Hi @charlton,
Pragmatic trials are interventional in nature, as they involve random assignment to the treatment groups. Given this fact, it would be inaccurate to refer to them as ‘observational’ studies. However, they differ in important ways from randomized control trials (RCTs).
RCTs are designed in a manner that prioritizes a highly controlled environment. While this comes at the cost of an artificial environment that can differ from the actual surrounding environment, it reduces the potential for biases that could impair the validity of causal inferences. While their results should be generalized cautiously to the real-world, they have high internal validity when designed, executed, analyzed, and reported properly.
Pragmatic trials are designed to prioritize reflecting the surrounding environment. This is why they are said to provide insights about “effectiveness”, which is the benefit that the treatment produces in routine clinical practice. For example, pragmatic trials tend to include patients with comorbidities, the methodology aims to reflect the lower adherence levels that occur in clinical practice, blinding is rarely utilized (as it does not reflect real-life practice), and investigators have more clinical freedom to make therapeutic decisions (i.e., less protocol standardization than in RCTs).
Due to the above, PAAB considers pragmatic trials to fall within the realm of RWEs. Please see our RWE guidance document on parameters for acceptability and presentation format.