An Introduction to R for EpidemiologistsSession Chair(s): Steve MooneyDate: 2019-06-18 Time: 8:30 am - 12:30 pmLocation:
This workshop will introduce participants to the R statistical computing platform for use in epidemiologic analysis. It is not intended to transform untested novices into R wizards in a mere half-day; rather, the goal will be to introduce the conceptual underpinnings, tools, and external resources that participants will need to overcome barriers to using R that they might encounter on their own, later. The material is designed for epidemiologists who are already familiar performing analyses using other statistical software (e.g. SAS/Stata/SPSS) but who have no first-hand experience with the R language. More specifically, the course will cover 1) basic R syntax, 2) importing data, 3) constructing, cleaning, and manipulating data objects, 4) loading and using external packages, 5) simple statistical modeling, and 6) graphics. Participants must bring a laptop with R installed; the instructor will be available by email beforehand to assist with R installation if difficulties arise.
Session Chairs:Steve Mooney, University of Washington
Confounding control for estimating causal effects: Looking under the hoodSession Chair(s): Nicolle Gatto, Ulka CampbellDate: 2019-06-18 Time: 8:30 am - 12:30 pmLocation:
This workshop introduces concepts of causal inference and confounding control for causal effect estimation. We will introduce potential outcomes, and articulate the conceptual basis and assumptions for two g-methods - standardization via g-computation and inverse probability weighting. Starting with a simple point-treatment setting we will explore how these methods estimate a causal effect, comparing them to more conventional techniques such as multivariable regression and propensity score control. We will then build to the more complex scenario of time-dependent confounding. Participants will learn how to apply these methods in SAS and R using an observational dataset with the primary goal of unpacking any “black boxes” to clarify the links among the causal effect of interest, the mechanics of these g-methods, and the programming code.
Session Chairs:Nicolle Gatto, Pfizer Inc.Ulka Campbell, Pfizer Inc.
Critical review and preparation of manuscripts reporting epidemiologic findingsSession Chair(s): Moyses SzkloDate: 2019-06-18 Time: 8:30 am - 12:30 pmLocation:
Overview abstract: In this half-day workshop, participants will critically review a paper as initially submitted to the American Journal of Epidemiology (AjE), but not yet published. The paper will be sent to participants in advance of the workshop for their critical review. During the workshop, a presentation will be made on some of the main points to be considered when preparing or reviewing a manuscript. Small group work will follow the presentation so that participants can compare their reviews and prepare a consolidated list of critical comments on the paper. Each group will designate a rapporteur, who will present the group’s review of the paper and decision letter to the whole group of participants. At the end of the workshop, students will receive copies of the manuscript’s AjE reviews, the initial editorial decision and the final version of the paper. Attendance will be limited to 30 participants.
Session Chairs:Moyses Szklo, Johns Hopkins School of Public Health
Implementing G Methods in Complex Longitudinal DataSession Chair(s): Ashley NaimiDate: 2019-06-18 Time: 8:30 am - 12:30 pmLocation:
Applied health scientists are increasingly dealing with complex data structures to answer questions about exposure effects and mediation. In such settings, feedback between confounders, exposures, and mediators render standard adjustment methods (regression, restriction, stratification, matching) inappropriate. G methods, which consist of the parametric g formula, g estimation of structural nested models, and inverse probability weighted marginal structural models, can be used to quantify a variety of exposure effects with both simple and complex data structures. After a brief introduction to the potential outcomes framework, I will review obstacles to effect estimation with complex longitudinal data. All three g methods will then be introduced in a series of examples using actual data and software code (SAS, Stata, R). No prior knowledge of causal modeling, counterfactuals, or g methods is required.
