Welcome to the BiosBeat communication tool for the Department of Biostatistics at the UNC Gillings School of Global Public Health! Here you will find a collection of the latest department news, special features, dates to save, and much more. So, read on, enjoy, and be sure to regularly check back for updates!
COVID19 Special Message from our Biostatistics Leadership
We hope you are staying safe and healthy during this time. We encourage you to continue taking all the preventative measures recommended by UNC and the CDC.
Please visit the School of Public Health’s official Coronavirus Information Portal for the most recent information and updates: https://sph.unc.edu/global-health/2019-coronavirus-info-portal/
HIGHLIGHTS OF THIS ISSUE:
- HEAR FROM DEPARTMENT CHAIR PROFESSOR LAVANGE
- BIOS STUDENTS RESEARCH POTENTIAL COVID-19 TREATMENT
- APP HELPS TO CALCULATE OPTIMAL POOL SIZE FOR COVID-19 TESTING
- SUPPORT STAFF KEEPS THE DEPARTMENT RUNNING VIRTUALLY
Greetings from your new chair! I am very excited to be writing my first corner for BiosBeat and welcome the opportunity to contribute to these newsletters as a way to stay in touch with our Biostatistics family throughout the year. The winter holidays are behind us, the lunar new year’s celebration just ended and the year of the ox has begun, and we are in the midst of our third semester affected by the Covid-19 pandemic. My thanks go out to both our faculty and our students for adapting to the remote learning experience once again this semester. I am looking forward to our eventual return to normal, or to a new post-pandemic normal, and am optimistic that we will see more of each other on campus during the 2021-2022 academic year. Or at least, that is my hope!
Graduation, like so many other activities, will look different this spring. The Gillings School ceremonies will be virtual and will feature Biostatistics alumna, Dr. Cicely Mitchell, as commencement speaker. Dr. Mitchell is Director of Biostatistics at Syneos Health and co-founder of the Art of Cool Project, a non-profit organization in Durham, NC with a mission to introduce broad audiences to jazz music. She will share her wisdom and vision with our students, drawing not only on her time spent pursuing her masters and doctoral degrees at UNC Biostatistics but also on a 20-year career that fuses biostatistics and music. We are looking forward to welcoming Dr. Mitchell back to campus for this milestone event in the lives of our graduates!
There will also be a series of in-person ceremonies in Kenan Stadium for those students wishing to attend. Two commencement speakers were recently announced. Dr. Anthony Fauci, Director of the National Institute on Allergy and Infectious Diseases (NIAID) needs no introduction, having led the nation in providing evidence-based advice on how to combat this pandemic for the past year. The second speaker may also be known to some of you. Dr. Kizzmekia (“Kizzy”) Corbette, PhD is an Orange County, NC native who graduated from Orange High School, received her undergraduate degree from the University of Maryland, and then returned home for her master’s and doctoral degrees in microbiology and immunology at UNC. She was on the team of researchers at the NIAID that contributed to the development of the Moderna vaccine and is now using her voice to encourage demographic groups most severely impacted by Covid-19 to get vaccinated – a real public health hero! Even with a virtual platform, all three commencement speakers, at this place and time, will be hard to forget.
Under the heading of faculty news, Professor Hongtu Zhu is back in the department full time and has lots of interesting data-science related experiences to share with us from his leave of absence. Welcome home, Hongtu! Our newest faculty member, Professor Tanya Garcia, welcomed her second child in January and will be returning to work later this month (mid-March). Congratulations, Tanya! We look forward to seeing you again as co-host of the Thursday seminar series.
Professor Michael Hudgens is our new Associate Chair, and I cannot say enough about how excited I am to work with him in this new position. His experience in the department and knowledge about all things biostatistics will be such an asset for us going forward. Congratulations, Michael, and welcome to the biostatistics leadership team! In addition to serving as Associate Chair, Michael will continue to be heavily involved in methodological and applied research related to infectious diseases. Recently he has been working with several BIOS students on Covid-19 related projects as described below.
