Critical reviews in medical informatics最新文献

A review of statistical methods in medical diagnosis is presented. Research has focused on three distinct tasks: classification of subjects into probable diagnostic categories on the basis of presenting clinical indicators (discriminant analysis), assessment of diagnostic test characteristics, and relation of diagnostic testing to subsequent patient management. Although many sophisticated models have been developed for discriminant analysis, recent empirical comparisons indicate that standard methods such as linear discrimination and logistic regression work very well. More research is needed to overcome practical difficulties that are not accommodated in the conventional assumptions. Research on the assessment of diagnostic tests has been oriented more toward selection biases and practical problems. There is a need to develop generalized models for the problem of differential diagnosis. The relation of testing to subsequent management of the patient is a topic that has only recently been explored. It represents an important task in the cost-effective management of health resources.

The development of computers has provided a potential tool to assist in the management of the information explosion in medicine. The field of medical diagnosis is intellectually challenging and has attracted the attention of computer scientists interested in building expert systems using artificial intelligence techniques. This paper reviews some of the problems of medical diagnosis and discusses examples of programs representing different approaches to solving these problems. The programs developed in our laboratory, INTERNIST-1/CADUCEUS, are discussed in some detail.

Medical education and the acquisition of professional credentials do not guarantee that medical knowledge (information that is taught, apart from the reality of practice, or gleaned from the literature) will be coupled rigorously to the decision-making process of everyday clinical practice. The limitations of the unaided human mind in a memory-based educational system must be forthrightly acknowledged by those who would be responsible for curriculum reform, so the need for new premises and new tools will be recognized and implemented. Because of their knowledge of the many variables that are unique to them, the patients themselves must be given a much more central role in the process of medical care and medical education. Weaknesses of specialization and credentialing in the present obsolete system are analyzed. The behavior of a well-defined system of education and medical care, and the function of the performers within it, are described. Causes of resistance to curricular reform founded on new premises and use of new tools, such as computers, are considered. New computer tools, as components of a problem-solving decision support system (knowledge couplers, knowledge networks, the coupler editor and documentation system, and the computerized patient record), are described. How these might be incorporated into a new type of medical education curriculum is presented. Finally, new goals, within the context of the new premises being implemented into a new system of education and medical care, are outlined.

The purpose of this paper is threefold: to review the status of Health Information Science (Health Informatics) as reported in the literature; to suggest a taxonomy to integrate the many varied concepts being discussed under the rubric of Health Information Science; and to propose a framework into which research can be classified and potential research hypotheses generated. It is suggested that such a framework is needed if generalizations are to be made to the end that the knowledge produced from research efforts may be cumulative with that from other studies. A framework for research is needed if the field is to become an established academic discipline; whether or not it takes on the mantel of a "science" is another matter.