Repeating: information lies in events (deaths), not enrollees. Extending to 42 days may capture 10% to 20% more deaths - it's like increasing sample size by that much but only adds two weeks to the schedule.