Looking ahead, we will host Professor Xihong Lin from Harvard University as our Greenberg Lecturer May 20-21, 2021. She will give three different lectures in this series. We will provide more details soon about attending what are sure to be very informative talks. Have a wonderful spring semester, and don’t forget to take advantage of the remaining wellness days.
Stay safe and healthy,
MEASURING EFFICACY OF POTENTIAL COVID-19 TREATMENT
Gillings biostatistics PhD students Ann Marie Weideman and Taylor Krajewski are on the frontlines of the fight to find a treatment for COVID-19. They are both involved in the EIDD-2801-2003 study, designed to measure the safety, tolerability, and efficacy of the EIDD-2801 drug as compared to a placebo. The drug’s aim is to eliminate SARS-CoV-2 viral RNA detection in symptomatic adults with COVID-19. The study, which is randomized, double-blind, multi-site phase 2 and currently has over 200 participants. The study enrolled its first patient here at UNC, but has quickly expanded its enrollment nationwide to sites as far as the West Coast.
Ann Marie Weideman is a 3rd-year biostatistics PhD student at the Gillings School. Prior to her arrival at UNC, she worked as a researcher at the National Institutes of Health Clinical Center, which is currently the largest research hospital in the world. Ann Marie works for the Center for AIDS Research (CFAR) core and is also a predoctoral trainee on the Biostatistics for Research in Genomics and Cancer training grant.
Ann Marie first became involved in EIDD-2801 in its infancy in May 2020, four months after the first confirmed case of COVID-19 was reported in the United States. Ann Marie’s responsibilities included independently generating randomization schedules for each part of the study and designing an electronic randomization system. With the “click of a button,” clinical staff at all sites were able to efficiently randomize their patients while remaining blinded as to who received the drug versus the placebo.
Taylor Krajewski is also a 3rd-year biostatistics PhD student at the Gillings School. Before coming to UNC, she was the mathematics department chair and a mathematics teacher at a high school in the Philadelphia, PA area for several years. Taylor works for the CFAR Biostatistics core and is also a predoctoral trainee on the National Institute of Environmental Health Sciences training program in Environmental Biostatistics.
Taylor became involved in EIDD-2801 in the fall of 2020 as a statistician on the trial. She contributed to creating the statistical analysis plan for this study and managed data as the study progressed. Now that the study is coming to an end, Taylor is working with a team of statisticians to perform primary and secondary analyses. She is excited to see the results of this study as antiviral therapeutic interventions that reduce infectious virus production could play a key role in decreasing community transmission, decreasing the surge on healthcare facilities, and potentially reducing clinical complications of infection.
SPOTLIGHT ON FACULTY
Name: Quefeng Li
Position: Assistant Professor
Time at the Gillings School: I joined the faculty in 2015.
What I do (and why I love it): My research focuses on developing new statistical methods for high-dimensional data. High dimensional data is a new data structure that emerges in much biomedical research, such as genetics data, imaging data and electronic health records data. Currently, I am working on developing new methods to integrate multiple types of data to inform clinical decisions. There are many interesting problems remaining open at this stage, like how to integrate data of different scales, how to quantify the contribution of each data type in predicting the clinical outcome, etc. Solving these problems may help us have a better understanding on why an integrative analysis can help make better decisions.
First job or Internship I had was: My first job was serving as a lecturer at the Department of Statistics, UW-Madison when I was a PhD student there. I taught a probability course for engineering students for two years. It’s interesting to see how probability/statistics can be useful in other fields, like engineering.
