Symposia

 

2019 Symposia – specific dates/times will be made available in April.

 




Beyond “Is race a cause?” – Bringing Critical Race Theory into Epidemiology

Session Chair(s):   John Pamplin II,   Zinzi Bailey
Date:       Time: 12:00 am - 12:00 am
Location:


     

Conversations about race at SER often rehash outmoded debates about whether race can be considered a cause under the Potential Outcomes framework. While acknowledging that race is socially constructed, causal questions are nonetheless framed around the possibility of estimating intervention effects of race, which implicitly or explicitly devolve into biological essentialism. The absence of a shared theoretical foundation for conceptualizing race and racism may contribute to this inconsistency. Fortunately, Critical Race Theory, a framework that has animated research in other social sciences for over three decades, has begun to make its way into quantitative health research. The proposed symposium will introduce Critical Race Theory to epidemiologists and feature research informed by this approach. The goal of the symposium is to move past old debates, and explore how Critical Race Theory can inform causal links between racism and disease in epidemiology.

Session Chairs:
John Pamplin II, Columbia University
Zinzi Bailey, University of Miami


Presenters:
Chandra Ford,
"If Race is not Biology, What is it?: Racialization, Social Constructs and Epidemiology"
Zinzi Bailey,
"The Potential of Critical Race Theory in Epidemiology: Using Historical Imagination to Construct Counterfactual Racial Formation and Dynamics"
Sharrelle Barber,
"Embodying Racism: Empirical Evidence from the United States and Brazil"


Diversity & Inclusion: Epidemiology and community stakeholders: Impressions and Experiences

Session Chair(s):   Martha Werler,   Elisa Jean-Baptiste
Date:       Time: 12:00 am - 12:00 am
Location:


     

Toward the reduction of health disparities, epidemiologic research questions are influenced by many factors. One such set of factors is the experiences, impressions, and priorities of community members who would most likely be affected by research results. As epidemiologists, we too often by-pass stakeholders’ views when forming research questions. In this symposium, a panel would comprise 4-5 speakers, each of whom would share their stories about community member inclusion and exclusion. Panel members would include community stakeholders and community participatory epidemiology researchers. A moderator would then lead discussion, with questions and comments from the audience and responses and comments from panel members. This symposium seeks to open conversations on how diversity and inclusion issues can be considered as we frame research agendas and research questions.

Session Chairs:
Martha Werler, BUSPH Epidemiology
Elisa Jean-Baptiste, Student University of Montreal


Presenters:
Ruby Nguyen, University of Minnesota
"Is It Time to Develop an Epidemiologic Triad Model to Support the Theory of Community-Engaged Scholarship?"
Mary Roary, NIH
"Who Is Responsible For Eliminating Health Disparities? What Are You Doing About It?"
Shawnita Sealy-Jefferson, Ohio State
"Listen To and Trust Black Women"
Joseph Yracheta, Missouri Breaks Industries Research, Inc.
"Intersections of Race, Culture, & Sovereignty Hurt American Indian Epidemiology"
Discussants:
Yvette Cozier, Boston University


Family Matters: using family designs to control confounding in observational studies

Session Chair(s):   Mollie Wood,   Katherine Keyes
Date:       Time: 12:00 am - 12:00 am
Location:


     

Family-based study designs are methods that leverage shared, time-stable familial factors to reduce confounding, both known sources of confounding that are difficult to measure, such as genetics and personality, and unknown confounders. The collection of data linking health information within families is proliferating, which has led to a rapid uptake in the epidemiologic literature. However, the assumptions for casual identification using family-based designs are often unarticulated, and these designs are vulnerable to biases from nonshared confounders and measurement error, indeed in some cases more so than designs of unrelated persons. This symposium brings together perspectives from epidemiology, statistics, and behavioral genetics to illustrate the promises and pitfalls of family designs for epidemiological research. The proposed format includes an overview of family methods, followed by three talks from researchers working with these study designs, with ample time for discussion.

