If a company and site have no data of their own, consultants will create a template of similar sites in the same country and
give it a suitable range of variability to accommodate margin of error. For example, if you were to look at doing training
and site recruitment in Germany, the parameters, regulations, and cycle times that are specific to that country would need
to be considered in order to achieve the most accurate results possible. Both approaches share the same goal: to ensure data
is accurate and dynamic so that companies can arrive at realistic and actionable answers. 
Clinical Trial Milestone Summary Table
|
Distributions are then built around the data and put into a decision support model. They capture a company's designated recruiting
process and timeline and ensure that it fits within the boundaries of a given tool set.
Push and pull Once the protocol is prepared and approved and the data fully entered, testing begins in one of two ways. The "push" method
involves putting the information in the simulation technology tool and running multiple scenarios to determine project deadlines.
This produces site enrollment and subject recruitment curves with future confidence intervals. The "pull" method, on the other hand, tells managers the best set of inputs for reaching a given goal and achieving cost savings.
First, project leads define all possible sites and related costs. Then they run a genetic goal-seeking algorithm optimizer
to blend the effects of site qualification requirements, screening volume and losses and startup and subjects costs. This step helps determine the optimum test center combination for a given protocol and often at the lowest possible cost or
the earliest required date since these two measurements are often mutually exclusive. It also takes into account the delay
time from site selection until the start of subject enrollment due to regulatory requirements for each site (see Figure 1).
The information is typically relayed on a scaled up recruitment prediction curve that pinpoints the cycle's milestones and
goals. This includes first subject first visit, first dosage final subject, final dosage, and the end point event or the point
when the final subject exits the study population. Decision support technology can be expanded to include the subject screening protocol for most therapeutic areas. This option
is helpful in cases where data is available on the different criteria that can result in subjects being disqualified from
participating in the clinical trial. This data can be combined with the projected patient screening volume to projected subject
enrollment in the trial. The technology then tracks this population of randomized subjects until they complete the designated
requirements for treatment and monitoring (see Table 1 and Figure 2). Return on investment Simulation will provide the maximum value in cases where there is a choice regarding the selection of trial site enrollment.
Simulation can aid in identifying the group of sites that will best achieve the trial objectives for subject enrollment using
both the trial completion date and the trial cost as the objective functions of the optimization process. Once the enrolling
sites have been chosen and the trial is underway, this tool can continue to provide value by combining both actual and projected
subject enrollment to ensure that the projected schedule is met or exceeded. 
Effect of Nonproducing Test Centers on Subject Enrollment
|
If problems occur and result in subject enrollment falling below expectations, simulation can aid in minimizing its impact
on the overall trial. The comparison between actual and projected enrollment will not only make it possible to identify problems
at an early stage but such a comparison can also can be used to evaluate the best option for recovering from the problem,
by evaluating whether it would be helpful to supplement the original list of enrolling sites.
| 
