Integrated decision support in a hospital cancer registry.

In this paper we present (a) a shell for integrated knowledge-based functions that is destined to support decision processes of the users of the Giessener Tumordokumentationssystem (GTDS) and (b) some results we obtained during a 6-month observation period at one of the customers of the GTDS. A special characteristic of the provided decision support is the high degree of integration in the underlying information system GTDS, i.e. the functions are triggered by events in the patient database, existing patient data is reused as input for the reasoning process and generated alerts are presented instantly to the end-user. The first routine field of application was supporting registrars to adhere to integrity constraints as defined by the International Agency of Research on Cancer (IARC) during the documentation process. This information is important for the registrars since the checks of the IARC are an accepted standard for data quality in cancer registries. The expected benefit of this application area is less effort in achieving adherence to the specification of the IARC by preventing the costly rectification at a later time. During the last 5 months of the observation period 164 alerts were displayed. About 65% of the assessed alerts were considered to be correct. Especially, the analysis of the incorrect alerts revealed some shortcomings in the knowledge behind some of the integrity constraints of the IARC. The general feedback from the end-users indicate positive user satisfaction. Currently, the shell is in use in six hospital cancer registries.

XML-based application interface services--a method to enhance integrability of disease specific systems.

Disease specific systems usually offer excellent functionality for the management of the covered diseases. But the restriction to a certain disease also often hampers their wide spread use since they aren't optimised for clinical workflow. The Giessener Tumordokumentationssystem (GTDS) is such a disease specific system. It is not only designed for the use in tumour registries but also to support clinical care. In order to integrate it into hospital information systems, we implemented standard communication interfaces, but this measure is not sufficient since it doesn't consider aspects of the normal workflow of a clinical user. Therefore, we developed a strategy that should ease the access to the system in the environment of existing systems. From the technical point of view, XML with its capabilities to represent even complex data in a rather simple way helped to implement this strategy. We use it to communicate with API-like services and also created a WWW environment to demonstrate the access to these services. Since this environment itself is a means to integrate systems, we intend to expand this environment to an appropriate region based means to improve the communication with registries. multidisciplinary environments [3]. The large amount of useful functions and its adaptability has made GTDS (http://www.gtds.de) a successful system in more than 30 various registries.

Value of XML in the implementation of clinical practice guidelines--the issue of content retrieval and presentation.

The purpose of guidelines in clinical practice is to improve the effectiveness and efficiency of clinical care. It is known that nationally or internationally produced guidelines which, in particular, do not involve medical processes at the time of consultation, do not take local factors into account, and have no consistent implementation strategy, have limited impact in changing either the behaviour of physicians, or patterns of care. The literature provides evidence for the effectiveness of computerization of CPGs for increasing compliance and improving patient outcomes. Probably the most effective concepts are knowledge-based functions for decision support or monitoring that are integrated in clinical information systems. This approach is mostly restricted by the effort required for development and maintenance of the information systems and the limited number of implemented medical rules. Most of the guidelines are text-based, and are primarily published in medical journals and posted on the internet. However, internet-published guidelines have little impact on the behaviour of physicians. It can be difficult and time-consuming to browse the internet to find (a) the correct guidelines to an existing diagnosis and (b) and adequate recommendation for a specific clinical problem. Our objective is to provide a web-based guideline service that takes as input clinical data on a particular patient and returns as output a customizable set of recommendations regarding diagnosis and treatment. Information in healthcare is to a very large extent transmitted and stored as unstructured or slightly structured text such as discharge letters, reports, forms, etc. The same applies for facilities containing medical information resources for clinical purposes and research such as text books, articles, guidelines, etc. Physicians are used to obtaining information from text-based sources. Since most guidelines are text-based, it would be practical to use a document-based solution that preserves the original cohesiveness. The lack of structure limits the automatic identification and extraction of the information contained in these resources. For this reason, we have chosen a document-based approach using eXtensible Markup Language (XML) with its schema definition and related technologies. XML empowers the applications for in-context searching. In addition it allows the same content to be represented in different ways. Our XML reference clinical data model for guidelines has been realized with the XML schema definition. The schema is used for structuring new text-based guidelines and updating existing documents. It is also used to establish search strategies on the document base. We hypothesize that enabling the physicians to query the available CPGs easily, and to get access to selected and specific information at the point of care will foster increased use. Based on current evidence we are confident that it will have substantial impact on the care provided, and will improve health outcomes.

