIGN: Within-subjects randomized controlled trial using observed structured clinical examination methods to assess the impact of use of an EHR on communication. SETTING: A large academic internal medicine training program. POPULATION: First-year internal medicine residents. INTERVENTION: Residents interviewed, diagnosed, and initiated treatment of simulated patients using a paper chart or an EHR on a laptop computer. Video recordings of interviews were rated by three trained observers using the Four Habits scale. RESULTS: Thirty-two residents completed the study and had data available for review (61.5% of those enrolled in the residency program). In most skill areas in the Four Habits model, residents performed at least as well using the EHR and were statistically better in six of 23 skills areas (p<0.05). The overall average communication score was better when using an EHR: mean difference 0.254 (95% CI 0.05 to 0.45), p = 0.012, Cohen's d of 0.47 (a moderate effect). Residents scoring poorly (>3 average score) with paper methods (n = 8) had clinically important improvement when using the EHR. LIMITATIONS: This study was conducted in first-year residents in a training environment using simulated patients at a single institution. CONCLUSIONS: Use of an EHR on a laptop computer appears to improve the ability of first-year residents to communicate with patients relative to using a paper chart.

The Office of the National Coordinator will be defining the architecture of the Nationwide Health Information Network (NWHIN) together with the proposed HealtheWay public/private partnership as a development and funding strategy. There are a number of open questions--for example, what is the best way to realize the benefits of health information exchange? How valuable are regional health information organizations in comparison with a more direct approach? What is the role of the carriers in delivering this service? The NWHIN is to exist for the public good, and thus shares many traits of the common law notion of 'common carriage' or 'public calling,' the modern term for which is network neutrality. Recent policy debates in Congress and resulting potential regulation have implications for key stakeholders within healthcare that use or provide services, and for those who exchange information. To date, there has been little policy debate or discussion about the implications of a neutral NWHIN. This paper frames the discussion for future policy debate in healthcare by providing a brief education and summary of the modern version of common carriage, of the key stakeholder positions in healthcare, and of the potential implications of the network neutrality debate within healthcare.

The AMIA biomedical informatics (BMI) core competencies have been designed to support and guide graduate education in BMI, the core scientific discipline underlying the breadth of the field's research, practice, and education. The core definition of BMI adopted by AMIA specifies that BMI is 'the interdisciplinary field that studies and pursues the effective uses of biomedical data, information, and knowledge for scientific inquiry, problem solving and decision making, motivated by efforts to improve human health.' Application areas range from bioinformatics to clinical and public health informatics and span the spectrum from the molecular to population levels of health and biomedicine. The shared core informatics competencies of BMI draw on the practical experience of many specific informatics sub-disciplines. The AMIA BMI analysis highlights the central shared set of competencies that should guide curriculum design and that graduate students should be expected to master.

In the midst of a US $30 billion USD investment in the Nationwide Health Information Network (NwHIN) and electronic health records systems, a significant change in the architecture of the NwHIN is taking place. Prior to 2010, the focus of information exchange in the NwHIN was the Regional Health Information Organization (RHIO). Since 2010, the Office of the National Coordinator (ONC) has been sponsoring policies that promote an internet-like architecture that encourages point to-point information exchange and private health information exchange networks. The net effect of these activities is to undercut the limited business model for RHIOs, decreasing the likelihood of their success, while making the NwHIN dependent on nascent technologies for community level functions such as record locator services. These changes may impact the health of patients and communities. Independent, scientifically focused debate is needed on the wisdom of ONC's proposed changes in its strategy for the NwHIN.

