[Should every innovation be evaluated in an RCT? Considerations for the selection of patients for proton therapy]

Elsewhere in this journal an important question is raised: 'Are randomized trials (RCTs) necessary to demonstrate the superiority of radiation treatment, more specifically proton therapy?' We recognize that RCTs are considered to be the highest degree of certainty (evidence) to substantiate differences, as they allow to demonstrate or reject causality between an intervention and an outcome. However, in this comment we also raise some concerns about RCTs, and explain why especially in technological innovations, RCTs are not always the best approach. We explain why we consider the model-based approach as currently applied in The Netherlands for proton therapy, as complementary to RCTs, just as epidemiological or real-life 'big data' are, each with their advantages and disadvantages. We compare this approach with the FDA's Fast Track program, since proton therapy is a treatment that 'reasonably likely' causes less serious side effects with the same tumour effects.