The IMIA History Working Group: Inception through the IMIA History Taskforce, and Major Events Leading Up to the 50th Anniversary of IMIA.

Background: The 50th Anniversary of IMIA will be celebrated in 2017 at the World Congress of Medical Informatics in China. This takes place 50 years after the International Federation of Information Processing (IFIP) Societies approved the formation of a new Technical Committee (TC) 4 on Medical Information Processing, which was the predecessor of IMIA, under the leadership of Dr. Francois Grémy. The IMIA History Working Group (WG) was approved in 2014 to document and write about the history of the field and its organizations. Objectives: The goals of this paper are to describe how the IMIA History WG arose and developed, including its meetings and projects, leading to the forthcoming 50th Anniversary of IMIA. Methods: We give a chronology of major developments leading up to the current work of the IMIA History WG and how it has stimulated writing on the international history of biomedical and health informatics, sponsoring the systematic compilation and writing of articles and stories from pioneers and leaders in the field, and the organization of workshops and panels over the past six years, leading towards the publication of the contributed volume on the 50th IMIA Anniversary History as an eBook by IOS Press. Conclusions: This article leads up to the IMIA History eBook which will contain original autobiographical retrospectives by pioneers and leaders in the field, together with professional organizational histories of the national and regional societies and working groups of IMIA, with commentary on the main themes and topics which have evolved as scientific and clinical practices have changed under the influence of new insights, technologies, and the changing socio-economic, cultural and professional circumstances around the globe over the past 50 years.

Collaboration using Internet for the development of case-based teaching files: report of the Computer and Instrumentation Council Internet Focus Group.

UNLABELLED: The Internet and particularly the World-Wide-Web is becoming a useful tool for the nuclear medicine community. METHODS: The Computer and Instrumentation Council of the Society of Nuclear Medicine convened an Internet Focus group to discuss collaboration using the Internet. The prototype application considered was development of case-based teaching files using the World-Wide-Web. Teaching file cases (clinical history, images, description of findings and discussion) on World-Wide-Web servers at different institutions are integrated using the Internet. The user can navigate from case to case using point-and-click hypertext linking. RESULTS: The initial experience with collaboration has been encouraging. An etiquette to help foster collaboration has been proposed. Development of quality control mechanisms and introduction of peer review were identified as issues needing further work. CONCLUSION: The World-Wide-Web offers great potential for new forms of collaboration. There is, however, a need to learn how to make best use of this new resource.

Assessment of diagnostic technologies. Methodology for unbiased estimation from samples of selectively verified patients.

Concern with the efficacy of diagnostic technologies has stimulated numerous studies aimed at quantifying the discriminatory properties of various tests and procedures. These have focused principally on estimations of the result conditional probabilities, given disease status, eg, the sensitivity and specificity or the ROC curve. A source of bias in estimating these probabilities that is often unavoidable is created by the existence of a nonrandom selection mechanism for determining which patients initially tested will receive definitive verification of disease status. Correction for verification bias requires frequency data on the test results and any symptoms or other factors that influence selection for verification, both in the verified sample and in the source sample of patients tested.

Strategic planning activities of the American Medical Informatics Association.

The American Medical Informatics Association (AMIA) has begun the process of long-range strategic plan development. The AMIA Board of Directors established an Ad Hoc Strategic Planning Task Force, with the goal of initiating such planning in November 1992. In January 1993, the Task Force convened a group of AMIA members in order to develop an initial set of goals and objectives. The group consisted of past and present AMIA Board members, AMIA Committee chairpersons, representative AMIA Working Group chairpersons, the AMIA Executive Director and members of the AMIA office staff, and a number of AMIA members-at-large. The group created a draft strategic plan, which was refined by the Task Force after circulation among two focus groups and through a mailing to the AMIA membership. This report of the AMIA strategic planning process is intended to create a historical record and to stimulate further discussion of a working plan that will evolve over time. AMIA will continue the strategic planning process through its Ad Hoc Strategic Planning Committee as it begins to implement aspects of the strategic plan over the next several years.

Audacious goals for health and biomedical informatics in the new millennium.

