Directors
DIRECTOR
Michael Hughes, PhD; Professor of Biostatistics and Director of the Statistical and Data Management Center for the AIDS Clinical Trials Group. His research involves a variety of issues concerning HIV, with particular emphasis on statistical methods for the design and analysis of HIV clinical trials. One area concerns methhods for the design and analysis of phase I/II studies for “special populations” such as infants, children and pregnant women. Such studies require novel dose-finding methods involving multiple outcome measures, including pharmacokinetic, anti-HIV activity and toxicity outcomes, and complexities related to long-term outcomes. More generally, this area extends to the design and analysis of bridging studies that allow translation of results from large clinical trials in one population to a second population (e.g. from the U.S. to sub-Saharan Africa). Another area of interest concerns the development of semi-parametric methods for longitudinal data analysis including informative missingness and censoring (as with repeated measurements of HIV RNA). These methods are critical for evaluating virologic and immunologic outcomes, as well as growth and development outcomes in HIV-infected children.
PRECEPTORS
Tianxi Cai, ScD; Professor of Biostatistics. Her current research interests are mainly in the area of biomarker evaluation; model selection and validation; prediction methods; personalized medicine in disease diagnosis, prognosis and treatment; statistical inference with high dimensional data; and survival analysis. In addition to her methdological research, Dr. Cai also collaborates with the I2B2 (Informatics for Integrating Biology and the Bedside) center on developing a scalable informatics framework that will bridge clinical research data and the vast data banks arising from basic science research in order to better understand the genetic bases of complex diseases.
Wafaie Fawzi, DPH; Professor of Nutrition and Epidemiology. Dr. Fawzi’ research focuses on examining the role of nutritional and other factors in the etiology of adverse health outcomes among populations in developing countries, with emphasis on infectious and perinatal outcomes among mothers and children. Dr. Fawzi and collaborators are implementing several large randomized controlled trials to examine the efficacy of various micronutrient supplements on the incidence and severity of a number of infectious diseases including pneumonia, diarrhea, tuberculosis, and HIV infection. In collaboration with colleagues at Muhimbili University in Dar es Salaam, Tanzania, the team completed a trial that documented a significant beneficial effect of periodic vitamin A supplementation on child mortality. In another large clinical trial prenatal multivitamin supplementation of HIV-infected women resulted in large and significant reductions in the risk of fetal loss, low birth weight, and severe prematurity. Currently, the group is examining whether the latter findings are generalizable to the larger population of HIV-negative women. As part of the HIV Prevention Trials Network at NIH the team is engaged in examining strategies for reducing perinatal and heterosexual transmission of infection.
Sebastien Haneuse, PhD; Associate Professor of Biostatistics at HSPH. His methodologic research follows two general themes, the first of which focuses on the development of novel study designs that help address bias encountered in the analysis of data from observational studies. He looks to augment the data collection process with supplementary information that can then be used to directly address the various biases. The simultaneous development of statistical tools that ensure valid and efficient estimation and inference is a crucial aspect of this research. The second general theme of his research involves the development and use of flexible, so-called non- parametric, prior distributions for semi-parametric Bayesian analyses. Two key components of this research are (i) exploiting the flexibility of these specifications to gain additional insights into mechanisms and/or etiology, and (ii) overcoming the consequences of model misspecification, particularly in the analysis of correlated or longitudinal data.
Bethany Hedt-Gauthier, PhD; Associate Professor of Global Health and Social Medicine and Associate Professor in the Department of Biostatistics. As a biostatistician working in global health, Dr. Hedt-Gauthier has over 15 years of experience supporting the design and analysis of health systems and clinical outcomes research in nine low- and middle-income countries. She is well positioned to serve as a faculty and mentor for the HCSPH AIDS Training Grant, having led several research grants, including two previous and one current NIH grant related to evaluating using community health workers and mHealth technologies to support postpartum care after c-section. She also leads statistical methods development and application, with expertise in complex survey sampling, surveillance, and methods for program monitoring and evaluation.
