Using Pharsight® Trial Simulator™

Workflow in Pharsight Trial Simulator

The flow chart below illustrates one way of working with the Trial Simulator.

In the flow chart above, the first stage is represented by steps 1 through 4, the second stage by steps 5 through 10.

At step 1, you enter settings into the drug model editor, using the Drug Model Debug features to confirm appropriate model behavior, then create a model for the patient population and study protocol. In step 2, you test the settings by running a small-scale simulation. Next, examine the data (step 3) and correct any mistakes (step 4). You might repeat the cycle, correcting, simulating and checking output until satisfied that the settings in the editors accurately reflect current information on the trial.

You can then vary the design and try different assumptions. How will the trial behave if the dropout rate is higher? What will be the effect of a wider standard deviation in efficacy? What if drug potency is lower than expected? Higher? How much more reliability can be purchased by increasing the trial enrollment, or using a different analysis plan?

After collecting all such scenarios to explore (step 5), you run a batch of simulations, generally 100 or more replicates per scenario (step 6). This generates a simulation database to analyze (step 7). Depending on your needs, you might compare the percentage of successful trials for each scenario (probability of success) or the proportion of trials producing a certain result with 95% certainty. The results could lead to modifications of the underlying trial design (step 9) and a repeat of all or portions of the process for further exploration.

Synthesis of the study is ongoing; the model, population, protocol, and protocol deviations can be updated throughout the development process to reflect new information, assumptions, and hypotheses about drug action.

»Back to Trial Simulator Guided Tour