The 1998 Scientific Symposium of the American College of Medical Informatics (ACMI) was devoted to developing visions for the future of health care and biomedicine and a strategic agenda for health and biomedical informatics in support of those visions. This symposium focus was prompted by the many major changes currently underway in health care delivery, education, and research, as well as in our health and biomedical enterprises, and by the constantly increasing role of information technology in both shaping and enabling these changes. The three audacious goals developed for 2008 are a virtual health care databank, a national health care knowledge base, and a personal clinical health record.

Experiments in concept modeling for radiographic image reports.

OBJECTIVE: Development of methods for building concept models to support structured data entry and image retrieval in chest radiography. DESIGN: An organizing model for chest-radiographic reporting was built by analyzing manually a set of natural-language chest-radiograph reports. During model building, clinician-informaticians judged alternative conceptual structures according to four criteria: content of clinically relevant detail, provision for semantic constraints, provision for canonical forms, and simplicity. The organizing model was applied in representing three sample reports in their entirety. To explore the potential for automatic model discovery, the representation of one sample report was compared with the noun phrases derived from the same report by the CLARIT natural-language processing system. RESULTS: The organizing model for chest-radiographic reporting consists of 62 concept types and 17 relations, arranged in an inheritance network. The broadest types in the model include finding, anatomic locus, procedure, attribute, and status. Diagnoses are modeled as a subtype of finding. Representing three sample reports in their entirety added 79 narrower concept types. Some CLARIT noun phrases suggested valid associations among subtypes of finding, status, and anatomic locus. CONCLUSIONS: A manual modeling process utilizing explicitly stated criteria for making modeling decisions produced an organizing model that showed consistency in early testing. A combination of top-down and bottom-up modeling was required. Natural-language processing may inform model building, but algorithms that would replace manual modeling were not discovered. Further progress in modeling will require methods for objective model evaluation and tools for formalizing the model-building process.

Clinical decision support systems for the practice of evidence-based medicine.

BACKGROUND: The use of clinical decision support systems to facilitate the practice of evidence-based medicine promises to substantially improve health care quality. OBJECTIVE: To describe, on the basis of the proceedings of the Evidence and Decision Support track at the 2000 AMIA Spring Symposium, the research and policy challenges for capturing research and practice-based evidence in machine-interpretable repositories, and to present recommendations for accelerating the development and adoption of clinical decision support systems for evidence-based medicine. RESULTS: The recommendations fall into five broad areas--capture literature-based and practice-based evidence in machine--interpretable knowledge bases; develop maintainable technical and methodological foundations for computer-based decision support; evaluate the clinical effects and costs of clinical decision support systems and the ways clinical decision support systems affect and are affected by professional and organizational practices; identify and disseminate best practices for work flow-sensitive implementations of clinical decision support systems; and establish public policies that provide incentives for implementing clinical decision support systems to improve health care quality. CONCLUSIONS: Although the promise of clinical decision support system-facilitated evidence-based medicine is strong, substantial work remains to be done to realize the potential benefits.

The guideline interchange format: a model for representing guidelines.

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.

Computer-aided diagnostic strategy selection.

Determination of the optimal diagnostic work-up strategy for the patient is becoming a major concern for the practicing physician. Overlap of the indications for various diagnostic procedures, differences in their invasiveness or risk, and high costs have made physicians aware of the need to consider the choice of procedure carefully, as well as its relation to management actions available. In this article, the author discusses research approaches that aim toward development of formal decision analytic methods to allow the physician to determine optimal strategy; clinical algorithms or rules as guides to physician decisions; improved measures for characterizing the performance of diagnostic tests; educational tools for increasing the familiarity of physicians with the concepts underlying these measures and analytic procedures; and computer-based aids for facilitating the employment of these resources in actual clinical practice.

Radiology systems architecture.

This article focuses on the software requirements for enterprise integration in radiology. The needs of a future radiology systems architecture are examined, both at a concrete functional level and at an abstract system-properties level. A component-based approach to software development is described and is validated in the context of each of the abstract system requirements for future radiology computing environments.

The era of health care reform and the information superhighway. Implications for radiology.

