- statslausanne - lists.sib.swiss

two short courses in statistics
by Rousson Valentin 26 Oct '16

26 Oct '16

Dear list members, We shall have two short courses in Statistics involving Stata by Dr. Rose Medeiros on Friday, November 18 on the Biopôle site in Lausanne-North (see the flyer enclosed). First course (10:30-12:00): Analyzing Data with Missing Values in Stata Using Multiple Imputation Second course (14:15-15:45) : Multilevel/Mixes-effects Models in Stata Place : Centre des laboratoires d'Epalinges, chemin des Boveresses 155, 1066 Epalinges Participation is free of charge but registration is mandatory until November 10 via email at formations.sante(a)chuv.ch. Hope to see you there! Patrick and Valentin

1 0

Statistical forum on November 10 at 11:00
by Rousson Valentin 26 Oct '16

26 Oct '16

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 10 November 2016 at 11h00, in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, 1010 Lausanne, M2: Vennes). The speaker will be Dr. Kaspar Rufibach, statistician by Roche, Basel, Switzerland. Title: Bayesian predictive power: update after a blinded interim analysis and choice of prior Abstract: Bayesian predictive power is the expectation of the probability to meet the primary endpoint of a clinical trial, or any statistical test, at the final analysis, where expectation is computed with respect to a prior distribution over the true underlying effect. Bayesian predictive power is thus a way of quantifying the success probability for the trial sponsor while the trial is still running. The existing framework typically assumes that once the trial is not stopped at an interim analysis, Bayesian predictive power is updated with the resulting interim estimate. However, in blinded Phase III trials, typically an independent committee looks at the data and no effect estimate is revealed to the sponsor after passing the interim analysis. Instead, the sponsor only knows that the effect estimate was between predefined futility and efficacy boundaries. In the first part, we show how Bayesian predictive power for a time-to-event endpoint can be updated based on such knowledge only. In the literature it is often suggested to use a Normal prior. In the second part, we show that, unless the prior sample size is large compared to the Phase III trial, using a Normal prior leads to potentially undesired features of Bayesian predictive power that seem to be generally overlooked in applications. This is joint work with Markus Abt, Hans Ulrich Burger, and Paul Jordan (all Roche, Basel). Hope to see you there! Valentin (for the organizers: Jérôme Goudet, Valentin Rousson, Frédéric Schütz) CHUV centre hospitalier universitaire vaudois Valentin ROUSSON - Professeur mathematicien/statisticien Département universitaire de medecine et santé communautaires (DUMSC) Médecine sociale et préventive (IUMSP) Biostatistique et méthodes quantitatives +41 (0)21 314 73 28 TEL Valentin.Rousson(a)chuv.ch<mailto:Valentin.Rousson@chuv.ch> www.chuv.ch<http://www.chuv.ch/>

1 0

Statistical forum on October 6 at 11:00
by Rousson Valentin 26 Sep '16

26 Sep '16

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 6 October 2016 at 11h00, in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, 1010 Lausanne, M2: Vennes). The speaker will be Prof. Rafael Lalive from the University of Lausanne (HEC) Title: How do Pension Wealth Shocks affect Working and Claiming? Abstract: We study how people respond to pension wealth shocks in Switzerland, where working while claiming a pension is not discouraged. A 1997 reform increased women's full retirement age (FRA) by two years, from 62 to 64, lowering pension wealth by 10.4%, and the reform introduced early claiming, at age 62, reducing pensions by 3.4%, then by 6.8%, the actuarially fair rate. We find that women delay pension claiming by about 7 to 8 months, and delay labor market exit by 5 to 6 months. Doubling the price of early retirement delays pension claiming by 4 to 5 months and has no effects on labor market exit. Women buffer about 35% of the wealth shock triggered by an FRA increase, mostly through working longer, and 100% of the wealth shock due to doubling the early retirement price. Most women either delay both, labor market exit and benefit claiming, or none of the two, suggesting that costs to delay claiming and labor market exit are positively correlated. This is a joint work with Stefan Staubli (University of Calgary) Hope to see you there! Valentin (for the organizers: Jérôme Goudet, Valentin Rousson, Frédéric Schütz) CHUV centre hospitalier universitaire vaudois Valentin ROUSSON - Professeur mathematicien/statisticien Département universitaire de medecine et santé communautaires (DUMSC) Médecine sociale et préventive (IUMSP) Biostatistique et méthodes quantitatives +41 (0)21 314 73 28 TEL Valentin.Rousson(a)chuv.ch<mailto:Valentin.Rousson@chuv.ch> www.chuv.ch<http://www.chuv.ch/>

