NAME: Michael J. Daniels BIOGRAPHICAL SKETCH era COMMONS USER NAME: MDANIELSUFL POSITION TITLE: Professor and Chair, Department of Statistics & Data Sciences; Professor, Department of Integrative Biology EDUCATION/TRAINING INSTITUTION AND LOCATION DEGREE Completion Date MM/YYYY FIELD OF STUDY Brown University, Providence, RI AB 05/1991 Applied Math Harvard University, Boston, MA ScD 08/1995 Biostatistics A. Personal Statement I am currently a tenured full professor in the Department of Statistics & Data Sciences and the Department of Integrative Biology (and chair of the former department) at the University of Texas at Austin. My areas of expertise are the development of methods for the analysis of longitudinal data and missing data in longitudinal studies, Bayesian causal inference, and estimation of dependence. I am currently the PI on an R01 entitled Bayesian methods for missingness and causality in cancer and behavior studies. I have had continuous funding as PI since 2001 (through 2019). I am also the PI on a subcontract for an R01 awarded to University of Pennsylvania entitled Nonparametric Bayesian approaches for causal inference among numerous funded collaborative proposals. I will serve on the Advisory Committee for the Scientific and Technical Core of the PRC, providing my expertise in statistical modeling and analysis and data science in general. My research over the next few years will continue to focus on missing data and causal inference via current funded NIH projects as well as a pending application to specifically address missing data in electronic health records. I am also working on a project (and have one pending proposal) regarding understanding the impact of CMS rules and preventable hospital complications as well as understanding the causes of differential fall rates in hospitals. Both involve large observational databases. My work fits within the PRC primary research area of Population Health. B. Positions and Honors Positions and Employment 1991-1995 Teaching Fellow in Biostatistics, Harvard School of Public Health 1995-1997 Visiting Assistant Professor, Carnegie-Mellon University 1997-2002 Assistant Professor, Iowa State University 2002-2007 Associate Professor of Statistics, University of Florida 2005-2012 Executive Committee and Leader of Biostatistics, Data Management, and Methodology Core, Institute on Aging, University of Florida 2006-2007 Associate Professor of Biostatistics, University of Florida 2006-2008 Program Director of Biostatistics, College of Public Health and Health Professions 2007-2008 Professor of Biostatistics and Statistics 2008-2009 Interim Chair, Department of Statistics 2009-2012 Professor and Chair, Department of Statistics, University of Florida 2012-present Professor, Department of Integrative Biology and Department of Statistics & Data Sciences University of Texas at Austin 2014-present Chair, Department of Statistics & Data Sciences Other Experience and Professional Memberships 1997-2000 Consultant, HIV RNA Surrogate Marker Collaborative Group, Harvard University 1998-2000 Consultant, Perinatology Group in Des Moines 1998-2002 Corresponding Editor for IMS Bulletin 1999-2001 Member of Nextran Microbiologic Safety Board for Xenotransplantation 1999-2001 Consultant, Nextran 1999 Member of EPA FIFRA Scientific Advisory Panel 1999 Review Committee for ENAR Student Paper Awards
2000 Co-organized ENAR invited session on informatively missing data 2001 NIH Study Section (Small grants program for cancer) 2001- Consultant for EPA SAB on water quality 2002 Organized ENAR invited session on longitudinal data 2003-2014 Associate Editor for Biometrics 2004 Ad hoc Member of NIH BMRD study section 2004-2005 HPSS Section Representative to ENAR 2005 2005 Organized JSM invited session 2005-2012 Associate Editor for Journal of the American Statistical Association 2006 American Statistical Association Biometrics Section Program Chair for JSM 2007-2010 American Statistical Association Biometrics Section Council of Sections Representative 2007-2014 Associate Editor for Statistics & Probability Letters 2008 American Statistical Association Section on Bayesian Statistical Science Program Chair-Elect for JSM 2009 American Statistical Association Section on Bayesian Statistical Science Program Chair for JSM 2009 Consultant, Asthmatx 2009-present Member of KL2 Mentoring Committee, University of Florida CTSI 2010 Program