Skip to main content

Table 6 Original and revised sample size substantiation

From: A pivotal registration phase III, multicenter, randomized tuberculosis controlled trial: design issues and lessons learnt from the Gatifloxacin for TB (OFLOTUB) project

  Original calculation Revised calculation
Outcome Relapses at 24 months after end of treatment ‘Unfavourable’ events at 24  months after end of treatment (MITTa) ‘Unfavourable’ events 24  months after end of treatment (PPb)
Events in control arm (%) 5% 20% 14%
1-sided significance level 2.5% 2.5% 2.5%
Power 80% 80% 80%
Non-inferiority margin δ 3% 6% 6%
Patients overall (n) 1,656 1,394 1,050
Adjustment for LTFUc 20% 15% d 27% d
Patients after adjustment for LTFU (n)d 2,070 1,640 1,438
  1. aModified intent to treat.
  2. bPer protocol.
  3. cLost to follow-up.
  4. dIn the revised sample size calculation, patients LTFU are re-classified as unfavorable or non-assessable. Therefore LTFU for the revised calculation refers to the adjustment made on the percentage of patients non-assessable.