1 0

statistical forum on September 22 at 11:00
by Rousson Valentin 07 Sep '16

07 Sep '16

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 22 September 2016 at 11h00, in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, 1010 Lausanne, M2: Vennes). The speaker will be Dr. Qing Li from the University of Sherbrooke, Canada. Title: Disentangling causal effect of socio-economic status on health for aged Quebeckers Abstract: While the socio-economic status-health gradient is well-known, it is more difficult to estimate causality. Using data of Quebeckers aged 67+, we model observed income as a sum of time-invariant permanent income and other components that vary with time, and estimate causal effect of permanent income (and other exogenous covariates) on four health dimensions namely physiology, chronic disease, depression and cognition. We find that permanent income have the strongest effect on cognition and the effect is stronger for the low income group. The causality from permanent income to health exists in other health dimensions though it is not as strong as in the cognition dimension. The causality from health to observed income is not statistically significant in general. We compare our results with a model where individual fixed effect is eliminated with a differencing approach; the results support our modeling of time-invariant permanent income for aged people. This seminar is organized in collaboration with the platform "Economie de la Santé" and the seminar « Health and Labour Economics » from the Faculty of Business and Economics, University of Lausanne. Hope to see you there! Valentin (for the organizers: Jérôme Goudet, Valentin Rousson, Frédéric Schütz) CHUV centre hospitalier universitaire vaudois Valentin ROUSSON - Professeur mathematicien/statisticien Département universitaire de medecine et santé communautaires (DUMSC) Médecine sociale et préventive (IUMSP) Biostatistique et méthodes quantitatives +41 (0)21 314 73 28 TEL Valentin.Rousson(a)chuv.ch<mailto:Valentin.Rousson@chuv.ch> www.chuv.ch<http://www.chuv.ch/>

1 0

two statistical seminars on June 27 and June 30
by Rousson Valentin 20 Jun '16

20 Jun '16

Dear all, I would like to draw your attention to two statistical seminars which will take place next week in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, 1010 Lausanne, M2: Vennes). ------------------------------------------------------------------------------------------------------------------------------------ Monday 27 June 11:00: The "Missing Cause" Approach to detect and reduce bias due to unmeasured confounding in pharmacoepidemiology. By Prof. Michal Abrahamowicz, Department of Epidemiology and Biostatistics, McGill University, Montréal, Canada (organized by Arnaud Chiolero, IUMSP). Objectives: To propose a new method for reducing the impact of unobserved confounding in large observational pharmacoepidemiological studies [1], validate it in simulations and apply it in a real-life drug safety study. Methods: Under assumptions similar to the prescribing preferences-based instrumental variable (IV) approach, our new "missing cause" method relies on discrepancies between a) treatment actually received by individual patients and b) treatment they would be expected to receive based on their observed characteristics and their physicians' prescribing preferences. Specifically, we use the treatment-by-discrepancy interaction to test for presence of unmeasured confounding and correct the treatment effect estimate for the resulting bias. We implement the method within risk difference (RD) multivariable linear regression modeling of the effect of a binary exposure on a binary outcome. We use simulations to compare our missing-cause estimates with (i) conventional estimates, adjusted only for measured covariates, and (ii) IV estimates. We apply the method to compare gastrointestinal safety of COX-2 inhibitors vs. traditional NSAIDs in a large cohort of elderly NSAID users in Quebec, Canada, and to assess the potential benefits of influenza vaccine in preventing influenza. Results: In simulations, our estimates had four times smaller bias than conventional estimates, much smaller variance than unbiased but unstable IV estimates, and uniformly best overall accuracy (lowest mean square error). In the first application, our method suggested a slight reduction of gastrointestinal risks for COX-2 inhibitors (RD=-2.6% (-8.4, 3.2%)), in contrast to risk increase suggested by IV estimates, with much wider CIs. In the second application, the method detected very serious bias, due to unmeasured confounding by indication, which resulted in the conventional estimates, adjusted only for the confounders measured in the study database, suggesting that influenza vaccine is associated with an increase risk of influenza. In contrast, once we used the missing cause method to correct for this bias, the corrected estimate indicated a significant risk reduction. Conclusion: Our "missing cause" method provides a potentially useful alternative to IV approaches for reducing unmeasured confounding bias in pharmacoepidemiology. ------------------------------------------------------------------------------------------------------------------------------------ Thursday 30 June, 11:00: Appropriate methods for rare outcomes and/or rare exposures. By Prof. Anirban Basu, Departments of Pharmacy, Health Services and Economics, University of Washington, Seattle (organized by Prof. Mark Dusheiko, IUMSP, in collaboration with the platform "Economie de la Santé" and the seminar « Health and Labour Economics » from the Faculty of Business and Economics, University of Lausanne). Abstract: Using Monte-Carlo simulations, we compare the two-stage least-squares (2SLS) estimator with two-stage residual inclusion (2SRI) estimators, with varying forms of residuals, to estimate the local average treatment effect parameter for a binary outcome and endogenous binary treatment model in the presence of binary covariates and a binary instrumental variable. We vary the rarity of both the outcome and the treatment and find different estimators to produce the least bias in different settings. We develop guidance for applied researchers and illustrate the utility of this guidance with estimating the effects of long-term care insurance on a variety of binary health care use outcomes among the near-elderly using the Health and Retirement Study. We would be pleased to see you there! Valentin CHUV centre hospitalier universitaire vaudois Valentin ROUSSON - Professeur mathematicien/statisticien Département universitaire de medecine et santé communautaires (DUMSC) Médecine sociale et préventive (IUMSP) Biostatistique et méthodes quantitatives +41 (0)21 314 73 28 TEL Valentin.Rousson(a)chuv.ch<mailto:Valentin.Rousson@chuv.ch> www.chuv.ch<http://www.chuv.ch/>

