Effects to Be Assumed in Each Study to Reach Power of 90%
| Binary outcome* | Time-to-event outcome† | Continuous outcome‡ | |||||||
| Study | Failure rate control arm | n | RR (90%) | Study | n | HR (90%) | Study | n | Cohen d (90%) |
| de Bree§ (14) | 0.38 | 75/75 | 0.36 | Beanlands (21) | 136/418 | 0.56 | Herder (20) | 233/232 | 0.30 |
| Fischer (25) | 0.42 | 91/98 | 0.46 | Sobhani (18) | 44/130 | 0.34 | |||
| Monteil║ (19) | 0.72 | 33/36 | 0.44 | Siebelink (22) | 24/103 | 0.20 | |||
| Mullani║ (16) | 0.71 | 105/105 | 0.68 | Tsai (23) | 25/129 | 0.21 | |||
| Maziak║ (17) | 0.93 | 162/167 | 0.86 | ||||||
| Ruers (26) | 0.45 | 75/75 | 0.43 | ||||||
| Viney (24) | 0.98 | 92/91 | 0.85 | ||||||
| van Tinteren (9) | 0.41 | 96/92 | 0.46 | ||||||
↵* Data are failure rate in control arm, number of patients in control arm and number in PET arm, and relative risk (RR) to be assumed.
↵† Data are number of events observed and number in overall sample, and hazard ratio (HR) to be assumed.
↵‡ Data are number of patients in control arm and number in PET arm, and Cohen d to be assumed. One study with crossover design (17) is not included in this table.
↵§ Assumptions are according to published study protocol.
↵║ Study reports success rate, which is transformed to failure rate.