Mark Bolton, Radiation Oncologist, shared a post on X:
“Comparison of TACE vs EBRT for BCLC A-B HCC
Assuming skeptical prior for treatment effect, the probability that EBRT offers better survival than TACE is a whopping 94.2%
Direct inference = directly clinically actionable; can mention at tumor board and to the patient; easily understood. Look at the confusion that arises in Frequentist context: what does “not statistically significantly different mean”? What does one do with that?
Is a reason why TACE is still being done as local control modality in trials simply because of errors of inference? “No difference in survival” is NOT the case; TACE should be kicked to the ash heap because it very very likely results in worse survival than EBRT.”