Session Chairs:Ashley Naimi, University of Pittsburgh
Data manipulation, visualization, and reproducible documents with R and the TidyverseSession Chair(s): Malcolm Barrett, Corinne RiddellDate: 2019-06-18 Time: 8:30 am - 12:30 pmLocation:
Recent developments by the R community have revolutionized the data analysis pipeline in R, from manipulating and visualizing data to communicating results. Our workshop will provide hands-on training in tools from the tidyverse ecosystem, using real epidemiologic data. In the first section, we will teach data manipulation with dplyr, a package that makes data cleaning easy, flexible, and enjoyable. In the next section, we will teach data visualization with ggplot2, the most popular plotting package in R, with a focus on creating publication-quality plots. We will then put these tools together to make reproducible documents. Using R Markdown, we will weave code and text together and learn to write papers and reports, exported to PDF, Word, or HTML, entirely in R. This workflow easily propagates upstream changes to data or analyses throughout a document and eliminates copy and paste errors. Together, these tools form a data analysis pipeline for reproducible, publication-ready work.
Session Chairs:Malcolm Barrett, University of Southern CaliforniaCorinne Riddell, University of California, Berkeley
Hone your soft skills and own the job marketSession Chair(s): Timothy Lash, Lauren McCulloughDate: 2019-06-18 Time: 8:30 am - 12:30 pmLocation:
“As a growing number of people with similar talents and education compete for the same jobs…soft skills become the differentiating factor separating one person from another” (Smith, 2011). Epidemiologists are well-trained in the knowledge and skills of the profession, yet have few opportunities to learn the soft skills that will most influence their career advancement. This workshop’s objectives are to introduce the importance of soft skills for career success, provide initial training in three soft skills, and suggest tools for self-study. The three soft skills will be: (a) introducing yourself (the “who am I” elevator pitch), (b) planning and time management, and (c) negotiation. For each, we will explain the importance of the skill for career advancement and conduct an exercise to initiate learning about the skill. The workshop will end with a description of additional soft skills, the reasons they are important, and resources for self-learning after the workshop concludes.
Session Chairs:Timothy Lash, Emory UniversityLauren McCullough, Emory University
E-values, Unmeasured Confounding, Measurement Error, and Selection BiasSession Chair(s): Tyler VanderWeeleDate: 2019-06-18 Time: 1:00 pm - 5:00 pmLocation:
The workshop will consider sensitivity analysis for different forms of bias in epidemiology. It will begin with confounding, focusing on a new metric to evaluate sensitivity to unmeasured confounding called the E-value. The E-value is the minimum strength of association, on the risk ratio scale, that an unmeasured confounder would need to have with both the exposure and the outcome, conditional on the measured covariates, to fully explain away the exposure-outcome association. E-value calculations for risk ratios, outcomes differences, odds ratios, and hazard ratios will be discussed. The E-value can be calculated in a straightforward way from study results and its use could help unify assessment of unmeasured confounding. The workshop will proceed by describing very recent analogous easy-to-implement approaches to also address differential measurement error and selection bias. The methods, taken as a whole, will constitute a straightforward comprehensive approach to bias analysis.
Session Chairs:Tyler VanderWeele, Harvard T.H. Chan School of Public Health
Communicating research findings to the media and publicSession Chair(s): Jennifer LoukissasDate: 2019-06-18 Time: 1:00 pm - 5:00 pmLocation:
Science advances when independent researchers publish their results. In addition to the written discourse maintained by the scientific journals, a parallel discussion often takes place in traditional (print, radio, television) and social media (Twitter, Reddit, etc.). Reporters are often interested in covering findings from epidemiological studies. To be effective in the important job of giving a good interview, scientists must prepare for interviews by writing 'key messages', participating in media training and mock interviews. This 4-hour workshop will review key principles of science communication as they relate to the particular challenges presented in epidemiological studies; participants will engage in several hands-on activities where they will practice skills to best communicate about their own research findings.
Session Chairs:Jennifer Loukissas, National Institute of Health
Using Monte Carlo Simulations for Quantitative Bias AnalysisSession Chair(s): Hailey Banack, Elizabeth Rose MayedaDate: 2019-06-18 Time: 1:00 pm - 5:00 pmLocation:
This workshop will provide a comprehensive overview of using Monte Carlo simulations as a tool for quantitative bias analysis. The content covered in the workshop is intended to strengthen participants’ understanding of the theoretical concepts of bias and technical details of running Monte Carlo simulations. To accomplish this objective we will work through three examples of bias analysis using Monte Carlo simulations: (1) an example of using Monte Carlo simulation procedures for record-level correction for exposure misclassification bias, (2) an simple example of Monte Carlo simulations for unmeasured confounding, and (3) a more complex example of Monte Carlo simulations for quantifying collider stratification bias. At the end of the workshop, participants should walk away with a greater understanding of the 'why' and 'how' of quantitative bias analysis (i.e., why bias analysis is important and how to use Monte Carlo simulation techniques to implement the bias analysis).
