The Comprehensive Health Enhancement Support System (CHESS) Is an Interactive computer system containing information, social support, and problem-solving tools. It was developed with Intensive input from potential users through needs-assessment surveys and field testing. CHESS had previously been used by women in the middle and upper socioeconomic classes with high school and college education. This article reports on the results of a pilot study Involving eight African-American women with breast cancer from impoverished neighborhoods in Chicago. CHESS was very well received; was extensively used; and produced feelings of acceptance, motivation, understanding, and relief.
WWW-EMRS is an architecture that provides a set of abstractions of electronic medical record systems (EMRS). This architecture has enabled us to implement interfaces to heterogeneous EMRS via the World-Wide Web that provide consistent visual presentations and functionality. We describe WHAM!, a program that allows rapid generation of customized interfaces to WWW-EMRS that therefore deliver the customized functions across multiple EMRS.
Clinical trials constitute one of the main sources of medical knowledge, yet trial reports are difficult to find, read, and apply to clinical care. Reasons for these difficulties include the lack of a common, standardized, structure for trial reports; the restricted length of reports; and limited computer support for use of the literature. We propose a new model of reporting clinical trials, in which trials are published as both prose commentary and as data in electronic "trial banks." The prose will allow authors to discuss their trials in writing; the electronic database will allow readers easy access to well-defined data about the trials. We are developing a formal conceptual model of the clinical trials domain for integrating the use of multiple trial banks. We will then focus on validating this conceptual model with clinical literature users.
The overall goal of MENELAS is to provide better access to the information contained in natural language patient discharge summaries (PDSs), through the design and implementation of a prototype able to analyse medical texts. The approach taken by MENELAS is based on the following key principles: (i) to maximise the usefulness of natural language analysis and the usability of its results, the output of natural language analysis must be a normalised conceptual representation of medical information; and (ii) to maximise the reuse of resources, language analysis should be domain-independent and conceptual representation should be language-independent. This paper discusses the results obtained and the issues raised when implementing these principles during the project.
Written protocols are often employed to guide patient care. For treatment within a clinical trial, compliance with the trial protocol may be critical in ensuring efficacy and safety. Previous empirical work has established generic safety principles for reasoning about adverse events in clinical trials and their formalisation has been applied in a decision support system for managing treatment plans in oncology. The same generic knowledge can be reused to generate specific safety clauses when designing new treatment plans. Typically, clinicians devise trial protocols relatively infrequently and so software aids, especially those assisting with regulatory/safety conformance, will encourage more effective use of their time. A similar approach to the formalisation of safety knowledge in the control of hazardous industrial processes is discussed.
PROTEGE-II is a methodology and a suite of tools that allow developers to build and maintain knowledge-based systems in a principled manner. We used PROTEGE-II to reconstruct the well-known INTERNIST-I system, demonstrating the role of a domain ontology (a framework for specification of a model of an application area), a reusable problem-solving method, and declarative mapping relations in creating a new, working program. PROTEGE-II generates automatically a domain-specific knowledge-acquisition tool, which, in the case of the INTERNIST-I reconstruction, has much of the functionality of the QMR-KAT knowledge-acquisition tool. This study provides a means to understand better both the PROTEGE-II methodology and the models that underlie INTERNIST-I.
To facilitate networked discovery and information retrieval in the biomedical domain, we have designed a system for automatic assignment of Medical Subject Headings to documents retrieved from the World-Wide Web. Our prototype implementations show significant promise. We describe our methods and discuss the further development of a completely automated indexing tool called the "Web-MeSH Medibot."
Despite a marked increase in computer-assisted instructional applications (CAI) over the past few years, little attention has been paid to revising traditional approaches toward educational testing. This paper reports a CAI project that emphasizes integrating the testing and training of visual judgmental capacities of health care professionals. It takes advantage of the computer's ability to display digital video segments and to record and compare user learning accomplishment and at the same time a normative performance scale can be developed. The program uses a method which, in addition to validating the efficacy of the project itself, collects data and stratifies users' level of proficiency by integrating pre-test and post-test modules. Routine incorporation of these principles in CAI may provide a more effective means of correctly evaluating the individual's mastery of a topic.
