Drug Safety Meta-Analysis: Promises and Pitfalls
Tuesday, January 19, 2016
2:00 p.m.-3:00 p.m.
FDA White Oak Location, Bldg. 2 Room 2031
301 405 0285
Michael A. Stoto, PhD, Georgetown University
Michael A. Stoto, PhD, a Professor of Health Systems Administration and Population Health at Georgetown University, is a statistician, epidemiologist, and health services researcher. Dr. Stoto’s research includes methodological topics in epidemiology and statistics including systematic reviews/meta-analysis and other analytical methods for comparative effectiveness research, community health assessment, evaluation methods, and performance measurement. His substantive research interests include public health practice, especially with regard to emergency preparedness; drug and vaccine safety; infectious disease policy; and ethical issues in research and public health practice. Before coming to Georgetown on a full-time basis in August 2006, Dr. Stoto was a Senior Statistician at the RAND Corporation and the Associate Director for Public Health in the Center for Domestic and International Health Security. He previously served as the director of the Institute of Medicine’s (IOM), Board on Health Promotion and Disease Prevention. Dr. Stoto is also an Adjunct Professor of Biostatistics at the Harvard School of Public Health and an adjunct faculty member of the Georgetown Public Policy Institute. He previously served on the faculty of Harvard’s John F. Kennedy School of Government and the George Washington University School of Public Health and Health Services. Dr. Stoto is a Fellow of the American Statistical Association.
Dr. Stoto is a recognized expert on population health and public health assessment. He is a co-editor of the 1997 IOM report Improving health in the community: A role for performance monitoring. His work in this area has included systems-oriented evaluations of public health surveillance systems at the local to global level, addressing both statistical methods and public health practice issues. Dr. Stoto has developed methods for evaluating community health assessments and performance measures and worked with state and local health departments, especially in the Washington DC metropolitan area, to implement these methods. He served on the National Quality Forum’s Population Health Steering Committee and AcademyHealth’s Population Health Advisory Committee. Along with a team of Georgetown students, Dr. Stoto is currently working with MedStar Georgetown University Hospital, and the other hospitals in the MedStar system, to prepare their 2015 Community Health Needs Assessments.
Dr. Stoto is also an expert in public health systems research (PHSR), focusing on applying and developing rigorous mixed-methods approaches to studying and evaluating federal, state, and local public health systems. He has chaired AcademyHealth’s PHSR Interest Group and its methods advisory committee. Much of Dr. Stoto’s recent PHSR work has focused on public health emergency preparedness, and he was the co-Principal Investigator of the CDC-funded Preparedness and Emergency Response Research Center based at the Harvard School of Public Health. Dr. Stoto’s work in this area has focused on regionalization in public health, the evaluation of biosurveillance methods, and the development of methods for assessing emergency preparedness capabilities based on exercises and actual events. Along with Melissa Higdon, HCMP 2007, Dr. Stoto is the co-editor of The Public Health Response to 2009 H1N1: A Systems Perspective, Oxford University Press, 2015 (http://bit.ly/1J3YuG9
Meta-analysis has increasingly been used to identify adverse effects of drugs and vaccines, but the results have often been controversial. In one respect, meta-analysis is an especially appropriate tool in these settings. Efficacy studies are often too small to reliably assess risks that become important when a medication is in widespread use, so meta-analysis, which is a statistically efficient way to pool evidence from similar studies, seems like a natural approach. But, as the examples in this paper illustrate, different syntheses can come to qualitatively different conclusions, and the results of any one analysis are usually not as precise as they seem to be. There are three reasons for this: the adverse events of interest are rare, standard meta-analysis methods may not be appropriate for the clinical and methodological heterogeneity that is common in these studies, and adverse effects are not always completely or consistently reported. To address these problems, analysts should explore heterogeneity and use random-effects or more complex statistical methods, and use multiple statistical models to see how dependent the results are to the choice of models.