Читать книгу Randomised Clinical Trials - David Machin - Страница 87

The primary endpoint is survival from the date of randomisation. In addition, although these measures will be optional, Child‐Pugh score and quality of life as assessed by the EORTC QLQ‐C30 will be recorded immediately prior to randomisation and monthly thereafter. Changes in these scores over time will also be compared between the patients in the three treatment groups.

We emphasise that whenever possible all trials should explicitly declare one, or possibly two, so‐called primary outcomes. Thus, for what appears as a several component primary objective, the PRESSURE (2000) trial in Example 3.5 requires a more precise definition to be included somewhere in the protocol. We also note in AHCC01 (1997) that the use of EORTC QLQ‐C30 implies 15 outcomes (scales), each outcome measured at multiple time points potentially resulting in a myriad of cross‐sectional or longitudinal analyses. Again, a more specific indication as to how this information will be summarised would be a necessary addition somewhere in the protocol.