Dr. Murray is a Professor in the Department of Biostatistics. She was the Senior Scientific Registry of Transplant Recipients Biostatistician who developed the Lung Allocation Score (LAS), implemented in 2005. She is currently a Biostatistical Editor for the American Journal of Respiratory and Critical Care Medicine and has served previously as an Associate Editor of Biometrics and Lifetime Data Analysis. She is a current member of the Cystic Fibrosis Foundation's Data and Safety Monitoring Board. Teaching awards include a University of Michigan School of Public Health Excellence in Teaching Award in 2003 and a 2001 Teacher of the Year Award from students in the On Job/On Campus Master's in Clinical Research Design and Statistical Analysis Program. Her research interests include development of methodology for correlated censored survival data, nonparametric survival analysis with an informative censoring mechanism, group sequential analysis methods for censored survival and correlated censored survival data, transplantation related survival methodology and quality-of-life adjusted survival analysis. She collaborates heavily with breast cancer, thoracic transplantation and pulmonary researchers both nationally and within the University of Michigan medical community.
- Sc.D., Biostatistics, Harvard University, 1994
- M.S., Biostatistics, Harvard University, 1992
- B.A., Statistics, Mathematical Sciences and English, Rice University, 1990
Research Interests & Projects
My current research interests touch on topics in nonparametric survival analysis, missing data issues, quality of life research, sequential monitoring of survival endpoints and lung allocation. These interests often converge. For instance in survival analysis, missing data problems occur when relevant failure time information is incomplete. This can be especially problematic when health status is at the root of the missing survival information. During my time working as a Senior Statistician with the Scientific Registry of Transplant Recipients, a new lung allocation policy was implemented nationwide that removed lung candidates for transplant based on estimated patient urgency and transplant benefit. As a result, estimation of waitlist survival became more complex due to the nature of the missing waitlist outcomes for those who had been transplanted. In determining the best approach for estimation in this missing data setting, my students and I considered integrated inverse-weighted Cox models as well as models we developed for dependently censored restricted means via pseudo-observations. So far, a multiple imputation approach we developed for dependent censoring has resulted in the best inferences.
I am also engaged in research on correlated survival endpoints. The problem first came to my attention in the context of quality-of-life research through a popular quality-of-life analysis tool called Q-TWiST. This statistic simultaneously analyzes length of time experiencing toxicity, time without symptoms or toxicity and time in relapse in order to assess the tradeoffs between quality-of-life and length of life. In determining the variability of this statistic, I came to understand variability structures that I've since extended for use in other situations. For instance I've developed methodology for nonparametrically analyzing paired censored survival data at a single analysis time or as part of a group sequential monitoring procedure in clinical trials.
I enjoy numerous collaborations with pulmonary researchers in the Medical School. These collaborations began from conversations with OJ/OC students taking a course in Clinical Trials. Since then, I've been involved in pulmonary research nationally with my proudest achievent being the development of the Lung Allocation Score (LAS) that is used in lung allocation today. (Yes, the same score that generated dependent censoring research problems described above!)
- Tayob,N., Murray,S. (2017). Statistical Consequences of a Successful Lung Allocation System – Recovering Information and Reducing Bias in Models for Urgency. Statistics in Medicine. July 10;36 (15):2435-2451.
- Tayob,N., Murray,S. (2016).Nonparametric Restricted Mean Analysis Across Multiple Follow-up Intervals. Statistics and Probability Letters. Feb 1; 109: 152-158. PMID: 26858468.
- Tayob, N., Murray, S. (2015). Nonparametric Tests of Treatment Effect Based on Combined Endpoints for Mortality and Recurrent Events. Biostatistics 16 (1): 73-83. doi: 10.1093/biostatistics/kxu013. Epub 2014 Apr 8. PMID: 24719282.
- Xiang, F., Murray, S. (2012). Restricted Mean Models for Transplant Benefit and Urgency. Statistics in Medicine 561-76.
- Liu, X., Murray, S., Tsodikov, A. (2011). Multiple Imputation Based on Restricted Mean Models for Censored Survival Data Statistics in Medicine 1339-1350 Andrei, A., Murray, S. (2007). Regression models for the mean of the quality-of-life-adjusted restricted survival time using pseudo-observations. Biometrics 398-404.
- Andrei, A., Murray, S. (2006). Estimating the Quality-of-Life-Adjusted Gap Time Distribution of Successive Events Subject to Censoring. Biometrika 343-55.
- Egan, T.M., Murray, S., Bustami, R.T., Shearon, T.H., McCullough, K.P., Edwards, L.B., Coke, M.A., Garrity, E.R., Sweet, S.C., Heiney, D.A., Grover, F.L. (2006). Development of the new lung allocation system in the United States. American Journal of Transplantation 1212-1227.
- Murray, S. (2001). Using Weighted Kaplan-Meier Statistics in Nonparametric Comparisons of Paired Censored Survival Outcomes. Biometrics 361-368.
- Murray, S. (2000). Nonparametric Rank-Based Methods for Group Sequential Monitoring of Paired Censored Survival Data. Biometrics 984-990.
Biostatistical Editor for American Journal of Respiratory and Critical Care Medicine
- Former Associate Editor of Biometrics, Lifetime Data Analysis
- The International Biometric Society/ENAR
- American Statistical Association
- Cystic Fibrosis Foundation Data and Safety Monitoring Board