Session Chairs:
Mollie Wood, Harvard School of Public Health
Katherine Keyes, Columbia University


Presenters:
Mollie Wood,
"The Anna Karenina Principle in family designs"
Jackie Cohen,
"Paternal medication use as a negative control exposure for pregnancy drug safety studies"
Thomas Frisell,
"Introducing or amplifying bias by using sibling controls: a cautionary note"
Matt McGue,
"On subject of methods concerns in twin designs"
Discussants:
Katherine Keyes,


Future of Epidemiology: Emerging Trends and Perspectives

Session Chair(s):   Andrew Olshan,   Brooke Anderson
Date:       Time: 12:00 am - 12:00 am
Location:


     

Building upon the ongoing collaboration between the Society for Epidemiologic Research and American Journal of Epidemiology collaboration on the future of epidemiology, we propose a symposium that addresses emerging and innovative issues in our field with a perspective on our discipline’s vision for the future of research, practice, and training. The symposium will feature commentaries solicited for a special SER-AJE issue on the “Future of Epidemiology”, addressing the most critical areas for our field and for public health including the role of observational epidemiology, systems modeling, climate change, emerging infections, the nature of our profession, and social factors. Thought leaders in these areas will provide a brief current state-of-knowledge overview, followed by specific future priorities, challenges and recommendations, including a focus on training. The symposium will end with a panel question-and-answer session including all the speakers.

Session Chairs:
Andrew Olshan, University of North Carolina
Brooke Anderson, Colorado State University


Presenters:
Kirsten Bibbins-Domingo, University of California, San Francisco
"Future of Epidemiology: What is the future of observational epidemiology?"
Magdalena Cerda, New York University
"Future of Epidemiology: Systems modeling"
Brooke Anderson, Colorado State University
"Future of Epidemiology: Climate Epidemiology"
Allison Aiello, University of North Carolina
"Future of Epidemiology: An outlook on measurement and methods for studying social determinants"
Art Reingold, University of California, Berkeley
"Future of Epidemiology: Future of Epidemiology: Infectious Disease Epidemiology in the 21st Century"


Genetic Epidemiology in the Causal Era: How Do Genes Affect the Health of Populations and How do we Know?

Session Chair(s):   Jay Kaufman,   Stephanie London
Date:       Time: 12:00 am - 12:00 am
Location:


     

The most significant revolutions in epidemiology over the last 3 decades have been the rise of formal causal reasoning in design and analysis and the growing hegemony of genetic epidemiology over scientific hypotheses and funding. But how do these two paradigms intersect?  What does formal causal reasoning tell us about how to study the effects of genetic variants, and how do concepts such as confounding, selection and information biases apply?   Are SNPs well-defined exposures? Can “Genome-Wide Association Studies” yield causally interpretable effects when they do not necessarily snag causal variants? This symposium will explore and debate population stratification, estimating and interpreting genetic risk scores as exposures, causal effect interpretation under environmental heterogeneity, and other challenges of designing and interpreting studies of the causal effects of variations in DNA sequence.  It’s time to put nature and nurture into the same DAG.

Session Chairs:
Jay Kaufman, McGill University
Stephanie London, NIEHS


Presenters:
Peter Kraft, Harvard University
"Genetic epidemiology as an art of the possible: design and interpretation of genetic association studies"
Christy Avery, UNC
"How studying racially/ethnically diverse populations informs causal inference."
John Witte, UCSF
"Polygenic risk scores: prediction without causality"
Brandon Pierce, University of Chicago
"Studying biological mechanisms of genetic susceptibility using causal reasoning"
Discussants:
Stephanie London, NIEHS


Good epi, bad epi: Is short birth spacing really that bad?

Session Chair(s):   Gavin Pereira,   Katherine Ahrens
Date:       Time: 12:00 am - 12:00 am
Location:


     

For both high and lower income countries the WHO and most obstetric societies recommend waiting at least 18 months to 2 years after giving birth before trying for another child. The harms of long and short intervals have been debated after the recent publication of several matched studies that cast doubt on the extent of these effects. This symposium will bring together researchers with opposing views on the evidence to date and the design of such studies. Birth spacing is an intuitively simple topic but requires application of many epidemiological concepts. The research is leading to new WHO recommendations and national guidelines. An innovative yet simple sibling matched study will be used to demonstrate systemic bias in previous studies (BMJ 2014). We will present conflicting results after stillbirth implying that adherence to current WHO guidelines could cause harm (Lancet 2018). The symposium will close on effects in low income countries, where further research is needed most.

