Please activate JavaScript!
Please install Adobe Flash Player, click here for download


ISCB 2014 Vienna, Austria • Abstracts - Poster Presentations 131Wednesday, 27th August 2014 • 15:30-16:00 Monday25thAugustTuesday26thAugustThursday28thAugustAuthorIndexPostersWednesday27thAugustSunday24thAugust group internally, trial-level imputation of correlation and solutions at the meta-analysis level. For each method, we note whether it calculates the mean or the SD, list which summary statistics are needed to implement it, and summarise any assumptions.   P4.1.190 Effect of lifestyle and metformin for delaying or preventing type 2 diabetes: a network meta-analysis K Yamaoka1 , T Tango1,2 1 Teikyo University Graduate School of Public Health, Tokyo, Japan, 2 Center for Medical Statistics, Tokyo, Japan   Background: A sedentary lifestyle accompanied by dietary issues is a central theme in the development of type 2 diabetes (T2DM), and ame- lioration of related factors is a central theme in the prevention of T2DM. Lifestyle modifica-tions are considered effective means of delaying or preventing T2DM. However, the effect of different types of therapies on resolution of this is controversial. Objective: To assess the effects of different types of therapies on delaying or preventing T2DM in high risk for T2DM. Methods: A systematic review and network meta-analysis of randomized controlled trials. Electronic literature search of PubMed, Medline, and the Cochrane Library for studies published up to January 2014. Randomized controlled trials of lifestyle education and other therapies (metformin) in high risk of T2DM with a follow-up of at least 6 months, reporting onset of T2DM, levels of 2hPD, and HbA1c. Pairwise meta-analyses and Bayesian network meta-analysis combined direct and indirect evidence to estimate the relative effects between treatments were used. Results and discussion: Overall estimates by the traditional meta-anal- ysis were calculated using a random-effects model. Under the several as- sumptions, the network meta-analysis was conducted. Those estimates were examined by several models. In the presentation, results of further analyses will be also presented.   P4.2 Modeling infectious diseases P4.2.24 A predictive model for HIV spreading in hyper-endemic settings S Blaizot1,2,3 , B Riche1,2,3 , D Maman4 , J-F Etard4,5 , R Ecochard1,2,3 1 Hospices Civils de Lyon, Service de Biostatistique, Lyon, France, 2 Université de Lyon, Université Lyon 1, Lyon, France, 3 CNRS UMR5558, LBBE, Equipe Biostatistique-Santé, Lyon, France, 4 Epicentre, Paris, France, 5 Institut de Recherche pour le Développement/UMI 233, Montpellier, France   We present a modelling approach of HIV spreading in hyper-endemic set- tings taking into account the sex and age structures of the population, the natural progression of HIV through different CD4-cell-count stages, and the use of antiretrovirals. The specificity of this approach is the use of the same data for both estimation and prediction. The approach involves estimating the model parameters with a statistical method that uses likelihood decomposition. In short, the complete model is split into several simple sub-models. The predictive model was formu- lated as a system of sex- and age-specific differential equations and was adapted to include prospective medical interventions such as increasing the coverages of antiretroviral therapy or male circumcision. The use of the same data for estimation and prediction, by avoiding trans- portability bias, allows obtaining, to some extent, more reliable results. The model is able to take into account the heterogeneity in volume and parameter values by sex and age, which is of great interest when inter- ventions such as circumcision or pre-exposure prophylaxis are modelled. The approach can be applied to more complex models that would like to capture additional information and to data from Demographic and Health Surveys, available in many Sub-Saharan African countries. In the near future, the model will be applied to a recently designed house- hold survey: the “Ndhiwa HIV Impact in Population Survey” (NHIPS). This survey is a district-representative cross-sectional population survey con- ducted in the district of Ndhiwa (Nyanza Province, Kenya). P4.2.58 A dynamic regression analysis of pulmonary tuberculosis incidence D Gomes1 , AL Brás2 , PA Filipe1 , B de Sousa3 , C Nunes4 1 University of Évora, School of Sciences and Technology, Évora, Portugal, 2 Faculty of Veterinary Medicine, University of Calgary, Calgary, Canada, 3 University of Coimbra, Fac. of Psych. and Educ. Sciences, Coimbra, Portugal, 4 Nova University of Lisbon, National School of Public Health, Lisbon, Portugal   Tuberculosis is a global health concern, being the second cause of death from an infectious disease worldwide. Tuberculosis infection is highly as- sociated with airborne transmission, hence, Pulmonary Tuberculosis (PTB) being the most common form is of special importance in Public Health. Understanding what characterizes patients who suffer from Pulmonary Tuberculosis is crucial when establishing screening strategies to better control of the disease. Monthly PTB incidence rates for Portugal mainland were analyzed, from 2000 to 2010. Two official data sources were used: the National Program for Tuberculosis Control (PNT) provided the number of PTB cases by date of diagnose, and the Statistics Portugal provided estimates of the annual population at risk for the period at study. Dynamic Regression analysis was applied to model PulmonaryTuberculosis incidence. Variables such as sex, age, being alcoholic, diagnosed with HIV, being a smoker, an inmate and/ or a homeless are some of the factors that can contribute for the diagnosis of PulmonaryTuberculosis and thus were used as covariates in our models. The key goal of all Tuberculosis control programs is reducing the trans- mission of the infectious agent within the community. The transmission of Tuberculosis from one individual to another occurs mainly due to sputum positive Pulmonary Tuberculosis form. Therefore, understanding Pulmonary Tuberculosis cases and comprehend the patterns of the dis- ease is especially important for the control of a Tuberculosis endemic. P4.2.66 Modelling infectious disease parameters using serological data SA Herzog1 , A Berghold1 1 Inst Med Info, Stat & Docu, Medical University of Graz, Graz, Austria   A challenge in modelling and investigating infectious diseases is to use the “correct” force of infection. A special issue arises for infectious agents which are transmitted only indirectly through environmental sources like contaminated food, animal, or ground. Infected individuals can possibly be identified but these individuals are not infectious. However, serologi- cal data provide information about whether or not a person has been in- fected before the time point at which the serological sample was taken. Age-specific prevalence and force of infection can be estimated using serological data under the assumption of lifelong immunity and that the epidemic is at equilibrium. If the epidemic is not at equilibrium age- and time-specific prevalence and force of infection should be estimated. We will apply and compare these methods to serological data on toxoplasma infection which are routinely collected by screening pregnant women in Austria (1991-2012). Furthermore, we investigate how repeated measure- ments and information of seroconversion can influence the results.

Pages Overview