The Final THC/CBD Chemotherapy Trial Analyzed First-Cycle Groups Rather Than Repeating Its Earlier Crossover Estimate
The final chemotherapy-nausea report used a parallel-group analysis of the first study cycle. It was not simply the earlier crossover result with additional participants attached.
The trial evolved across stages
The phase II component switched treatments between cycles, whereas the phase III component initially retained treatment assignment across cycles. The final report’s prespecified primary comparison used cycle A, before those later treatment changes, across recruited participants.
The estimates have different comparison structures
A crossover analysis compares periods within participants under additional assumptions. A first-cycle parallel analysis compares separately allocated groups. The earlier and final estimates should not be merged as though they were two independent trials involving entirely different people or identical estimators of the same dataset.
The final primary result retains its own definition
Complete response required no vomiting or retching and no rescue medication over 120 hours. The final first-cycle comparison favored THC/CBD, while several component or stricter outcomes were nonsignificant. The correct account specifies the analysis stage, cycle and response definition. This prevents an earlier favorable preference or crossover finding from being counted again as independent confirmation of the final trial’s primary treatment effect.
Related analysis: the trial’s main results and limitations.