BACKGROUND: There is growing interest in the use of technology to enhance the tracking and quality of clinical information available for patients in disaster settings. This paper describes the design and evaluation of the Wireless Internet Information System for Medical Response in Disasters (WIISARD). MATERIALS AND METHODS: WIISARD combined advanced networking technology with electronic triage tags that reported victims' position and recorded medical information, with wireless pulse-oximeters that monitored patient vital signs, and a wireless electronic medical record (EMR) for disaster care. The EMR system included WiFi handheld devices with barcode scanners (used by front-line responders) and computer tablets with role-tailored software (used by managers of the triage, treatment, transport and medical communications teams). An additional software system provided situational awareness for the incident commander. The WIISARD system was evaluated in a large-scale simulation exercise designed for training first responders. A randomized trial was overlaid on this exercise with 100 simulated victims, 50 in a control pathway (paper-based), and 50 in completely electronic WIISARD pathway. All patients in the electronic pathway were cared for within the WIISARD system without paper-based workarounds. RESULTS: WIISARD reduced the rate of the missing and/or duplicated patient identifiers (0% vs 47%, p<0.001). The total time of the field was nearly identical (38:20 vs 38:23, IQR 26:53-1:05:32 vs 18:55-57:22). CONCLUSION: Overall, the results of WIISARD show that wireless EMR systems for care of the victims of disasters would be complex to develop but potentially feasible to build and deploy, and likely to improve the quality of information available for the delivery of care during disasters.

Health care providers are legally obliged to report cases of specified diseases to public health authorities, but existing manual, provider-initiated reporting systems generally result in incomplete, error-prone, and tardy information flow. Automated laboratory-based reports are more likely accurate and timely, but lack clinical information and treatment details. Here, we describe the Electronic Support for Public Health (ESP) application, a robust, automated, secure, portable public health detection and messaging system for cases of notifiable diseases. The ESP application applies disease specific logic to any complete source of electronic medical data in a fully automated process, and supports an optional case management workflow system for case notification control. All relevant clinical, laboratory and demographic details are securely transferred to the local health authority as an HL7 message. The ESP application has operated continuously in production mode since January 2007, applying rigorously validated case identification logic to ambulatory EMR data from more than 600,000 patients. Source code for this highly interoperable application is freely available under an approved open-source license at http://esphealth.org.

