Skip to main content

Advertisement

Fig. 3 | BMC Cancer

Fig. 3

From: Protocol for a phase III RCT and economic analysis of two exercise delivery methods in men with PC on ADT

Fig. 3

Illustration of Four Possible Scenarios in Non-Inferiority Testing for Proposed 2-Arm Trial. We will compare groups by means of ANCOVA using the baseline score as the covariate. After fitting a Bayesian model to each outcome, we will evaluate the probability in each of the 4 regions in the figure above. We can calculate the probability that: 1. Both the 6MWT and FACT-F fatigue scores are non-inferior in the home-based arm. This is the green region (b). 2. Both the 6MWT and FACT-F fatigue scores are inferior in the home-based arm. This is the grey region (c). 3. The 6MWT score is non-inferior in the home-based arm. This is the green region (b) plus the blue region (a). 4. The FACT-F Fatigue score is non-inferior in the home-based arm. This is the green region (b) plus the gold region (d). If either probability in (3) or (4) is above 97.5%, then we will declare non-inferiority of the home-based intervention.

Back to article page