An Algorithm Using Twelve Properties of Antibiotics to Find the Recommended Antibiotics, as in CPGs.

BACKGROUND: Clinical Decision Support Systems (CDSS) incorporating justifications, updating and adjustable recommendations can considerably improve the quality of healthcare. We propose a new approach to the design of CDSS for empiric antibiotic prescription, based on implementation of the deeper medical reasoning used by experts in the development of clinical practice guidelines (CPGs), to deduce the recommended antibiotics. METHODS: We investigated two methods ("exclusion" versus "scoring") for reproducing this reasoning based on antibiotic properties. RESULTS: The "exclusion" method reproduced expert reasoning the more accurately, retrieving the full list of recommended antibiotics for almost all clinical situations. DISCUSSION: This approach has several advantages: (i) it provides convincing explanations for physicians; (ii) updating could easily be incorporated into the CDSS; (iii) it can provide recommendations for clinical situations missing from CPGs.

Design factors for success or failure of guideline-based decision support systems: an hypothesis involving case complexity.

Computer-based decision support systems (CDSSs) are currently mostly reminder systems. However, the effectiveness of such systems to modify physician behavior is not always observed. We assume that this approach is appropriate when physicians think they know how to prescribe and consider they don't need to be helped, i.e. for simple clinical cases. On the opposite, on-demand approaches allowing for flexibility in the interpretation of patient conditions are more appropriate for more complex cases, e.g. in chronic disease management. ASTI is a CDSS operating in two modes, a critiquing mode working as a reminder-based system and a user-initiated guiding mode. Using a clinical case complexity score, a pre/post-intervention experiment with 10 GPs and 15 cases of hypertensive patients has been performed. Preliminary results tend to indicate that reminder-based interaction is appropriate for simple cases and that physicians are willing to use on-demand systems as clinical situations become complex, making both modes complementary.

Improving model robustness with bootstrapping -- application to optimal discriminant analysis for ordinal responses (ODAO).

OBJECTIVE: Recent results published by Coste et al. in discriminant analysis with ordinal responses showed the superiority of optimal discriminating analysis for ordinal responses (ODAO) both in terms of classification and simplicity of implementation compared to classic methods (Fisher's discrimination, logistic regression) applied to medical data (prognostics of burns) and to simulated data. Nevertheless, the solutions obtained by ODAO may be sensitive to re-sampling (i.e the estimated coefficients by ODAO may show excessive sensitivity to the training sample). This study proposes some solutions to control the fluctuations of sampling and to ensure model stability. METHODS: We used intensive computational methods and bootstrapping, at the outset of model building in order to reduce the sampling variability of estimated coefficients. Thus, the estimation of the coefficients was not based on the minimization of a classification criterion of the training sample, but on the minimization of an aggregate criterion of bootstrapped replications of a classification criterion. Five aggregate criteria were studied. RESULTS: The improvement in terms of robustness appeared in 30% of the test cases with moderate training sample size and 55% of those with small training sample size. CONCLUSION: Simulated test cases showed that bootstrapping can help construct more robust models in difficult classification situations and small training samples which are particularly frequent.

Identification of a developmental chemoattractant in Myxococcus xanthus through metabolic engineering.

Fruiting body formation of Myxococcus xanthus requires the ordered migration of tens of thousands of cells by using a form of surface motility known as gliding and chemical signal(s) that have yet to be elucidated. Directed movement is regulated by phosphatidylethanolamine (PE) purified from M. xanthus cell membranes. Because the purified PE preparation contains a remarkably diverse mixture of fatty acids, metabolic engineering was used to elucidate the biologically active fatty acid component. The mutational block in an esg mutant, which renders it defective in producing primers for branched-chain fatty acid biosynthesis, was bypassed with one of a series of primers that enriches for a particular family of branched-chain fatty acids. Each PE enrichment was observed for chemotactic activity by using an excitation assay and for fatty acid content. The excitation activity of a PE preparation was generally proportional with the concentration of the fatty acid 16:1 omega 5c. 1,2-O-Bis[11-(Z)-hexadecenoyl]-sn-glycero-3-phosphoethanolamine (PE-16:1 omega 5c/16:1 omega 5c) was synthesized and elicited an excitation peak at 2 ng. This peak activity occurred at a 1,000-fold lower concentration than dilauroyl PE (PE-12:0/12:0) and the peak magnitude was 2-fold higher. PE containing 16:1 omega 5c is likely to play a role in development because it is active at physiological concentrations and only under developmental conditions.

Distance mapping of protein-binding sites using spin-labeled oligosaccharide ligands.

