不良研究所

tibor.schuster [at] mcgill.ca (Tibor Schuster, PhD)

Assistant Professor, Primary Care Biostatistics

Biography: Dr Schuster accomplished his early academic and professional education at the Ludwig Maximilian University (LMU) of Munich and the Institute for Medical Statistics and Epidemiology at the Technical University of Munich (TUM). He obtained his doctorate in Biostatistics from the Faculty of Mathematics, Informatics and Statistics at the LMU. Subsequently, he received a post-doctoral award from the Canadian Network of Observational Drug Effect Studies (CNODES) and carried out a post-doctoral fellowship in pharmacoepidemiology at the Department of Epidemiology, Biostatistics and Occupational Health, 不良研究所 and the Centre for Clinical Epidemiology, Lady Davis Institute for Medical Research in Montreal. He continued with a research fellowship at the Murdoch Childrens Research Institute in Melbourne where he was acting Director of Biostatistics at the newly established Melbourne Children鈥檚 Trial Centre in 2015. In August 2016, Dr Schuster started a tenure-track faculty position as Assistant Professor at the Department of Family Medicine. He is nominated for the Tier II Canada Research Chair in Biostatistical Methods for Primary Care Research. Dr Schuster taught biostatistical methods at renowned institutions in Germany, Canada and Australia. He acted as supervisor and mentor for graduate and doctoral students in the field of biostatistics, epidemiology and bio-medical research.

Research Interests: Dr Schuster鈥檚 main methodological interests are in the development and application of causal inference methods for the design and analysis of cluster randomized controlled trials and observational research studies based on administrative or electronic medical / health record data.

Randomised controlled trials are considered to be the gold standard for inference on intervention effects in bio-medical research and health sciences. If rigorously conducted, such trials yield unbiased and consistent estimates of average intervention effects in relevant target populations.

However, systematic patient drop-out and missing data issues occur frequently and can lead to substantial bias in effect estimation if not considered appropriately. Furthermore, treatment cross-over, non-adherence or non-compliance as well as subsequent (often event-driven) changes of individual treatment protocols require sophisticated analysis strategies to enable estimation of meaningful population-level effects.

Recent methodological developments, in particular so called causal inference approaches, provide promising solutions to these problems. However, for an effective implementation, consideration of relevant data to be collected is compulsory at the design stage, which is a shortcoming of many past and currently ongoing research studies. Furthermore, the immense amount of emerging data due to modern electronic sources requires computational and algorithmic intelligence that goes beyond conventional statistical modelling. Dr Schuster therefore encourages the incorporation and application of modern Machine Learning techniques in conjunction with fundamental principles of Causal Inference.

His specific methodological interests are in:

Design and analysis of Cluster Crossover Trials, in particular Stepped Wedge Designs
Causal Inference methods such as Marginal Structural Models and Targeted Learning
Theory and applications of Personalized Medicine and Dynamic Treatment Regimens such as Sequential, Multiple Assignment, Randomized Trial (SMART) designs
Bayesian adaptive and sequential study designs, in particular Internal Pilot Studies and so called Platform Trials
Confounder selection and adjustment in high dimensional covariate settings
Modern methods for Statistical and Machine Learning and Data Visualization

Keywords: Cluster Randomized Controlled Trials, Electronic Medical / Health Record Data, Causal Inference, (Pharmaco-) Epidemiology, Biostatistics, Data Science

Selected Recent Publications: Dr. Schuster has contributed with numerous developments and applications of quantitative methods covering a broad range of diagnostic, clinical therapeutic and epidemiological studies in different disciplines such as cardiology, oncology, internal medicine, anaesthesiology, toxicology, psychology, psychiatrics, surgery, medical engineering and primary care. Over the last ten years, he has (co-) authored over 250 publications in peer-reviewed international scientific journals. A selection of recent contributions is listed below

Pang M, Schuster T, Filion KB, Schnitzer ME, Eberg M, Platt RW.
Effect Estimation in Point-Exposure Studies with Binary Outcomes and High-Dimensional Covariate Data - A Comparison of Targeted Maximum Likelihood Estimation and Inverse Probability of Treatment Weighting.
Int J Biostat. 2016 Nov 1;12(2). PMID: 27889705.
~
Schuster T, Lowe WK, Platt RW.
Propensity score model overfitting led to inflated variance of estimated odds ratios.
J Clin Epidemiol. 2016 Dec;80:97-106. PMID: 27498378.
~
Pang M, Schuster T, Filion KB, Eberg M, Platt RW.
Targeted Maximum Likelihood Estimation for Pharmacoepidemiologic Research.
Epidemiology. 2016 Jul;27(4):570-7. PMID: 27031037.
~
Davidson AJ, Disma N, de Graaff JC, Withington DE, Dorris L, Bell G, Stargatt R, Bellinger DC, Schuster T et al.
Neurodevelopmental outcome at 2 years of age after general anaesthesia and awake-regional anaesthesia in infancy (GAS): an international multicentre, randomised controlled trial.
Lancet. 2016 Jan 16;387(10015):239-50. PMID: 26507180.
~
Austin PC, Schuster T, Platt RW.
Statistical power in parallel group point exposure studies with time-to-event outcomes: an empirical comparison of the performance of randomized controlled trials and the inverse probability of treatment weighting (IPTW) approach.
BMC Med Res Methodol. 2015 Oct 15;15:87. PMID: 26472109.
~
Chevance A, Schuster T, Steele R, Tern猫s N, Platt RW.
Contour plot assessment of existing meta-analyses confirms robust association of statin use and acute kidney injury risk.
J Clin Epidemiol. 2015 Oct;68(10):1138-43. PMID: 26092287.
~
Schuster T, Pang M, Platt RW.
On the role of marginal confounder prevalence - implications for the high-dimensional propensity score algorithm.
Pharmacoepidemiol Drug Saf. 2015 Sep;24(9):1004-7. PMID: 25866189.