|
Abstract:
|
Having good predictive models is important for planning in clinical research. Simulation methods for enrollment modeling provide the ability to include rich detail into the model, but are slow to fit. Analytical methods, such as those based on the Poisson-gamma distribution, are attractive because they are quick to fit which allows real-time scenario planning with project teams. We will discuss the trade-offs of the different approaches. Additionally, we will discuss approaches to modeling historic performance data to derive quality input values for our predictive models.
|