Session Chairs:Hailey Banack, McGill UniversityElizabeth Rose Mayeda, University of California, Los Angeles
Deep learning for epidemiologists: An introduction to neural networksSession Chair(s): Kathryn RoughDate: 2019-06-18 Time: 1:00 pm - 5:00 pmLocation:
Deep learning methods are increasingly being applied to problems in the health and medical domains; this workshop is designed to offer an accessible introduction to the basics of deep learning for epidemiologists. There will be a review of fundamental machine learning concepts, including model capacity, underfitting, overfitting, regularization, and performance metrics. This will be followed by an overview of several commonly-used deep learning model architectures, including Convolutional Neural Networks (CNNs) for images and Recurrent Neural Networks (RNNs) for sequences. Finally, we will discuss the evaluation of deep learning models, including critical challenges such as interpretability, transportability, fairness, and implementation. Participants will leave the workshop with the ability to engage with and critically evaluate medical research using deep learning methods.
Session Chairs:Kathryn Rough, Google
Introduction to causal inference for multiple time point exposuresSession Chair(s): Laura Balzer, Maya PetersenDate: 2019-06-18 Time: 1:00 pm - 5:00 pmLocation:
This workshop applies the "causal roadmap" to longitudinal data structures, including survival-type outcomes subject to right-censoring, and to the joint causal effects of multiple interventions, including cumulative effects of treatments over time and controlled direct effects. We will cover longitudinal causal models, identification in the presence of time-dependent confounding; and estimation of joint treatment effects using G-computation, inverse probability weighting, and targeted maximum likelihood estimators (TMLE). During the workshop session, participants will work through the roadmap using an applied example and implement these estimators with the ltmle R package.
Session Chairs:Laura Balzer, University of Massachusetts-AmherstMaya Petersen, University of California, Berkeley
Mendelian randomization: a practical guideSession Chair(s): Neil DaviesDate: 2019-06-18 Time: 1:00 pm - 5:00 pmLocation:
Mendelian randomization (MR) is a method that uses genetic variants as instrumental variables to test the causal effect of a (non-genetic) risk factor on a disease or health-related outcome. This course provides an introduction to the conduct, assumptions, strengths and limitations of MR, including the latest sensitivity analyses. It uses a mixture of lectures and practicals in R and Stata. This is a beginner to intermediate level course. Students will learn about one-sample and two-sample MR, including their assumptions and get practical experience of applying these methods to real data. They will also learn about a range of sensitivity analyses that explore likely violation of the assumptions of MR. Prior experience of using MR is not required, but participants should have an understanding of aetiological epidemiological principles and ideally be working on causal population health questions. Provisional Tutors : Debbie Lawlor , Neil Davies , Rebecca Richmond , Kaitlin Wade , Carolina Borge
Session Chairs:Neil Davies, University of Bristol
How to make a picture worth a thousand words: Effectively communicating your research results using statistical graphicsSession Chair(s): Michael JacksonDate: 2019-06-18 Time: 5:30 pm - 7:30 pmLocation:
Epidemiologists can use statistical graphics to understand our data and to guide us toward correct inferences. Well-designed graphics can also be powerful tools for communicating our study findings. However, while statistical software makes it easy to produce certain types of figures, the default options leave much to be desired. Too often, the result is figures that distract, confuse, or even distort data. In this workshop, participants will first learn the fundamentals of effective data visualization. This includes selecting appropriate chart types, drawing attention to the relevant data, using effective visual cues, and providing helpful context. We will discuss how to put these principles into practice, leading viewers to make comparisons, identify trends, and find meaningful correlations. Finally, we will walk through techniques for going beyond the default settings of various software packages to produce well-designed figures.