APP CALCULATES OPTIMAL POOL SIZE FOR GROUP COVID-19 TESTING
As the number of COVID-19 cases continues to rise throughout the United States, there has been an increased demand for tests to diagnose individuals that may be infected with the virus. Researchers at the Gillings School have been searching for a way to enhance the efficiency and accuracy of these tests without dramatically reducing test sensitivity, or the probability that a truly positive individual is correctly classified as positive by that test. One way to achieve this is with pooling algorithms, where group testing is used to identify individuals with significant viral load of SARS-CoV-2. The app is designed for users to find the optimal pool size from three commonly used pooling algorithms: two-stage hierarchical, three-stage hierarchical, and a two-stage square array pooling method. The app also provides visualizations and statistics for each pooling algorithm.
Pooling can dramatically reduce the number of tests needed for large groups of people. This is done by splitting the large number of people into smaller subgroups and then combining samples from all of those in a given subgroup into a single batch for testing. Members of a subgroup whose batch tests are all negative are considered negative, while any members of a subgroup whose batch may test positive are all re-tested individually to obtain results for each specific person.
“The overall goal of the calculator is to provide the user with the optimal pool size which, for a given sensitivity, specificity, and prevalence, minimizes the total number of COVID-19 tests used. So, employing these pooled testing algorithms would enable people in charge of large-scale testing efforts to use the tests more effectively” Sam shared.
The application is currently available to the public and can be found on the Gillings COVID-19 Dashboard.
BIOS BEHIND THE SCENES
DEPARTMENT SUPPORT TEAM KEEPS US RUNNING VIRTUALLY
When the University moved to a fully remote format, faculty and students were forced to adapt to a virtual classroom style with little moment’s notice. Naturally, this required a learning curve for all involved as they embraced new technologies. The Bios Support team, consisting of David Hill and Scott Zentz, were there every step of the way to assist with technical difficulties or confusion that any faculty, staff, or student may have had as they adjusted to the new normal.
Scott and David have worked with the department for many years, and are accustomed to providing support with various technology-based requests. When the pandemic shifted everyone to an online format, the nature of the support did not change, but the way the team was able to provide assistance had to be altered.
“I think supporting the students and faculty remotely, first and foremost has been more challenging in the fact that communication is so delayed. Everything is done through email, video chat and on some occasions over the phone. The biggest challenge was to adapt to not being able to address issues in person and basically going to phone support which presents a lot of blind situations in terms of troubleshooting.” David shared. “A lot to the work relies on the end user to do the pointing/clicking and telling us what they see so we can present them with guidance and solutions.”
While there have been some challenges with this new format, there have also been plenty of newly found positives. The support team is able to help more people in more locations, and are able to assist at whatever time they are needed. “Everyone that works remotely has a more open schedule so with that I can address an issue in the early morning or the late evening, whatever they need as not to interrupt their workflows” David said.
Throughout the past year, the Bios Support team has remained passionate and dedicated to the work they do and the positive impact they make. “We are all going through this together and to be able to help our department continue to be successful is a positive for everyone involved.” David shared. “That is the sole reason I work to find ways to make my support of the department as seamless as possible given the remote constraints that we are under.”
GILLINGS REPRESENTED WELL AT ENAR 2021
The Gillings School was represented admirably at the Eastern North American Region 2021 Spring Meeting Virutal Event with two students from Gillings Biostatistics winning the prestigious ENAR Distinguished Student Paper Award.
Sarah Reifeis, one of the winning students, is a current doctoral student pursuing a PhD in Biostatistics, Professor Hudgens is her advisor. Her paper, entitled “On variance of the Treatment Effect in the Treated Using Inverse Probability Weighting”, focuses on estimating the variance of the IPW ATT estimator. In the analysis of observational studies, inverse probability weighting (IPW) is commonly used to consistently estimate the average treatment effect (ATE) or the average treatment effect in the treated (ATT). The variance of the IPW ATE estimator is often estimated by assuming the weights are known and then using the so-called “robust” (Huber-White) sandwich estimator, which results in conservative standard error (SE) estimation. Here it is shown that using such an approach when estimating the variance of the IPW ATT estimator does not necessarily result in conservative SE estimates. That is, assuming the weights are known, the robust sandwich estimator may be conservative or anti-conservative. Thus confidence intervals of the ATT using the robust SE estimate will not be valid in general. Instead, stacked estimating equations which account for the weight estimation can be used to compute a consistent, closed-form variance estimator for the IPW ATT estimator. The two variance estimators are compared via simulation studies and in a data analysis of the effect of smoking on gene expression.