Session Chairs:
Gavin Pereira, Curtin University, Austrailia
Katherine Ahrens, University of Southern Maine


Presenters:
Mark Klebanoff, The Ohio State University
"Why do we think there’s a causal effect?"
Gavin Pereira, Curtin University, Australia
"Is it all due to confounding and selection bias?"
Katherine Ahrens, University of Southern Maine
"Response – No it’s not!"
Annette Regan, Texas A&M
"It is all due to effect modification."
Amanuel Gebremedhin, Curtin University Australia/Jimma University Ethiopia
"Response – Effect modification might exist but we have forgotten about low income countries, where it matters the most"


Is it my brain or my senses? Measuring and interpreting cognitive function in the context of sensory deficit.

Session Chair(s):   Alison Abraham,   Bonnie Swenor
Date:       Time: 12:00 am - 12:00 am
Location:


     

Over the past decade, research examining the relationship between sensory impairment and cognitive function has expanded. However, the sensory-cognitive relationship is complex: while there is mounting evidence that sensory impairments are a risk factor for cognitive decline in older adults, the observed sensory-cognitive relationships may also be driven by common causes of late-life cognitive decline and sensory impairment. Moreover, measuring cognitive function using tests that rely on auditory and/or visual presentation may be problematic in individuals with hearing and/or vision impairments.

Session Chairs:
Alison Abraham, Johns Hopkins School of Medicine
Bonnie Swenor, Johns Hopkins School of Medicine


Presenters:
Frank Lin,
"Do I really need to worry about this and why? Understanding the relationship between sensory impairment and cognitive outcomes"
Bonnie Swenor,
"Apples and oranges. Inconsistencies in assessing cognitive functioning in individuals with sensory impairment"
Jennifer Deal,
"Who gets tested? Selection issues and missing data in cognitive function testing in individuals with sensory impairment"
Aiden Gross,
"What am I measuring? Disentangling cognitive status from sensory impairment"
Walter Wittich,
"The horizon: how to get better at assessing cognitive function in individuals with sensory impairment"


Novel and innovative measures of fetal growth and development

Session Chair(s):   Anne Marie Jukic
Date:       Time: 12:00 am - 12:00 am
Location:


     

Fetal growth is relevant for understanding pregnancy outcomes, pediatric development, and, if the in utero origins of disease hypothesis is applied, subsequent adult health. Fetal growth is often characterized by size at birth, which is known to be misclassified. In recent years technological advances have allowed for more sophisticated measures and analytical techniques. However, controversy exists across clinical disciplines - maternal-fetal medicine, obstetrics, and pediatrics - regarding the most appropriate ways to measure and quantify fetal growth and development. Specifically, it is not clear 1) which fetal parameters should be measured, 2) what tools should be used for those measurements, 3) and how the data should be analyzed. This symposium will address each of these questions with presenters who are leaders in the field. Each of the following presenters would be allotted about 18 minutes with the final 18 minutes used for a panel-style summary/questions/discussion.

Session Chairs:
Anne Marie Jukic, NIEHS


Presenters:
Anne Eskild,
"Novel tools for measuring fetal and placental growth: 3D ultrasound as validated by MRI measurements"
Katherine Grantz,
"What to measure? Fetal growth and fetal growth velocity in relation to preterm delivery, preeclampsia and birthweight"
Russel Deter,
"Analyzing fetal growth data with Individualized Growth Assessment (IGA): A more accurate and sensitive method for detecting growth restriction"
Jason Gardosi,
"Why stop at birth? Building ethnically-diverse standards for fetal plus infant growth"


Open Epidemiology, Data Transparency and Reproducibility: An egg that can’t be unbroken

Session Chair(s):   Neal Goldstein
Date:       Time: 12:00 am - 12:00 am
Location:


     

Increasingly the field of epidemiology is moving towards an open science paradigm: a barrier free approach to sharing data and analyses. Yet given the current environment of public health and the nature of epidemiological data, an open epidemiological framework faces specific challenges before it is routinely employed in the evidence base for public health policy and decision making. New technology, such as data sharing platforms, combined with a culture of transparency may in fact expand our ability to identify new public health solutions faster. While other scientific disciplines have embraced the open science concept, epidemiologist have argued that our data are not necessarily amenable to sharing and a focus on reproducibility detracts from methodological advancement. This symposium will tackle these broad challenges by framing the controversy, discussing solutions and methods towards open epidemiology, and concluding with a collaborative discussion of the way forward.

