American Statistical Association
In comparative effectiveness studies, we are interested in estimating the causal effect of a new treatment relative to a standard one on patientsí outcomes, such as the health related quality of life (HRQOL), after a certain time period of taking the treatment. In these studies, some patients may die before their outcomes can be measured, and hence, their measures are not well defined.
For example, in a randomized trial comparing a new drug with a traditional drug, we are interested in estimating the causal effect of the new drug relative to the old drug on the health related quality of life (HRQOL) after a certain time period of taking the treatment. But some of the patients in the study may die before their outcomes are measured. One main issue with estimation of such the causal effect is parameter identifiability. We first show that the causal effect of interest is not identifiable non-parametrically under the commonly made regularity conditions in the causal inference literature. We then introduce a concept of using one additional baseline covariate associated with principal strata to make the causal effect identifiable. After we derive the sufficient conditions for identifiability of the causal effect, we then propose a non-parametric method for estimating the causal effect of interest. Our simulation studies show the proposed estimation methods work well in finite-sample sizes. Finally, we apply our approach to a data set from Southwest Oncology Group (SWOG) clinical trial on the effectiveness of the treatment of docetaxel and estramustine (DE) with mitoxantrone and prednisone (MP) in patients with metastatic, androgen-independent prostate cancer.
Dr. Zhou received his B.Sc. in mathematics from Sichuan University in 1984, his M.Sc. in statistics from the University of Calgary in 1987, and Ph.D. in biostatistics from Ohio State University in 1991. He finished a two-year post-doc fellowship in biostatistics at Harvard University Medical School in 1993. He joined the Division of Biostatistics at Indiana University School of Medicine as an Assistant Professor in 1993, and was promoted to an Associate Professor in 1997. He then took the Director of Biostatistics Unit position in the U.S. Department of Veterans Affairs Puget Sound Health Care System. He was promoted to a Full Research Professor in 2003 and became a Full Professor in 2005 in the Department of Biostatistics at University of Washington. He was elected to the International Statistical Institute in 1998 and became a Fellow of the American Statistical Association in 2004. He was the Chair of the Section on Statistics in Epidemiology of the American Statistical Association in 2004. Currently, Dr. Zhou is Chair-Elect of the Section on Health Policy Statistics of the ASA. He served on a Study Section of National Institute of Health and on an Advisory Committee of the Center for Devices and Radiological Health at the U.S. Food and Drug Administration (FDA). He is a member of the Oncologic Drugs Advisory Committee in Center for Drug Evaluation and Research, U.S. Food and Drug Administration (FDA). He received a Research Career Scientist Award (RCS) from the U.S. Federal Government Department of Veterans Affairs in 2007. Along with two colleagues, he wrote the first textbook on statistical methods in diagnostic medicine, published by Wiley & Sons, Statistical Methods in Diagnostic Medicine.
|Date:||Thursday, April 15, 2010|
|Time:||4:00 - 5:00 P.M.|
Mailman School of Public Health
Department of Biostatistics
722 West 168th Street
Biostatistics Computer Lab
6th Floor - Room 656
New York, New York