The binding of a nitroxide spin-labeled analog of N-acetyllactosamine to galectin-3, a mammalian lectin of 26 kD size, is studied to map the binding sites of this small oligosaccharide on the protein surface. Perturbation of intensities of cross-peaks in the (15)N heteronuclear single quantum coherence (HSQC) spectrum of full-length galectin-3 owing to the bound spin label is used qualitatively to identify protein residues proximate to the binding site for N-acetyllactosamine. A protocol for converting intensity measurements to a more quantitative determination of distances between discrete protein amide protons and the bound spin label is then described. This protocol is discussed as part of a drug design strategy in which subsequent perturbation of chemical shifts of distance mapped amide cross-peaks can be used effectively to screen a library of compounds for other ligands that bind to the target protein at distances suitable for chemical linkage to the primary ligand. This approach is novel in that it bypasses the need for structure determination and resonance assignment of the target protein.

ASTI: a guideline-based drug-ordering system for primary care.

Existing computer-based ordering systems for physicians provide effective drug-centered checks but offer little assistance for optimizing the overall patient-centered treatment strategy. Evidence-based clinical practice guidelines have been developed to disseminate state-of-the-art information concerning treatment strategy but these guidelines are poorly used in routine practice. The ASTI project aims to design a guideline-based ordering system to enable general practitioners to avoid prescription errors and to improve compliance with best therapeutic practices. The " critic mode " operates as a background process and corrects the physician's prescription on the basis of automatically triggered elementary rules that account for isolated guideline recommendations. The " guided mode " directs the physician to the best treatment by browsing a comprehensive guideline knowledge base represented as a decision tree. A first prototype, applied to hypertension, is currently under development.

Design of a computerized system for the retrospective analysis by the physician of his own drug prescriptions.

The quality of drug prescription is known to be poor in many medical domains. Feedback information may help the physician to improve this component of his medical activity. The increasing use of computerized drug prescription systems makes possible the retrospective processing of prescriptions made over a given time period. This paper describes the design of a computerized system for providing feedback information to the physician about the quality of his drug prescribing activity. It is designed to be used by the physician himself. The analysis is based on two groups of quantitative indicators, simple and composite. The constraints that must be satisfied to calculate these indices automatically are given. They are related to the coding systems used for drug prescription, knowledge on drugs and patient data. The system design is based on modeling according to UML formalism. We give also a concrete view of the main functions of the system, based on a pictorial illustration with a component of the user interface. Some limitations and the possible extensions of this system for analyzing simultaneously the prescriptions made by set of physicians are discussed.

Structured representation of drug indications: lexical and semantic analysis and object-oriented modeling.

No standardized representation of drug indications is currently available that could be used in drug knowledge bases. We describe an object-oriented representation of indications that should make it possible to develop new tools for selecting drugs and checking prescriptions in computerized drug prescription systems. The model was developed using the results of a lexical and semantic analysis of drug indications, collected into a single file and processed using natural language processing software. It distinguishes both the diseases for which the drug may be given and the efficiency of the drug for a given indication. Two aspects of the model were evaluated: the differences if two independent evaluators filled the attributes independently and the loss of information induced by the use of the model. A system based on this model, making it possible for the physician to select all the drugs satisfying various criteria, is also presented.

An epidemiologic approach to drug prescribing quality assessment: a study in primary care practice in France.

OBJECTIVE: To assess drug prescribing by primary care physicians in France for various types of conditions, and to identify patterns and risk factors for poor prescribing quality. METHODS: The orders (n = 23,080) written for patients with five target diseases (acute nasopharyngitis, acute tonsillitis, essential hypertension, osteoarthrosis, and back and periarticular disorders), by primary care physicians (n = 1,049) were extracted from a nationwide prescription database and analyzed according to 17 quantitative indicators of drug prescribing quality constructed on explicit a priori criteria. RESULTS: Ineffective drugs were prescribed in 32% to 88% of orders according to the target disease. Six percent to 40% of orders resulted in drug interactions, age problems, and overdosage. A consistent pattern of associations between indicators was found, which suggests that drug prescribing quality is multidimensional and is composed of at least five dimensions: placebo, novelty, exoticism, misdosage, and interaction. Several factors associated with indicators were also identified, some of them defining groups of patients at risk (women, elderly, and less educated), physicians at risk (women, aged, and isolated), and contexts at risk (patient's home and disease frequently treated by the physician) of poor drug prescribing quality. CONCLUSIONS: Drug prescribing by French primary care physicians appears nonoptimal, in terms of both risk of iatrogeny and waste of money. This study further documents the complexity and the multidimensionality of drug prescribing quality. It suggests that more attention must be paid to patients' and physicians' risk factors for poor drug prescribing quality if educational programs and regulatory processes are to succeed in promoting safer and more cost-effective practices.

Methodology for the analysis and representation of the medical information about drugs in the Summary of Product Characteristics (SPC).