Session Chairs:
Neal Goldstein, Drexel University


Presenters:
Tim Lash, Emory University
"Open Data: Ideology Meets Reality"
Tonya White, Erasmus Medical Center
"Overcoming obstacles for data sharing within epidemiological research: Hey, it’s my data!"
Bryan Shepherd, Vanderbilt University
"A pragmatic approach for reproducible research with sensitive data"
Harrison Quick, Drexel University
"Using spatiotemporal models to generate synthetic data for public use"
Discussants:
Tyler VanderWeele, Harvard University


Population-based inter-generational studies

Session Chair(s):   Enrique Schisterman
Date:       Time: 12:00 am - 12:00 am
Location:


     

Animal studies show that numerous chemical and environmental exposures, such as endocrine disruptors (EDCs), act via epigenetic reprogramming of the germline, and thus may have multi- or inter-generational. Given the myriad EDCs people are exposed to every day (e.g. pesticides and common plastics), and with increased access to large cohorts that include multiple generations, a new wave of epidemiologic studies to better understand the impacts of these exposures on future generations is emerging. In this session 3 speakers will present on cutting-edge research on inter-generational exposure effects in population-based studies. The first two will present recently-conducted substantive studies of: (1) in-utero maternal caffeine intake and ovarian reserve in adult offspring; and, (2) smoking during pregnancy and ADHD in grandchildren. The third talk will be on methodological considerations/challenges in the analysis of population-based inter-generational studies, along with some solutions.

Session Chairs:
Enrique Schisterman,


Presenters:
Sunni Mumford,
"Inter-generational effects of maternal dietary and lifestyle behaviors on reproductive health of the offspring"
Marc Weisskopf, Harvard University
"Multi-generational association between smoking during pregnancy and ADHD"
Sebastien Haneuse, Harvard University
"Methodological considerations and challenges in the analysis of population-based inter-generational studies."


PrEPidemiology: innovations in methods for biomedical HIV-prevention research

Session Chair(s):   Julia Marcus
Date:       Time: 12:00 am - 12:00 am
Location:


     

With 5000 new HIV infections each day worldwide, HIV-prevention strategies are a public health priority. Antiretroviral preexposure prophylaxis (PrEP), typically a once-daily pill, is up to 99% effective in preventing HIV acquisition. However, PrEP uptake has been slow in high-priority populations, potentially limiting its impact at a population level. This session will highlight innovations in methods for biomedical HIV-prevention research, including novel approaches to optimize PrEP implementation and measure its impact. Specific topics include the use of traditional and machine learning approaches to identify PrEP candidates in settings with low and high HIV incidence; estimation of the population-level effects of PrEP on incidence of HIV and other sexually transmitted infections, using both mathematical modeling and emulation of randomized trials in observational data; and the creative use of clinical trials data to estimate the effectiveness of emerging PrEP delivery methods.

Session Chairs:
Julia Marcus, Harvard Medical School and Harvard Pilgrim Health Care Institute


Presenters:
Julia Marcus, Harvard University
"PrEP implementation: challenges and opportunities"
Douglas Krakower, Beth Israel Deaconess Medical Center
"Integrating electronic health record data to identify potential PrEP candidates in U.S. primary care settings"
Laura Balzer, University of Massachusetts Amherst
"Super learning vs. traditional approaches for targeting PrEP in rural East Africa"
Samuel Jenness, Emory University
"Innovations and opportunities in modeling HIV PrEP: network-based targeting, co-circulating STIs, racial and geographic disparities, and long-acting formulations"
David Glidden, University of California, San Francisco
"Bridging clinical trials data to estimate public health impact for novel PrEP delivery methods"


Remove the hesitation from your validation

Session Chair(s):   Lisa Bodnar,   Matthew Fox
Date:       Time: 12:00 am - 12:00 am
Location:


     

With the development of bias analysis methods, more emphasis has been placed on collecting validation data yet few epidemiologists receive training in conducting validation studies. This lack of training is concerning as poor validation study methodology can lead to more biased adjusted estimates than conventional estimates. Validation studies can suffer from many of the same biases as etiologic studies, including measurement error and selection bias. Further, it is not well known which parameters can be estimated from a validation study are determined by the sampling approach. Before the session, we will survey a sample of SER members on their understanding of validation study design. We will use these responses to guide the content of our symposium. We seek to illustrate concepts of validation study design and implementation of bias analysis in the context of 3 research questions. We will discuss implications of validation design choices, including potential for biased findings.