1 0

Statistical forum on May 19 at 11:00
by Rousson Valentin 09 May '16

09 May '16

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 19 May 2016 at 11h00, in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, 1010 Lausanne, M2: Vennes). The speaker will be Dr. Arthur Allignol from the Institute of Statistics, University of Ulm, Germany. Title: Application of multistate models in hospital epidemiology Abstract: The use of multistate models has started to rise in hospital epidemiology. This is motivated by the fact that they provide the tools both to show that timing of events must be modelled and to actually model this timing. Indeed, when considering hospital-acquired infections (HI), patients are usually free of infection when they are admitted to hospital but may acquire one as time progresses. HI constitutes a major complication that can be severe both in terms of mortality and morbidity and often leads to a prolonged length of stay (LoS) in the hospital, which is one of the main driver for extra costs due to infections. An estimator of the excess LoS associated with an infection can be obtained as a functional of the matrix of transition probabilities, which is in turn based on the transition hazards. The use of the flexible pseudo observation regression technique for the excess LoS was recently proposed. We illustrate the use of these methods with a recent study on HI consisting of more than 600,000 patients admitted in 10 west-European hospitals between 2011 and 2012. Hope to see you there! Valentin (for the organizers: Jérôme Goudet, Valentin Rousson, Frédéric Schütz) CHUV centre hospitalier universitaire vaudois Valentin ROUSSON - Professeur mathematicien/statisticien Département universitaire de medecine et santé communautaires (DUMSC) Médecine sociale et préventive (IUMSP) Biostatistique et méthodes quantitatives +41 (0)21 314 73 28 TEL Valentin.Rousson(a)chuv.ch<mailto:Valentin.Rousson@chuv.ch> www.chuv.ch<http://www.chuv.ch/>

1 0

statistical forum on April 7 at 11:00
by Rousson Valentin 29 Mar '16

29 Mar '16

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 7 April 2016 at 11h00, in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, 1010 Lausanne, M2: Vennes). The speaker will be Prof. Michael Lechner from the Department of Economics, University of St. Gallen, Switzerland. Title: The finite sample performance of inference methods for propensity score matching and weighting estimators Abstract: This paper investigates the finite sample properties of a range of inference methods for propensity score-based matching and weighting estimators frequently applied to evaluate the average treatment effect on the treated. We analyze both asymptotic approximations and bootstrap methods for computing variances and confidence intervals in our simulation design, which is based on large scale labor market data from Germany and varies with respect to treatment selectivity, effect heterogeneity, the share of treated, and the sample size. The results suggest that in general, the bootstrap procedures dominate the asymptotic ones in terms of size and power for both matching and weighting estimators. Furthermore, the results are qualitatively quite robust across the various simulation features. Hope to see you there! Valentin (for the organizers: Jérôme Goudet, Valentin Rousson, Frédéric Schütz) CHUV centre hospitalier universitaire vaudois Valentin ROUSSON - Professeur mathematicien/statisticien Département universitaire de medecine et santé communautaires (DUMSC) Médecine sociale et préventive (IUMSP) Biostatistique et méthodes quantitatives +41 (0)21 314 73 28 TEL Valentin.Rousson(a)chuv.ch<mailto:Valentin.Rousson@chuv.ch> www.chuv.ch<http://www.chuv.ch/>

