Dear all,
Another seminar in our series of statistics talks for people interested in quantitative methods in biology and medicine will take place on Tuesday 15 May 2018 at 11:00, in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, 1010 Lausanne, M2: Vennes).
The speaker will be Aziz Chaouch from the Institute of Social and Preventive Medicine, University of Lausanne.
Title: Meta-analysis of heterogeneous predictive models based on Monte Carlo simulations and the concept of a meta-model
Abstract: Predictive models are being increasingly used in clinical practice e.g. to predict future drug plasma concentrations in subpopulations of patients as a function of the dosage regimen. When different studies developed predictive models for the same outcome, one would like to combine evidence from these models to optimize the quality of predictions. One problem is that the different studies may have used different predictive models with e.g. different sets of predictors. We shall present a simulation approach to allow a meta-analysis of such heterogeneous predictive models. The method postulates a hypothetic model (called the meta-model) which supposedly approximates the data generating mechanism common to all studies. By generating pseudo-observations from the meta-model and refitting the possibly misspecified models to these pseudo-data, it is possible to estimate meta-model parameters using an iterative procedure. Statistical properties of the method will be presented in the context of linear and linear mixed models.
This is a joint work with Thierry Buclin and Valentin Rousson
Hope to see you there!
Valentin
(for the organizers: Zoltán Kutalik, Valentin Rousson, Frédéric Schütz)
CHUV
centre hospitalier universitaire vaudois
Valentin ROUSSON - Professeur associé, mathematicien/statisticien
Département universitaire de medecine et santé communautaires (DUMSC)
Médecine sociale et préventive (IUMSP)
Biostatistique et méthodes quantitatives
BIO 2/2ème/103
Rte de la Corniche 10, CH - 1010 Lausanne
+41 (0)21 314 73 28 TEL
Valentin.Rousson(a)chuv.ch<mailto:Valentin.Rousson@chuv.ch>
www.chuv.ch<http://www.chuv.ch/>
Dear all,
Another seminar in our series of statistics talks for people interested in quantitative methods in biology and medicine will take place on Monday 16 April 2018 at 15:00, in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, 1010 Lausanne, M2: Vennes).
The speaker will be Prof. Tianxi Cai from the Department of Biostatistics at the Harvard School of Public Health.
Title: Enabling Imprecise EHR Data for Precision Medicine
Abstract: While clinical trials remain a critical source for studying disease risk, progression and treatment response, they have limitations including the generalizability of the study findings to the real world and the limited ability to test broader hypotheses. In recent years, due to the increasing adoption of electronic health records (EHR) and the linkage of EHR with specimen bio-repositories, large integrated EHR datasets now exist as a new source for translational research. These datasets open new opportunities for deriving real-word, data-driven prediction models of disease risk and progression as well as unbiased investigation of shared genetic etiology of multiple phenotypes. Yet, they also bring methodological challenges. For example, obtaining validated phenotype information, such as presence of a disease condition and treatment response, is a major bottleneck in EHR research, as it requires laborious medical record review. A valuable type of EHR data is narrative free-text data. Extracting accurate yet concise information from the narrative data via natural language processing is also challenging. In this talk, I'll discuss various statistical and informatics methods that illustrate both opportunities and challenges when using EHR data for translational research. These methods will be illustrated using EHR data from Partner's Healthcare.
Hope to see you there!
Valentin
(for the organizers: Zoltán Kutalik, Valentin Rousson, Frédéric Schütz)
CHUV
centre hospitalier universitaire vaudois
Valentin ROUSSON - Professeur associé, mathematicien/statisticien
Département universitaire de medecine et santé communautaires (DUMSC)
Médecine sociale et préventive (IUMSP)
Biostatistique et méthodes quantitatives
BIO 2/2ème/103
Rte de la Corniche 10, CH - 1010 Lausanne
+41 (0)21 314 73 28 TEL
Valentin.Rousson(a)chuv.ch<mailto:Valentin.Rousson@chuv.ch>
www.chuv.ch<http://www.chuv.ch/>
Dear all,
The next seminar in our series of statistics talks for people interested in quantitative methods in biology and medicine will take place on Thursday 12 April 2018 at 11:00, in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, 1010 Lausanne, M2: Vennes).
The speaker will be Elisabeth Dahlqwist, PhD Student at the « Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm », currently visiting Zoltan Kutalik's group at the IUMSP.
Title: Regression standardization and attributable fraction estimation with between-within frailty models for clustered survival data
Abstract: In epidemiological studies the relationship between an exposure and the time to some outcome is usually presented as a hazard ratio. However, the hazard ratio is a relative measure that does not give any information about the impact of the exposure on the time to the outcome on a population level. In public health and policy making interest is usually in how the prevalence of some disease can be reduced by some intervention targeted at reducing an exposure that we expect to cause the disease. The attributable fraction (or attributable risk) and standardization are useful measures in public health since it inform about the population impact of an exposure on an outcome. In this study we analyze the relationship between preterm birth (born before week 37 of gestation) and the time-to ADHD diagnosis. We expect that the relationship between preterm birth and ADHD is confounded by unobserved familial factors such as genes and childhood environment. In the presence of unobserved confounding we cannot estimate the causal effect of preterm birth on ADHD. A solution is to use family data and compare siblings born by the same parents but with discordant status in preterm birth. The main method that is used to adjust for family shared confounders in survival data is the stratified Cox model. However, the stratified Cox model cannot estimate absolute effects and thus, we cannot use the model to estimate standardized survival curves or the attributable fraction function. An alternative method is the frailty model (which parallels the random effect model for point outcomes). The limitation with the ordinary frailty model is that it does not adjust for family shared unobserved confounding. However, the between-within model adjust for family shared unobserved confounding and can be used for standardization. In this work we present how the between-within model can be used for standardization and the estimation of the attributable fraction. As an illustration of the difference between the stratified Cox model, the frailty model and the BW model we use data from Swedish registries in a within-mother-between-pregnancy analysis of preterm birth on the time-to-ADHD diagnosis/medication. We also estimate standardized survival curves and the attributable fraction function for preterm birth and ADHD.
Hope to see you there!
Valentin
(for the organizers: Zoltán Kutalik, Valentin Rousson, Frédéric Schütz)
CHUV
centre hospitalier universitaire vaudois
Valentin ROUSSON - Professeur associé, mathematicien/statisticien
Département universitaire de medecine et santé communautaires (DUMSC)
Médecine sociale et préventive (IUMSP)
Biostatistique et méthodes quantitatives
BIO 2/2ème/103
Rte de la Corniche 10, CH - 1010 Lausanne
+41 (0)21 314 73 28 TEL
Valentin.Rousson(a)chuv.ch<mailto:Valentin.Rousson@chuv.ch>
www.chuv.ch<http://www.chuv.ch/>
