A rational request behavior: the development of prediction instruments regarding thyroid function tests in primary care.

A prospective study was performed to encourage a rational thyroid-stimulating hormone (TSH) test request behavior of physicians, using prediction instruments. The latter give direct feedback about request adequacy on the basis of pretest probabilities of hyperthyroidism using patients' signs and symptoms. For instrument design, stepwise logistic regression was used on diagnostic data acquired through questionnaires, answered by 80 physicians and 668 patients for whom physicians requested a TSH test. Instruments were designed for clinical and subclinical hyperthyroidism and for clinical hyperthyroidism alone. Use of the instrument for clinical or subclinical hyperthyroidism on the selected group, at a 5% probability threshold, can result in a 37% reduction of unnecessary TSH test requests. With the instrument for clinical hyperthyroidism at a 5% probability threshold, the number of unnecessary test requests can be reduced by 57%. Therefore, it can be concluded that the instruments can determine TSH test request adequacy and encourage a rational TSH test request behavior of physicians at low pretest probability thresholds.

Evaluation of GPs' medical knowledge by a decision support system.

This paper deals with the EValuator which is a multipurpose DSS. EV can be applied by creators of automated disease and therapy models to get insight into knowledge necessary for diagnosis and treatment of diseases. GPs, trainees/interns, hospital clinicians, and students can test their medical knowledge with EV. If necessary, EV offers decision support. In a functionality test, EV was used by GPs to assess their knowledge of thyroid disorders. Test conclusions referred to the GPs' level of medical knowledge of thyroid disorders, differences in EV usage by urban and rural GPs, use of thyroid disorder indices and the GPs' commendations that EV is capable of providing automated decision support to medical professionals in daily practice, and that it can very well be used in the education of students and in GP refresher courses.

User satisfaction of general practitioners with HIOS+, a medical decision support system.

In this article, a user satisfaction study is described. The Formal Interpreter (FI) and the Free-text Formalizer (FF) are modules of the medical DSS HIOS+. With FF, medical experts can formalize topics selected from free-text ICHPPC-II-defined criteria into queries. FI gives decision support to GPs in their daily work by interpreting queries belonging to entered ICPC codes and presenting advice/warnings concerning the validity (correctness) of the diagnoses in patient cases. In the study, FI was used in field and laboratory test situations by GPs. User satisfaction was assessed using questionnaires consisting of an adapted version of the Bailey questionnaire. Answers given were analysed using Bailey's approach and the 'two-sample t-test'. The most important result of the FI evaluation was that users given active decision support were more satisfied with FI than users given none or only textual decision support.

Computer-based assistance in family medicine.

In this paper, the fully integrated General Practitioner (GP) Information and Research System HIOS and Decision Support System HIOS+ are described. Both systems were developed with the Phased Development Methodology (PDM), which is also presented. HIOS consists of an administrative module, medical modules, and a research module to support GPs in daily practice activities and research. It was implemented at a general practice and a health centre, and was evaluated positively. HIOS+ was developed by extending HIOS with modules which offer several types of decision support to assist GPs in (diagnostic) decision making. The modules offer passive and active decision support. HIOS+ can present textbook information, generate advice and warnings, support the creation of models, generate reports, and can be used in education. Some HIOS+ modules were evaluated by GPs in evaluation studies, all with positive results.

The use of the Arden Syntax for MLMs in HIOS+, a decision support system for general practitioners in The Netherlands.

This paper covers query creation and evaluation in HIOS+, a General Practitioner Decision Support System. The International Classification of Primary Care (ICPC), free-text International Classification of Health Problems in Primary Care (ICHPPC)-II-Defined criteria, and the Arden Syntax for medical logic modules (MLMs) are building blocks in query creation with the free-text formalizer (FF) module. The backward-reasoning formal interpreter (FI) module evaluates queries by checking them against medical patient data. It can advise or warn users on the suitability of a registered diagnostic ICPC code in a patient's case. In all, 101 queries were created in MLM form on a wide range of diseases. The MLMs and FI were evaluated in a field and a laboratory test.

Decision support for health professionals.

In this contribution two decision support tools are discussed. One is used for reducing the number of inadequately requested tests by GPs and the other to support neurologists in prescribing medications for epilepsy patients. The experiences with both programs will be discussed.

Evaluation of a decision support system in a medical environment.

This paper describes the impact of a decision support system on the quality of recorded diagnoses and the completeness of medical records. The assumption is that for quality assessment purposes enough data have to be recorded in an electronic medical record so that diagnostic decisions can be justified. The hypotheses were tested that active decision support will lead to better quality recorded diagnoses and more complete medical records. Three groups of ten GPs were presented with 10 cases each. The GPs had to enter the data about these cases in a GP information system. One group of GPs was not supported. The second group was presented with the ICHPPC-II-Defined criteria that had to be fulfilled when a diagnosis was entered. In a third group, the GPs were asked those data that were needed to justify an entered diagnosis (active support). It could be shown that the last group of GPs entered better quality diagnoses than the other two groups. These latter groups also entered so little data that most of their diagnoses could not be justified. It is concluded from the study that only active decision support will lead to better diagnoses and a more complete medical record that can be used for quality assessment.

A research tool for general practitioners.

In this paper a research tool for general practitioners is described, that will be used in a network of community-based primary care practices in order to collect the necessary data for research purpose. The research tool is coupled to a commercial information system for general practitioners. The research tool supports coding strongly. The way in which the research tool is designed is described and examples of the use of windowing techniques are given.