Larry Han, another winning student, graduated from the Gillings School in 2016 with a BSPH in Biostatistics. He is now in his third year as a PhD student at Harvard. His winning paper is entitled “On the Evaluation of Surrogate Markers in Real World Data Settings”.
BIOS STUDENTS HONORED AT ASA STUDENT PAPER COMPETITION
Three of our bios students, Hunyong Cho, Ethan Alt, and Nate Bean, have received awards in the American Statistical Association Biopharmaceutical Section Student Paper Competition. Hunyong, a fifth-year PhD student, has received a travel award, and Ethan and Nate, both fourth-year PhD students, received honorable mentions.
Hunyong’s research interests include precision medicine, machine learning, and genomics. His paper, entitled “Multi-stage Optimal Dynamic Treatment Regimes for Survival Outcomes with Dependent Censoring”, provides individualized treatment recommendations to patients over multiple rounds of therapies based on patient history. As a result, this enhances the patient survival time or survival probability. Hunyong and his co-authors, Professor Kosorok and NCSU’s Dr. Shannon Holloway, addressed unique challenges of multi-stage survival data by generalizing the Q-learning approach and the random survival forests. They demonstrated the optimality of their method by showing its large-sample behavior through a leukemia patient data example.
Nate, working under the direction of co-advisors Professors Psioda and Ibrahim, holds primary research interests in Bayesian clinical trials and oncology clinical trials. His paper, entitled “Bayesian Multi-Regional Clinical Trials Using Model Averaging”, focuses on the effects of multi-regional clinical trials, or MRCTs. These trials provide the benefit of more rapidly introducing drugs to the global market, however, small regional sample sizes pose limitations to the reliability of current statistical methods when conducting subgroup analyses and estimating regional treatment effects. His team develops novel methodology for estimating regional and global treatment effects from MRCTs using Bayesian model averaging, and they propose new methods to evaluate the consistency of the treatment effect across regions.
Ethan’s research interests include Bayesian methods for the design and analysis of clinical trials. In his paper, entitled “Bayesian multivariate probability of success using historical data with family-wise error rate control”, his team develops methods to robustly estimate sample size in the presence of multiple endpoints (e.g., co-primary endpoints) based on a Bayesian version of statistical power. They also develop a fully Bayesian approach to jointly test hypotheses across different outcomes that attains exact type I error control. Simulation results suggest that the method is more powerful than the most popular methods to control for multiplicity, such as the Bonferroni-Holm method.
PROFESSOR LI PROMOTED TO FULL PROFESSOR
Congratulations to Professor Yun Li on promotion to full professor of biostatistics! Professor Li has a primary appointment in the Department of Genetics and a secondary appointment in the Department of Biostatistics. Through her impressive research and publication record, her reputation as a team scientist, and her contributions to teaching and mentoring, she has successfully met the promotion requirements within both departments.
- April 5 – Wellness Day
- April 29 – 2021 Department of Biostatistics Awards Day
- May 5 – Last day of classes
- May 15 – Gillings School Commencement featuring Gillings Biostatistics Alumna Cicely Mitchell, DrPH
- May 20 and 21 – SAVE THE DATE – 2021 Greenberg Lectures – Harvard Professor Xihong Lin
If you have news or a story idea you feel would fit BiosBeat, please submit them to Jeff Oberhaus.
PLEASE NOTE: Given the recent events regarding COVID-19, the annual BiosRhythms is on hold. Issue 31 will not be sent until further notice.