Table 1. Number of (A) negative trials and (B) positive trials.
|
(A) Number of negative trialsa
| |||||
|---|---|---|---|---|---|
| RR (n=112) |
Power
|
TCR (n=96) |
Power
|
||
| HA: pA | 80% | 90% | HA: pA | 80% | 90% |
| ⩾22.6% | 32 (29%) | 25(22%) | ⩾57.3% | 25 (26%) | 15 (16%) |
| ⩾30.0% | 56 (50%) | 40(36%) | ⩾65.0% | 36 (38%) | 31 (32%) |
| ⩾35.0% | 66 (59%) | 58(52%) | ⩾70.0% | 52 (54%) | 39 (41%) |
| ⩾40.0% | 77 (69%) | 67(60%) | ⩾75.0% | 69 (72%) | 53 (55%) |
|
(B) Number of positive trialsb
| |||||
| RR (n=112) |
Alpha
|
TCR (n=96) |
Alpha
|
||
| H0: p0 | 0.05 | 0.10 | H0: p0 | 0.05 | 0.10 |
| ⩽10.0% | 47 (42%) | 58 (52%) | ⩽45.0% | 35 (36%) | 43 (45%) |
| ⩽15.0% | 35 (31%) | 36 (32%) | ⩽50.0% | 21 (22%) | 33 (34%) |
| ⩽20.0% | 21 (19%) | 28 (25%) | ⩽55.0% | 11 (11%) | 18 (19%) |
| ⩽22.6% | 10 (9%) | 20 (18%) | ⩽57.3% | 9 (9%) | 13 (14%) |
RR=Response Rate; TCR=Tumour Control Rate.
a
Number of trials analysed, which would have been negative, if tested with (RR or TCR as primary end point against four different alternative hypotheses pA with two different powers.
b
Number of trials analysed, which would have been positive, if tested with (RR) or TCR as primary end point against four different null hypotheses p0 with two different significance levels (alpha).