In a recent letter to the editor,1 Dr. Peter Szolovits called for “distinct names for distinct ideas” when discussing grid. His suggestion for better “precision of language” arose from listening to numerous talks at the 2006 American Medical Informatics Association (AMIA) Symposium “where speakers describe grids that have little in common.”1 At this early stage in the maturation of grid, as with any emerging technology, lack of clarity is natural and, in fact, stimulates an ever-improving “precision of thought.”1 Specifically, this ambiguity has already begun to generate discussion, as demonstrated by Dr. Szolovits' asking “What Is a Grid?” Most would agree that ambiguity is part of the normal lifecycle of adoption of new ideas, concepts, and technologies. Our reading and interpretation of the 2006 Gartner Hype Curve2 illustrates this point quite graphically; it articulates the ambiguity with regard to “grid computing” by placing it midway between the “peak of inflated expectations” and the “trough of disillusionment,” on its way towards the “slope of enlightenment” and within 2–5 years of mainstream adoption.2 What could be more ambiguous than living somewhere in-between “expectations” and “disillusionment?” Another aspect of ambiguity is the role of perspective. The spectrum of ambiguity ranges from the least (for those actively working with grid) to the most (for those peripherally involved or newly learning about the field). To borrow from a well-known metaphor to help explain the ambiguity around the grid-focused presentations at the 2006 AMIA Symposium, the speakers unintentionally actualized the American poet John Godfrey Saxe's poem, “The Blind Men and the Elephant.”3 In the poem, each blind man described the elephant differently from his own perspective—the elephant's side felt like a wall, his tusk like a spear, his trunk like a snake, his knee like a tree, his ear like a fan and his tail like a rope. The AMIA speakers presented different and important views on the very large domain of grid, resulting in, unavoidably, a somewhat unclear overall picture. In time, with further discussion and education, this picture can, and should, become much clearer. In fact, at this year's Symposium a tutorial was held on the topic “HealthGrid” for those interested in the area. Grid computing has unambiguously functioned as a distributed and cost-effective way to boost computational power to solve large-scale mathematical and data-bound problems for the physics and the bioinformatics communities, and continues to excel in these areas today. However, grid was, from the beginning, a more mature and complex concept that addressed a specific problem: “The real and specific problem that underlies the grid concept is coordinated resource sharing and problem solving in dynamic, multi-institutional virtual organizations.”4 As envisioned in the initial concept, grid has matured far beyond computational augmentation to include inter-organizational, distributed data management and exchange, industrial-strength security, distributed services and workload orchestration. To be more precise, grid represents a robust framework (based upon the principle of service-oriented architecture (SOA)) for performing distributed computing tasks on the scale of the Internet which can enable “Service-Oriented Science [and Medicine].”5 To this end, grid may be defined as four distinct technology layers (resources, middleware, services, and applications) and one implicit (social) layer.4 This infrastructure assumes a lower layer of grid-accessible resources (widely dispersed storage, computational power (CPU cycles), devices, sensors, data and relational databases amongst different enterprises), a middle layer of grid-specific middleware (as provided by an open-source project known as the Globus Alliance (http://www.globus.org), and an upper layer of grid-enabled services and applications distributed across the Internet (the Internet itself enabling connectivity) that are useful to combine into arbitrary virtual combinations to achieve some common goal. The first two technology layers (resources and middleware), provide the core, domain-independent infrastructure. The grid-enabled upper layers represent the algorithmic stack which encapsulates the business logic required to support the shared processes (services and applications) within a specific virtual organization or “community of practice,” (e.g., the healthcare domain), and defines the domain-dependent grid-based services (data, computational, visualization, etc.). It is notable that these services harmonize with Web Service standards. The social layer, then, encompasses the governance, coordination and policy activities required to assure an efficient and effective community of practice while the technology layers provide a robust framework to execute such shared policies and goals. In other words, the community of practice could be thought of as the social manifestation of the grid. Fundamentally, grid computing interconnects people, organizations, processes, application silos and data silos in completely innovative ways, and has the potential to significantly augment the efficiency and effectiveness of virtual organizations (flexible, secure, coordinated resource sharing among dynamic collections of individuals, institutions, and resources). In order to lessen the ambiguity one degree more, a brief comparison of grid computing with another distributed architecture should add both historical and technical depth to this conversation. The most similar architectural pattern relative to grid, at least on the surface, is the Common Object Request Broker Architecture, also known as CORBA. Both approaches attempt to integrate data, resource sharing, communications, conventions and organizational exchanges; nevertheless, CORBA contrasts more readily with web services than grid computing per se. First and foremost, grid provides a secure overarching framework in which web services may interoperate; in other words, web services are an integral part of the grid fabric. Secondarily, grid supplies a rich set of additional computational, federated data and collaborative services above and beyond existing distributed systems. The CORBA may interoperate and coexist with web services but by no means are web services endemic to its architecture. Furthermore, while CORBA is distinctly tightly coupled, object-oriented and stateful, web services are loosely coupled, utilize a message exchange model and are stateless. These differences give web services an inherent flexibility and simplicity unavailable in CORBA implementations. Grid, therefore, is unique from other methods of distributed computing by its ability to coalesce the technology layers with the social layer through its commercial-strength security model (X.509 certificates), thus enabling the dynamic creation of multiple virtual organizations and providing unprecedented technical and social agility. Grid and CORBA represent different models for building distributed systems and are complementary rather than competing technologies and architectures. So, while it may appear that the 2006 AMIA Symposium presenters spoke of different definitions of the same grid, demonstrating a lack of precision of thought, we would argue that they are, in fact, accurate definitions of different aspects and layers of the same grid framework, as described above. Thus, grid (as a framework), in fact: provides a technical infrastructure that enables “a community of common interests,”1 requires a governance model that supports “a social and funding infrastructure to encourage data sharing,”1 delivers a comprehensive set of tools and consequently “a technical approach to what we used to call federated databases,”1 provides a flexible, scaleable and secure technical framework that facilitates “standardization and ontology construction for specific fields, and of course, offers distributed computing far beyond its “original meaning.”1 In this broader context, grid computing has the potential to offer a vast set of production-tested software tools designed specifically for building virtual organizations beyond the limits and silos of institutional boundaries. To further extend the metaphor, one could ask, “what is an interconnected national healthcare system if not a virtual organization?”4 At the Centers for Disease Control & Prevention's (CDC) National Center for Public Health Informatics (NCPHI), we are attempting to answer this and other questions by engaging in active preliminary research to discover the potential role that grid-based technologies may provide in interconnecting and augmenting the virtual organizations known as public health and clinical care. Clearly the task is challenging. Fortunately, there are many worldwide initiatives from which to learn. In contrast to the US, Europe and other countries around the world have made significant advancements in the development and utilization of grid computing technologies across health and other domain boundaries. In addition, the five-year-old International HealthGrid organization (http://community.healthgrid.org) and the newly formed HealthGrid.US Alliance (http://www.healthgrid.us) are leading the way towards exploring and discovering the capability of grid computing for healthcare. Altogether, grid is clearly a very broad domain, and the ambiguity Dr. Szolovits addressed is a challenge to be expected and addressed. To minimize ambiguity and accelerate its introduction into the healthcare domain, we propose that grid be conceived of as an overall framework with layers of core technologies, processes and sociology. Education and collaboration are critical to quickly mature its place in the US biomedical informatics field. We look forward to continued research and discussion on this topic.

