Table 1.
Randomised questions corresponding to number needed to benefit, harm and inconclusive relative to threshold NNT by haematological cancer type
| NNT* | Haematological cancer randomised questions (n=45) | |||
| ALL (n=24) |
Lymphoma (n=13) |
AML (n=6) |
Mixed diagnoses† (n=2) |
|
| NNTB | 7 (29.2%) | 0 (0%) | 3 (50.0%) | 1 (50.0%) |
| NNTB<threshold NNT | 2 (28.6%) | 0 (0%) | 1 (33.3%) | 1 (100.0%) |
| NNTB lower confidence limit≥threshold NNT | 2 (100.0%) | 0 (0%) | 1 (100.0%) | 0 (0%) |
| NNTB>threshold | 5 (71.4%) | 0 (0%) | 2 (66.7%) | 0 (0%) |
| NNTB upper confidence limit≤threshold NNT | 4 (80.0%) | 0 (0%) | 1 (50.0%) | 0 (0%) |
| NNTB=threshold NNT | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Inconclusive NNT | 16 (66.7%) | 11 (84.6%) | 3 (50.0%) | 1 (50.0%) |
| NNTH | 1 (4.2%) | 2 (15.4%) | 0 (0%) | 0 (0%) |
Threshold NNT corresponds to the inverse of the absolute difference (ie, delta value) as reported in the sample size calculation.
*Denominator for indented corresponds to above row.
†Mixed diagnoses refer to RCTs where more than one haematological cancer was included.
ALL, acute lymphoblastic leukaemia; AML, acute myeloid leukaemia; NNT, number needed to treat; NNTB, number needed to treat to benefit; NNTH, number needed to treat to harm; RCTs, randomised controlled trials.