Claire Heffernan
Senior Statistician, Merck
Senior Statistician, Merck
LinkedIn: www.linkedin.com/in/claire-heffernan
I am a Senior Statistician at Merck, working on Late Stage Oncology clinical trials. I completed my PhD in Biostatistics at Johns Hopkins University, where I worked with Dr. Abhirup Datta and Dr. Roger Peng on spatial statistical methods for analyzing data from low-cost air pollution monitoring networks, and on a method to estimate the causal effect of COVID-19 lockdowns on air pollution concentrations. I graduated from McGill University in Montreal, Canada, in 2019 with a B.S. in Honours Mathematics.
Ph.D.: Biostatistics (2019-2024)
Johns Hopkins University, Baltimore, MD
Advisor: Abhirup Datta
Co-advisor: Roger Peng
cGPA: 4.0
B.S.: Honours in Mathematics (2015-2019)
McGill University, Montreal, QC, Canada
Exchange student at the University of Edinburgh, Edinburgh, Scotland (2018)
GPA: 4.0; Dean's Honor List
PhD research
Spatial filtering:
Low-cost networks are increasingly being used to measure air pollution in cities. Since each sensor is cheaper than the gold standard regulatory sensors, more low-cost sensors can be deployed, giving increased spatial resolution of air pollution. However, these low-cost measurements are biased and noisy and must be calibrated.
We developed a spatial filtering method to calibrate such measurements that (a) does not underestimate peaks in air pollution, as the commonly used methods do; (b) incorporates spatial correlation between sites across the network; and (c) is dynamic, using the current timepoints' gold standard measurements in calibration.
We present both a frequentist and Bayesian implementation of the method, and we apply the method to PM2.5 as measured by a low-cost network in Baltimore.
Causal impact of policy interventions on air pollution:
COVID-19 lockdowns changed people's behaviors, such as mobility and traffic patterns. Many studies have shown that these changes resulted in a decrease in air pollution, but the methods have several limitations.
We propose a flexible machine-learning based approach to estimate the effect of any sharp policy intervention on air pollution. This method addresses the limitations of commonly used approaches.
We also propose a validation framework to assess the validity of the estimates.
The approach is applied to estimate the change in NO2 due to COVID-19 lockdowns in several US cities.
Multi-network spatial filtering:
We are extending our spatial filtering method to incorporate data from multiple networks, each with different regression equations and noise levels.
Other Research Assistant work
Compared the performance of different interpolators for air pollution in Baltimore to assess exposure in an asthma study
AstraZeneca Internship
Worked with Dr. Michael Sweeting, Dr. Dan Jackson, and Dr. Binbing Yu
Investigated properties of mixture cure survival analysis models used for estimating treatment effects in clinical trials as part of the Oncolocy Biometrics Statistical Innovation Team
Margaret Merrell Award for outstanding research (2024)
EnviBayes Section Student Paper Award winner (2023)
Helen Abbey Award for excellence in teaching (2023)
ENVR Section Student Paper Award winner (2023)
Washington Statistical Society Travel Award (2023)
National Science Foundation Graduate Research Fellowships Program award recipient (2021 – present)
Award for best performance on the first year PhD Biostatistics comprehensive exam (2020)
Scholarship from Fonds de Recherche du Québec – Nature et technologies (2019 – 2021)
Heffernan C, Peng R., Gentner D.R., Koehler K., Datta A. (2023). A dynamic spatial filtering approach to mitigate underestimation bias in field calibrated low-cost sensor air pollution data. Annals of Applied Statistics. link.
Heffernan C, Koehler K., Levy Zamora M., Buehler C., Gentner D.R., Peng R., Datta A. (2024). A machine learning based interrupted time series framework for studying causal changes in pollutant concentrations due to policy interventions: A case study in COVID-19 lockdowns. In revision at American Journal of Epidemiology.
December 2024. CFE-CMStatistics, London, UK. Title: Spatial filtering for unified calibration of air pollution data from multiple low-cost sensor networks
March 2024. Eastern North American Region (ENAR) Spring Meeting, Baltimore, MD. Title: A causal machine-learning framework for studying policy impact on air pollution: a case-study in COVID-19 lockdowns
October 2023. Women in Statistics and Data Science (WSDS), Bellevue, WA. Title: A dynamic spatial filtering approach to mitigate underestimation bias in field calibrated low-cost sensor air-pollution data.
September 2023. EnviBayes Workshop on Complex Environmental Data, Fort Collins, CO. Title: A dynamic spatial filtering approach to mitigate underestimation bias in field calibrated low-cost sensor air-pollution data.
August 2023. Joint Statistical Meetings (JSM), Toronto, Canada. Title: Did the COVID-19 Lockdowns Improve Air Quality? Machine-Learning Based Robust Estimation of Effects of Policy Interventions on Air Pollution.
March 2023. Eastern North American Region (ENAR) Spring Meeting, Nashville, TN. Title: Did the COVID-19 Lockdowns Improve Air Quality? Machine-Learning Based Robust Estimation of Effects of Policy Interventions on Air Pollution .
August 2022. Joint Statistical Meetings (JSM), Washington D.C. Title: Gaussian Process Filtering of Low-Cost Air-Pollution Sensor Networks.
March 2022. Eastern North American Region (ENAR) Spring Meeting, Houston, TX. Title: Gaussian Process Filtering of Low-cost Air-pollution Sensor Networks.
Spring 2023 - Spring 2024: Statistical Methods in Public Health III in the Biostatistics department, JHU.
Instructors: Marie Diener-West, Leah Jager
~600 students. Teach lab sessions ~3 times per week to teach students (mainly masters students in the School of Public Health) how to apply course concepts to solve problems
Fall 2022: Statistical Methods in Public Health I and II in the Biostatistics department, JHU.
Instructors: Marie Diener-West, Karen Bandeen-Roche
~600 students. Held office hours
Spring 2022: Real Analysis I, TAing for the Biostatistics department, JHU.
Supervised by Michael Rosenblum to design this new TA position
~5 students. Designed weekly lab sessions to go over real analysis topics that would be important for biostatistics PhD students taking measure theoretic probability the following year
Fall 2021: Probability Theory I and II in the Biostatistics department, JHU.
Instructors: Cristian Tomasetti, Michael Rosenblum
~10 students. Held office hours and designed bi-weekly lab sessions for problem solving
Spring 2021: Probability Theory III and IV in the Biostatistics department, JHU.
Instructors: Cristian Tomasetti, Abhirup Datta
~5 students. Held office hours and designed bi-weekly lab sessions for problem solving
Fall 2020: Probability Theory 1 in the Applied Mathematics and Statistics department, JHU.
Instructor: Dr. Jim Fill
~25 students. Held office hours
For the biostatistics PhD students in the class (7 students): designed weekly lab sessions to go over important course topics and do problems in groups