The emergence of personal mobile computing and ubiquitous wireless networks enables powerful field applications of video streaming, such as vision-enabled command centers for hazardous materials response. However, experience has repeatedly demonstrated both the fragility of the wireless networks and the insatiable demand for higher resolution and more video streams. In the wild, even the best streaming video mechanisms result in low-resolution, low-frame-rate video, in part because the motion of first-person mobile video (e.g., via a head-mounted camera) decimates temporal (inter-frame) compression. We introduce a visualization technique for displaying low-bit-rate first-person video that maintains the benefits of high resolution, while minimizing the problems typically associated with low frame rates. This technique has the unexpected benefit of eliminating the "Blaire Witch Project" effect-the nausea-inducing jumpiness typical of first-person video. We explore the features and benefits of the technique through both a field study involving hazardous waste disposal and a lab study of side-by-side comparisons with alternate methods. The technique was praised as a possible command center tool, and some of the participants in the lab study preferred our low-bitrate encoding technique to the full-frame, high resolution video that was used as a control

OBJECTIVE: The aim of this study was to determine whether an automated e-mail messaging system that sent individually timed educational messages (ITEMs) increased the effectiveness of an Internet smoking cessation intervention. DESIGN: Using two consecutive series of participants, the authors compared two Web-based self-help style smoking cessation interventions: a single-point-in-time educational intervention and an enhanced intervention that also sent ITEMs timed to participants' quit efforts. Outcomes were compared in 199 participants receiving the one-time intervention and 286 receiving ITEMs. MEASUREMENTS: Demographic factors, number of cigarettes smoked, nicotine addiction, depressive symptoms, and confidence in ability to quit were measured at entry. Twenty-four-hour quit attempts and seven-day point-prevalence of abstinence (nonrespondents assumed to smoke) were measured 30 days after each subject's self-selected quit date. RESULTS: The one-time and ITEMs groups differed in some demographics and some relapse risk factors but not in factors associated with 30-day quit rates. ITEMs appeared to increase the rate at which individuals set quit dates (97% vs. 91%, p = 0.005) and, among the respondents to follow-up questionnaires (n = 145), the rate of reported 24-hour quit efforts (83% vs. 54%, p = 0.001). The 30-day intent-to-treat quit rates were higher in the ITEMs group: 7.5% vs. 13.6%, p = 0.035. In multivariate analyses controlling for differences between groups, receiving ITEMs was associated with an increase in the odds ratio for quitting of 2.6 (95% confidence interval = 1.3-5.3). CONCLUSION: ITEMs sent on strategic days in smokers' quit efforts enhanced early success with smoking cessation relative to a single-point-in-time Web intervention. The effect appears to be mediated by ITEMs' causing smokers to plan and undertake quit efforts more frequently.

