Lee Kennedy-Shaffer, PhD

My research has focused on methods for the design and analysis of infectious disease studies, with specific attention to the role of correlated outcomes, properly assessing variance, and identifying estimands that are relevant for public health. This work has drawn on traditional biostatistics and clinical trials methodology, epidemiologic study designs, econometric methods, specific features of infectious disease epidemiology, viral kinetics and transmission, and mathematical modeling of infections.

Infectious disease studies face many statistical and epidemiological challenges, of which three of the most significant and unique are: (1) clustering and correlation of outcomes among individuals; (2) high spatiotemporal variation, especially in outbreak and epidemic settings; and (3) different estimands at individual and population levels. These make common statistical assumptions—including independence, parallel trends, exchangeability, and homogeneous effects—unlikely to hold for a variety of exposures and outcomes. Understanding infectious disease interventions thus requires the development of new study designs and new methods for analysis, or the modification of existing ones to address these challenges. These engender tradeoffs, however, on the epidemiological, statistical, and policy levels.

Specific areas of research include:

• Developing cluster-randomized trial and stepped-wedge trial designs and analysis methods,
  Quantifying and using spatiotemporal variation in infectious disease outbreaks to improve surveillance and studies of interventions,
  Designing studies to understand the full spectrum of effects of vaccines and other infectious disease interventions,
  The communication, education, and history of statistics, and
  Exploring the policy implications of statistical methods.