Statistical Modelling in Clinical Trials
Vladimir Anisimov (GlaxoSmithKline)
Friday 18th May, 2012 15:00-16:00 Maths 203
In the talk the advanced analytic statistical techniques for modelling and predicting stochastic processes describing the behaviour in time of different stages in Phase III multicentre clinical trials are discussed. The methodology for predictive patient's recruitment modelling is developed [1,3]. The patient's flows are modelled by using Poisson processes with random delays and gamma distributed rates. The predictive bounds for the number of recruited patients over time are derived using ML and Bayesian techniques and asymptotic approximations. The optimal number of clinical centres and trial performance can be also evaluated.
The technique for predicting in time the number of different events in trials with waiting time to response is developed . The expressions for predictive distributions are derived and used in oncology trials. The technique for predicting in time randomisation processes in centres/regions is developed and the impact of stratification on sample size is investigated [2,3]. Using these results, a risk-based supply modelling tool is developed.
Software tools in R supporting these techniques are created. These tools are implemented on R&D GSK level and already led to significant benefits and cost savings.
- V. Anisimov, V. Fedorov, Modeling, prediction and adaptive adjustment of recruitment in multicentre trials. Statistics in Medicine, Vol. 26, No. 27, 2007, pp. 4958-4975.
- V. Anisimov, Effects of unstratified and centre-stratified randomization in multicentre clinical trials. Pharmaceutical Statistics, v. 10, iss. 1, 2011, pp. 50-59.
- V. Anisimov, Statistical modeling of clinical trials (recruitment and randomization), Communications in Statistics - Theory and Methods, 40: 19-20, 2011, pp. 3684-3699.
- V. Anisimov, Predictive event modelling in multicentre clinical trials with waiting time to response, Pharmaceutical Statistics, v. 10, iss. 6, 2011, pp. 517-522.