Relatively little is known about how to use the Internet to promote health behavioral change. This article describes a multiple-contact Internet smoking cessation program with an 8-week web-based course, online tools for self-monitoring of behaviors, and computer-tailored e-mail messages timed to enrollees' quit efforts. In a pilot study in 49 smokers, we found that enrollees returned to the website a median of 2 times and completed an average of 2 of 8 educational modules. In follow-up, respondents (n = 26) rated e-mail and web components of the intervention as equally valuable (5.9 vs. 5.5 of 10, p = 0.44). While site had potentially important effects on smoking behaviors (34% of enrollees either quit smoking or had a 50% reduction in cigarette use), we were not able hold the interest of the majority of enrollees over the intervention period. Problems with the design of the site are discussed.

In the past, both health professionals and the public have assumed that the commitment of health professionals to obtaining the best possible outcomes for their patients would guarantee safe and effective health care. Through the work of the Institute of Medicine (IOM) and other leading organizations, however, it has become obvious both to researchers and to society as a whole, that commitment to the betterment of health does not guarantee safe health care practices. As highlighted in “To Err is Human,” 1 a shockingly large number of errors occur, despite the best intentions of providers, and many errors have fatal results. Although there is some debate about the exact number of people who die each year as a result of errors, 2 by and large, deaths do not occur because of individual negligence—rather they result from flawed systems in which people fail to deliver the quality of a care to which they aspire. How can health professionals perform as they both wish and must? A second IOM report released in 20001, “Crossing the Quality Chasm,” 3 identified the critical role that information technology will play in engineering health care systems that produce care that is “safe, effective, patient-centered, timely, efficient, and equitable” (p. 164). As is well recognized by most readers of JAMIA , the needed information technologies are not yet readily available or easily implementable. Improving safety through information technology is an inherently cross-disciplinary effort that requires close collaboration among computer scientists, engineers, and social scientists for a successful endeavor. Where could such a diverse group of people come together to present work and advance this nascent field? This special issue presents papers, posters, and panel discussions from the 2001 AMIA Fall Symposium addressing the issues of patient safety. Three additional papers provide guidance for integrating patient safety concepts into continuing medical education, undergraduate medical education, and nursing informatics curricula. This meeting was one of the first major scientific meetings devoted to medical computing after the new funding initiatives to develop specific research support for the safety area. Papers were part of the first-ever meeting track devoted to the issue of patient safety. These results represent the new wine—first fruits of researchers coming to the safety arena as a result of national initiatives—as well as the continuing contributions of stalwarts in the field. As is typical of Fall Symposium meetings, the papers in this supplement present conceptual models, technological developments that are evolving and still being evaluated, and a snapshot of ongoing research related to computer safety systems. The breadth of work highlighted in this issue reveals that the pursuit of safer practices requires a multifaceted approach. Work presented in the track included work on clinical systems, human factors, knowledge representation, and protecting confidentiality. As of result of the 9-11 terrorist attacks, we have expanded the focus of safety work presented in this issue to include work related to detection of bioterrorist incidents and other issues relevant to homeland security. It is our hope that this issue provides clinicians and educators with a glimpse into the future of the role of technology in improvement of safety and quality of care. As efforts continue to expand funding for use of informatics to enhance safety, we hope that this issue also provides important feedback to policy makers about the types of programs ongoing in the research community and the needs of that community for research support.