We present a methodology for the representation of the medical knowledge in the drug SPCs. It includes four steps, the two first of which are automated. All instances of a particular SPC text are gathered into a single file. Lexical analysis of the content of this file is performed and a lexicon with the occurrence of words and groups of words is built. Semantic analysis is carried out considering the concepts underlying each word of the lexicon and the most important concepts are kept. This semantic analysis results in a list of attributes which are then included in an object-oriented model. We have used this method to structure drug indications. This application clearly illustrates the advantages of this method over purely manual analysis. This method could be generalized for all categories of medical information about drugs.

Electronic prescribing for the elderly: will it improve medication usage?

Drug treatment of the elderly is problematic for various reasons. These include the risks of drug interactions, contraindications, incorrect dose and adverse events. Electronic prescribing is a promising tool which should help solve many of the problems associated with the use of medications in the elderly by providing information on drug selection, prescription checks, and information about drugs and prescriptions. Nevertheless, electronic prescribing must be improved by providing a user-friendly interface, structured drug data-bases, and the capacity to generate both criticism and suggestions, if it is to progressively replace pen and paper in the drug prescription process.

Autoantibodies directed against the ribosomal P proteins are not only directed against a common epitope of the P0, P1 and P2 proteins.

The autoantibodies (aAbs) directed against the ribosomal P proteins (RPP aAbs) are known to react mainly against epitopes localized within the common C-terminal sequence of the three acidic ribosomal P proteins, P0, P1 and P2. In order to investigate the opportunity to select short recombinant peptides of this common C-terminal sequence to detect the RPP-aAbs, the location of the epitopes recognized by ribosomal proteins (RP) aAb(+)sera of systemic lupus erythematosus patients (SLE) was investigated. Immunoblotting and ELISA techniques using extracted or recombinant, entire or cleaved RPP showed that 55% of the RP aAbs were directed against the three ribosomal P0, P1, and P2 proteins. The epitopes recognized by the RPP aAbs are located not only within the C-terminal sequence common to the three proteins but also within the N-terminal sequence of the P2 or P1 protein. The other RP aAbs sera (45%) did not react with all three proteins but with some of them, and showed the following pattern: P0(+)P1(+); P1(+); P2(+); P0(+)and P1(+). They recognized epitopes located in the region of the C-terminal sequence of the protein but not common to the three proteins. In addition two out of the six monoclonal Abs produced by immunization of mice using the P1 protein did not react with the peptide N-65 or N-71 of the P2 protein or with the C-terminal sequence of the three proteins. In conclusion, this study showed that the RPP aAb in SLE patients are not only directed against epitopes within the C-terminal sequence shared by the three acidic ribosomal P proteins. In view of these data it seems necessary to be cautious in using only a C-terminal peptide of ribosomal P proteins in tests performed to detect RPP aAb in human sera.

Crystal state and solution conformation of the B blood group trisaccharide alpha-L-Fucp-(1-->2)-[alpha-D-Galp]-(1-->3)]-beta-D-Galp-OCH3.

The crystal structure of the human B blood group related trisaccharide alpha-L-Fucp-(1-->2)-[alpha-D-Galp]-(1-->3)-beta-D-Galp-OCH3 (1) has been determined. The solution structure of 1 was studied by two-dimensional NMR techniques at 600 MHz in D2O solution and the conformational properties were analyzed in terms of the torsional angles phiH and psiH, derived from 3JCH coupling constants, and 10 inter-residue proton-proton distances. 3JCH could be accurately measured by a recently introduced two-dimensional heteronuclear correlation experiment (EXSIDE). The nuclear Overhauser enhancement-derived distances and the calculated torsion angles were compared with the same information available from the crystal structure. The agreement is excellent, indicating that the trisaccharide adopts a restricted conformation in solution, which was also predicted by the Hard Sphere Exo-Anomeric forcefield. The data of 1 are complemented by NMR studies of the closely related alpha-L-Fucp-(1-->2)-[6-deoxy-alpha-D-Galp]-(1-->3)-beta-D-Galp O-(CH2 )7CH3 trisaccharide (2).

Indicators for the automated analysis of drug prescribing quality.

Irrational and inconsistent drug prescription has considerable impact on morbidity, mortality, health service utilization, and community burden. However, few studies have addressed the methodology of processing the information contained in these drug orders used to study the quality of drug prescriptions and prescriber behavior. We present a comprehensive set of quantitative indicators for the quality of drug prescriptions which can be derived from a drug order. These indicators were constructed using explicit a priori criteria which were previously validated on the basis of scientific data. Automatic computation is straightforward, using a relational database system, such that large sets of prescriptions can be processed with minimal human effort. We illustrate the feasibility and value of this approach by using a large set of 23,000 prescriptions for several diseases, selected from a nationally representative prescriptions database. Our study may result in direct and wide applications in the epidemiology of medical practice and in quality control procedures.

