CP - I think I have figured out where you're confused regarding my model. The end date for full enrollment is fixed at Dec 2015, the start is fixed at December 31, 2013, the site activation profile is fixed because it is known. The only variable that can adjust is the average time it takes to enroll a patient. The key is to end with 592 patients enrolled in December 2015.
So as extreme examples, if all sites opened 1 January 2014, the average enrollment time would have to be much longer to only have 592 enrolled in December 2015. If all sites opened in July 2015, that average would have to be much smaller to get everyone enrolled by December 2015.
you say:
You do understand that a site that has been open 7.6 months now, it probably doesn't have an event? Again, site opens, wait 4 months for first patient, wait ANOTHER 7.6 months if the patient is assigned to the control group, and additional 12 months if they are assigned the investigational drug.
I'm a visual guy, so please provide a time plot of your enrollment and events. That might help.