Session Chairs:Michael Jackson, Kaiser Permanente
Infographics 101 for public healthSession Chair(s): Sujani SivanantharajahDate: 2019-06-18 Time: 5:30 pm - 7:30 pmLocation:
"Infographics 101 for public health" provides public health professionals the confidence to design infographics. This workshop is a condensed 2-hour hands-on experience of the full online course available on PHSPOT.ca. The modules of the workshop include: a crash course on design (elements of design, core design principles); infographics 101 (elements of an infographic, layouts, public health examples), and the designing process (a 9 step process). Participants will have the opportunity to apply the concepts and recreate an infographic using PowerPoint (a widely available software on all computers) during the workshop, and be introduced to a suite of tools and resources. The workshop is designed for public health professionals interested in using infographics as a knowledge translation product but are not confident in design.
Session Chairs:Sujani Sivanantharajah, PH Spot
Bayesian Multilevel/Longitudinal ModelingSession Chair(s): Chuck HuberDate: 2019-06-18 Time: 5:30 pm - 7:30 pmLocation:
Bayesian statistical methods are increasing in popularity in epidemiology. This talk will briefly review the concepts of multilevel/longitudinal models, introduce the concepts and jargon of Bayesian statistics, and then demonstrate how to fit Bayesian multilevel/longitudinal models using Stata's -bayes- prefix. We will focus on practical issues such as selection of priors for random effects, how to fit models with cross-classified random effects, and Bayesian model selection.
Session Chairs:Chuck Huber, StataCorp
An introduction to predicting exposures and outcomes using omic biomarkersSession Chair(s): Paul Yousefi, Matthew SudermanDate: 2019-06-18 Time: 5:30 pm - 7:30 pmLocation:
Deeply phenotyping participants over several classes of omic markers, including genetic, epigenetic, and metabolomic, is increasingly feasible for epidemiological studies and provides opportunities to develop biomarkers with improved prediction of exposure and ill health. Optimizing the utility of these high dimensional, integrated data sources requires a toolkit of statistical approaches distinct from those typically used in epidemiology to assess correlation and causation. We provide a primer on the essential considerations (e.g. bias, variance, calibration) and evaluation procedures (e.g. performance metrics, cross-validation) required when developing such predictors. We offer context on how such problems have been approached to date (e.g. polygenic risk scores, epigenetic clocks), as well as common pitfalls. Lastly, we introduce how several machine learning techniques can be harnessed to design generalizable and interpretable predictions of complex biological phenomena.
Session Chairs:Paul Yousefi, University of BristolMatthew Suderman, University of Bristol
An introduction to directed acyclic graphs: What you never wanted but needed to know about bias and didn’t even know to ask.Session Chair(s): Ian ShrierDate: 2019-06-18 Time: 5:30 pm - 7:30 pmLocation:
This workshop will introduce participants to directed acyclic graphs (DAGs). We will review the basic principles and show how they can be used to determine appropriate sets of variables for estimating total causal effects of exposure (treatment). Participants will work through concrete examples of increasing complexity. We will also introduce how DAGs can be used in more advanced applications, including natural and controlled direct and indirect effects and study design.
Session Chairs:Ian Shrier, McGill University
A primer on electronic health record research design and analysisSession Chair(s): Neal GoldsteinDate: 2019-06-18 Time: 5:30 pm - 7:30 pmLocation:
Increasingly data mined from the electronic health record (EHR) are being used in epidemiological research. But more data does not equate to better quality research. In this workshop, we will cover the basics of working with EHRs and designing valid epidemiological analyses with these data. Specifically, we will broadly cover the following topics through didactic lecture and interactive group exercises: 1. Architecture of the electronic health record 2. Understanding the clinical population and how this relates to a target/general population 3. Designing epidemiological studies using EHR data 4. Obtaining data from the EHR, including data export, linkage, and variable manipulation (e.g. parsing data from free text) 5. Appropriate analytic designs from these data, including case-control, retrospective cohort, and longitudinal study design 6. Common pitfalls in working with EHR data and resources for additional reference.