Chair for ENAR 2010 2010-2011 Treasurer, ENAR 2010-2014 Associate Editor for Biostatistics 2011-2013 Treasurer, International Society for Bayesian Analysis (ISBA) 2014 Chair, Biometrics Section of the ASA 2014-2016 Member of ENAR RECOM (elected) 2015 Past Chair, Biometrics Section of the ASA 2015-2017 Co-editor of Biometrics 2015-present Member, Committee on Applied and Theoretical Statistics (CATS), National Academies of Science Honors 1992-1995 Howard Hughes Medical Institute Predoctoral Fellowship in Biological Sciences 1995-1997 National Research Service Award in Psychiatric Statistics (NIMH) 2001 Co-author of paper awarded the ASA Statistics in Epidemiology Section Young Investigator's Award 2001 Young Investigator Travel Award for VIII Latin American Congress on Probability and Mathematical Statistics 2002 Provost Research 'Star' Award, Iowa State University 2004 College of Liberal Arts and Sciences Travel Award for travel to Chile 2007 American Statistical Association fellow (elected) 2014 Lagakos Distinguished Alumni Award, Department of Biostatistics, Harvard 2014 Appointed as co-editor of Biometrics C. Contributions to Science (over 110 total publications) 1. My early work sparked from my dissertation focused on construction of hierarchical models in the settings of health services research, understanding environmental and genetic contributions to IQ, and metaanalysis for the assessment of potential surrogate markers. The latter two have been quite influential and well cited. I have also started a new collaboration extending this surrogate marker work. a. Devlin B, Daniels M, Roeder K. (1997) Heritability of IQ. Nature. 388:468-471. b. Daniels M, Hughes M. (1997) Meta-Analysis for the evaluation of potential surrogate markers. Statistics in Medicine. 16:1965-1982. c. Hughes MD, Daniels MJ, Fischl MA, Kim S, Schooley RT. (1998) CD4 cell count as a surrogate endpoint in HIV clinical trials: a meta-analysis of studies of the AIDS Clinical Trials Group. AIDS. 12:1823-1832. d. Daniels M, Gatsonis C. (1999) Hierarchical Generalized Linear Models in the Analysis of Variations in Health Care Utilization. Journal of the American Statistical Association. 94:29-42. 2. After graduating from Harvard, I spent two years at Carnegie Mellon and five years at Iowa State
University. During this time, I started to work (first with Rob Kass, then with Mohsen Pourahmadi) on prior distributions for covariance matrices focusing on different parameterizations (often in the setting of longitudinal data). In particular, the modified Choleski decomposition of a covariance matrix and the partial autocorrelation parameterization of a correlation matrix. This work has also been very influential and well cited and formed a key component of my NIH grant CA85295. Estimation of these matrices is sometimes of interest in and of itself but is also very important in estimating mean parameters in settings with missing data and outcome-dependent follow-up. Very recent work in this area has extended some of these approaches to estimation of high dimensional covariance matrices and constructed priors using Bayesian nonparametric approaches. a. Daniels M.J., Kass R.E. (2001) Shrinkage estimators for covariance matrices. Biometrics, 57: 1173-1184. b. Daniels, M. Pourahmadi, M. (2002) Bayesian analysis of covariance matrices and dynamic models for longitudinal data. Biometrika, 89, 553-566. c. Daniels, M.J. and Pourahmadi, M. (2009) Modeling covariance matrices using partial autocorrelations. Journal of Multivariate Analysis, 100, 2352-2363. PMC2748961 d. Wang, Y. and Daniels, M.J. (2014) Computationally efficient banding of large covariance matrices for ordered data and connections to banding the inverse Cholesky factor. Journal of Multivariate Analysis, 130, 21-26. PMC 4136395 3. In the late 1990's, I also started to engage in some important research in air pollution modeling, both spatial models and health effects models. The latter has been part of important evidence that impacts changing air pollution regulations. This work was done with collaborators at Iowa State and researchers at Johns Hopkins University. a. Daniels M, Dominici F, Samet J., Zeger, S. (2000) Estimating particulate matter-mortality doseresponse curves and threshold levels: An analysis of daily time series data for the 20 largest U.S. cities (with invited commentary). American Journal of Epidemiology, 152, 397-406. b. Dominici, F., Daniels, M., Zeger S., Samet J. (2002) National models for estimating the effect of particulate matter on mortality in U.S. cities. Journal of the American Statistical Association, 97:100-111. c. Dominici F., McDermott A., Daniels, M., Zeger S.L, Samet J.M. (2005) Revised analyses of the National Morbidity, Mortality, and Air Pollution Study: mortality among resident of 90 cities. Journal of Toxicology and Environmental Health Part A., 68, 1071-1092. d. Daniels, M., Zhou, Z, and Zou, H. (2006) Conditionally specified space-time models for multivariate processes. Journal of Computational and Graphical Statistics, 15, 157-177. 