Table 1.

$\theta =$ a weighted average of stratum-specific log hazard ratios, each comparing BiV versus RiV pacing; PP₀$=$ posterior probability that the study objective has been met, PRR $=$ posterior probability the study objective has not been met.

Decision boundaries
	Conclude objective is met and stop study early	Conclude that sample size is sufficient to continue	Determine that sample size is insufficient but elect not to increase sample size	Conclude that sample size must be increased in increments of 175	Stop study for safety
First Interim Analysis	$P{P}_0>0.99$	$0.90\le P{P}_0\le 0.99$	PRR > 0.9	$P{P}_0<.90$ and $\mathit{PRR}\le 0.9$	P ($\theta >0|\text{data}, \text{prior}$) $\ge$ 0.90
Sample Size Re-estimation Phase	N/A	$0.90\le P{P}_0$	PRR > 0.9	$P{P}_0<.90$ and $\mathit{PRR}\le 0.9$	N/A
Second Interim Analysis	$P{P}_0>0.99$	If neither the outcome in column 2 nor the outcome in column 6 occurs, then the study will continue with the current sample size.	P ($\theta >0|\text{data}, \text{prior}$) $\ge$ 0.90

NOTE: See Curtis et al. (2013) for details.