Q-switched ruby laser treatment for postsclerotherapy hyperpigmentation.

BACKGROUND: One of the most common adverse sequelae of sclerotherapy is cutaneous hyperpigmentation. OBJECTIVE: We used the Q-switched ruby laser to treat postsclerotherapy hyperpigmentation. METHODS: Eight patients developed pigmentation lasting more than 1 year (1-2 years on average) after sclerotherapy treatment for reticular and telangiectatic veins in the legs ranging in size from 0.2 to 4 mm in diameter. All patients were treated with a Q-switched ruby laser at 694 nm, 4 mm beam size, and fluence range of 5.6-10.5 J/cm2. RESULTS: Ninety-two percent of lesions lightened after treatment. There was significant (75-100%) resolution of hyperpigmentation in 58% of treated areas, 25% improvement in 33% of treated areas, and no improvement in the remaining areas. CONCLUSION: Our impression is that given a choice of lasers, the Q-switched ruby laser provides the greatest efficacy for treating postsclerotherapy hyperpigmentation.

Development of an oncology data network in Germany.

In the 1990ies, an oncology data network has been set up mainly in the New States of Germany. Although not formally planned and established as a whole, it consists of a number of initiatives, that co-operate well and gain added value from this co-operation. From the technological view, the centre of the network is the Giessener Tumordokumentationssystem (GTDS), that was developed at Giessen University. We present important basic conditions in which this development took place, show some results and describe future directions of the development.

Combining dictionary techniques with extensible markup language (XML)--requirements to a new approach towards flexible and standardized documentation.

In oncology various international and national standards exist for the documentation of different aspects of a disease. Since elements of these standards are repeated in different contexts, a common data dictionary could support consistent representation in any context. For the construction of such a dictionary existing documents have to be worked up in a complex procedure, that considers aspects of hierarchical decomposition of documents and of domain control as well as aspects of user presentation and models of the underlying model of patient data. In contrast to other thesauri, text chunks like definitions or explanations are very important and have to be preserved, since oncologic documentation often means coding and classification on an aggregate level and the safe use of coding systems is an important precondition for comparability of data. This paper discusses the potentials of the use of XML in combination with a dictionary for the promotion and development of standard conformable applications for tumor documentation.

Integrated knowledge-based functions in a hospital cancer registry--specific requirements for routine applicability.

The background of the presented work is the design, realization, and routine use of integrated knowledge-based functions in the context of a hospital cancer registry. The first field of application was supporting registrars to detect data inconsistencies and incompleteness timely during the documentation process. Especially, we focused on the acceptance of the administrator of the underlying information system and on the phenomenon of duplicate and outdated messages. These aspects are specific for integrated knowledge based functions and a precondition for obtaining a routine applicability and acceptance.

Data warehousing as a tool for quality management in oncology.

At present, physicians are constrained by their limited skills to integrate and understand the growing amount of electronic medical information. To handle, extract, integrate, analyse and take advantage of the gathered information regarding the quality of patient care, the concept of a data warehouse seems to be especially interesting in medicine. Medical data warehousing allows the physicians to take advantage of all the operational data they have been collecting over the years. Our purpose is to build a data warehouse in order to use all available information about cancer patients. We think that with the sensible use of this tool, there are economic benefits for the Society and an improvement of quality of medical care for patients.

A model for integration and continuous development of standards for tumour documentation using relational database techniques and extensible markup language.

In oncology various international and national standards exist for different aspects of a disease. These standards, maintained by different organisations, have multiple relationships with each other. A common data dictionary like UMLS would facilitate the reorganisation of such relationships when a new version of a standard is published. While the modelling of relationships usually is restricted to types having a relevant frequency, there are often relationships which are expressed in texts like definitions or explanations. Such texts are a very important supplement for the acceptance and the safe use of coding systems, but often are neglected when implementing coding systems in computerised systems, because they are costly to implement. This paper discusses potentials when integrating various sources in a common, database based dictionary enhanced by XML (Extensible Markup Language) techniques.

Plug and play--fiction or reality?

This paper discusses some experiences with the integration of a tumor documentation system into a distributed healthcare environment using a new European middleware technology. Although the middleware already offers a considerable support to the software engineer, further facilities have been suggested and partially implemented in order to accelerate integration scenarios.

Survey among physicians by means of dynamic access to an Internet information server database.