4. In the early 2000's, with my colleagues Joe Hogan and Dan Scharfstein, I started to work on Bayesian models for missing data, culminating in an important research monograph in 2008 that outlines Bayesian approaches for missing data in longitudinal studies. This book has been quite influential and well cited. We have also begun to examine how to use the factorization proposed in the book to construct Bayesian nonparametric models for the observed data distribution and restrictions, with sensitivity parameters, to identify the 'extrapolation' distribution, for which the observed data provides no information. This work on missing data was funded by competing continuations of NIH R01 CA 85295 and a current R01, CA 183854. a. Daniels M, Hogan J. (2000) Reparameterizing the pattern mixture model for sensitivity analysis under informative dropout in longitudinal studies. Biometrics, 56, 1241-1249. b. Daniels, M.J. and Hogan, J.W. (2008) Missing data in longitudinal studies: Strategies for Bayesian Modeling and Sensitivity Analysis. Chapman & Hall (CRC Press). c. Wang, C., Daniels, M.J., Scharfstein, D.O., and Land, S. (2010) A Bayesian Shrinkage Model for Incomplete Longitudinal Binary Data with Application to the Breast Cancer Prevention Trial. Journal of the American Statistical Association, 105, 1333-1346. PMC3079242. d. Linero, A. and Daniels, M.J. (2015) A flexible Bayesian approach to monotone missing data in longitudinal studies with informative missingness with application to an acute schizophrenia clinical trial. Journal of the American Statistical Association, 110. PMC 4517693 5. My work in missing data naturally led me to begin to work on problems in causal inference, which shares many similar features and identification strategies. My work here has used principal stratification to properly make inferences on quality of life data in (complex) cancer clinical trials. We have also started to work on new identification strategies for causal mediation, which is very important in behavioral clinical trials, and
using Bayesian nonparametric models in these settings. This research has been funded by my previous R01, CA 85295, my current, CA 183854, another R01 GM 112327 (PI: Jason Roy at UPENN), and a previous challenge grant, RC1 AA019186 (PI: Joe Hogan at Brown University). a. Lee, K., Daniels, M.J., and Sargent, D. (2010) Causal effects of treatments for informative missing data due to progression/death. Journal of the American Statistical Association, 105, 912-929. PMC3035160. b. Daniels, M.J., Roy, J., Kim, C., Hogan, J.W., and Perri, M.G. (2012) Bayesian inference for the causal effect of mediation. Biometrics, 68, 1028-1036. PMCID 3927554 c. Lee, K. and Daniels, M.J. (2013) Causal Inference for Bivariate Longitudinal Quality of Life Data in Presence of Death Using Global Odds Ratios. Statistics in Medicine, 32, 4275-4284. PMC3935993 d. Josefsson, M., de Luna, X., Daniels, M.J., and Nyberg, L. (2015) Causal inference with longitudinal outcomes and non-ignorable dropout: Estimating the effect of living alone on cognitive decline. JRSS-C, 65, 131-144. PMC 4733472. D. Research Support Ongoing Research Support T32LM012414 (M.J. Daniels, PI) 04/01/16-03/31/21 National Library of Medicine Predoctoral Training in Biomedical Big Data Science The major goals of this pre-doctoral training program is for the trainee to become an expert in Statistics, Computer Science, Computational Science, Engineering, and Mathematics (CSEM), or Biology while also obtaining essential training in all three core areas (statistics, computer science, and biology). Responsibilities: Oversee the graduate training program, including review of trainees and selection of fellowship participants. R01CA183854 (M.J. Daniels, PI) 04/18/14-02/28/19 