The participants in this panel described major federal initiatives aimed at improving patient safety. Summaries of the panelists' remarks follow in the order of presentation at the Symposium. Medication errors have multiple causes including: poor communication, name confusion, abbreviations, poor techniques, knowledge deficit, inexperience, and confirmation bias. In addition to the patient, hospitals, practitioners, manufacturers, and governments can also be victims of medication errors. The FDA has collected more than 20,000 medication errors since 1993 and data reflect a 10% mortality rate. In order to combat medication errors, the FDA has initiatives in the area of pre-marketing, post-marketing, risk management, and research ( Table 1 ). For more information see < http://www.fda.gov/ >. ... The FDA is involved in numerous risk communication activities including publishing in professional journals such as AHSP, JAMA, and NEJM and through the consumer-oriented FDA Today. Other risk minimization activities include improved labeling, letters to physicians, education, restricted access to a particular drug, and withdrawal from the market. In addition, the FDA collaborates with other organizations such as the AMA, AHSP, USP, and Institute for Safe Medication practices.

OBJECTIVE: While the use of multimedia methods in medical education and decision support can facilitate learning, it also has certain hazards. One potential hazard is the inadvertent triggering of racial and gender bias by the appearance of actors or patients in presentations. The authors hypothesized that race and gender affect preferences. To explore this issue they studied the effects of actors' race and gender on preference ratings for health states that include symptoms of schizophrenia. DESIGN: A convenience sample of patients with schizophrenia, family members of patients, and health professionals was used. Participants were randomly assigned to rate two health states, one portrayed by either a man of mixed race (Hispanic-black) or a white man and the second portrayed by either a white woman or a white man. MEASUREMENTS: Visual analog scale (VAS) and standard gamble ratings of health state preferences for health states that include symptoms of mild and moderate schizophrenia. RESULTS: Studies of the effects of the race of the actor (n = 114) revealed that racial mismatch between the actor and the participant affected the participant's preferences for health states. Ratings were lower when racial groups differed (mean difference, 0.098 for visual analog scale ratings and 0.053 lower in standard gamble, P = 0.006 for interactions between the race of the subject and the actor). In studies of the effects of a female actress on ratings (n = 117), we found no evidence of a corresponding interaction between the gender of the actor and the study participant. Rather, an interaction between actor's gender and method of assessment was observed. Standard gamble ratings (difference between means, 0.151), but not visual analog scale ratings (difference, 0.005), were markedly higher when the state was portrayed by the actress (P = 0.003 for interactions between actor's gender and method of preference assessment). Differential effects on standard gamble ratings suggest that an actor's gender may influence the willingness of viewers to gamble to gain health benefits (or risk attitude). CONCLUSIONS: Educators and researchers considering the use of multimedia methods for decision support need to be aware of the potential for the race and gender of patients or actors to influence preferences for health states and thus, potentially, medical decisions.

OBJECTIVES: Describe and evaluate an Internet-based approach to patient decision support using mathematical models that predict the probability of successful treatment on the basis of meta-analytic summaries of the mean and standard deviation of symptom response. DESIGN: An Internet-based decision support tool was developed to help patients with benign prostatic hypertrophy (BPH) determine whether they wanted to use alpha blockers. The Internet site incorporates a meta-analytic model of the results of randomized trials of the alpha blocker terazosin. The site describes alternative treatments for BPH and potential adverse effects of alpha blockers. The site then measures patients' current symptoms and desired level of symptom reduction. In response, the site computes and displays the probability of a patient's achieving his objective by means of terazosin or placebo treatment. SETTING: Self-identified BPH patients accessing the site over the Internet. MAIN OUTCOME MEASURES: Patients' perceptions of the usefulness of information. RESULTS: Over a three-month period, 191 patients who were over 50 years of age and who reported that they have BPH used the decision support tool. Respondents had a mean American Urological Association (AUA) score of 18.8 and a desired drop in symptoms of 10.1 AUA points. Patients had a 40 percent chance of achieving treatment goals with terazosin and a 20 percent chance with placebo. Patients found the information useful (93 percent), and most (71 percent) believed this type of information should be discussed before prescribing medications. CONCLUSIONS: Interactive meta-analytic summary models of the effects of pharmacologic treatments can help patients determine whether a treatment offers sufficient benefits to offset its risks.