Most real-life decisions require the decision maker to make trade-offs in order to fulfill multiple conflicting objectives. This is especially true in medical decision making while selecting the optimal therapy plan from among competing therapy plans for a patient. Multi-attribute utility theory provides a framework to specify these trade-offs for optimal decision making based on the preferences of the decision maker. However traditional preference-assessment techniques are difficult to implement and rarely elicit the true preferences of the decision maker. We describe a new preference-assessment method based on the concept of knowledge maintenance where the preference model is changed each time it makes an incorrect recommendation. The method is implemented in a decision-theoretic system to evaluate competing three-dimensional radiation treatment plans. The preference-assessment method leads to preference models which perform better than preference models elicited using traditional assessment techniques.
This study compares two classification models used to predict survival of injured patients entering the emergency department. Concept formation is a machine learning technique that summarizes known examples/cases in the form of a tree. After the tree is constructed, it can then be used to predict the classification of new cases. Logistic regression, on the other hand, is a statistical model that allows for a quantitative relationship for a dichotomous event with several independent variables. The outcome (dependent) variable must have only two choices, e.g. does or does not occur, alive or dead. etc. The result of this model is an equation which is then used to predict the probability of class membership of a new case. The two models were evaluated on a trauma registry database composed of information on all trauma patients admitted in 1992 to a Level I trauma center. A total of 2155 records, representing all trauma patients admitted for more than 24 h or who died in the Emergency Department, were grouped into two databases as follows: (1) discharge status of 'died' (containing 151 records), and (2) any discharge status other than 'died' (containing 2004 records). Both databases contained the same variables.
Although the concept of distributed systems for the storage of patient data is more and more commonly accepted, for some considerable time yet most patient data will be stored in centralized rather than departmental systems. An important advantage of storage in a central system is hospital-wide access to much of the patient data. Disadvantages are however that these data cannot be reviewed through one user interface, and that the structure of the data does not lend itself to exploitation for other purposes. We describe the implementation of an Andrology Research Information System in which these data are integrated in a well-structured database facilitating multiple views on the patient data through a graphical user interface, and clinical research, quality control and summary reports. The data can be analyzed directly using the Hermes workstation. In this way the strengths of the centralized system are combined with those of the dedicated ARIS system.
A considerable amount of research has been concerned with the development of natural language systems to automate the encoding of clinical information that occurs in textual form. The task is very complex, and not many language processors are used routinely within clinical information systems. Those systems that are operational, have been implemented in narrow domains for particular applications. For a system to be truly useful, it should be designed so that it could be widely used within the clinical environment. This paper examines architectural requirements we have identified as being necessary for portability and describes the architecture of the system we developed. Our system was designed so that it could be used in different domains to serve a variety of applications. It has been integrated with the clinical information system at Columbia-Presbyterian Medical Center where it routinely encodes clinical information from radiological reports of patients.
Concurrent Engineering Research Center (CERC), under the sponsorship of NLM (National Library of Medicine) is in the process of developing a computerized patient record system for a clinical environment distributed in rural West Virginia. This realization of the CCN (Community Care Network), besides providing computer-based patient records accessible from a chain of clinics and one hospital, supports collaborative health care processes like referral and consulting. To evaluate the effectiveness of the system, a study was designed and is in the process of being executed. Three surveys were designed to provide subjective measures, and four experiments for collecting objective data. Data collection is taking place in several phases: baseline data are collected before the system is deployed; the process is repeated with minimal changes three, then six months later or as often as new versions of the system are installed. Results are then to be compared, using whenever possible matching techniques (i.e. the preliminary data collected on a provider will be matched with the data collected later on the same provider). Surveys are conducted through questionnaires distributed to providers and nurses and person-to-person interviews of the patients. The time spent on patient-chart related activities is measured by work-sampling, aided by a computer application running on a laptop PC. Information about missing patient record parts is collected by the providers, the frequency by which new features of the computerized system are used will be logged by the system itself and clinical outcome measures will be studied from the results of the clinics' own patient chart audits. Preliminary results of the surveys and plans for the immediate and distant future are discussed at the end of the paper.
Clinical decision analysis seeks to identify the optimal management strategy by modelling the uncertainty and risks entailed in the diagnosis, natural history, and treatment of a particular problem or disorder. Decision trees are the most frequently used model in clinical decision analysis, but can be tedious to construct, cumbersome to use, and computationally prohibitive, especially with large, complex decision problems. We present a new method for clinical decision analysis that combines the techniques of decision theory and artificial intelligence. Our model uses a modular representation of knowledge that simplifies model building and enables more fully automated decision making. Moreover, the model exploits problem structures to yield better computational efficiency. As an example we apply our techniques to the problem of management of acute deep venous thrombosis.
