Search

This paper describes a prototype for research to evaluate the impact of diagnostic decision support systems on the behavior of physicians. Several indices that can be used to quantify the magnitude of impact are proposed. A large medical diagnostic knowledge base in internal medicine (the Iliad knowledge base) was used in this evaluation. The impact on behavior when different inference models are run against this knowledge base is evaluated for two different case domains and physician's specialties.

Information management is central to modern patient care. Computerization of information management has resulted in both departmental systems which serve information needs in locations such as the Radiology Department and in hospital-wide information systems which seek to integrate management of clinical data from many departments. For each of these systems to achieve the goal of maximizing both the effectiveness of health care workers and the quality of patient care, they need to share the data that they capture. Below we discuss a variety of applications, both currently available and in the realm of research protocols, that depend on a high level of communication between Radiology Information Systems and Hospital Information Systems. These examples suggest the benefits of integrating the medically relevant data collected by all of the computer-based information systems in the hospital setting.

We tested a new model of the HELP patient database that makes use of relational tables to store patient data and provides access to data using SQL (Structured Query Language). The SQL database required more storage space and had many more physical records than the HELP database, but it was faster and more efficient in storing data than the standard HELP utilities. The HELP utilities used disk space more efficiently and were faster than the SQL tools when retrieving data for typical clinical reports. However, the SQL model provides networking capabilities, general report writing tools, detailed user documentation, and an ability for creating secondary indexes that offset its poorer performance.

The Arden Syntax for sharing medical knowledge bases is described. Its current focus is on knowledge that is represented as a set of independent modules that can provide therapeutic suggestions, alerts, diagnosis scores, etc. The syntax is based largely upon HELP and the Regenstrief Medical Record System. Each module, called a Medical Logic Module or MLM, is made of slots grouped into maintenance, library, and knowledge categories. The syntax has provisions for querying a clinical database and representing time. Several clinical information systems were analyzed and appear to be compatible with the syntax. The syntax has been tested for syntactic ambiguities using the tools lex and yacc. Four institutions are currently in the process of adopting the Arden Syntax for their decision-support systems.

Sharing medical logic has traditionally occurred in the form of lectures, conversations, books and journals. As knowledge based computer systems have demonstrated their utility in the health care arena, individuals have pondered the best way to transfer knowledge in a computer based representation (1). A simple representation which allows the knowledge to be shared can be constructed when the knowledge base is modular. Within this representation, units have been named Medical Logic Modules (MLM's) and a syntax has emerged which would allow multiple users to create, criticize, and share those types of medical logic which can be represented in this format. In this paper we talk about why standards exist and why they emerge in some areas and not in others. The appropriateness of using the proposed standards for medical logic modules is then examined against this broader context.

A new order entry system has been developed on the HELP system utilizing the extensive HELP knowledge base capabilities. Rather than use tables to control site specific characteristics of the ordering system, the new order entry system uses Medical Logic Modules (MLM's) to store and execute site specific features. Using the MLM the user is not only able to set traditional parameters, but enter medical logic to control patient/hospital specific requirements and execute critiquing logic on the order. Simple knowledge editors are provided which make the creation of the order MLM's no more difficult than those programs used to define the site specific parameters of traditional systems.

The HELP system, designed and developed at the LDS Hospital in Salt Lake City, is a data driven hospital and medical information system. In contrast to most hospital information systems the HELP system is primarily designed to implement medical decision-making in real time for all aspects of patient care. The current HELP system is implemented on a Tandem “non stop” computer system consisting of 4 CPUs, disc storage and 240 communication ports. The major subsystems used in development of the application programs consist of a data base management subsystem, a decision-making subsystem, a manual data entry subsystem and a communications subsystem. Applications running on the HELP system include ADT, order entry, pharmacy, laboratory, x-ray, patient monitoring, medical records, clinical research and teaching.

Design of the software architecture for a knowledge driven HIS is presented. In our design the frame has been used as the basic unit of knowledge representation. The structure of the frame is being designed to be sufficiently universal to contain knowledge required to implement not only expert systems, but almost all traditional HIS functions including ADT, order entry and results review. The design incorporates a two level format for the knowledge. The first level as ASCII records is used to maintain the knowledge base while the second level converted by special knowledge compilers to standard computer languages is used for efficient implementation of the knowledge applications.

Medical knowledge in a computer processable representation can be shared. The issues involved in distribution include: 1. the representational format (syntax) of the knowledge, 2. availability of a comprehensive editor, 3. development of convenient methods for translating the distributed knowledge base into the native language and database environments which vary from institution to institution, 4. quality control, 5. the need for users to review and assume responsibility for the content of any logic which they obtain, 5. willingness of authors to share knowledge, 6. willingness of institutions to act as clearing houses (libraries) for such collections of knowledge, 7. techniques for knowledge base management, and 8. liability. We describe some of the problems and potential solutions associated with each of these issues.

Patient acuity measurements are utilized for several functions in a hospital. This automating of acuity from nursing computer documentation was developed at LDS Hospital in Salt Lake City, utilizing HELP, a total hospital information system. The goals are to reduce paperwork responsibilities for the nurse and to define a patient's acuity measurement objectively from a documented nursing record.

This paper describes an examination of using bedside terminals versus one computer terminal per 4 patients (pod terminal). The study of this concept was conducted on a 46 bed acute care nursing unit. Bedside terminals were installed on one half of the nursing unit, and pod terminals were installed on the other half of the nursing unit. A six month pilot study was conducted which measured the following aspects

Ongoing efforts have been made to improve the efficiency and quality of care available in ICU's at LDS Hospital. ICU's have been computerized, and patient data collection, storage and presentation have been improved. Nurse care plans and charting have been computerized, and effects of these changes have been evaluated.

At LDS Hospital in Salt Lake City, Utah, development and implementation of a Nursing Information System has been underway for the last seven years. The current endeavor is implementation of computer charting at the bedside. A most important aspect of this gradual implementation is evaluation of the effects of this change in charting method and behavior. This study investigates the differences in three areas before and after implementation of computerized charting; nursing time distribution, nurse attitudes toward computerization, and compliance with charting standards. Computerization brings hopes of time savings for the nurse so more time is spent with the patient, positive attitudes towards computer use, and an increase in the quality of charting and its relationship to the plan of care for the patient. The results of this study point towards positive changes in these areas.

To address the issue of facilitating transfer and integration of the variety of computer-based programs which contain medical expertise, a retreat was held at Columbia University's Arden Homestead conference center June 16-18, 1989. The focus of this retreat was to explore ways in which the medical expertise contained in knowledge-based systems could be shared and expanded. During the three day meeting, the eighteen attendees from ten institutions discussed: (a) the need for better ways of mapping terminology used in one setting or program to terms with similar meaning that have been used in other programs, (b) the need for catalogues which list the variety of programs which are available, (c) a representational syntax and format for sharing modular medical knowledge, (d) the possibility of developing standards for interfacing program modules so that they could be “snapped” into place in a variety of systems, (e) methods for evaluating, validating and testing knowledge based systems, and (f) the legal and financial aspects of sharing systems which influence the care that is given to a patient.