Radiology is undergoing dramatic change, along with the rest of the health care system, in its mode of organization, financing, and delivery. Information technology is becoming central to health care delivery and will enable a higher degree of integration of imaging practice with the rest of the health care system. Radiology will need to address the requirements for achieving this integration to continue to be successful in the future.

Interactive microcomputer-based graphical tools for physician decision support. Aids to test selection and interpretation and use of Bayes' theorem.

Three microcomputer programs have been developed to give the practicing physician graphic aid in assessing the usefulness of obtaining a diagnostic test, to assist in interpreting the result after a test has been obtained, and to facilitate exploration of, and familiarity with, the use of Bayes' theorem in clinical practice. Each program uses a different representation for depicting the transformation of pretest to post-test probability produced by the various results of the test. Choice of representation depends on whether multiple tests or results are evaluated concurrently, on the probability range of interest, and on which model the user finds most intuitively appealing. A library of data on sensitivities and specificities of various tests and procedures can be used to select tests of interest from a menu, or data may be entered by the user.

Improving clinical guidelines with logic and decision-table techniques: application to hepatitis immunization recommendations.

Clinical practice guidelines can be clarified, verified, and simplified by the use of logical analysis and the application of decision-table techniques. This methodology is applied to a CDC guideline for the prevention of perinatal transmission of hepatitis B by immunization. Relevant clinical variables are identified and possible values for each variable are defined. An exhaustive enumeration of value combinations is generated. Logically impossible combinations are identified and eliminated. The guideline's recommendations are next translated into a set of rules and compared with the remaining value combinations. Variable combinations that are not covered by guideline recommendations represent incomplete guideline specifications. Inconsistency can be identified by finding identical condition sets in two or more rules. The procedure demonstrates that the hepatitis guideline is incomplete. Logical analysis can improve the quality of clinical practice guidelines by assuring comprehensiveness and consistency.

Construction of receiver operating characteristic curves when disease verification is subject to selection bias.

The estimates of ROC curves, which are frequently used in the assessment of diagnostic tests, may be biased if the sample is restricted to subjects whose disease status has been definitively verified. A method to provide an unbiased estimate of the ROC curve under these sample conditions is proposed. The new method requires information on the probability distribution of test results in the population from which the verified sample is drawn. It is illustrated using data from a study of computed tomography for fever of uncertain origin.

Diagnosis of cholestasis: an analytic view.

The authors analyzed two invasive procedures used to visualize the biliary tree, endoscopic retrograde cholangiopancreatography (ERCP) and percutaneous transhepatic cholangiography (PTC), and also explored the utility of preinvasive workups for patients with suspected cholestasis. For this analysis they used published ranges for success, fatality, complications, diagnostic accuracies of the procedures, and prognostic information about the underlying diseases. The choice between ERCP and PTC was found to be a "close call," but ERCP was generally favored as the first-choice procedure. The results suggest that noninvasive imaging does not help decide between ERCP and PTC. Although noninvasive imaging may identify those patients with common duct dilation, the higher success rate with PTC in these patients is offset by a slightly higher mortality rate. Consequently, the choice between ERCP and PTC remains close even if ultrasound has shown that biliary ducts are dilated. Furthermore, it is shown that these noninvasive tests are most useful when they can conclusively determine the presence or absence of biliary obstruction. For many patients, noninvasive imaging will not obviate the need for invasive tests.

Patient-oriented performance measures of diagnostic tests. 3. U-Factor.

When a clinician is faced with the problem of deciding whether to order a specific diagnostic test, the ideal information would be the utility of the "perform test" branch of the decision tree versus the utility of the "do not perform test" branch. This difference in utility is termed the Expected Utility of the Test ( EUT ). We propose a new performance measure of a test called the U-Factor (UF) which is related to EUT by a simple formula. UF depends on the prior probability of disease and on the decision thresholds at which one would be indifferent between any two immediately subsequent management options, and can be presented as a two-dimensional nomogram. UF is useful as a computational aid in a formal decision analysis, and may be useful as an informal measure of the value of a diagnostic test when a formal analysis is not feasible.

Patient-oriented performance measures of diagnostic tests. 1. Tools for prospective evaluation of test order decisions.