Nima Hejazi, Phd; Assistant Professor of Biostatistics. Dr. Hejazi’s research explores how advances in causal inference, statistical machine learning, and computational statistics can empower discovery in the biomedical and health sciences. He focuses primarily on the development of model-agnostic, assumption-lean statistical inference procedures, doing so while emphasizing a science-first, translational philosophy that stresses the rich interplay between the applied sciences and statistical methodology–how emerging questions in the former spur advances in the latter, which, in turn, help to refine the scientific discovery process. This approach leverages causal inference as a framework for the translation of scientific questions into interpretable statistical estimands, and then aims to formulate analytic methods that incorporate flexible learning techniques (i.e., machine learning), draw upon semi-parametric efficiency theory, and impose only those modeling restrictions justified by domain knowledge. He is also deeply interested in high-performance statistical computing and the role that open-source software and programming play in the responsible practice of applied statistics and statistical data science, especially as these relate to the promotion of transparent, reproducible, and replicable science.
Miguel Hernan, PhD. Kolokotrones Professor of Biostatistics and Epidemiology. Dr. Hernan’s research is focused on methodology for causal inference, including comparative effectiveness of policy and clinical interventions. His team works to combine observational data, mostly untestable assumptions, and statistical methods to emulate hypothetical randomized experiments. They emphasize the need to formulate well defined causal questions, and use analytic approaches whose validity does not require assumptions that conflict with current subject-matter knowledge. For example, in settings in which experts suspect the presence of time-dependent confounders affected by prior treatment, we do not use adjustment methods (e.g., conventional regression analysis) that require the absence of such confounders. While causal inferences from observational data are always risky, an appropriate analysis of observational studies often results in the best available evidence for policy or clinical decision-making. At the very least, the findings from well designed and properly analyzed observational studies may guide the design of future randomized experiments. His applied work is focused on optimal use of antiretroviral therapy in persons infected with HIV, lifestyle and pharmacological interventions to reduce the incidence of cardiovascular disease, and the effects of erythropoiesis-stimulating agents among dialysis patients.
Sonia Hernandez-Diaz, MD, DrPH is a Professor of Epidemiology at the Harvard T.H. Chan School of Public Health. Her area of interest is drug safety evaluation from non-randomized data, with a special emphasis on the design, conduct, and analysis of studies in pregnant women and their infants. Examples of her work include inquiries of the comparative safety of psychotropics for pregnant women and their offspring using real world evidence from both pregnancy registries and large healthcare databases. She is Past-President of the International Society for Pharmacoepidemiology and the Society for Perinatal and Pediatric Epidemiology Research; and serves as a Special Government Employee for the FDA Drug Safety and Risk Management Advisory Committee (current Chair), as a member of the NICHD Pregnancy & Neonatology (PN) Study Section, and as member of the Teratogenic Information Services (TERIS) Advisory Board. Through her service to public health institutions she has contributed to the translation of research into policy and actionable recommendations for stakeholders.
Michael Hughes, PhD; Professor of Biostatistics and Director of the Statistical and Data Management Center for the AIDS Clinical Trials Group. His research involves a variety of issues concerning HIV, with particular emphasis on statistical methods for the design and analysis of HIV clinical trials. One area concerns methhods for the design and analysis of phase I/II studies for “special populations” such as infants, children and pregnant women. Such studies require novel dose-finding methods involving multiple outcome measures, including pharmacokinetic, anti-HIV activity and toxicity outcomes, and complexities related to long-term outcomes. More generally, this area extends to the design and analysis of bridging studies that allow translation of results from large clinical trials in one population to a second population (e.g. from the U.S. to sub-Saharan Africa). Another area of interest concerns the development of semi-parametric methods for longitudinal data analysis including informative missingness and censoring (as with repeated measurements of HIV RNA). These methods are critical for evaluating virologic and immunologic outcomes, as well as growth and development outcomes in HIV-infected children.