We are performing a randomized, controlled trial of a Physician's Workstation (PWS), an ambulatory care information system, developed for use in the General Medical Clinic (GMC) of the Palo Alto VA. Goals for the project include selecting appropriate outcome variables and developing a statistically powerful experimental design with a limited number of subjects. As PWS provides real-time drug-ordering advice, we retrospectively examined drug costs and drug-drug interactions in order to select outcome variables sensitive to our short-term intervention as well as to estimate the statistical efficiency of alternative design possibilities. Drug cost data revealed the mean daily cost per physician per patient was 99.3 cents +/- 13.4 cents, with a range from 0.77 cent to 1.37 cents. The rate of major interactions per prescription for each physician was 2.9% +/- 1%, with a range from 1.5% to 4.8%. Based on these baseline analyses, we selected a two-period parallel design for the evaluation, which maximized statistical power while minimizing sources of bias.
CompuHx* is an Interactive Health Appraisal System (IHAPS) used in the examining room at Kaiser-Permanente's San Diego Department of Preventive Medicine to record patient information, assist in diagnosis, and provide a legible summary of findings. The purpose of the present project was to examine the impact of computer use in the examining room on patient satisfaction with the Health Appraisal experience. Survey results showed no significant differences in patient satisfaction between patients whose examiners used CompuHx and those whose examiners did not. These findings indicate that, in the eyes of the patients surveyed, clinician use of a computer in the examining room did not depersonalize their relationship with the clinician, nor did it enhance satisfaction with the thoroughness of the exam or confidence in the examiner's findings.
This evaluation looks at the use of templates for entering structured text nursing notes that generate both a legal text note that is the chart record and an underlying coded form of the note to support analysis and research. This study reflects the first phase of a prototype project of an integrated, computerized health record. Templates are notes that have been prewritten using a standard clinical vocabulary. Templates can be used as the basis of a new clinical note and can be either signed unchanged or modified to represent variations in clinical presentation. The prototype setting is a Primary Care clinic where both physicians and nurses are using the computer to enter clinical notes. In the prototype clinic team, nursing utilized the CPR for 100% of all documentation from day one. Use of templates was found to be the most frequent method of initiating a note.
The ability to manage information with regard to changes in a database is critical for quality control. This information can also provide audit trails about the time of the change and the person who made the change. In addition, historical information can provide the proper context in which to interpret the relationships between the current and past data. In most genomic databases, only the most recent copy of the information is presented to the user, thereby losing the audit trail and the historical context. Therefore, we have constructed a delivery mechanism for the historical information in the Mendelian Inheritance in Man database. Furthermore, this feature was designed to optionally display only the changes so that the user can bypass the unchanged portions of the text. It was anticipated that technical problems would influence the acceptance of this information delivery. However, the involvement of the editorial staff became the critical factor.
The Military Health Service System (MHSS) provides health care for the Department of Defense (DOD). This system operates on an annual budget of $15 Billion, supports 127 medical treatment facilities (MTFs) and 500 clinics, and provides support to 8.7 million beneficiaries worldwide. To support these facilities and their patients, the MHSS uses more than 125 different networked automated medical systems. These systems rely on a heterogeneous telecommunications infrastructure for data communications. With the support of the Defense Medical Information Management (DMIM) Program Office, our goal was to identify the network requirements for DMIM migration and target systems and design a communications infrastructure to support all systems with an integrated network. This work used tools from Business Process Reengineering (BPR) and applied it to communications infrastructure design for the first time. The methodology and results are applicable to any health care enterprise, military or civilian.
We present a 3D graphical system that allows users to visualize different penetration path hypotheses for (multiple) gunshot or stab wounds, using a 3D graphical model of a human body with appropriate anatomical structures. The system also identifies the anatomical structures associated with each hypothesis. The various penetration path hypotheses follow from a combinatorial analysis of the set of surface wounds. The affected structures are determined by performing a detailed interpenetration analysis between 3D models of a penetration path and each anatomical structure within the body.