National Cancer Institute Bayesian approaches for missingness and causality in cancer and behavior studies The major goals of this project are to develop new Bayesian methods for missingness and causal inference problems in cancer and behavioral trials. Responsibilities: Lead the development and coding of new methods, and lead dissemination. R01HS023783 (T.M. Waters, PI) 04/01/15-03/31/18 Agency for Healthcare Research and Quality Hospital Responses to Medicare Readmission Penalties The goal of this project is to assess the impact of the Hospital Readmissions Reduction Program. Role: Co-Principal Investigator R01AI108441 (J.W. Hogan, PI) 09/15/14-10/31/19 National Institute of Allergy and Infectious Disease Optimizing HIV Treatment Monitoring under Resource Constraints The proposed research will develop, evaluate, and implement methods to optimize monitoring of antiretroviral therapy in resource-limited settings (RLS) that have diverse HIV viral load (VL) testing availability. Role: Co-investigator U01GM087719 (A.P. Galvani, PI) 06/01/09-04/31/19 Dynamic Data-Driven Decision Models for Infections Disease Control The goal is to use models to study impacts of individual-level decision on disease dynamics and control Role: Senior Personnel R01GM112327 (J.A. Roy, PI) 09/10/14-06/30/18 National Institute of General Medical Sciences Non-Parametric Bayesian Methods for Causal Inference The major goals of this project are to develop Bayesian nonparametric approaches to several problems in
causal inference Role: Co-Principal Investigator R18HL112720 (M.G. Perri, PI) 08/15/13-04/30/18 National Heart, Lung, and Blood Institute Rural lifestyle eating and activity program (Rural LEAP) The major objective of this study is to conduct a randomized controlled trial evaluating the effects of three interventions to lose and maintain body weight loss in obese adults from underserved rural areas. R01AR056973 (K.H. Vandenborne, PI) 04/01/09-08/31/20 National Institute of Arthritis and Musculoskeletal and Skin Diseases Magnetic resonance imaging and biomarkers for muscular dystrophy The overall objective of this proposal is to validate the potential of noninvasive magnetic resonance imaging (MRI) and spectroscopy (MRS) to monitor disease progression and to serve as a surrogate outcome measure for clinical trials in Duchenne muscular dystrophy (DMD). Completed Research Support R01CA85295 (M.J. Daniels, PI) 05/01/00-11/30/14 National Cancer Institute Bayesian methods for (incomplete) longitudinal cancer data The major goals of this project were to develop new models and methodology for correlation matrices for incomplete longitudinal data and to develop new Bayesian methods for various settings with informative censoring. Responsibilities: Oversee all aspects of the data collection, analysis of data, report writing, and supervision of research staff. R01HS020627 (T. M. Waters) 08/01/11-07/31/14 Agency for Healthcare Research and Quality Responses to Medicare s Nonpayment for Preventable Hospital Complications This project examined the impact of Medicare's new nonpayment rule on hospital behavior related to four of the eight conditions identified by CMS as preventable: catheter-associated urinary tract infection rates, central line- associated blood stream infection rates, falls, and hospital-acquired pressure ulcers. Investigators also assessed how hospital responses may vary depending on particular circumstances, such as financial health and market conditions. R18 HL087800-01 (M.G. Perri, PI) 06/01/08-05/31/13 National Heart, Lung, and Blood Institute Rural lifestyle intervention treatment effectiveness trial (Rural LITE) The major objective of this study was to conduct a randomized controlled trial evaluating the effects of three doses of lifestyle treatment on 2-year changes in body weight in obese adults from underserved rural areas. P30AG028740 (M. Pahor, PI) 07/01/06-03/31/17 National Institute on Aging Claude D. Pepper Older Americans Independence Center (OAIC) The major goals of this program are to assess the mechanisms leading to sarcopenia and functional decline, and to develop and test interventions for the treatment and prevention of physical disability in older adults. Role: Leader Biostatistics and Data Management Core (2007-2012) Responsibilities: Lead the core to ensure proper and efficient designs and analyses of all studies, train junior investigators.