1 0

statistical forum on March 24 at 15:00
by Rousson Valentin 14 Mar '16

14 Mar '16

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 24 March 2016. Exceptionally, it will take place at 15h00, in the room C4 at the level -1 from the Biopôle 1 (Route de la Corniche 2, 1066 Epalinges, M2 stop: Les Croisettes). The speaker will be Prof. Martin Huber from the Department of Economics, University of Fribourg, Switzerland. Title: Endogeneity and non-response bias in treatment evaluation - nonparametric identification of causal effects by instruments Abstract: This paper proposes a nonparametric method for evaluating treatment effects in the presence of both treatment endogeneity and attrition/non-response bias, using two instrumental variables. Making use of a discrete instrument for the treatment and a continuous instrument for non-response/attrition, we identify the average treatment effect on compliers as well as the total population and suggest non- and semiparametric estimators. We apply the latter to a randomized experiment at a Swiss University in order to estimate the effect of gym training on students'self-assessed health. The treatment (gym training) and attrition are instrumented by randomized cash incentives paid out conditional on gym visits and by a cash lottery for participating in the follow-up survey, respectively. Hope to see you there! Valentin (for the organizers: Jérôme Goudet, Valentin Rousson, Frédéric Schütz) CHUV centre hospitalier universitaire vaudois Valentin Rousson - Division biostatistique et méthodes quantitatives IUMSP - Institut universitaire de médecine sociale et préventive BIO 2/2ème/103 Rte de la Corniche 10, CH - 1010 Lausanne +41 (0)21 314 7328 valentin.rousson(a)chuv.ch www.iumsp.ch<http://www.iumsp.ch/>

1 0

STATISTICAL FORUM - CHANGE OF DATE!
by Rousson Valentin 13 Nov '15

13 Nov '15

Dear all, There is a change of date about the next seminar in our series of statistics talks for people interested in quantitative methods in biology and medicine! It will ultimately take place on Thursday 10 December 2015 at 11h00, in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, Lausanne, M2 stop: Vennes). The speaker will be Mark Hannay from the Research Center for Statistics, School of Economics and Management, University of Geneva. Title: Confidence sets in model selection Abstract: At first glance the goals of model selection might seem clear. Out of a list of possible models we want to select the "best" or a subset of "best" models. This notion of "best" however is not well defined, since it obviously depends on the initial goals of the selection. In order to study the variability of model selection, we introduce a new definition of a model, where they are no longer defined through exact zeroes components but through negligible components. Then inspired by confidence intervals for estimation of parameters, we propose a method to build confidence sets for model selection in a parametric setting, i.e. create sets of models within which the "true"model is included with a certain confidence. This allows us to perform inference on model selection. We discuss the computational challenges with such a method, how to find p-values (for the model), consistency in model selection and through a data set show the implications of this new method. Key words: confidence sets of model selection, consistent model selection, variability in model selection Hope to see you there! Valentin (for the organizers: Jérôme Goudet, Valentin Rousson, Frédéric Schütz) CHUV centre hospitalier universitaire vaudois Valentin Rousson - Division biostatistique et méthodes quantitatives IUMSP - Institut universitaire de médecine sociale et préventive BIO 2/2ème/103 Rte de la Corniche 10, CH - 1010 Lausanne +41 (0)21 314 7328 valentin.rousson(a)chuv.ch<mailto:valentin.rousson@chuv.ch> www.iumsp.ch<http://www.iumsp.ch/>

1 0

statistical forum on November 19 at 11:00
by Rousson Valentin 13 Nov '15

13 Nov '15

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 19 November 2015 at 11h00, in the Salle Delachaux (Biopôle 2, first floor, Route de la Corniche 10, Lausanne, M2 stop: Vennes). The speaker will be Mark Hannay from the Research Center for Statistics, School of Economics and Management, University of Geneva. Title: Confidence sets in model selection Abstract: At first glance the goals of model selection might seem clear. Out of a list of possible models we want to select the "best" or a subset of "best" models. This notion of "best" however is not well defined, since it obviously depends on the initial goals of the selection. In order to study the variability of model selection, we introduce a new definition of a model, where they are no longer defined through exact zeroes components but through negligible components. Then inspired by confidence intervals for estimation of parameters, we propose a method to build confidence sets for model selection in a parametric setting, i.e. create sets of models within which the "true"model is included with a certain confidence. This allows us to perform inference on model selection. We discuss the computational challenges with such a method, how to find p-values (for the model), consistency in model selection and through a data set show the implications of this new method. Key words: confidence sets of model selection, consistent model selection, variability in model selection Hope to see you there! Valentin (for the organizers: Jérôme Goudet, Valentin Rousson, Frédéric Schütz) CHUV centre hospitalier universitaire vaudois Valentin Rousson - Division biostatistique et méthodes quantitatives IUMSP - Institut universitaire de médecine sociale et préventive BIO 2/2ème/103 Rte de la Corniche 10, CH - 1010 Lausanne +41 (0)21 314 7328 valentin.rousson(a)chuv.ch www.iumsp.ch<http://www.iumsp.ch/>

1 0