Prospective assessment of a test for possible use in evaluation of a patient ideally should be based on the test's ability to affect subsequent patient management beneficially, in relation to both the costs of the test itself and the costs of misclassification of disease status. This requires specification of the costs and benefits of subsequent actions and the cost of the test, estimation of the probability of disease, knowledge of the discriminatory properties of the test, and formal decision analysis. Often, however, the physician has less complete information with which to make a test order decision. A spectrum of performance measures exists for characterising a diagnostic test, ranging from measures that are largely patient-independent to measures that are highly patient-dependent. Two measures from the patient-dependent part of the spectrum, Assignment Potential and Assignment Strength, can be useful in decision making when formal decision analysis is not feasible. A third measure, the U-Factor, is the product of the other two, and is a computational tool that facilitates formal decision analysis.

Data assessing the usefulness of screening obstetrical ultrasonography for detecting fetal and placental abnormalities in uncomplicated pregnancy: effects of screening a low-risk population.

To investigate the usefulness of screening in low-risk populations, the authors evaluated the yield of ultrasonography for detecting abnormalities in 678 clinically uncomplicated pregnancies. The yield of ultrasonography in high-risk women who were referred for amniocentesis was remarkably similar to the yield in other women. All four diagnoses of twins were correct, but overall only six of 12 initial ultrasound diagnoses of fetal demise or fetal anomalies were confirmed at delivery. Of the eight major fetal anomalies present at delivery, including two cases of Down's syndrome, three had been detected by ultrasonography; none of the nine minor anomalies had been detected, usually because they were too small or might be detectable only at a later gestational age. Although ultrasonography may have a nearly perfect predictive value for certain anomalies, on average, in this study, positive ultrasonography increased the probability of an adverse outcome of pregnancy from 5.3% to 36%, while a normal ultrasound examination decreased the probability to 4.4%. These data, which emphasize the implications of screening a low-risk population, suggest that recommendations regarding routine screening obstetrical ultrasonography should await sufficiently large controlled trials demonstrating consistent clinical benefit, in terms of reassurance or of providing a baseline for future comparison or in terms of improved outcome at a reasonable cost.

Patient-oriented performance measures of diagnostic tests. 2. Assignment potential and assignment strength.

Assignment Potential (AP) is a performance measure of a diagnostic test, characterizing the chance that, as a consequence of performing the test, the probability of disease will exceed a decision threshold, thereby permitting a management action to be taken. Another performance measure, Assignment Strength (AS) characterizes the average extent to which a decision threshold will be exceeded when the post-test probability of disease does exceed the threshold. Both AP and AS are functions of prior probability of disease and decision threshold, and can be represented as two-dimensional contour maps indicating their behavior throughout the entire probability and threshold space. AP and AS can be determined for both discrete-valued tests and tests with continuous spectra of results. The contour map displays facilitate determination of the values of these measures at any prior probability and threshold, as well as visual sensitivity analysis for ranges of prior probability and/or threshold. AP and AS may be useful to the clinician in prospective evaluation of a diagnostic test in situations where formal decision analysis is not feasible.

Education in the workplace for the physician: clinical management states as an organizing framework.

Medical educators are interested in approaches to making selected relevant knowledge available in the context of problem-based care. This is of value both during the process of care and as a means of organizing information for offline self-study. Four trends in health information technology are relevant to achieving the goal and can be expected to play a growing role in the future. First, health care enterprises are developing approaches for access to information resources related to the care of a patient, including clinical data and images but also communication tools, referral and other logistic tools, decision support, and educational materials. Second, information for patients and methods for patient-doctor interaction and decision making are becoming available. Third, computer-based methods for representation of practice guidelines are being developed to support applications that can incorporate their logic. Finally, considering patients as being in particular "clinical management states" (or CMSs) for specific problems, approaches are being developed to use guidelines as a kind of "predictive" framework to enable development of interfaces for problem-based clinical encounters. The guidelines for a CMS can be used to identify the kinds of resources specifically needed for clinical encounters of that type. As the above trends converge to produce problem-specific environments, professional specialty organizations and continuing medical education course designers will need to focus energies on organizing and updating medical knowledge to make it available in CMS-specific contexts.