OBJECTIVE: Develop the methodological foundation for interactive use of Markov process decision models by patients and physicians at the bedside. DESIGN: Monte Carlo simulation studies of a decision model comparing two treatments for benign prostatic hypertrophy: watchful waiting (WW) and transurethral prostatectomy (TUR). MEASUREMENTS: The 95% confidence interval (CI) for the mean of the Markov model; the correlation of a linear approximation with the full Markov model; the predictive performance of the approximation; the information index of specific utilities in the model. RESULTS: The 95% CI for the gain in utility with initial TUR was -1.4 to 19.0 quality-adjusted life-months. A multivariate linear model had an excellent fit to the predictions of the Markov model (R2 = 0.966). In an independent data set, the linear model also had a high correlation with the full Markov model (R2 = 0.967); its predictions were unbiased (p = 0.597, paired t-test); and, in 96.4% of simulated cases, its treatment recommendation was the same. CONCLUSION: Using the linear model, it was possible to efficiently compute which health state had the largest contribution to the variance of the decision model. This is the most informative utility value to elicit next. The most informative utility at any point in a sequence changed depending on utilities previously entered into the model. A linear model can be used to approximate the predictions of a Markov process decision model.

OBJECTIVE: To assess the practicality of an automated computer interview as a method to assess preferences for use in decision making. To assess preferences for outcomes of deep vein thrombosis (DVT) and its treatment. STUDY DESIGN: A multimedia program was developed to train subjects in the use of different preference assessment methods, presented descriptions of mild post-thrombotic syndrome (PTS), severe PTS and stroke and elicited subject preferences for these health states. This instrument was used to measure preferences in 30 community volunteers and 30 internal medicine physicians. We then assessed the validity of subject responses and calculated the number of quality-adjusted life years (QALYs) for each individual for each alternative. RESULTS: All subjects completed the computerized survey instrument without assistance. Subjects generally responded positively to the program, with volunteers and physicians reporting similar preferences. Approximately 26.5% of volunteers and physicians had preferences that would be consistent with the use of thrombolysis. Individualization of therapy would lead to the most QALYs. CONCLUSIONS: Utilization of computerized survey instruments to elicit patient preferences appears to be a practical and valid approach to individualize therapy. Application of this method suggests that there may be many patients with DVT for whom treatment with a thrombolytic drug would be optimal.

OBJECTIVE: The World Wide Web (WWW) is a new communications medium that permits investigators to contact patients in nonmedical settings and study the effects of disease on quality of life through self-administered questionnaires. However, little is known about the feasibility and, what is more important, the validity of this approach. An on-line survey for patients with ulcerative colitis (UC) and patients whose UC had been treated with surgical procedures was developed. To understand how patients on the WWW might differ from those in practice and the potential biases in conducting epidemiological research in volunteers recruited on the Internet, post-surgery patients who responded to the WWW survey were compared with those in a surgical practice. SETTING: The Internet and private practice surgical clinic. MAIN OUTCOMES: Scores from the Short form 36 (SF-36) Health Assessment Questionnaire and the Self-Administered Inflammatory Bowel Disease Questionnaire (IBDQ). RESULTS: Over a 5-month period, 53 post-surgery patients enrolled in the Internet study; 47 patients from a surgical clinic completed the same computer-based questionnaire. Surgically treated patients on the WWW were younger than their clinic counterparts (median age category 35-44 years vs. 45-54 years, p = 0.01) but more ill with a lower summary IBDQ score (168 vs. 186, p = 0.019) and lower health status across almost all dimensions of the SF-36 (p = 0.016). CONCLUSIONS: It is feasible to conduct epidemiological research on the effects of UC on quality of life on the Web; however, systematic differences in disease activity between volunteer patients on the WWW and "in the clinic" may limit the applicability of results.