Table 1.

Example trial demonstrating data gathered at each interim analysis, probability each treatment arm offers the highest and lowest response rate, randomization probabilities for the next 100 patients, and the predictive probability of identifying a ‘best’ or ‘worst’ treatment at the maximum sample size.

N	Observed Responses/Randomized			Probability tmax Probability tmin			Randomization Probs for Next 100 Pts			Pred Prob @ Max
	fPHT	LVT	VPA	fPHT	LVT	VPA	fPHT	LVT	VPA
300	51/100 51%	55/100 55%	64/100 64%	0.025 0.70	0.092 0.29	0.88 0.014	0.12	0.22	0.66	0.71
New	6/11 55%	19/26 73%	39/63 62%
400	57/111 51%	74/126 59%	105/163 64%	0.010 0.87	0.16 0.13	0.83 0.008	0.094	0.34	0.57	0.50
New	5/12 42%	20/38 53%	34/50 68%
500	62/123 50%	94/164 57%	139/213 65%	0.004 0.88	0.056 0.12	0.94 0.002	0.080	0.23	0.69	0.59
New	3/3 100%	17/28 61%	55/69 80%
600	65/126 52%	111/192 58%	194/282 69%	0.000 0.87	0.008 0.13	0.992 0.00	--	--	--	---