Published: Thursday, August 20, 2020
Updated: Thursday, December 31, 2020
Faculty in the Department of Statistics at the University of Washington are actively engaged in research related to COVID-19. Here we focus on a few of these efforts.
Boeing International Professor Adrian Raftery has been appointed as a founding member of the Executive Committee of the National Academy of Sciences Societal Experts Action Network (SEAN) set up in collaboration with NSF to coordinate the NAS response to COVID-19 questions in the area. He was recently lead author of the National Academy report on Evaluating Data Typers: A Guide for Decision Makers using Data to Understand the Extent and Spread of COVID-19.
Professor of Statistics and Computer Science Sham Kakade led the CovidSafe app project which has recently become part of the CommonCircle suite. This app, which will be available on both the Android and IOS platforms, will help contact tracers slow the spread of the virus. You can listen to a recent interview that Sham gave with KUOW’s Bill Radke here.
Associate Professor of Statistics and Sociology Tyler McCormick is involved in a multidisciplinary proposal for a community and healthcare facility based testing strategy, aimed at helping policymakers who are forced to make decisions in the face of uncertainties about the coronavirus. The strategy uses a combination of risk-stratified testing and sequential decision-making to quickly identify changes in COVID-19 cases within small geographic areas.
Together with Associate Professor Rachel Heath of the Department of Economics, McCormick has also begun to develop and implement a sampling strategy for workers in manufacturing settings. The project focuses on manufacturing workers in low-resource settings and uses a mobile-phone based network sampling approach to quickly obtain representative samples of workers when traditional in-person data collection is not possible, with the goal of obtaining insights about the scope and scale of COVID-19 responses in the workplace. This work is funded by a UW Population Health Initiative rapid response grant.
Assistant Professor Alex Luedtke is collaborating with researchers at Johns Hopkins University, Emory, and Weill-Cornell to identify statistical methods for the analysis of COVID-19 therapeutic trials that require smaller sample sizes compared to standard statistical approaches. They have shown that their proposed approaches can achieve the same power as competing approaches using 10-20% fewer samples. Some of the results have been recently published in Biometrics as a discussion paper. You can read more about their efforts here.
Luedtke has been involved in a number of COVID-19 vaccine efficacy trials. He co-wrote the statistical analysis plan for the Solidarity Vaccine Trial by the World Health Organization (WHO), which will evaluate the efficacy of all COVID-19 vaccines to be tested by the WHO. He is also a lead statistician for the NIH-funded COVID Prevention Network (CoVPN) on an upcoming Phase 3 trial evaluating Johnson & Johnson's vaccine. His responsibilities include co-writing an analysis plan describing the methods to be used to evaluate the relationship between vaccine-induced immune responses and efficacy. A version of this plan will also be used for several other trials funded by Operation Warp Speed.
Recently, Luedtke has worked on two manuscripts with a group of co-authors from academia, government, and industry. One was published in the Annals of Internal Medicine, which proposed endpoints to be used in trials for COVID-19 vaccines and discussed the possibility that vaccine protection against symptomatic COVID-19 could be accompanied by a shift towards more asymptomatic infections --- this question is still under investigation for all vaccines, including those with an Emergency Use Authorization from the U.S. Food and Drug Administration (FDA). The second manuscript on medRxiv describes a statistical method for estimating the durability of vaccine efficacy even in the trials where all placebo recipients are crossed over to vaccine. This work is currently under review at a medical journal.
Associate Professor of Statistics & Computer Science and Engineering Emily Fox has used Apple Maps mobility data to develop an epidemiological model to infer the impact of human mobility on SARS-CoV-2 transmission in the United States. This framework enables real-time tracking of the effective reproductive number and forecasting of transmission under different mobility policies, informing both intervention evaluations and surge planning for healthcare systems. A pre-print of Fox’s work can be found on MedRxiv.
Professor Jon Wakefield is working with Department of Statistics alum Zehang Richard Li and epidemiology graduate student Julianne Meisner on estimating the excess mortality that may be attributable to COVID-19. They are using space-time smoothing with a delayed reporting model and are looking at county level data in the United States. Wakefield is also working on prevalence surveys in the United States, and on gauging the impact of school closures, with researchers in Switzerland, Germany and Belgium.
Assistant Professor Fang Han is working with Department of Statistics graduate student Yanran Cui on a project to develop robust nonparametric methods for non-stationary point process models. The methods that they develop will be used to study the U.S. spread pattern of COVID-19, in order to provide early detection of pattern changes.
Assistant Professor Yen-Chi Chen and Department of Statistics graduate student Steven Wilkins-Reeves are working on a project related to harmonizing COVID-corrupted datasets in the National Alzheimer's Coordinating Center (NACC) database. The data collection process of the NACC data has been largely impacted by the COVID-19; the data collected during the pandemic is different from the data collected before the pandemic. Collaborating with Dr. Gary Chan in Biostatistics department and researchers in the NACC, they are developing new tools for harmonizing the corrupted data due to COVID-19 with the existing NACC data.
Professor Emeritus Peter Guttorp has been maintaining the International Statistical Institute (ISI) COVID-19 page as Chair of the ISI Committee on Public Voice. He has also created a weekly blog, “Statisticians React to the News,” which includes entries on COVID-19.
Associate Professor of Political Science and Adjunct Associate Professor of Statistics Chris Adolph is leading a team of political science and public health scholars tracking state-level policy responses to the pandemic in the United States. Since March 2020, this group has tracked mandates and recommendations for social distancing, travel restrictions, mask policies, and more. These data are freely available to the public via their Github, and are presently used as inputs for public-facing policy engagement or modeling endeavors by a number of groups, including COVID Exit Strategy, COVID Mobility Network, Imperial College London, Institute for Health Metrics and Evaluation, and Tableau.
Affiliate Associate Professor of Statistics and Genome Sciences Erick Matsen is currently working on two projects related to COVID-19. In collaboration with Julie Overbaugh, he is using phage immunoprecipitation sequencing data to find the regions of the virus that are bound by antibodies. In collaboration with Jesse Bloom, he is developing methods to predict the functional properties of unseen variants from deep mutational scanning of the SARS-CoV-2 receptor binding domain. Matsen recently worked on two preprints pertaining to PhIP-seq work described in bioRxiv, which can be found here and here.
We are grateful to these and other members of the Department of Statistics who have rapidly pivoted their research to serve our broad community as we cope with the COVID-19 pandemic.