Fig. 9: Case-count VE substantially underestimates VE in a substantial portion of trials; PoDBAY eliminates most of these outliers.

Point estimates of CoP-based VE and case-count VE in 1000 simulated trials, with true VE set to 40% and a number of subjects N = 3760. Simulated trials with case-count VE < 9% are shown in red.