Session Chairs:
Lisa Bodnar, University of Pittsburgh
Matthew Fox, Boston University


Presenters:
Matthew Fox,
"What do SER members know about validation studies? Results of an informal survey"
Hailey Banack,
"Battle of the bias parameters: comparing methods for validating BMI in postmenopausal women"
Timothy Lash,
"The specific importance of specificity"
Penny Howards,
"Cancer treatment and pregnancy outcomes: challenges in validation and bias analyses"


Saving lives as an off-target effect: the impact of economic and social policies on population health

Session Chair(s):   Anjum Hajat,   Steve Mooney
Date:       Time: 12:00 am - 12:00 am
Location:


     

Public health researchers have focused on individual and community based health interventions to improve population health. Although often effective, these interventions may not scale to benefit the full population. Exploring the impacts of policies is an alternative approach that may yield wide-reaching health improvement.

Session Chairs:
Anjum Hajat, University of Washington
Steve Mooney, University of Washington


Presenters:
James Buszkiewic, University of Washington
"Effects of minimum wage increases on health and health behavior in working-age-adults"
Nathan Nickel, University of Manitoba
"Health impacts of an unconditional prenatal cash transfer in the Canadian province of Manitoba"
Pia Mauro, Columbia University
"Age differences in the effects of medical and recreational marijuana laws across the US"
Ellicott Matthay, University of California, San Francisco
"Local policy and firearm violence: A quasi-experimental study of the Operation Peacemaker Fellowship in Richmond, California"
Discussants:
Arjumand Siddiqi, University of Toronto


Spotlight on aging in postmenopausal women: Twenty-five years of research in the Women’s Health Initiative (WHI)

Session Chair(s):   Hailey Banack
Date:       Time: 12:00 am - 12:00 am
Location:


     

Since 1993, the WHI has focused on understanding the major causes of morbidity and mortality in older women. The study has made seminal contributions to our knowledge of health outcomes in postmenopausal women, perhaps most notably highlighting the risks associated with hormone replacement therapy. At its inception, the WHI was innovative because it represented the first cohort study of postmenopausal women, who, at the time, were an underrepresented group in scientific research. This symposium will feature research by the ‘next generation’ of WHI investigators. In addition to presentations by a diverse group of early career investigators, the symposium will include a discussion of findings from 25 years of research in women’s health by senior WHI investigators and present opportunities for epidemiologists use the WHI data. The symposium is relevant to the membership of SER because the WHI is a landmark epidemiologic study and an incredibly rich resource for research on women’s health.

Session Chairs:
Hailey Banack, University at Buffalo


Presenters:
Andrew Odegaard,
"Leveraging old data for new insight into the obesity-cardiometabolic health relationship"
Wendy Barrington,
"Diving deeper into racial disparities: unique opportunities in WHI"
John Bellettiere,
"Using intensive longitudinal measures of body movement to study the link between sedentary behaviors and incident CVD, diabetes and mortality"
Hailey Banack,
"Can small bugs create big problems? Understanding the relationship between the microbiome and aging in postmenopausal women."
Discussants:
JoAnn Mason,
Jean Wactawski-Wende,


Spotlight on the controls: interesting real-world challenges faced with control groups and the implications for randomized controlled trials.

Session Chair(s):   Jennifer Ahern,   Maya Petersen
Date:       Time: 12:00 am - 12:00 am
Location:


     

The randomized controlled trial is considered to provide the highest level of evidence for causal effects. The focus is typically on the treatment arm, and less attention has been paid to the control group. As the comparator to the intervention, this group has as much impact on the results and interpretation of the trial as the treatment. Recent randomized controlled trials have encountered issues that include changes in local policy and standards of care, spillover, and major unexpected disruptions during the course of the trial. Bringing issues encountered with controls into the spotlight can provide insights for how to proceed in the face of real-world challenges and opportunities for trials. This discussion will include practical and ethical challenges faced, how trade-offs were balanced, what was learned along the way, and what the implications are for trial results. Recent experiences with control groups from individual, family and community randomized trials will be discussed.

