Pharsight^® Trial Simulator^™

Test trial designs against expected drug and subject characteristics to predict the probability of success.

Trial Simulator models predict subject responses from subject, drug and disease characteristics.

• Your modeling team and/or expert consultants (like Pharsight Consulting Services) build the drug model based on real data—your internal studies, literature data and public databases—using Phoenix® NLME or Phoenix® WinNonlin® and similar model fitting tools.
• Probability distributions in the models capture variability and uncertainty in key model features, including trial-to-trial variation, between- and within-subject variation, measurement error and other unexplained variability, as well as confidence intervals on expected outcomes.

Repeated, Monte-Carlo simulation of each Trial Simulator scenario generates a distribution of the most likely outcomes, so you know not only the most likely subject and trial outcomes, but also the confidence bounds on those outcomes.

Trial Simulator’s simulation scenarios and replicate-level variability allow your modeling team to simulate outcomes based on different trial design and model settings, reducing the likelihood of failed trails.

• Compare different dosing or sampling schemes, trial sizes or analyses.
• Test your designs against different possibilities for drug potency, subject recruitment or disease severity.

New data are easily incorporated by editing the drug model, or swapping in alternate model segments.

As knowledge accumulates in your Trial Simulator models, the uncertainty in outcomes decreases.

• Know how likely your trial is to succeed.
• Avoid surprises by simulating a variety of possible subject, disease and drug scenarios.
   

• Specify treatment protocols, inclusion criteria, and observations.
• Define realistic subject populations through covariate distribution models.
• Define compliance and dropouts based on drug effects or time.
• Model variability in the actual time doses are taken or observations are made.

• Model drug effects as a function of dose(s), disease processes, time and subject characteristics.
• Estimate measurement error as well as biological and unexplained variability.
• Use the patented, component-driven graphical interface for rapid, intuitive modeling.
• Rely on a broad library of pre-defined model components or write your own FORTRAN.
• Automate simulations through Visual Basic or other scripts.
• Save models as editable text files or XML.
• Swap and test model segments using "include sets."
• Compile simulation models through Compaq Digital Fortran, or Intel Visual Fortran.

• Special version of S-PLUS Professional for Trial Simulator included.
• Perform descriptive statistics, ANOVA, survival analysis, and custom S-PLUS analyses on each trial replicate and as meta-analysis across replicates.
• Plot results through a connection to S-PLUS (provided with the tool).
• Save simulation output in a Microsoft Access database or performance-enhancing binary format.
• Export data to use with R, SAS, S-PLUS®, Phoenix® NLME, Phoenix® WinNonlin®, NONMEM® and more.