Object-oriented modeling and terminologies for drug contraindications.

There is no available standardized representation of contraindications that can be used in drug knowledge bases. In this paper, an object-oriented representation of contraindications is proposed that would allow computerized checking of drug prescription safety. It distinguishes four types of contraindications: pathological state, physiological state, findings of investigation procedures, and diagnostic or therapeutic procedures. The vocabulary usable for expressing the contraindications appropriate for automated prescription checking was also investigated. ICD10 (International Classification of Diseases, tenth revision), SNOMED III (Systematized Nomenclature of Medicine), ICPC (International Classification of Primary Care), and ATC (Anatomic Therapeutic and Chemical Classification) were studied as potential sources of standardized vocabulary. A system to support entering data on contraindications into a drug knowledge base, compatible with the proposed structure and using the identified vocabulary sources, is also presented.

Cost-related information to be provided by computerised drug-prescription systems to promote cost-effective prescribing.

Drug prescriptions are a source of steadily increasing healthcare expenditure in most developed countries. As the use of Computerised Drug Prescription Systems (CDPS) increase both in hospital and community-care settings, the potential of such systems to promote cost-effective prescribing and help contain prescription costs should be considered. This article describes the cost-related information that could be provided by a CDPS, namely decision-support information to be provided on-line during the prescription entry, and retrospective information made available by processing stored prescription records. The review also describes design and specification requirements for building a cost-information module that can be used in various health-delivery systems. These are: (i) adequate and well-organised data; (ii) pertinent background knowledge of the domain; and (iii) algorithms that allow adaptation to site-specific features. The propositions expounded in this article result from a part of the work performed during the Optimisation of Drug Prescription using Advanced Informatics (OPADE) European project.

Optimal discriminant analysis for ordinal responses.

Optimal classification formulation is adapted to the context of discrimination when the response is ordinal. The resulting method, optimal discriminant analysis for ordinal responses (ODAO), is presented and compared with two reference discrimination techniques used in this context (proportional-odds ordinal logistic regression and normal discrimination) using a study of prognosis following burn injuries and simulated data. The ODAO method clearly outperforms both reference methods in terms of classification accuracy (in training and validation samples), robustness to outliers, simplicity of use and applicability in clinical settings. ODAO is a promising method for improving classification performance in discrimination with ordinal responses and merits further investigation.

Predicting mortality in adult burned patients: methodological aspects of the construction and validation of a composite ratio scale.

This article describes the methodology of construction and validation of a composite measurement scale (CMS) to predict the risk of death for burned patients, with severity of burn considered as a continuous phenomenon. Three large data sets of burned patients hospitalized in France were analyzed. Logistic regression was used to construct a prognostic model, based on age and initial evaluation of total body surface area burned. The resulting model appears to be a valid clinical tool for predicting the risk of death. In addition, the devised CMS has satisfactory content and construct validity and reliability, and provides a high measurement level (logistic ratio level). Moreover, its simplicity of use (the score is integer based) is appropriate for the daily activities of burn unit physicians, emergency medical technicians, and public health professionals.

The Asthma Impact Record (AIR) index: a rating scale to evaluate the quality of life of asthmatic patients in France.

Asthma is a chronic disease, which affects patients' daily lives. The goal of this study was the development of a disease-specific scale to evaluate quality of life in asthmatic patients in France: The Asthma Impact Record Index (AIR Index). The study was conducted with the participation of 486 asthmatic patients using the following steps: 1) selection of dichotomous items; 2) reduction of the number of items; 3) study of reproducibility of the questionnaire; 4) weighting of items; 5) study of the reliability and validity of the final version of the AIR Index. The final version of the AIR Index contains 63 unweighted items. The items were classified into subscales representing the main dimensions of quality of life: 1) physical, which was itself split into two subscales: a) physical activities; and b) symptoms; 2) psychological; and 3) social or relational. The internal consistency, measured by Cronbach's alpha coefficients, was found to be high, for the global scale and all subscales (range 0.79-0.94). The concurrent validity, evaluated by studying the relationship between the score values on the global scale and the subscales, and the parameters reflecting disease severity, was also high. We conclude that the AIR Index might represent a useful evaluative and discriminant instrument in studying quality of life in asthma in French populations.

A new approach to selection and weighting of items in evaluative composite measurement scales.

Evaluative composite measurement scales (CMS) are increasingly used in medicine to measure complex constructs in the absence of a reference criterion or 'gold standard'. However, the level of measurement provided by these instruments has usually been given little attention. This paper explores the influence of the characteristics of the item set and weighting on the level of measurement of an evaluative CMS. The approach is illustrated with an application to the scoring system of the Nottingham Health Profile. Simulations are provided to indicate when the composite score of item responses, on an ordinal or pass-fail scale, could be considered to achieve the properties of an interval scale.