Session Chairs:
Jennifer Ahern, University of California, Berkeley
Maya Petersen, University of California, Berkeley


Presenters:
Maya Petersen,
"Cluster randomized trial design and interpretation in a rapidly evolving policy and implementation context: Insights from the Sustainable East Africa Research on Community Health (SEARCH) Trial"
Jade Benjamin-Chung,
"Measuring spillover effects in the treatment and control arms in cluster-randomized trials"
Mi-Suk Kang-Dufour,
"Adapting a community randomized trial design in response to a regional outbreak event: A case study of malaria elimination in Namibia"
Theresa Osypuk,
"What happened to the controls in the Moving To Opportunity housing experiment?"
Brandon Marshall,
"Usual and unusual care: Selecting control groups in behavioral interventions"
Discussants:
Enrique Schisterman,


The Promise and Peril of Randomness

Session Chair(s):   Eric Lofgren
Date:       Time: 12:00 am - 12:00 am
Location:


     

Randomness – and random error – are inescapable elements of the design, analysis and interpretation of epidemiologic studies, but the exact nature of the impact of randomness and its effects on our inference are often elided, or assumed to be captured in relatively simplistic representations of uncertainty. This symposium will delve deeper into the role randomness plays in epidemiology, exploring both threats to the validity of our studies as well as ways to exploit randomness to improve and validate our methods.

Session Chairs:
Eric Lofgren, Washington State University


Presenters:
Jeanette Stingone,
"Do these results seem random to you? A validation approach for investigating multiple exposures using tree-based methods"
Laura Balzer,
"Biologically or socially transmitted outcomes? Be (un)certain of your uncertainty!"
Daniel Westreich,
"Causal identification and randomness: why 'chance confounding' is a contradiction in terms"
Eric Lofgren,
"Bad Doctors or Bad Luck?: Stochastic threats to common quality of care metrics"
Jon Zelner,
"Accounting for selection bias in the measurement of outbreak intervention efficacy"


The baddest of the bad: Ranking the most pernicious biases facing observational studies

Session Chair(s):   Catherine Lesko,   Matthew Fox
Date:       Time: 12:00 am - 12:00 am
Location:


     

There are many threats to the generation of valid and precise knowledge useful for improving population health. Because the metaphysical, physical, practical and mental resources available to epidemiologists are limited to draw inference, epidemiologists tend to focus on a limited number of sources of bias when prioritizing resources. Furthermore, (when done) sensitivity analyses tend to focus on one source of bias at a time. Within any one study, one source of bias may dominate, but this symposium will consist of debate-style presentations that address which source of bias is the most commonly present but un- or under-addressed threat to validity in epidemiologic research. We will include ample time for discussion and interaction between the audience and speaker panel. We will use online voting programs (PollEverywhere.com) to assess pre- and post- discussion ranking of threats to valid inference.

Session Chairs:
Catherine Lesko, Johns Hopkins School of Public Health
Matthew Fox, Boston University


Presenters:
Robert Platt,
"Misalignment of person-time (inappropriate time 0, immortal person-time, etc)"
Maria Glymour,
"Residual confounding or unmeasured confounders"
Matthew Fox,
"Information Bias"
Jessie Edwards,
"Missing Data"
Ashley Naimi,
"Model Misspecification"
Chanelle Howe,
"Selection Bias"
Catherine Lesko,
"Lack of Generalizability"
Discussants:
Jay Kaufman,


The elephant in the room: Causal inference in the face of competing events

Session Chair(s):   Jessie K Edwards,   Jessica G Young
Date:       Time: 12:00 am - 12:00 am
Location:


     

Competing events preclude occurrence of the event of interest. Epidemiologists must frequently make choices about how to handle competing events in estimating the effect of a treatment or exposure on the event of interest. Debate has surrounded various analytic options in this setting and confusion pervades as to the causal interpretation of resulting estimates along with the underlying assumptions needed to justify a given interpretation. The goal of this symposium is to help demystify choices surrounding these analytic options. Speakers will elucidate choices of causal parameters that may be defined in a competing events setting, the assumptions needed to estimate them in real-world data and how these jointly lead to an analytic approach. Applied examples will be given illustrating how these choices ultimately have implications in practice. The session will close with an interactive discussion of the merits and drawbacks of each approach.