Curtis Huttenhower, PhD; Professor of Computational Biology and Bioinformatics. Dr. Huttenhower’s research focuses on computational biology at the intersection of microbial community function and human health. The Huttenhower group works on a variety of computational methods for data mining in microbial communities, model organisms, pathogens, and the human genome. In practice, this entails a combination of computational methods development for mining and integrating large multi’omic data collections, as well as biological analyses and laboratory experiments to link the microbiome in human populations to specific microbiological mechanisms. The lab has worked extensively with the NIH Human Microbiome Project to help develop the first comprehensive map of the healthy Western adult microbiome, and it currently co-leads one of the “HMP2” Centers for Characterizing the Gut Microbial Ecosystem in Inflammatory Bowel Disease. This is one of many open problems in understanding how human-associated microbial communities can be used as a means of diagnosis or therapeutic intervention on the continuum between health and disease.
Jeffrey W. Imai-Eaton, PhD; Associate Professor of Epidemiology. Dr Imai-Eaton’s research involves developing new mathematical models and statistical methods for estimating HIV epidemic trends, understanding HIV transmission dynamics, and quantifying the demographic impacts of HIV epidemics, particularly in sub-Saharan Africa. He has extensive experience in the collection and analysis of population HIV surveillance data. He is a co- investigator of the ALPHA Network of general population HIV cohort studies (http://alpha.lshtm.ac.uk) and the Manicaland Centre for Public Health Research in northeastern Zimbabwe, where he analyses HIV epidemic trends, leads management and analysis for a cluster-randomised controlled trial of conditional cash transfers to improve health and development of orphans and vulnerable children in Manicaland province, Zimbabwe, and analyses data about sources and causes of adolescent HIV.
Rajarshi Mukherjee, PhD; Assistant Professor in the Department of Biostatistics. Dr. Mukherjee is generally interested in understanding broad aspects of causal inference in observational studies with a focus on learning about fundamental challenges in the statistical analysis of environmental mixtures and their effects on the cognitive development of children. His research is also motivated by applications in large-scale genetic association studies, developing statistical methods to quantify the effects of climate change on human health, and understanding the effects of homelessness on human health.
Rachel Nethery, PhD; Assistant Professor of Biostatistics. Dr. Nethery’s research is focused on the development of statistical methods that enable rigorous and impactful analyses of public health data to quantify health effects of environmental, climate, and social exposures and to characterize and identify drivers of health inequities. Some of her key scientific contributions include co-leading the first paper on the association between air pollution and COVID- 19 outcomes, and conducting an early investigation of the impacts of the US census’s adoption of differential privacy on disease mapping and health inequity studies using census data. Her methodological work is broadly focused on spatio-temporal methods, causal inference, and machine learning, and she has developed methods to address a wide range of challenges in these domains presented by real data applications, including spatial misalignment, instability due to rare outcomes, measurement error, and strong confounding. These methodological developments have enabled detailed characterization and high-resolution prediction of the health impacts of tropical cyclones; estimation of the health effects of large-scale air pollution regulations using nationwide health records data; and estimation of causal effects in the presence of strong confounding. They have also yielded important insights about how standard spatio-temporal statistical methods used in health inequity studies can fail and provide misleading results in the context of error-prone or outdated population count data and/or when applied to data from populations in which the socio-demographic groups of interest are highly residentially segregated.
JP Onella, PhD; Associate Professor of Biostatistics. Dr. Onnela’s research involves two interrelated research themes. In statistical network science, the study of network representations of physical, biological, and social phenomena, his lab develops quantitative methods for studying social and biological networks and their connection to health. In digital phenotyping, or “the moment-by-moment quantification of the individual-level human phenotype in situ using data from personal digital devices, in particular smartphones,” he develops quantitative methods for studying social, behavioral, and cognitive phenotypes. The focus of the team in both statistical network science and digital phenotyping is development of new statistical and quantitative methods, but also co-leading or supporting several applied studies ranging from central nervous system disorders to women’s health. His group has developed and maintains the open source Beiwe research platform for high-throughput smartphone-based digital phenotyping.
John Quackenbush, PhD; Henry Pickering Walcott Professor of Computational Biology and Bioinformatics. Dr. Quackenbush’s research group focuses on methods spanning the laboratory to the laptop that are designed to use genomic and computational approaches to reveal the underlying biology. In particular, he has been looking at patterns of gene expression in cancer with the goal of elucidating the networks and pathways that are fundamental in the development and progression of the disease.