Table 3. Study design characteristics stratified by type of trial population.
| Trials in ARV-naïve patients (n = 23) | Trials in ARV-experienced patients (n = 19) | |
| NI margin (%) | 7 (1) | |
| 10 (6) | 10 (3) | |
| 12 (11) | 12 (8) | |
| 13 (1) | 12.5 (1) | |
| 15 (3) | 14 (1) | |
| 25 (1) | 15 (3) | |
| NA* (1) | 20 (1) | |
| NA** (1) | ||
| Method of selection of NI margin | Guidelines (2) | |
| Investigator’s assumption (2) | ||
| Other publications or reviews (2) | Other publications or reviews (2) | |
| Calculated by investigator based on previous trials’results (2) | Guidelines and calculated by investigator based on previous trials’ results (1) | |
| Investigator’s assumption and other publications orreviews (1) | Not clear (16) | |
| Not clear (14) | ||
| Sample size calculation used NI margin | No (15) | No (6) |
| Yes (8) | Yes (13) | |
| 1 or 2 sided confidence intervals | 1-sided (1) | 1-sided (4) |
| 2-sided (22) | 2-sided (15) | |
| Blinding method | Open label (15) | Open label (18) |
| Double blind (8) | Double blind (1) | |
| Statistical analysis | Intention-to-treat (10) | Intention-to-treat (9) |
| Per protocol (1) | Per protocol (1) | |
| Intention-to-treat and per protocol (10) | Intention-to-treat and per protocol (9) | |
| Not clear (2) | ||
| Main conclusion based on | Intention-to-treat (19) | Intention-to-treat (12) |
| Per protocol (2) | Per protocol (2) | |
| Not clear(2) | Intention-to-treat and per protocol (5) |
*
Upper 95% confidence limit for Hazard ratio was no greater than 1.18; **Upper 95% confidence limit for the Hazard ratio was less than 1.40.