Session Chairs:
Jessie K Edwards, UNC Chapel Hill
Jessica G Young, Harvard Medical School


Presenters:
Jessica Young,
"Making sense of analytic choices in competing events data using counterfactual casual models"
Jess Edwards,
"Choices about competing events, illustrated"
Mats Stensrud,
"Separable effects: new estimands for causal inference in competing risk settings"
Discussants:
Miguel Hernan,
Stephen Cole,


Triangulation of Causal Effect Estimates

Session Chair(s):   Sonja Swanson
Date:       Time: 12:00 am - 12:00 am
Location:


     

Causal effect estimation is typically conducted within the scope of one study, and inferences need to be interpreted with respect to that study’s strengths, limitations and underlying assumptions. Causal inferences can be strengthened by leveraging multiple approaches that require different assumptions or that have different strengths and limitations. This practice, referred to as triangulation, is an ideal part of all scientific endeavors. However, it is often not highlighted in epidemiologic training or in research. In this symposium we will discuss the role of triangulation in our own research lines. Real epidemiologic examples will be emphasized, including applications in studying maternal pregnancy exposures, HIV treatment, Alzheimer's disease and dementia, and cardiovascular disease.

Session Chairs:
Sonja Swanson, Erasmus MC


Presenters:
Jeremy Labrecque,
"Using counterfactuals to formalize triangulation"
Debbie Lawlor,
"Using triangulation to explore causal effects of maternal exposures on offspring outcomes"
Catherine Oldenburg,
"Regression discontinuity and inverse probability weighting methods to estimate causal effects in HIV epidemiology"
Maria Glymour,
"Triangulation in Alzheimer’s disease and dementia research"


Use of simulation in non-infectious disease epidemiology

Session Chair(s):   Melinda Power
Date:       Time: 12:00 am - 12:00 am
Location:


     

Simulation studies long been used in other contexts, but use in non-infectious disease epidemiology is limited. This symposium highlights situations where simulation may be useful, illustrates how simulation can address pressing questions that may be unanswerable with traditional methods or available data, and discusses best practices and helpful approaches to implementation. Dr. Mayeda will discuss use of simulation to quantify bias due to selective survival in lifecourse epidemiology, with application to the question of sex differences in Alzheimer’s disease incidence. Dr. Power will discuss use of simulation to bridge the gap between observational studies and RCT design, with application to the example of diabetes management and dementia risk. Dr. Cerdá will discuss use of agent-based modeling to answer policy-relevant questions we can’t answer with real data, with a focus on firearm violence. Finally, Drs. Burke and Levine will discuss novel strategies to ease use of simulation.

Session Chairs:
Melinda Power, George Washington University


Presenters:
Teresa Filshtein,
"Use of a simulation framework to combine data from multiple observational data sources"
Elizabeth Rose Mayeda,
"Simulations for quantifying selection bias in lifecourse epidemiology"
Magdalena Cerda,
"Simulating the bounds of plausibility: using agent-based models to estimate the impact of high-risk versus population-based approaches to prevent firearm violence"
Luciana Copeland and James Burke,
"Strategies to Optimize Simulation Fidelity and Code Readability — Perspectives, Pair Programming, and Open Source"


iEpidemiology: Innovative exposure and outcome assessment using advanced technologies

Session Chair(s):   Keewan Kim,   Carrie Nobles
Date:       Time: 12:00 am - 12:00 am
Location:


     

Given the diversity of exposures and their complex interactions, the need for new methods of exposure assessment is a pressing concern. We feel this symposium would be of interest to SER members given the need to accurately identify exposures during critical windows of susceptibility in the field of epidemiology. New technologies (e.g., smartphone app) are being developed to more accurately and precisely capture personal exposures to a range of health factors, including physical activity, reproductive, and environmental exposures. The goal of this symposium is to introduce new technologies being implemented in the field. Furthermore, we aim to foster a discussion of the potential to learn new methods across various disciplines and the costs/benefits of incorporating emerging technologies for exposure/outcome assessment into research studies.

Session Chairs:
Keewan Kim, NICHD
Carrie Nobles, NICHD


Presenters:
Danielle Bradley,
"Leveraging data collected using a digital mobile app to improve women’s health: the Ovia Health experience"
Christina Troutner,
"HealthWatch 360: Mobile app for population and interventional nutrition research"
Marc Epstein,
"A new tool for epidemiological research: MyExposome silicone wristbands"