Current database management systems, client-server architecture and the internet infrastructure are simplifying the exchange of information. Large and widespread electronic medical record systems are accessible via platform-independent browsing applications. The following brief summary shows one of the manifold conceivable applications of these technologies in medicine. It describes a survey among physicians with the scope of quality assurance in medicine. The dynamic, platform-independent, world-wide access to databases offers interesting aspects in medical informatics.

Application of a standard methodology for the development of messages and aspects of realization in the area of tumour documentation.

The "Methodology for the Development of Health Care Messages" was published in 1995 by CEN TC251. This contribution describes the application of the methodology in a project of developing messages for the exchange of tumour patient data. A standard data set, the "Basisdokumentation für Tumorkranke" was already available for the data of tumour documentation. The development of the messages as well as the design and implementation of a communication interface in a documentation system could therefore be carried out straight forward. Problems arise due to the import of data from different sources. Up to now, there is no standardized method for identifying patients and numbering of different tumours. An automated matching of new data to a specific patient or tumour is therefore not always possible. Methods are presented which try to make a compromise between fully automatic data import and complete user control.

GTDS--a tool for tumor registries to support shared patient care.

German hospital tumor registries play an active role in the treatment of cancer patients. Besides the documentation of the course of a disease, they directly support medical treatment and follow-up care over a long period. As the treatment of oncologic patients is a shared multidisciplinary task, the availability of information is one of the most valuable outputs. Therefore, the documentation has to be integrated into the medical treatment, which only can be achieved when useful services are based on it. Since 1983 the basis of the documentation has been a uniform basic data set which was revised in 1990 and, according to be requirements mentioned above, allows detailed documentation, especially of therapy. During the last four years a new documentation system for tumor diseases has been developed and was implemented in 30 hospital tumor registries by the "Arbeitsgruppe zur Koordination Klinischer Krebsregister". The so-called "Giessener Tumordokumentationssystem" (GTDS) is the basis of the work in those registries. In this paper the functions and services which were implemented in order to support the individual treatment of oncologic patients and the methods of collecting and delivering that information to the physician are presented.

Selective alterations in gene expression for NMDA receptor subunits in prefrontal cortex of schizophrenics.

NMDA receptor antagonists can induce a schizophrenia-like psychosis, but the role of NMDA receptors in the pathophysiology of schizophrenia remains unclear. Expression patterns of mRNAs for five NMDA receptor subunits (NR1/NR2A-D) were determined by in situ hybridization in prefrontal, parieto-temporal, and cerebellar cortex of brains from schizophrenics and from neuroleptic-treated and nonmedicated controls. In the cerebral cortex of both schizophrenics and controls, mRNAs for NR1, NR2A, NR2B, and NR2D subunits were preferentially expressed in layers II/III, Va, and VIa, with much higher levels in the prefrontal than in the parieto-temporal cortex. Levels of mRNA for the NR2C subunit were very low overall. By contrast, the cerebellar cortex of both schizophrenics and controls contained very high levels of NR2C subunit mRNA, whereas levels for the other subunit mRNAs were very low, except NR1, for which levels were moderate. Significant alterations in the schizophrenic cohort were confined to the prefrontal cortex. Here there was a shift in the relative proportions of mRNAs for the NR2 subunit family, with a 53% relative increase in expression of the NR2D subunit mRNA. No comparable changes were found in neuroleptic-treated or untreated controls. These findings indicate regional heterogeneity of NMDA receptor subunit expression in human cerebral and cerebellar cortex. In schizophrenics, the alterations in expression of NR2 subunit mRNA in prefrontal cortex are potential indicators of deficits in NMDA receptor-mediated neurotransmission accompanying functional hypoactivity of the frontal lobes.

GABAA receptor subunit gene expression in human prefrontal cortex: comparison of schizophrenics and controls.

