G. Octo Barnett

The Unified Medical Language System: an informatics research collaboration.

In 1986, the National Library of Medicine (NLM) assembled a large multidisciplinary, multisite team to work on the Unified Medical Language System (UMLS), a collaborative research project aimed at reducing fundamental barriers to the application of computers to medicine. Beyond its tangible products, the UMLS Knowledge Sources, and its influence on the field of informatics, the UMLS project is an interesting case study in collaborative research and development. It illustrates the strengths and challenges of substantive collaboration among widely distributed research groups. Over the past decade, advances in computing and communications have minimized the technical difficulties associated with UMLS collaboration and also facilitated the development, dissemination, and use of the UMLS Knowledge Sources. The spread of the World Wide Web has increased the visibility of the information access problems caused by multiple vocabularies and many information sources which are the focus of UMLS work. The time is propitious for building on UMLS accomplishments and making more progress on the informatics research issues first highlighted by the UMLS project more than 10 years ago.

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 Womens 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.

Medical Intensive Care: Indications, Interventions, and Outcomes

To evaluate current practices regarding intensive-care units (ICUs), we collected data on 2693 consecutive admissions to a medical ICU during a two-year period and studied indications for admission, specific interventions, costs, and outcomes. The need for noninvasive monitoring rather than immediate major interventions prompted 77 per cent of the admissions. Only 10 per cent of monitored patients had subsequent indications for major interventions. The 23 per cent who required immediate interventions accounted for disproportionate shares of total charges (37 per cent) and deaths during hospitalization (58 per cent). Demographic and diagnostic data indicate that the aged and chronically ill have become the principal consumers of intensive care. Overall mortality during hospitalization was 10 per cent; cumulative mortality during follow-up study (mean duration, 15 months) was 25 per cent. We conclude that identification of sensitive predictors of complications and specific predictors of mortality can lead to more efficient and effective ICU practices.

The Application of Computer-Based Medical-Record Systems in Ambulatory Practice

In attempting to arrive at the truth, I have applied everywhere for information, but in scarcely an instance have I been able to obtain hospital records fit for any purposes of comparison. If they ...

Experience with a model of sequential diagnosis.

Abstract A physician must have available two different classes of information when he attempts to deal with a diagnostic problem: the presenting signs and symptoms of the patient, and the pertinent medical knowledge about the disease state. The physician must then use this information in arriving at certain decisions: what additional information about the condition of the patient should be collected; what disease states are most probable; what are the potential consequences of misdiagnosis; what are the therapeutic and prognostic implications of the particular clinical situation. The possibility of significant errors in a diagnostic problem coupled with the decision-making activities involved have motivated the development of a wide range of tools to improve the performance of the diagnostician. This paper reports on an effort to exploit computer technology as such a tool for assistance in the diagnostic activity.

Quality Assurance through Automated Monitoring and Concurrent Feedback Using a Computer-Based Medical Information System

A computer-based medical information system (COSTAR) has been used to support a quality assurance program where the data collection is an integral part of the patient care recording activity and, therefore, does not require a separate abstracting or encoding process. This program utilizes concurrent audit to detect deficiencies in patient care, and automatic rapid feedback to the responsible provider in time to allow the provider to correct the deficiency. This system has been demonstrated to improve follow-up of throat cultures, positive for Group A Beta hemolytic streptococcus. It is well accepted by the medical staff whose practice is being audited. Because the data are collected as part of the routine operation of COSTAR, the computer monitoring and feedback have only a small incremental cost.

Improving Physician Performance Through Peer Comparison Feedback

A project to improve physician performance in colorectal cancer screening was evaluated as part of an ambulatory quality assurance program. A minimum standard was adopted requiring a digital examination and stool test for occult blood at annual check-ups of patients aged 40 years and older. During a 3½-year period, three different intervention strategies for improved compliance with the standard were sequentially implemented and assessed: educational meeting, retrospective feedback of group compliance rate, and retrospective feedback of individual compliance rate compared with that of peers. A pretest/posttest design was employed in evaluating the first two intervention strategies. Neither strategy resulted in significant improvement in compliance. Monthly feedback of individual performance ranked with that of peers was then implemented in a randomized clinical trial utilizing a crossover design. During the first 6-month period, the physicians receiving feedback (group 1) improved from 66.0% to 79.9% (P <0.001), while the control group (group 2) also improved, from 67.5% to 76.6% (P <0.001), suggesting a spillover effect. During the second 6-month period, group 2 received feedback and group 1 did not. Group 1 stabilized at approximately 80% while group 2 continued to improve from 76.6% to 84.0% (P <0.001). Behavior changes persisted at 6 and 12 months after intervention.

Design and implementation of a clinical data management system

Abstract Increasing activity in the use of computers for acquisition, storage, and retrieval of medical information has been stimulated by the growing complexity of medical care, and the needs for standardization, quality control, and retrievability of clinical data. Criteria for the design of a clinical data management system include flexibility in its interface with its environment, the capability of handling variable length text string data, and of organizing it in tree-structured files, the availability of this data to a multi-user environment, and the existence of a high-level language facility for programming and development of the system. The scale and cost of the computer configuration required to meet these demands must nevertheless permit gradual expansion, modularity, and usually duplication of hardware. The MGH Utility Multi-Programming System (MUMPS) is a compact time-sharing system on a medium-scale computer dedicated to clinical data management applications. A novel system design based on a reentrant high-level language interpreter has permitted the implementation of a highly responsive, flexible system, both for research and development and for economical, reliable service operation.

A Computer-Based Monitoring System for Follow-Up of Elevated Blood Pressure

An automated surveillance system utilizing a computer-based medical record system (COSTAR) was designed to improve the follow-up of patients with newly identified elevated diastolic blood pressure. A population of patients was selected where, in the 6-month period following the initial measurement of an elevated diastolic blood pressure, there were fewer than two visits during which blood pressure was recorded. In a randomized controlled clinical trial, this poor follow-up population was divided into two groups, with computer-generated reminders being automatically generated for only patients in the experimental group. Follow-up was significantly improved in the group receiving the reminders, both in terms of rate of follow-up attempted or achieved by the responsible physician and in the repeated recording of blood pressure. We conclude that a computer-based reminder system improves follow-up of newly discovered elevation in diastolic blood pressure.

Sequential diagnosis by computer.

A computer program for sequential diagnosis has been developed for use in a variety of problem areas. The program employs sequential decision-making to balance the risk of making a diagnosis against the cost of further testing and the value of the evidence which can be obtained. Basically the program consists of an information structure which describes the problem area, an inference function which interprets signs and symptoms in terms of the medical knowledge in the information structure, and a test selection function which continually reevaluates the potential value of diagnostic tests in the light of increased information about the patient and the costs of tests and possible misdiagnoses.