The prefrontal cortex of schizophrenics is hypoactive and displays changes related to inhibitory, GABAergic neurons, and GABAergic synapses. These changes include decreased levels of glutamic acid decarboxylase (GAD), the enzyme for GABA synthesis, upregulation of muscimol binding, and downregulation of benzodiazepine binding to GABAA receptors. Studies in the visual cortex of nonhuman primates have demonstrated that gene expression for GAD and for several GABAA receptor subunit polypeptides is under control of neuronal activity, raising the possibility that similar mechanisms in the hypoactive prefrontal cortex of schizophrenics may explain the abnormalities in GAD and in GABAA receptor regulation. In the present study, which is the first of its type on human cerebral cortex, levels of mRNAs for six GABAA receptor subunits (alpha 1, alpha 2, alpha 5, beta 1, beta 2, gamma 2) and their laminar expression patterns were analyzed in the prefrontal cortex of schizophrenics and matched controls, using in situ hybridization histochemistry and densitometry. Three types of laminar expression pattern were observed: mRNAs for the alpha 1, beta 2, and gamma 2 subunits, which are the predominant receptor subunits expressed in the mature cortex, were expressed at comparatively high levels by cells of all six cortical layers, but most intensely by cells in lower layer III and layer IV. mRNAs for the alpha 2, alpha 5, and beta 1 subunits were expressed at lower levels; alpha 2 and beta 1 were expressed predominantly by cells in layers II, III, and IV; alpha 5 was expressed predominantly in layers IV, V, and VI. There were no significant changes in overall mRNA levels for any of the receptor subunits in the prefrontal cortex of schizophrenics, and the laminar expression pattern of all six receptor subunit mRNAs did not differ between schizophrenics and controls. Because gene expression for GABAA receptor subunits is not consistently altered in the prefrontal cortex of schizophrenics, the previously reported upregulation of muscimol binding sites and downregulation of benzodiazepine binding sites in the prefrontal and adjacent cingulate cortex of schizophrenics are possibly due to posttranscriptional modifications of mRNAs and their translated polypeptides.

Gene expression for glutamic acid decarboxylase is reduced without loss of neurons in prefrontal cortex of schizophrenics.

BACKGROUND: Up-regulation of gamma-aminobutyric acidA (GABAA) receptors and decreased GABA uptake in the cerebral cortex of schizophrenics suggest altered GABAergic transmission, which could be caused by primary disturbance of GABA synapses or by decreased production of the transmitter. Decreased production could be due to a shutdown in GABA production or to loss of GABA neurons caused by cell death or their failure to migrate to the cortex during brain development. METHODS: To discriminate between these possibilities, we quantified levels of messenger RNA (mRNA) for the 67-kd isoform of glutamic acid decarboxylase (GAD), the key enzyme in GABA synthesis, and the number and laminar distribution of GAD mRNA--expressing neurons in the dorsolateral prefrontal cortex (DLPFC) of schizophrenics and matched controls, using in situ hybridization-histochemistry, densitometry, and cell-counting methods. These data were compared with the total number of neurons, the number of small, round or ovoid neurons 8 to 15 microns in diameter, and overall frontal lobe volume. As a control, mRNA levels for type II calcium-calmodulin-dependent protein kinase (CamIIK) were quantified. RESULTS: Schizophrenics showed a pronounced decrease in GAD mRNA levels in neurons of layer I (40%) and layer II (48%) and an overall 30% decrease in layers III to VI. There were also strong overall reductions in GAD mRNA levels. The CamIIK mRNA levels showed no significant differences between samples. No differences were found in the total number of neurons nor in small, round or ovoid neurons, which should include a majority of the GABA cells. Prefrontal gray and white matter volume did not differ significantly between controls and schizophrenics. CONCLUSIONS: The prefrontal cortex of schizophrenics shows reduced expression for GAD in the absence of significant cell loss. This may be brought about by an activity-dependent down-regulation associated with the functional hypoactivity of the DLPFC. The lack of significant alterations in cell numbers in the DLPFC and frontal lobe volume in schizophrenics also implies that overall cortical neuronal migration had not been compromised in development. Previous reports of altered neuronal distribution in the subcortical white matter of schizophrenic brains in comparison with that of controls may indicate disturbances of migration or programmed cell death in the cortical subplate, leading to altered connection formation in the overlying cortex of schizophrenics and activity-dependent down-regulation of neurotransmitter-related gene expression.

Users conceptual views on medical information databases.

As information databases we consider all the kinds of information repositories that are handled by computer systems. When querying very large information databases, the end-users are often faced with the problem to parse their questions efficiently into the query languages of the computer systems. Conceptual graphs were initially designed for natural language analysis and understanding. Due to their closeness to semantic networks, their expressiveness is powerful enough to be applied to knowledge representation and use by computer systems. This work demonstrates that conceptual graphs are a suitable means to model both the information in patient databases and the queries to these databases, and that operations on graphs can compute the pattern matching process needed to provide the answers. A prototype that exploits this model is presented. Experiments have been made with the material furnished by the Unified Medical Language System project (version 2, 1992) of the National Library of Medicine, USA.